Plasma proteome and metabolites profiling reveals dynamics for adverse events and responses after neoadjuvant radiochemotherapy plus PDL1 blockade in microsatellite-stable locally advanced rectal cancer: A prospective longitudinal study

Plasma proteome and metabolites profiling reveals dynamics for adverse events and responses after neoadjuvant radiochemotherapy plus PDL1 blockade in microsatellite-stable locally advanced rectal cancer: A prospective longitudinal study | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Plasma proteome and metabolites profiling reveals dynamics for adverse events and responses after neoadjuvant radiochemotherapy plus PDL1 blockade in microsatellite-stable locally advanced rectal cancer: A prospective longitudinal study Jianmin Xu, Yang Lv, Zhehui Zhu, Peng Zheng, Dexiang Zhu, Qi Lin, and 5 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5509842/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Neoadjuvant radio-chemotherapy (nCRT) plus immune checkpoint inhibitors (ICIs) have emerged as an effective antitumor regimen for locally advanced rectal cancer. Yet, few effective biomarkers are developed to monitor the therapy response. Herein, we investigate the longitudinal plasma proteome and metabolites profiling including 117 longitudinal samples from 50 patients who received nCRT plus PDL1 blockade therapy. Notably, the cholesterol metabolism is activated in the disease non-response group during the therapy. Correspondingly, the 1,4-cholestadienone, 7-methyloctanoylcarnitine and 3-hydroxybutyrylcarnitine, ABCA13, RAB3IP, GBA2 show significantly positive association with the cholesterol metabolism. Furthermore, by integrating proteome and metabolites approach, we identify a candidate metabolite (phosphatidylcholine, cholest-5-en-23-yn-3beta-ol) and proteome (APBB1IP, OLFM4, DNAJC19) that can reflect nCRT plus PDL1 response. Above, we establish a machine learning model to predict response, and the model performance is validated by repeated group-to-group validation with accuracy is 0.954. Thus, the plasma proteome and metabolites profiling strategy evaluate the alteration of cholesterol metabolism and identifies a panel of biomarkers. Biological sciences/Immunology Health sciences/Biomarkers/Predictive markers Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Locally advanced rectal cancer (LARC) presents significant treatment challenges because of its potential for local invasion and distant metastasis 1 . Since 1990, neoadjuvant therapy, which is administered before surgical resection, has aimed to improve outcomes by reducing tumour size, facilitating surgical resection, and enhancing overall survival 2 , 3 . Traditionally, neoadjuvant treatment (nCRT) involves a combination of chemotherapy and radiotherapy to shrink tumours and improve resectability 4 , 5 . Immunotherapy, particularly immune checkpoint inhibitors (ICIs, such as pembrolizumab and nivolumab), has revolutionized the treatment of various cancers, including colorectal cancer (CRC) 6 , 7 . In LARC, checkpoint inhibitors have shown potential in patients with high microsatellite instability (MSI-H) or mismatch repair deficiency (dMMR), thus providing new avenues for treatment in patients who are less responsive to conventional therapies 8 . However, for proficient MMR (pMMR) or microsatellite stable (MSS) LARC patients, the combination of ICIs with standard radiochemotherapy is an area of active research. Preliminary studies have suggested that such combinations may enhance the immune response against tumour cells, thus potentially improving outcomes in LARC patients 9 , 10 , 11 . These studies demonstrated that ICIs plus radiochemotherapy (short-course or long-course) resulted in a significantly higher pCR rate (nearly 40%) with a well-tolerated safety profile. It is evident that this treatment model represents a logical next step in immunotherapy to improve response rates and increase cure rates and response duration. However, some issues related to nCRT plus ICIs in LARC patients still need to be addressed. For example, predictive biomarkers have not yet been developed to identify whether patients will be able to respond to this treatment model 11 . In ICI therapy, especially anti-PD1 (programmed cell death 1) or anti-PDL1 (programmed cell death ligand 1) therapy, some biomarkers, including PD-L1 status and the TMB, have strong predictive power before therapy 12 , 13 . Regrettably, for MSS/pMMR rectal cancer, the role of biomarkers in the response to ICIs has not been extensively studied. The exploration of response markers will considerably promote LARC clinical trials and the development of precision medicine 14 . Recently, blood plasma proteomic and metabolomic profiling has been reported as being a powerful approach for biomarker discovery, with the potential to identify the heterogeneous mechanisms of response and resistance to therapy 14 , 15 , 16 and aid in the development of personalized treatment strategies based on ICIs used in combination with other agents 17 . It is expected that the integration of plasma proteomics and metabolomics expression signatures will improve predictive biomarker algorithms 18 . Thus, in this study, we performed an integrated analysis of plasma proteomics and metabolomics in LARC patients before and during neoadjuvant radio-chemoimmunotherapy. Hence, the identification of multiomics biomarkers that can be readily evaluated through peripheral blood sampling is crucial for real-time implementation in routine clinical practice. To the best of our knowledge, we report the first large analysis of plasma proteome and metabolomics from LARC patients treated with neoadjuvant immune checkpoint blockers (ICBs). By integrating the multiomics of plasma and incorporating a machine learning algorithm, we can construct a robust model as a prediction tool for highly accurate responses. Methods and materials Patients and treatment regimen From June 20, 2020, to August 15, 2023, a total of 80 patients were enrolled, and 50 patients with complete evaluable specimen were included in this study. Patient disposition and samples collection are summarized in Figure S1 . This prospective study was performed at Zhongshan Hospital of Fudan University, in accordance with the Declaration of Helsinki. Ethical approval was obtained from the Institutional Review Board of Zhongshan Hospital of Fudan University (Shanghai, P.R. China), and written informed consent was provided by all subjects before sampling. The major eligibility criteria included: treatment-naive individuals with rectal adenocarcinoma; pMMR, as proven by immunohistochemistry (IHC) for mismatch repair proteins (MLH1, MSH2, MSH6, PMS2); clinical stage T3N + M0 or T4NanyM0; aged between 18 and 75 years old; an Eastern Cooperative Oncology Group (ECOG) performance status 0–1; and adequate organ function. Tumor staging was carried out using the 8th Edition of the American Joint Committee of Cancer (AJCC) tumor-node-metastasis (TNM) staging classification. To determine the TNM stage, enhanced CT scans of the chest and abdomen, as well as MRI scans of the rectum, were conducted. Deep Mass spectrometry analysis Nanomagnetic bead-mediated enrichment of low-abundance plasma proteins Using EasyPept DeeP low abundance protein enrichment and pre-treatment reagent kit (Omicron, Shanghai, China) to enrich low abundance proteins in plasma samples. According to the manufacturer's instructions, 1 mg (40 µL) of magnetic nanoparticles suspension was taken, and the magnetic beads were separated by magnetic separation to discard the supernatant. After multiple washes of the magnetic beads, they were resuspended and 100 µL of serum/plasma was added and placed in a flip mixer (360° rotation mixing) at 37℃ for 1 hour. The supernatant was separated by magnetic separation, 300 µL of washing solution was added and shaken for 5 minutes for washing, and the washing was repeated three times. After enzymatic hydrolysis to convert proteins into peptides, reductive alkylation and desalting were performed. The total peptide concentration was determined by nanodroplet method. Data independent acquisition(DIA)Mass spectrometry analysis The Proteomic data analysis was performed by Shanghai Luming biological technology co., LTD (Shanghai, China). TimsTOF Pro2 mass spectrometer (Bruker) and nanoElute (Bruker) were used for both shotgun proteomics and DIA experiments. Samples were loaded and separated by a C18 column (25 cm × 75 µm) on an EASY-nLCTM 1200 system (Thermo, USA). The flow rate was 300 nL/min and linear gradient was 60 min. The flow rate was 300 nL/min and linear gradient was set as follow: 0 ~ 45 min, 5–27% B;45 ~ 50 min, 27–46% B༛50 ~ 55min, 46–100% B; 55 min ~ 60 min, 100% B. For DIA, 56 DIA windows were acquired (automatic gain control target 3e6 and auto for injection time) and the collision energy was ramped linearly as a function of the mobility from 59 eV at 1/K0 = 1.6 Vs cm − 2 to 20 eV at 1/K0 = 0.6 Vs cm − 2.The MS/MS spectra were recorded from 100 to 1700 m/z. Database search The default factory settings were used for the Spectronaut Pulsar 18.4 (Biognosys, Swiss) search and library generation (including Trypsyin as enzyme, up to two missed cleavages allowed Oxidation of Me as variable modifications, carbamidomethyl as fixed modification, and 1% FDR for PSM, peptide and protein identification). The DIA data were analysed with Spectronaut searching the above constructed spectral library. Main parameters of the software were set as follows: Precursor Qvalue cutoff and Protein Qvalue cutoff were set as 0.01, Normalization Strategy was set as Local Normalization and use MS2 as Quantity MS-Level. Statistical analyses A total of 5611 proteins expressed were identified as belonging to the proteome of plasma in this study. The thresholds of fold change and P-value < 0.05 were used to identify differentially expressed proteins (DEPs). Annotation of all identified proteins was performed using GO ( http://www.blast2go.com/b2ghome ; http://geneontology.org/ ) and KEGG pathway ( http://www.genome.jp/kegg/ ). DEPs were further used for GO and KEGG enrichment analysis. Protein-protein interaction analysis was performed using the String ( https://string-db.org/ ). Metabolomics Sample Preparation Samples stored at -80 ℃ were thawed at room temperature. 50 µL of sample was added to a 1.5 mL Eppendorf tube with 5µL of L-2-chlorophenylalanine (0.06 mg/mL) dissolved in methanol as internal standard, and the tube was vortexed for 10 s. Subsequently, 5 µL of ice-cold mixture of methanol and acetonitrile (2/1, vol/vol) was added, and the mixtures were vortexed for 1 min, and the whole samples were extracted by ultrasonic for 10 min in ice-water bath, stored at -20 ℃ for 30 min. The extract was centrifuged at at 4°C (13000 rpm) for 10 min. 5 µL of supernatant in a glass vial was dried in a freeze concentration centrifugal dryer. 5 µL mixture of methanol and water (1/4, vol/vol) were added to each sample, samples vortexed for 30 s, extracted by ultrasonic for 3 min in ice-water bat, then placed at -20°C for 2 h. Samples were centrifuged at 4°C (13000 rpm) for 10 min. The supernatants (5 µL) from each tube were collected using crystal syringes, filtered through 0.22 µm microfilters and transferred to LC vials. The vials were stored at -80°C until LC -MS analysis. QC samples were prepared by mixing aliquot of all samples to be a pooled sample. LC-MS/MS analysis The metabolomic data analysis was performed by Shanghai Luming biological technology co., LTD (Shanghai, China). An ACQUITY UPLC I-Class plus(Waters Corporation, Milford, USA) fitted with Q-Exactive mass spectrometer equipped with heated electrospray ionization (ESI) source (Thermo Fisher Scientific, Waltham, MA, USA) was used to analyze the metabolic profiling in both ESI positive and ESI negative ion modes. An ACQUITY UPLC HSS T3 column (1.8 µm, 2.1 × 100 mm) were employed in both positive and negative modes. The binary gradient elution system consisted of (A) water (containing 0.1% formic acid, v/v) and (B) acetonitrile (containing 0.1% formic acid, v/v) and separation was achieved using the following gradient: 0.01 min, 5% B; 2min, 5% B; 4min, 30% B; 8min, 50% B; 10min, 80% B; 14min, 100% B; 15 min, 100% B; 15.1 min, 5% and 16 min, 5%B. The flow rate was 0.35 mL/min and column temperature were 45℃. All the samples were kept at 10℃ during the analysis. The mass range was from m/z 100 to 1,000. The resolution was set at 70,000 for the full MS scans and 17500 for HCD MS/MS scans. The Collision energy was set at 10, 20 and 40 eV. The mass spectrometer operated as follows: spray voltage, 3800 V (+) and 3200 V (−); sheath gas flow rate, 35 arbitrary units; auxiliary gas flow rate, 8 arbitrary units; capillary temperature, 320°C; Aux gas heater temperature, 350°C; S-lens RF level, 50. Data Preprocessing and Statistical Analysis The original LC-MS data were processed by software Progenesis QI V2.3 (Nonlinear, Dynamics, Newcastle, UK) for baseline filtering, peak identification, integral, retention time correction, peak alignment, and normalization. Main parameters of 5 ppm precursor tolerance, 10 ppm product tolerance, and 5% product ion threshold were applied. Compound identifications were based on precise mass-to-charge ratio (M/z), secondary fragments, and isotopic distribution using The Human Metabolome Database (HMDB), Lipidmaps (V2.3), Metlin, and self-built databases. The extracted data were then further processed by removing any peaks with a missing value (ion intensity = 0) in more than 50% in groups, by replacing zero value by half of the minimum value, and by screening according to the qualitative results of the compound. Compounds with resulting scores below 36 (out of 60) points were also deemed to be inaccurate and removed. A data matrix was combined from the positive and negative ion data. The matrix was imported in R to carry out Principal Component Analysis (PCA) to observe the overall distribution among the samples and the stability of the whole analysis process. Orthogonal Partial Least-Squares-Discriminant Analysis (OPLS-DA) and Partial Least-Squares-Discriminant Analysis (PLS-DA) were utilized to distinguish the metabolites that differ between groups. To prevent overfitting, 7-fold cross-validation and 200 Response Permutation Testing (RPT) were used to evaluate the quality of the model. V variable Importance of Projection (VIP) values obtained from the OPLS-DA model were used to rank the overall contribution of each variable to group discrimination. A two-tailed Student’s T-test was further used to verify whether the metabolites of difference between groups were significant. Differential metabolites were selected with VIP values greater than 1.0 and p-values less than 0.05. Differential metabolites were further used to for KEGG pathway ( http://www.genome.jp/kegg/ ) enrichment analysis. Construction of Machine learning models. Machine learning was conducted to identify the responses. The graphical machine learning model construction pipeline is illustrated in Fig. 6 A. The machine learning pipeline was built in Python (version 3.9.15) using the following libraries: scikit-learn (version 1.2.1), numpy (version 1.26.3), scipy (version 1.12.0), and pandas (version 1.5.3). Results Patient cohorts and treatment regimens The analysed samples were collected from 50 patients with LARC who were receiving neoadjuvant radiochemotherapy plus PDL1 treatment ( Table 1 ). Patients who responded to neoadjuvant therapy (pCR) (specifically, patients with 0% residual viable tumour (RVT) after resection) were included in the analysis. Compared with the pCR cohort, the nonpCR cohort did not have differences in clinical characteristics at baseline (Table 1) . Blood samples were collected at baseline, postradiotherapy and postimmunotherapy during the therapy cycle for haematological evaluation ( Figure S1 and Methods ). Ultimately, 117 plasma samples were collected from the 50 enrolled tumour patients before and during neoadjuvant therapy. Among the 117 samples, 35 samples were collected at the baseline stage, 46 samples were collected after radiotherapy, and 36 samples were collected after the PDL1i stage. Among the 50 patients, a total of 20 patients were pathologically recognized as having CR, reaching a pCR rate of 40%. The study design is shown in Fig. 1 A. Deep proteomics dynamics of plasma in LARC patients receiving neoadjuvant chemotherapy plus PDL1i To detect changes in the plasma proteome as a result of PDL1 treatment, we analysed serial pretrm and trm plasma samples. An EasyPept DeeP low-abundance protein enrichment and pretreatment reagent kit (Omicron, Shanghai, China) was used to enrich low-abundance proteins in the plasma samples 19 . For quality control of the performance of deep proteomics, mixtures of all of the plasma samples were measured every twenty samples, and this protocol was adopted in the proteomic studies. Pearson’s correlation coefficient was calculated for all of the quality-control runs, and the results are shown in Fig. 1 B. Proteomic analysis revealed 2023–5035 gene products (GPs) in each sample (median number of 4282 in Fig. 1 B). A total of 5611 gene products (GPs) were identified in all of the plasma samples of the recruited cohort, of which 83250 peptide GPs were identified ( Figure S2A and S2B ; Supplementary data 1 and data 2 ). The molecular weights, peptide lengths and numbers are also shown in Figure S2C to S2E . To explore the molecular differences between CRC patients and healthy controls, we performed a comparative proteomic analysis of response samples and nonresponse samples. Heatmap analysis revealed the top 50 proteins between the samples from the two cohorts (Fig. 1 C and Figure S3A ). We observed that 204 proteins were significantly upregulated and that 68 were significantly downregulated (adjusted P value 2 in Figure S3B and Supplementary data 3 ) according to the protein classes in the Human Protein Atlas (HPA) database 20 . Thus far, our study has presented a comprehensive overview of the plasma proteomic landscape of the CRC cohort. We then performed pathway enrichment analysis on the differentially expressed proteins (DEPs) via Gene Ontology (GO) 21 , Interpro 22 , Reactome 23 and Wikipathway 24 analyses. As shown in Fig. 1 D, 1 E, 1 F and 1 G, metabolism, fatty acid and lipoprotein transport mechanisms, phosphatidylinositol 3-kinase regulator activity and oxidative phosphorylation were enriched in the plasma samples of responsive patients. In addition, the KEGG analysis revealed genes related to neutrophil degranulation and the innate immune system, among other activities (adjusted P value < 0.05) ( Figure S3C ). Alterations in plasma metabolites in association with ICB response in LARC patients To investigate the metabolomic profile before and during neo-CRT plus PDL1 blockade, we also collected 117 plasma samples and performed untargeted liquid chromatography‒mass spectrometry (LC‒MS) metabolomics analysis (Fig. 1 A). Four levels and a total of 4,791 endogenous metabolites were identified ( Figure S4A ). A box plot of the metabolite intensity of each sample is shown in Fig. 2 A, which indicates that the skewness of the distribution was greater for further analysis. The results of the quality control analysis are shown in Figure S4B . The subclasses of the detected metabolites are shown in Fig. 2 B. Most of the classes included amino acids, peptides, and analogues (10.48%), followed by fatty acids and fatty acid conjugates (8.50%), carbohydrates and carbohydrate conjugates (3.99%), and bile acids (2.44%). PCA and OPLS-DA analyses revealed that the plasma metabolome was significantly different between the response and nonresponse cohorts (both p < 0.05; Fig. 2 C and Figure S4C ). To characterize the altered plasma metabolites according to response status, we conducted a heatmap comparison to identify metabolites with significantly differential peaks between different groups (Fig. 2 D, Figure S4D ; Supplementary data 4 ). Differential plasma metabolites with consistent shifts across the response and nonresponse cohorts were revealed, with 31 sequentially enriched metabolites, including glycoursodeoxycholic acid, (9E)-10-nitrophenyloctadec-9-enoylcarnitine, trans-3-octenedioic acid and multiple fatty acids (Fig. 2 E and Figure S4E ). We also observed 67 sequentially depleted metabolites, including Cholest-5-en-3-ol (3. beta.) butanoate, N-palmitoyl taurine and maracin A (Fig. 2 E). To gain insight into the functional significance of the altered metabolites, we conducted a correlation network analysis, as shown in Fig. 2 F, and the associated metabolites mostly included bile acids and lipid acids. Among the plasma metabolites, the top 3 differential pathways between the two response cohorts were (1) primary bile acid biosynthesis, (2) cholesterol metabolism, and (3) taurine and hypotaurine metabolism (KEGG analysis in Fig. 2 G). The top 3 differential pathways from the reactome database were (1) fructose biosynthesis, (2) digestion of dietary carbohydrates, and (3) fructose metabolism (Fig. 2 H). Collectively, metabolomics analysis demonstrated that fatty acid, bile acid, and lipid metabolism were involved in ICB resistance in LARC plasma. Analysis of the plasma proteome and metabolites revealed a significant relationship between cholesterol metabolism and response in LARC patients after nCRT plus PDL1i Another aim of this study was to investigate the effects of PDL1i on the plasma proteome and metabolites of patients. Therefore, we compared both proteomic and metabolomic characteristics among the baseline cohort, post-radiotherapy cohort, and post-immunotherapy cohort. Differential expression protein (DEP) analysis was performed among the different treatment stages, as shown in Fig. 3 A, Figure S5 and Supplementary Data 5 to 8 . A Venn diagram was drawn, and 92 response-specific proteins were screened out while 88 nonresponse-related proteins were revealed (Fig. 3 B). These genes include multiple cholesterol-related markers, such as ABCA13, RAB3IP, GBA2, BCS1L, FAH, ABHD17A and SREBF2. Protein N-linked glycosylation and peptidyl-asparagine modification were significantly enriched in nonresponsive proteins, whereas KEGG analysis revealed significant functional alterations in N-glycan biosynthesis and protein processing in the endoplasmic reticulum (Fig. 3 B). As shown in Fig. 3 C, protein-to-protein interaction (PPI) network analysis revealed dramatic alterations in proteome interactions, including cholesterol metabolism markers (p < 0.05). Pathway alterations in the response cohort after PDL1 blockade were revealed at baseline and post immunotherapy, and Fig. 3 D shows alterations in the drug metabolism signature. In addition, differential expression metabolite (DEM) analysis was performed to explore the significant alterations in LARC patients after nCRT plus PDL1i. As demonstrated in Fig. 3 E and Figure S6 , in a similar manner, the response in rectal cancer patients to nCRT plus PDL1i exhibited increased expression of many cholesterol-related markers, including 1,4-cholestadienone, 7-methyloctanoylcarnitine and 3-hydroxybutyrylcarnitine. Functional alterations from metabolomics were also compared between different response cohorts, as shown in Fig. 3 F. The biosynthesis of unsaturated fatty acids, as well as arginine and proline metabolism, were significantly enriched in the response cohort, whereas ether lipid metabolism was enriched in the nonresponse cohort. Metabolites in the response cluster, which suggested the recovery of cholesterol metabolism markers after nCRT plus PDL1 blockade, included 1,4-cholestadienone, 7-oxo-cholesterol and PA (18:0/0:0) (Fig. 3 G and 3 H). Collectively, these results demonstrated the strong impacts of PDL1i on cholesterol processes and oncogenic signalling. The correlation between paired plasma proteins and metabolites in response to ICBs in LARC patients was examined using Spearman correlation analysis. As shown in Fig. 4 A and Figure S7 , we identified the top 20 significantly correlated metabolites and proteins between multi-omics (correlation coefficient > 0.3, FDR < 0.05, p < 0.05). These included glycochenodeoxycholic acid , 14(R)-hydroxyretrovitamin A , PG (20:4(5Z,8Z,10E,14Z)-OH(12S)/i-24:0) and cholest-5-en-23-yn-3beta-ol . Accordingly, integrated proteomics and metabolomics analyses were performed at baseline, postradiotherapy, and posti-mmunotherapy(Fig. 4 B to 4 E). Notably, sedoheptulose and the splicing factor C9orf78 were regarded as being common altered factors between the treatments (Fig. 4 B). Functionally, we also noted that there were very few overlapping pathways between the proteomic and metabolomic data (Fig. 4 F) from the KEGG database, which included those related to cholesterol metabolism, bile secretion and galactose metabolism (Fig. 4 G). Collectively, these results demonstrate the potential role of the cholesterol-related pathway as a biomarker for the PDL1 inhibitor response in LARC. Circulating proteomic and metabolic signatures associated with nCRT plus PDL1i-related side effects in LARC patients Next, we identified circulating proteomic and metabolic differences between adverse events (AEs) and Nonadverse events (NAEs) and demonstrated more differential proteins or metabolites associated with AEs, which were highly biologically distinct. PCA between AE samples and NAE samples at metabolic and protein level were shown in Fig. 5 A. A broad overview of expression heatmap was also displayed in Fig. 5 B, volcano map for significantly expressed proteins was shown in Figure S8A-S8B and Supplementary data 9 . Protein to protein interaction between AE and NAE samples was shown in Fig. 5 C, which revealed that various kinds of interactions, including MYD88, SF3A3, and SNW1 were statistically correlated with AEs. Functional alteration at protein level was also shown in Fig. 5 D. In addition, metabolite alterations between the AE and NAE cohorts, including alterations in N-Palmitoyl Histidine , Chenodeoxycholylphenylalanine , 4,4'-Thiobis-2-butanone , Naratriptan , Chenodeoxyglycocholic acid and 11-deoxy-PGF1a were shown in Fig. 5 E, Figure S8C and Supplementary data 10 . Besides, functional alterations from KEGG database demonstrated Ferroptosis, cytochrome P450, Pyrimidine metabolism pathway were enriched in AE cohorts (Fig. 5 F). Also, correlation between paired plasma proteins and metabolites in AE to ICBs in LARC patients was examined using Spearman correlation analysis. The top 20 significantly correlated metabolites and proteins were identified between multi-omics (correlation coefficient > 0.3, FDR < 0.05, p < 0.05) in Fig. 5 G. Functionally, integrated analysis noted that there were very few overlapping pathways between the proteomic and metabolomic data from the KEGG database, which included those related to cytochrome P450 metabolism and Pyrimidine metabolism (Fig. 5 H). Collectively, our results revealed unique proteomic and metabolic profiles in different AE groups, which may be associated with the mechanisms underlying immunotherapy-related AEs in LARC patients. Integrated proteome-metabolic feature-based machine learning model provides an accurate measure for the nCRT plus PDL1i response In this study, we aimed to construct an integrated proteome‒metabolic predictive signature for responses. As shown in Fig. 6 A, we identified significant metabolites and proteomes via a machine learning framework. The models were based on a multistep processing pipeline (Fig. 6 A). Inside of the pipeline, for each feature or classifier combination, we compared different data preprocessing strategies along with machine learning models and selected the top N (such as N = 5) best models 25 , 26 . We screened the rfcv function, performed k-fold cross validation, repeated it 1000 times, and calculated the average error cv value 1000 times. This process was repeated N times, and a line graph was drawn on the basis of the average value of N times in Fig. 6 B. The top 30 metabolites and proteomes are shown in Fig. 6 C and were ranked by using the mean decrease Gini score. Via feature-based biomarker construction, the proteomes (APBB1IP, OLFM4, and DNAJC19) demonstrated better predictive power for PDL1 responses, whereas the proteome panels (APBB1IP, OLFM4, MRPL58, ERGIC3 and DNAJC19) were determined to have better predictive power via classifier-based statistics (Fig. 6 D and 6 E). Predictive metabolites were also screened out, as shown in Fig. 6 F. By overlapping the significant biomarkers, we identified metabolite (phosphatidylcholine, cholest-5-en-23-yn-3beta-ol) and proteome (APBB1IP, OLFM4, DNAJC19) features that were present in the longitudinal cohort after nCRT plus PDL1i. This result led to the use of a machine learning framework (Fig. 6 A) to integrate features into a predictive model of PDL1 responses to predict disease progression in rectal cancer patients. There are five optionally different feature combinations based on metabolic and proteome features: (1) phosphatidylcholine, (2) cholest-5-en-23-yn-3beta-ol, (3) APBB1IP, (4) OLFM4, (5) DNAJC19, and (6) all of the integrated features. To examine the robustness and efficacy of the diagnostic strategy, we conducted repeated group-to-group validation analyses of the baseline cohort and postradiotherapy cohort and calculated a model predictive ability of 0.954; in contrast, postimmunotherapy validation revealed a predictive ability of 0.877 (Fig. 6 G). Overall, the selected features from each group showed excellent potential for identifying LARC patients with pathological responses to neoadjuvant immunotherapy, thus indicating the great potential of this screening strategy for people at pretreatment stages. Discussion Neoadjuvant immunotherapy is designed to stimulate the immune system prior to surgical intervention, thus potentially improving outcomes by targeting the tumour microenvironment (TME) and enhancing systemic antitumour responses 27 . Recent studies have indicated that immune checkpoint inhibitors, such as PD-1/PD-L1 and CTLA-4 inhibitors, can lead to significant tumour regression in patients with dMMR or MSI-H rectal cancers 28 . For MSS rectal cancer, with the stimulated TMEs elicited by radiotherapy, neoadjuvant immunotherapy has emerged as a promising approach for rectal cancer treatment, particularly in cases where traditional modalities have shown limited efficacy 29 , 30 , 31 . However, most published studies have indicated that 40% of MSS rectal cancer patients achieve pathological complete response (pCR), which demonstrates that another 60% of patients have a limited response to nCRT plus PDL1 blockade. Responses can vary widely among individuals, thus highlighting the need for reliable biomarkers to predict which patients will benefit most from such therapies. In this context, the integration of plasma proteomics and metabolomics offers a novel framework for predicting therapeutic responses and enhancing treatment strategies 32 . Plasma proteomics focuses on the comprehensive analysis of proteins that are present in the blood plasma, which can reflect tumour biology and host immune responses. Recent advances in mass spectrometry and bioinformatics have enabled researchers to identify specific protein signatures associated with tumour progression and treatment response 16 , 33 . For example, certain cytokines and immune-related proteins have been linked to favourable outcomes in immunotherapy 34 . In addition, metabolites can provide insights into the metabolic states of tumours and their microenvironments. In rectal cancer, alterations in metabolic pathways have been associated with tumour aggressiveness and treatment resistance 35 , 36 . For example, specific metabolic signatures may indicate a greater likelihood of immune evasion or lower responsiveness to therapy 37 . The use of metabolomic profiling in conjunction with proteomic analysis can enhance our understanding of the biochemical changes associated with rectal cancer and improve the ability of these methods to predict therapeutic efficacy. In this work, we aimed to systematically describe comprehensive alterations in the plasma proteome and metabolites and explore potential biomarkers of PDL1 blockade in MSS rectal cancer patients. We collected longitudinal samples from rectal cancer patients treated with PDL1 blockade and performed advanced sample preparation, in-depth MS detection, MS data preprocessing, LC-MS sequencing and bioinformatic analysis, thus resulting in a comprehensive proteome and metabolite workflow 32 . This longitudinal study of PDL1 blockade revealed proteome and metabolite fluctuations, which may facilitate biomarker exploration and further speculation/research on the biological mechanisms. In this study, we described the link between the PDL1 blockade response and cholesterol metabolism. Furthermore, we proposed a metabolite (phosphatidylcholine, cholest-5-en-23-yn-3beta-ol) and proteome (APBB1IP, OLFM4, DNAJC19) as being potential response biomarkers for PDL1 blockade. In our cohort, we observed enrichment of cholesterol metabolism-related processes in nonresponsive samples after radiotherapy, whereas functional alterations in the cholesterol transport process were also observed via both proteomics and metabolomics. These results implied that, for nonresponsive cohorts, increased cholesterol was detected in the blood, which is consistent with the findings of a previously published study 38 , 39 . On the basis of these results, the increase in blood cholesterol was regarded as being the result of a nonresponse to nCRT plus PDL1 blockade, which demonstrated that anti-cholesterol metabolism treatment may be a drug target for overcoming PDL1 resistance; consequently, the specific biological mechanism needs to be further investigated. Another key strength of this study was the use of longitudinal blood samples from MSS rectal cancer patients, which allowed us to evaluate fluctuations in protein and metabolite levels by setting the treatment time as a covariate. Notably, the use of these samples in combination with proteomic and metabolomic features outperformed other methods for predicting the response of patients to nCRT plus PDL1 blockade. The combination of proteomics and metabolomics represents a powerful approach for developing predictive models for neoadjuvant immunotherapy. This integrative strategy allows for a more comprehensive understanding of tumour biology and its interactions with the host immune system. By correlating protein expression profiles with metabolic alterations, such models could significantly improve the precision of immunotherapy, thus enabling tailored approaches that consider both the proteomic landscape and the metabolic context of the tumour. This study has several limitations. First, although our plasma-based analysis provided valuable insights, the heterogeneity of patient responses could not be fully identified due to the inherent complexity of the TME. Second, another limitation of the study includes the relatively small sample size, which may not fully represent the broader patient population; however, the longitudinal study design can partly offer a more comprehensive understanding of treatment response over time. Another limitation is the potential for bias in patient selection, as our study cohort may not encompass the full spectrum of rectal cancer patients. This could affect the representativeness of the proteomic and metabolomic profiles that we identified. In conclusion, our findings suggest several avenues for future research. First, we introduced the proteomic and metabolomic landscape of the plasma and designed a longitudinal study with MSS rectal cancer patient cohorts receiving nCRT plus PDL1 blockade. Additionally, explorations of the mechanisms underlying the observed alterations in cholesterol metabolism could provide deeper insights into how immunotherapy exerts its effects and can help in identifying novel therapeutic targets. Moreover, this integrated model may ultimately lead to more effective and personalized treatment strategies for RC. Declarations Acknowledgement This manuscript has not been submitted to any other journal and is not currently being considered for publication by another journal. We Thank the doctors and nurses who performed the treatment. Ethics approval and consent to participate Written informed consent was obtained from all patients. The study protocol followed the ethical guidelines of the Declaration of Helsinki and was approved by the Ethics Committee of Zhongshan Hospital of Fudan University. The ethics approval ID was ZS2022-019. Consent for publication We have obtained consent to publish from the participant to report individual patient data. Availability of data and material The datasets used and/or analyzed during the current study are available from the corresponding authors upon reasonable request. Conflicts of interest The authors declare that they have no competing interests. Authors' contributions Prof. Xu JianMin and Prof. Ye LeChi contributed to the design of this study. Dr. Lv Yang, Zhu ZheHui, Tang WenTao, Ji MeiLing and Zheng Peng performed the research, and Dr. Lv Yang, Tang Wentao, Zhu DeXiang and Zheng Peng analyzed and interpreted the patient data. Dr. Lv Yang was a major contributor to the writing of the manuscript. Dr. Ji MeiLing provided research background and perspectives. Prof. Xu JianMin, Dr. Tang WenTao, and Prof. Ye LeChi were the corresponding authors and approved the final version of this manuscript for publication. Funding This work was supported by the Natural Science Foundation of China (82303889, 82172816), the Natural Science Foundation of Shanghai (23ZR1410400) and Shanghai Sailing Program (23YF1406100). The funding bodies had no role in the design of the study; collection, analysis, and interpretation of data; or in the writing of the manuscript. References Li Y , et al. A Review of Neoadjuvant Chemoradiotherapy for Locally Advanced Rectal Cancer. Int J Biol Sci 12 , 1022-1031 (2016). NIH consensus conference. Adjuvant therapy for patients with colon and rectal cancer. JAMA 264 , 1444-1450 (1990). Sauer R , et al. Preoperative versus postoperative chemoradiotherapy for rectal cancer. N Engl J Med 351 , 1731-1740 (2004). Kasi A , et al. Total Neoadjuvant Therapy vs Standard Therapy in Locally Advanced Rectal Cancer: A Systematic Review and Meta-analysis. JAMA Netw Open 3 , e2030097 (2020). Petrelli F , et al. Total Neoadjuvant Therapy in Rectal Cancer: A Systematic Review and Meta-analysis of Treatment Outcomes. Ann Surg 271 , 440-448 (2020). Malmberg R , et al. Alternative dosing strategies for immune checkpoint inhibitors to improve cost-effectiveness: a special focus on nivolumab and pembrolizumab. Lancet Oncol 23 , e552-e561 (2022). Matsumoto K , et al. Real-world outcomes of nivolumab plus ipilimumab and pembrolizumab with platinum-based chemotherapy in advanced non-small cell lung cancer: a multicenter retrospective comparative study. Cancer Immunol Immunother 73 , 4 (2024). Yu JH , et al. Long-Term Outcomes of dMMR/MSI-H Rectal Cancer Treated With Anti-PD-1-Based Immunotherapy as Curative-Intent Treatment. J Natl Compr Canc Netw 22 , (2024). Xiao WW , et al. Effect of neoadjuvant chemoradiotherapy with or without PD-1 antibody sintilimab in pMMR locally advanced rectal cancer: A randomized clinical trial. Cancer Cell , (2024). Xia F , et al. Randomized Phase II Trial of Immunotherapy-Based Total Neoadjuvant Therapy for Proficient Mismatch Repair or Microsatellite Stable Locally Advanced Rectal Cancer (TORCH). J Clin Oncol , JCO2302261 (2024). Lin ZY , et al. Neoadjuvant short-course radiotherapy followed by camrelizumab and chemotherapy in locally advanced rectal cancer (UNION): early outcomes of a multicenter randomized phase III trial. Ann Oncol , (2024). Bortolomeazzi M , et al. Immunogenomics of Colorectal Cancer Response to Checkpoint Blockade: Analysis of the KEYNOTE 177 Trial and Validation Cohorts. Gastroenterology 161 , 1179-1193 (2021). Shan J, Han D, Shen C, Lei Q, Zhang Y. Mechanism and strategies of immunotherapy resistance in colorectal cancer. Front Immunol 13 , 1016646 (2022). Meric-Bernstam F, Larkin J, Tabernero J, Bonini C. Enhancing anti-tumour efficacy with immunotherapy combinations. Lancet 397 , 1010-1022 (2021). Di Meo F , et al. A target discovery pipeline identified ILT3 as a target for immunotherapy of multiple myeloma. Cell Rep Med 4 , 101110 (2023). Loriot Y , et al. Plasma proteomics identifies leukemia inhibitory factor (LIF) as a novel predictive biomarker of immune-checkpoint blockade resistance. Ann Oncol 32 , 1381-1390 (2021). Lyu J , et al. Plasma proteome profiling reveals dynamic of cholesterol marker after dual blocker therapy. Nat Commun 15 , 3860 (2024). Benson MD , et al. Protein-metabolite association studies identify novel proteomic determinants of metabolite levels in human plasma. Cell Metab 35 , 1646-1660 e1643 (2023). Nusinow DP , et al. Quantitative Proteomics of the Cancer Cell Line Encyclopedia. Cell 180 , 387-402 e316 (2020). Colwill K, Renewable Protein Binder Working G, Graslund S. A roadmap to generate renewable protein binders to the human proteome. Nat Methods 8 , 551-558 (2011). Gene Ontology C. Gene Ontology Consortium: going forward. Nucleic Acids Res 43 , D1049-1056 (2015). Paysan-Lafosse T , et al. InterPro in 2022. Nucleic Acids Res 51 , D418-D427 (2023). Milacic M , et al. The Reactome Pathway Knowledgebase 2024. Nucleic Acids Res 52 , D672-D678 (2024). Frohlich K, Furrer R, Schori C, Handschin C, Schmidt A. Robust, Precise, and Deep Proteome Profiling Using a Small Mass Range and Narrow Window Data-Independent-Acquisition Scheme. J Proteome Res 23 , 1028-1038 (2024). Rueda F , et al. Protein-based cardiogenic shock patient classifier. Eur Heart J 40 , 2684-2694 (2019). Shen B , et al. Proteomic and Metabolomic Characterization of COVID-19 Patient Sera. Cell 182 , 59-72 e15 (2020). Ganesh K , et al. Immunotherapy in colorectal cancer: rationale, challenges and potential. Nat Rev Gastroenterol Hepatol 16 , 361-375 (2019). Chalabi M , et al. Neoadjuvant Immunotherapy in Locally Advanced Mismatch Repair-Deficient Colon Cancer. N Engl J Med 390 , 1949-1958 (2024). Xia F , et al. Randomized Phase II Trial of Immunotherapy-Based Total Neoadjuvant Therapy for Proficient Mismatch Repair or Microsatellite Stable Locally Advanced Rectal Cancer (TORCH). J Clin Oncol 42 , 3308-3318 (2024). Yang Z , et al. Efficacy and safety of PD-1 blockade plus long-course chemoradiotherapy in locally advanced rectal cancer (NECTAR): a multi-center phase 2 study. Signal Transduct Target Ther 9 , 56 (2024). Wang F , et al. Efficacy and safety of combining short-course neoadjuvant chemoradiotherapy with envafolimab in locally advanced rectal cancer patients with microsatellite stability: A phase II PRECAM experimental study. Int J Surg , (2024). Navas-Carrillo D, Rodriguez JM, Montoro-Garcia S, Orenes-Pinero E. High-resolution proteomics and metabolomics in thyroid cancer: Deciphering novel biomarkers. Crit Rev Clin Lab Sci 54 , 446-457 (2017). Gao Y , et al. Acetylation-dependent regulation of PD-L1 nuclear translocation dictates the efficacy of anti-PD-1 immunotherapy. Nat Cell Biol 22 , 1064-1075 (2020). Babacic H, Lehtio J, Pico de Coana Y, Pernemalm M, Eriksson H. In-depth plasma proteomics reveals increase in circulating PD-1 during anti-PD-1 immunotherapy in patients with metastatic cutaneous melanoma. J Immunother Cancer 8 , (2020). Kroemer G, Kepp O. Radiochemotherapy-induced elevations of plasma HMGB1 levels predict therapeutic responses in cancer patients. Oncoimmunology 10 , 2005859 (2021). Wang H , et al. Serum metabolic traits reveal therapeutic toxicities and responses of neoadjuvant chemoradiotherapy in patients with rectal cancer. Nat Commun 13 , 7802 (2022). Liu H , et al. ADORA1 Inhibition Promotes Tumor Immune Evasion by Regulating the ATF3-PD-L1 Axis. Cancer Cell 37 , 324-339 e328 (2020). Liu X , et al. Inhibition of PCSK9 potentiates immune checkpoint therapy for cancer. Nature 588 , 693-698 (2020). Yang W , et al. Potentiating the antitumour response of CD8(+) T cells by modulating cholesterol metabolism. Nature 531 , 651-655 (2016). Additional Declarations There is NO Competing Interest. Supplementary Files supplementarydata1proteinsofallsamples.xlsx supplementary data 1 supplementarydata2pepetidesofallsamples.xlsx supplementary data 2 supplementarydata3volcanoAvsB.xls supplementary data 3 Supplementarydata4metabolitesheatmapAvsB.xls supplementary data 4 Supplementarydata5volcanoA1vsA2.xls supplementary data 5 Supplementarydata6volcanoA1vsA3.xls supplementary data 6 Supplementarydata7volcanoB1vsB2.xls supplementary data 7 Supplementarydata8volcanoB1vsB3.xls supplementary data 8 Supplementarydata9AEvsNAEproteins.xlsx supplementary data 9 Supplementarydata10AEvsNAEmetabolites.xlsx supplementary data 10 Supplementarymaterials.docx Supplementary materials Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5509842","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":449176805,"identity":"e26eaba5-8b36-4e86-913e-9cff819c7d13","order_by":0,"name":"Jianmin Xu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAtElEQVRIiWNgGAWjYHCCBCC0YeBjJlFLGgMbKVpA4DADG9FqDW4kPN3wcMd5eTZ2HjPpAgY7Od0GAlokZySk3Ug8c9uwjRmoZQZDsrHZAQJa+CVAWtpuM4K18DAcSNxGSAsbRMs5e+K1QG05kEi8FsmeByAtycltzGzF1jwGRPjF4HhO2s2fbXa2/fyHN97mqbCTI6iFgYEnAcrgMACaQFA5CLDDTGV/QJT6UTAKRsEoGHkAAEKYOvUvSf+CAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0001-5441-8240","institution":"Zhongshan Hospital affiliated to Fudan University","correspondingAuthor":true,"prefix":"","firstName":"Jianmin","middleName":"","lastName":"Xu","suffix":""},{"id":449176806,"identity":"c08cddbb-3756-4955-af3e-53f32188211c","order_by":1,"name":"Yang Lv","email":"","orcid":"https://orcid.org/0000-0002-6013-0558","institution":"Zhongshan Hospital, Fudan University","correspondingAuthor":false,"prefix":"","firstName":"Yang","middleName":"","lastName":"Lv","suffix":""},{"id":449176807,"identity":"75289a34-bb1e-4cfa-966d-7bd4d4ed6a41","order_by":2,"name":"Zhehui Zhu","email":"","orcid":"","institution":"Xinhua Hospital, Shanghai Jiao Tong University School of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Zhehui","middleName":"","lastName":"Zhu","suffix":""},{"id":449176808,"identity":"87e54138-dc59-4b04-a84e-06c531e17594","order_by":3,"name":"Peng Zheng","email":"","orcid":"","institution":"Department of General Surgery, Zhongshan Hospital, Fudan University","correspondingAuthor":false,"prefix":"","firstName":"Peng","middleName":"","lastName":"Zheng","suffix":""},{"id":449176809,"identity":"e7909e41-b702-432c-9fef-4180f42bcaae","order_by":4,"name":"Dexiang Zhu","email":"","orcid":"","institution":"Department of General Surgery, Zhongshan Hospital, Fudan University, Shanghai 200032, China","correspondingAuthor":false,"prefix":"","firstName":"Dexiang","middleName":"","lastName":"Zhu","suffix":""},{"id":449176810,"identity":"a5a6a8d0-e566-49e1-9144-c3acb78b6cf9","order_by":5,"name":"Qi Lin","email":"","orcid":"","institution":"Zhongshan hospital，Fudan University","correspondingAuthor":false,"prefix":"","firstName":"Qi","middleName":"","lastName":"Lin","suffix":""},{"id":449176811,"identity":"25fa8b07-0124-461f-9241-164f3d3d2825","order_by":6,"name":"Qingyang Feng","email":"","orcid":"","institution":"Fudan affiliated Zhongshan Hospital","correspondingAuthor":false,"prefix":"","firstName":"Qingyang","middleName":"","lastName":"Feng","suffix":""},{"id":449176812,"identity":"262a8c11-1ab7-44af-80ba-7d517f2c6c02","order_by":7,"name":"HongYu Zhang","email":"","orcid":"","institution":"Zhongshan Hospital，Fudan Universiy","correspondingAuthor":false,"prefix":"","firstName":"HongYu","middleName":"","lastName":"Zhang","suffix":""},{"id":449176813,"identity":"be6f5466-edb9-418e-8338-ab671b13cc78","order_by":8,"name":"Meiling Ji","email":"","orcid":"","institution":"Fudan affiliated Zhongshan Hospital","correspondingAuthor":false,"prefix":"","firstName":"Meiling","middleName":"","lastName":"Ji","suffix":""},{"id":449176814,"identity":"bbe87335-f5ff-470d-b8aa-d5b6837cd094","order_by":9,"name":"Lechi Ye","email":"","orcid":"","institution":"The First Affiliated Hospital of Wenzhou Medical University","correspondingAuthor":false,"prefix":"","firstName":"Lechi","middleName":"","lastName":"Ye","suffix":""},{"id":449176815,"identity":"4ca98d2e-b78e-4e3e-bacd-00d0ffeecb76","order_by":10,"name":"Wentao Tang","email":"","orcid":"","institution":"Zhongshan Hospital Fudan University","correspondingAuthor":false,"prefix":"","firstName":"Wentao","middleName":"","lastName":"Tang","suffix":""}],"badges":[],"createdAt":"2024-11-23 12:00:23","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5509842/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5509842/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":82148058,"identity":"52c4384b-4179-4b96-841c-618e36232bab","added_by":"auto","created_at":"2025-05-07 07:06:02","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":689153,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eWorkflow of this study, plasma proteomic landscape from MSS type rectal cancer patients receiving nCRT plus PDL1 blockade.\u003c/strong\u003e \u003cstrong\u003eA, \u003c/strong\u003ea total of 50 MSS rectal cancer patients receiving nCRT plus PDL1 blockade were recruited in this study. Proteomics and metabolomics analysis for plasma samples were performed. \u003cstrong\u003eB,\u003c/strong\u003e Quality control of the deep proteomics and Pearson’s correlation coefficient were shown. \u003cstrong\u003eC,\u003c/strong\u003e Heatmap for comparative proteomic analysis of Response samples and non-response samples was analyzed and the top 50 proteins were shown. \u003cstrong\u003eD,\u003c/strong\u003e GO pathway enrichment for differential expressed proteins between Response and Non-Response cohort. \u003cstrong\u003eE,\u003c/strong\u003eInterpro pathway enrichment for differential expressed proteins between Response and Non-Response cohort. \u003cstrong\u003eF,\u003c/strong\u003e Reactome pathway enrichment for differential expressed proteins between Response and Non-Response cohort. \u003cstrong\u003eG,\u003c/strong\u003e Wikipathway enrichment for differential expressed proteins between Response and Non-Response cohort. \u003cstrong\u003eAberration\u003c/strong\u003e: MSS, microsatellite stable; nCRT, neoadjuvant chemotherapy; PDL1, Programmed cell death 1 ligand 1; GO, gene ontology.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-5509842/v1/900e58bf65eec96902afcd1a.png"},{"id":82148062,"identity":"bc49e700-751b-4516-a28e-f8a4333b6889","added_by":"auto","created_at":"2025-05-07 07:06:02","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":616453,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePlasma metabolomics dynamics of MSS rectal cancer patients receiving nCRT plus PDL1 blockade.\u003c/strong\u003e \u003cstrong\u003eA, \u003c/strong\u003emetabolite intensity of each sample was shown by boxplot. \u003cstrong\u003eB,\u003c/strong\u003e Sub-class of detected metabolite was shown. Most of the class was Amino Acids, Peptides, and Analogues (10.48%), followed by fatty acids and conjugates (8.50%), Carbohydrates and Carbohydrate Conjugates (3.99%) and Bile Acids (2.44%). \u003cstrong\u003eC, \u003c/strong\u003ePCA plot was drawn and showed that the plasma metabolome was significantly different among Response and Non-Response cohorts (p \u0026lt;0.05). \u003cstrong\u003eD,\u003c/strong\u003e heatmap comparison was conducted to identify metabolites with significantly differential peaks between different groups. \u003cstrong\u003eE,\u003c/strong\u003e volcano analysis of Differential plasma metabolites with consistent shift across Response and Non-Response cohort were revealed, 67 sequentially depleted metabolites were identified, including Cholest-5-en-3-ol (3. beta.) butanoate, N-palmitoyl taurine and Maracin A. \u003cstrong\u003eF,\u003c/strong\u003e metabolites correlation network was drawn, and the associated metabolites were mostly bile acid and lipid acids. \u003cstrong\u003eG,\u003c/strong\u003e KEGG analysis demonstrated pathways in between two response cohorts. G, reactome analysis demonstrated altered pathways between two response cohorts. \u003cstrong\u003eAberration\u003c/strong\u003e: MSS, microsatellite stable; nCRT, neoadjuvant chemotherapy; PDL1, Programmed cell death 1 ligand 1; PCA, Principal Component Analysis; gene ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-5509842/v1/1e8ced524f6e169fccb49a24.png"},{"id":82148060,"identity":"076f3cca-1a5b-4430-8e54-4b819de9d12a","added_by":"auto","created_at":"2025-05-07 07:06:02","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":545368,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePlasma proteome and metabolites revealed significant linkage between cholesterol metabolism and response in MSS rectal cancer after nCRT plus PDL1i. A, \u003c/strong\u003evolcano plot for\u003cstrong\u003e \u003c/strong\u003eDEPs of proteome and metabolites at baseline, post-radiotherapy and post-immunotherapy. Specific data related to\u003cstrong\u003e supplementary data 5 to 8\u003c/strong\u003e. \u003cstrong\u003eB,\u003c/strong\u003e Venn analysis was performed, and 92 response-specific proteins were screened out, while 88 non-response-related proteins were revealed. \u003cstrong\u003eC\u003c/strong\u003e, PPI revealed dramatic alterations in the proteome interaction, including cholesterol metabolism markers (\u003cem\u003ep\u0026lt;0.05\u003c/em\u003e). Pathway alteration in response cohort after PDL1 blockade was revealed among baseline and post-immunotherapy. \u003cstrong\u003eD, \u003c/strong\u003edrug metabolism signature alterations were revealed in plasma proteomics. \u003cstrong\u003eE,\u003c/strong\u003eDEMs analysis was performed to explore the significant alterations in MSS rectal cancer after nCRT plus PDL1 blockade. \u003cstrong\u003eF,\u003c/strong\u003e Functional alterations from metabolomics were also compared between different response cohorts and biosynthesis of unsaturated fatty acids, arginine and proline metabolism were significantly enriched in Response cohorts, while ether lipid metabolism was enriched in non-response cohorts. \u003cstrong\u003eG \u003c/strong\u003eand\u003cstrong\u003e H,\u003c/strong\u003e Metabolites in response cluster, implying the recovery of cholesterol metabolism markers after nCRT plus PDL1 blockade, participated in 1,4−cholestadienone, 7−oxo−cholesterol and PA (18:0/0:0). \u003cstrong\u003eAberration\u003c/strong\u003e: MSS, microsatellite stable; nCRT, neoadjuvant chemotherapy; PDL1, Programmed cell death 1 ligand 1; DEP, differential expressed proteins, PPI, protein to protein interaction; DEM, differential expressed metabolites.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-5509842/v1/76c9616712bd2e04000afc76.png"},{"id":82150179,"identity":"9851025c-3d7a-4638-bfce-14317120d06c","added_by":"auto","created_at":"2025-05-07 07:14:02","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":377269,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCombined correlation between paired plasma proteins and metabolites in response status of MSS rectal cancer to PDL1 blockade.\u003c/strong\u003e \u003cstrong\u003eA\u003c/strong\u003e, top 20 significantly correlated metabolites and proteins between multi-omics were identified (correlation coefficient \u0026gt; 0.3, FDR \u0026lt; 0.05, p\u0026lt;0.05), including Glycochenodeoxycholic acid, 14(R)-Hydroxy-retro-vitamin A, PG (20:4(5Z,8Z,10E,14Z)-OH(12S)/i-24:0) and cholest-5-en-23-yn-3beta-ol. \u003cstrong\u003eB\u003c/strong\u003e to \u003cstrong\u003eE\u003c/strong\u003e, integrated analysis of proteomics and metabolomics were performed at baseline, post-radiotherapy, post-immunotherapy. Metabolite Sedoheptulose and protein Splicing factor C9orf78 were regarded as common altered factors between treatment in Venn diagram. \u003cstrong\u003eF \u003c/strong\u003eand\u003cstrong\u003e G,\u003c/strong\u003e overlapped functional pathways between proteomics and metabolomics from KEGG database, including Cholesterol metabolism, Bile secretion and Galactose metabolism. \u003cstrong\u003eAberration\u003c/strong\u003e: MSS, microsatellite stable; PDL1, Programmed cell death 1 ligand 1; FDR, false discovery rate; KEGG, Kyoto Encyclopedia of Genes and Genomes.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-5509842/v1/9d4807e4e17be5e49d477a11.png"},{"id":82151775,"identity":"4645c715-7a87-45a9-bbd4-cd098b7fadf0","added_by":"auto","created_at":"2025-05-07 07:22:02","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":667452,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCirculating proteomic and metabolic signatures associated with nCRT plus PDL1i-related side effects in LARC patients. A, \u003c/strong\u003esamples were divided as AE and NAEscohorts; PCA at protein and metabolite levels revealed significant differences. \u003cstrong\u003eB,\u003c/strong\u003e expression heatmap for different proteins was drawn and displayed. \u003cstrong\u003eC,\u003c/strong\u003e PPI between AE and NAE samples wasshown\u003cstrong\u003e \u003c/strong\u003eand revealed thatvarious kinds of interactions, including MYD88, SF3A3, and SNW1 were statistically correlated with AEs. \u003cstrong\u003eD,\u003c/strong\u003e Functional pathways alteration based on GO term at was also shown, including NF−kappaB signal transduction and Nucleocytoplasmic transport. \u003cstrong\u003eE,\u003c/strong\u003eat metabolites level, alterations between the AE and NAE cohorts\u003cstrong\u003e, \u003c/strong\u003eincluding alterations in \u003cem\u003eN-Palmitoyl Histidine\u003c/em\u003e, \u003cem\u003eChenodeoxycholylphenylalanine\u003c/em\u003e, \u003cem\u003e4,4'-Thiobis-2-butanone\u003c/em\u003e, \u003cem\u003eNaratriptan\u003c/em\u003e, \u003cem\u003eChenodeoxyglycocholic acid\u003c/em\u003e and \u003cem\u003e11-deoxy-PGF1a\u003c/em\u003e were shown. \u003cstrong\u003eF,\u003c/strong\u003eBesides, metabolic pathways alterations from KEGG database demonstrated Ferroptosis, cytochrome P450, Pyrimidine metabolism were enriched in AE cohorts. \u003cstrong\u003eG \u003c/strong\u003eto\u003cstrong\u003e H\u003c/strong\u003e, paired analysis for significantly expressed signatures and biologically functions were performed. \u003cstrong\u003eAberration\u003c/strong\u003e: AE, adverse events; LARC, locally advanced rectal cancer; PPI, protein to protein interaction; GO, Gene Ontology; PDL1, Programmed cell death 1 ligand 1; FDR, false discovery rate; KEGG, Kyoto Encyclopedia of Genes and Genomes\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-5509842/v1/93cc95fb56b990679d665190.png"},{"id":82148063,"identity":"b096a41a-9412-4ced-9cdb-78ac52275c47","added_by":"auto","created_at":"2025-05-07 07:06:02","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":416273,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eIntegrated proteome-metabolic features-based machine learning model provide an accurate measure for nCRT plus PDL1 blockade response. A, \u003c/strong\u003eworkflow for the construction of response prediction model. The models were based on a multistepprocessing pipeline. Inside of the pipeline, for each feature or classifier combination, we compared different data preprocessing strategies along with machine learning models and selected the top N best models. \u003cstrong\u003eB,\u003c/strong\u003e the rfcv function was screened, k-fold cross validation was performed, repeated it1000 times, and calculated the average error cv value 1000 times. \u003cstrong\u003eC,\u003c/strong\u003e top 30 metabolites and proteomes are shown and were ranked by using the\u003cstrong\u003e \u003c/strong\u003emean decreaseGini score.\u003cstrong\u003e D \u003c/strong\u003eand\u003cstrong\u003e E, \u003c/strong\u003eVia feature-based biomarker construction, the proteomes (APBB1IP, OLFM4, and DNAJC19) demonstratedbetter predictive power for PDL1 responses, whereas theproteome panels (APBB1IP, OLFM4, MRPL58, ERGIC3 and DNAJC19) were determined to have better predictive power viaclassifier-based statistics. \u003cstrong\u003eF,\u003c/strong\u003e Predictive metabolites were also screened out and shown. \u003cstrong\u003eG,\u003c/strong\u003e by overlapping the significant biomarkers, we identified metabolite (phosphatidylcholine, cholest-5-en-23-yn-3beta-ol) and proteome (APBB1IP, OLFM4, DNAJC19) features that were present in the longitudinal cohort after nCRT plus PDL1i. and repeated group-to-group validation analyses were conducted. \u003cstrong\u003eAberration\u003c/strong\u003e: nCRT, neoadjuvant chemoradiotherapy; PDL1, Programmed cell death 1 ligand 1.\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-5509842/v1/f46093e67cd1fe9b53fa6096.png"},{"id":90421641,"identity":"4b957cf3-e43b-4391-8d83-c01347b05178","added_by":"auto","created_at":"2025-09-02 14:05:37","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":5012745,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5509842/v1/af680e03-deee-4873-99cc-96bc7b23f5ac.pdf"},{"id":82148078,"identity":"3abf19cf-cea4-4821-94e6-4e517eec5dbc","added_by":"auto","created_at":"2025-05-07 07:06:03","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":6982508,"visible":true,"origin":"","legend":"supplementary data 1","description":"","filename":"supplementarydata1proteinsofallsamples.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5509842/v1/0fbc1bb00c09fabcd12b7d10.xlsx"},{"id":82148092,"identity":"7d113466-fff9-4b64-a0b1-70f60852d48c","added_by":"auto","created_at":"2025-05-07 07:06:04","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":72390848,"visible":true,"origin":"","legend":"supplementary data 2","description":"","filename":"supplementarydata2pepetidesofallsamples.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5509842/v1/66709c3a06043c93a7baea58.xlsx"},{"id":82148067,"identity":"68ffdd1b-4801-4d3d-9e1d-cea881786750","added_by":"auto","created_at":"2025-05-07 07:06:03","extension":"xls","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":1066460,"visible":true,"origin":"","legend":"supplementary data 3","description":"","filename":"supplementarydata3volcanoAvsB.xls","url":"https://assets-eu.researchsquare.com/files/rs-5509842/v1/e98b7cf8bdff4880f31f6a75.xls"},{"id":82150181,"identity":"3e8368c7-0be0-4c9a-bfe7-b6232f27ef88","added_by":"auto","created_at":"2025-05-07 07:14:03","extension":"xls","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":220990,"visible":true,"origin":"","legend":"supplementary data 4","description":"","filename":"Supplementarydata4metabolitesheatmapAvsB.xls","url":"https://assets-eu.researchsquare.com/files/rs-5509842/v1/455ec968a9dd7a2d0e1f8382.xls"},{"id":82151776,"identity":"5aea836b-839f-439a-bf07-25db2239f85b","added_by":"auto","created_at":"2025-05-07 07:22:03","extension":"xls","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":1046284,"visible":true,"origin":"","legend":"supplementary data 5","description":"","filename":"Supplementarydata5volcanoA1vsA2.xls","url":"https://assets-eu.researchsquare.com/files/rs-5509842/v1/f2174f9357bfdc0036d5d193.xls"},{"id":82148077,"identity":"753b4b3b-2020-421c-b776-7798153abee0","added_by":"auto","created_at":"2025-05-07 07:06:03","extension":"xls","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":1058794,"visible":true,"origin":"","legend":"supplementary data 6","description":"","filename":"Supplementarydata6volcanoA1vsA3.xls","url":"https://assets-eu.researchsquare.com/files/rs-5509842/v1/3a753723e8b5fad7d628b86f.xls"},{"id":82148066,"identity":"123bc777-5366-4670-ba87-bbe05abe1926","added_by":"auto","created_at":"2025-05-07 07:06:03","extension":"xls","order_by":7,"title":"","display":"","copyAsset":false,"role":"supplement","size":1075468,"visible":true,"origin":"","legend":"supplementary data 7","description":"","filename":"Supplementarydata7volcanoB1vsB2.xls","url":"https://assets-eu.researchsquare.com/files/rs-5509842/v1/9d2c8ffcf2f49f59576b8363.xls"},{"id":82150182,"identity":"1525c483-e940-44ee-96cd-21dd19dba90a","added_by":"auto","created_at":"2025-05-07 07:14:03","extension":"xls","order_by":8,"title":"","display":"","copyAsset":false,"role":"supplement","size":1079586,"visible":true,"origin":"","legend":"supplementary data 8","description":"","filename":"Supplementarydata8volcanoB1vsB3.xls","url":"https://assets-eu.researchsquare.com/files/rs-5509842/v1/4eeb9f02a616f75488cbdb21.xls"},{"id":82148091,"identity":"8c003885-43b6-4dec-9c67-ba05dbede1e8","added_by":"auto","created_at":"2025-05-07 07:06:03","extension":"xlsx","order_by":9,"title":"","display":"","copyAsset":false,"role":"supplement","size":6765450,"visible":true,"origin":"","legend":"\u003cp\u003esupplementary data 9\u003c/p\u003e","description":"","filename":"Supplementarydata9AEvsNAEproteins.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5509842/v1/781cf58f820bca349b813559.xlsx"},{"id":82151777,"identity":"c2958928-a55c-44e9-8fdd-e32801efee6d","added_by":"auto","created_at":"2025-05-07 07:22:03","extension":"xlsx","order_by":10,"title":"","display":"","copyAsset":false,"role":"supplement","size":17323984,"visible":true,"origin":"","legend":"\u003cp\u003esupplementary data 10\u003c/p\u003e","description":"","filename":"Supplementarydata10AEvsNAEmetabolites.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5509842/v1/3c1bab0843f39551e917e4ea.xlsx"},{"id":82148087,"identity":"4142e4d7-942e-4432-bb52-66f581acd210","added_by":"auto","created_at":"2025-05-07 07:06:03","extension":"docx","order_by":11,"title":"","display":"","copyAsset":false,"role":"supplement","size":8323953,"visible":true,"origin":"","legend":"Supplementary materials","description":"","filename":"Supplementarymaterials.docx","url":"https://assets-eu.researchsquare.com/files/rs-5509842/v1/08f181488be99df2e0b903b2.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"\u003cp\u003ePlasma proteome and metabolites profiling reveals dynamics for adverse events and responses after neoadjuvant radiochemotherapy plus PDL1 blockade in microsatellite-stable locally advanced rectal cancer: A prospective longitudinal study\u003c/p\u003e","fulltext":[{"header":"Introduction","content":"\u003cp\u003eLocally advanced rectal cancer (LARC) presents significant treatment challenges because of its potential for local invasion and distant metastasis\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. Since 1990, neoadjuvant therapy, which is administered before surgical resection, has aimed to improve outcomes by reducing tumour size, facilitating surgical resection, and enhancing overall survival\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. Traditionally, neoadjuvant treatment (nCRT) involves a combination of chemotherapy and radiotherapy to shrink tumours and improve resectability\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eImmunotherapy, particularly immune checkpoint inhibitors (ICIs, such as pembrolizumab and nivolumab), has revolutionized the treatment of various cancers, including colorectal cancer (CRC)\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. In LARC, checkpoint inhibitors have shown potential in patients with high microsatellite instability (MSI-H) or mismatch repair deficiency (dMMR), thus providing new avenues for treatment in patients who are less responsive to conventional therapies\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. However, for proficient MMR (pMMR) or microsatellite stable (MSS) LARC patients, the combination of ICIs with standard radiochemotherapy is an area of active research. Preliminary studies have suggested that such combinations may enhance the immune response against tumour cells, thus potentially improving outcomes in LARC patients\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. These studies demonstrated that ICIs plus radiochemotherapy (short-course or long-course) resulted in a significantly higher pCR rate (nearly 40%) with a well-tolerated safety profile. It is evident that this treatment model represents a logical next step in immunotherapy to improve response rates and increase cure rates and response duration.\u003c/p\u003e \u003cp\u003eHowever, some issues related to nCRT plus ICIs in LARC patients still need to be addressed. For example, predictive biomarkers have not yet been developed to identify whether patients will be able to respond to this treatment model\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. In ICI therapy, especially anti-PD1 (programmed cell death 1) or anti-PDL1 (programmed cell death ligand 1) therapy, some biomarkers, including PD-L1 status and the TMB, have strong predictive power before therapy\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. Regrettably, for MSS/pMMR rectal cancer, the role of biomarkers in the response to ICIs has not been extensively studied. The exploration of response markers will considerably promote LARC clinical trials and the development of precision medicine\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eRecently, blood plasma proteomic and metabolomic profiling has been reported as being a powerful approach for biomarker discovery, with the potential to identify the heterogeneous mechanisms of response and resistance to therapy\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e and aid in the development of personalized treatment strategies based on ICIs used in combination with other agents\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. It is expected that the integration of plasma proteomics and metabolomics expression signatures will improve predictive biomarker algorithms\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThus, in this study, we performed an integrated analysis of plasma proteomics and metabolomics in LARC patients before and during neoadjuvant radio-chemoimmunotherapy. Hence, the identification of multiomics biomarkers that can be readily evaluated through peripheral blood sampling is crucial for real-time implementation in routine clinical practice. To the best of our knowledge, we report the first large analysis of plasma proteome and metabolomics from LARC patients treated with neoadjuvant immune checkpoint blockers (ICBs). By integrating the multiomics of plasma and incorporating a machine learning algorithm, we can construct a robust model as a prediction tool for highly accurate responses.\u003c/p\u003e"},{"header":"Methods and materials","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003ePatients and treatment regimen\u003c/h2\u003e \u003cp\u003eFrom June 20, 2020, to August 15, 2023, a total of 80 patients were enrolled, and 50 patients with complete evaluable specimen were included in this study. Patient disposition and samples collection are summarized in \u003cb\u003eFigure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e\u003c/b\u003e. This prospective study was performed at Zhongshan Hospital of Fudan University, in accordance with the Declaration of Helsinki. Ethical approval was obtained from the Institutional Review Board of Zhongshan Hospital of Fudan University (Shanghai, P.R. China), and written informed consent was provided by all subjects before sampling.\u003c/p\u003e \u003cp\u003eThe major eligibility criteria included: treatment-naive individuals with rectal adenocarcinoma; pMMR, as proven by immunohistochemistry (IHC) for mismatch repair proteins (MLH1, MSH2, MSH6, PMS2); clinical stage T3N\u0026thinsp;+\u0026thinsp;M0 or T4NanyM0; aged between 18 and 75 years old; an Eastern Cooperative Oncology Group (ECOG) performance status 0\u0026ndash;1; and adequate organ function. Tumor staging was carried out using the 8th Edition of the American Joint Committee of Cancer (AJCC) tumor-node-metastasis (TNM) staging classification. To determine the TNM stage, enhanced CT scans of the chest and abdomen, as well as MRI scans of the rectum, were conducted.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eDeep Mass spectrometry analysis\u003c/h3\u003e\n\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003eNanomagnetic bead-mediated enrichment of low-abundance plasma proteins\u003c/h2\u003e \u003cp\u003eUsing EasyPept DeeP low abundance protein enrichment and pre-treatment reagent kit (Omicron, Shanghai, China) to enrich low abundance proteins in plasma samples. According to the manufacturer's instructions, 1 mg (40 \u0026micro;L) of magnetic nanoparticles suspension was taken, and the magnetic beads were separated by magnetic separation to discard the supernatant. After multiple washes of the magnetic beads, they were resuspended and 100 \u0026micro;L of serum/plasma was added and placed in a flip mixer (360\u0026deg; rotation mixing) at 37℃ for 1 hour. The supernatant was separated by magnetic separation, 300 \u0026micro;L of washing solution was added and shaken for 5 minutes for washing, and the washing was repeated three times. After enzymatic hydrolysis to convert proteins into peptides, reductive alkylation and desalting were performed. The total peptide concentration was determined by nanodroplet method.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eData independent acquisition(DIA)Mass spectrometry analysis\u003c/h3\u003e\n\u003cp\u003eThe Proteomic data analysis was performed by Shanghai Luming biological technology co., LTD (Shanghai, China). TimsTOF Pro2 mass spectrometer (Bruker) and nanoElute (Bruker) were used for both shotgun proteomics and DIA experiments. Samples were loaded and separated by a C18 column (25 cm \u0026times; 75 \u0026micro;m) on an EASY-nLCTM 1200 system (Thermo, USA). The flow rate was 300 nL/min and linear gradient was 60 min. The flow rate was 300 nL/min and linear gradient was set as follow: 0\u0026thinsp;~\u0026thinsp;45 min, 5\u0026ndash;27% B;45\u0026thinsp;~\u0026thinsp;50 min, 27\u0026ndash;46% B༛50\u0026thinsp;~\u0026thinsp;55min, 46\u0026ndash;100% B; 55 min\u0026thinsp;~\u0026thinsp;60 min, 100% B. For DIA, 56 DIA windows were acquired (automatic gain control target 3e6 and auto for injection time) and the collision energy was ramped linearly as a function of the mobility from 59 eV at 1/K0\u0026thinsp;=\u0026thinsp;1.6 Vs cm\u0026thinsp;\u0026minus;\u0026thinsp;2 to 20 eV at 1/K0\u0026thinsp;=\u0026thinsp;0.6 Vs cm\u0026thinsp;\u0026minus;\u0026thinsp;2.The MS/MS spectra were recorded from 100 to 1700 m/z.\u003c/p\u003e\n\u003ch3\u003eDatabase search\u003c/h3\u003e\n\u003cp\u003eThe default factory settings were used for the Spectronaut Pulsar 18.4 (Biognosys, Swiss) search and library generation (including Trypsyin as enzyme, up to two missed cleavages allowed Oxidation of Me as variable modifications, carbamidomethyl as fixed modification, and 1% FDR for PSM, peptide and protein identification). The DIA data were analysed with Spectronaut searching the above constructed spectral library. Main parameters of the software were set as follows: Precursor Qvalue cutoff and Protein Qvalue cutoff were set as 0.01, Normalization Strategy was set as Local Normalization and use MS2 as Quantity MS-Level.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eStatistical analyses\u003c/h2\u003e \u003cp\u003eA total of 5611 proteins expressed were identified as belonging to the proteome of plasma in this study. The thresholds of fold change and P-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05 were used to identify differentially expressed proteins (DEPs). Annotation of all identified proteins was performed using GO (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.blast2go.com/b2ghome\u003c/span\u003e\u003cspan address=\"http://www.blast2go.com/b2ghome\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e; \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://geneontology.org/\u003c/span\u003e\u003cspan address=\"http://geneontology.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) and KEGG pathway (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.genome.jp/kegg/\u003c/span\u003e\u003cspan address=\"http://www.genome.jp/kegg/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). DEPs were further used for GO and KEGG enrichment analysis. Protein-protein interaction analysis was performed using the String (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://string-db.org/\u003c/span\u003e\u003cspan address=\"https://string-db.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eMetabolomics\u003c/h3\u003e\n\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003eSample Preparation\u003c/h2\u003e \u003cp\u003eSamples stored at -80 ℃ were thawed at room temperature. 50 \u0026micro;L of sample was added to a 1.5 mL Eppendorf tube with 5\u0026micro;L of L-2-chlorophenylalanine (0.06 mg/mL) dissolved in methanol as internal standard, and the tube was vortexed for 10 s. Subsequently, 5 \u0026micro;L of ice-cold mixture of methanol and acetonitrile (2/1, vol/vol) was added, and the mixtures were vortexed for 1 min, and the whole samples were extracted by ultrasonic for 10 min in ice-water bath, stored at -20 ℃ for 30 min. The extract was centrifuged at at 4\u0026deg;C (13000 rpm) for 10 min. 5 \u0026micro;L of supernatant in a glass vial was dried in a freeze concentration centrifugal dryer. 5 \u0026micro;L mixture of methanol and water (1/4, vol/vol) were added to each sample, samples vortexed for 30 s, extracted by ultrasonic for 3 min in ice-water bat, then placed at -20\u0026deg;C for 2 h. Samples were centrifuged at 4\u0026deg;C (13000 rpm) for 10 min. The supernatants (5 \u0026micro;L) from each tube were collected using crystal syringes, filtered through 0.22 \u0026micro;m microfilters and transferred to LC vials. The vials were stored at -80\u0026deg;C until LC -MS analysis. QC samples were prepared by mixing aliquot of all samples to be a pooled sample.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eLC-MS/MS analysis\u003c/h2\u003e \u003cp\u003eThe metabolomic data analysis was performed by Shanghai Luming biological technology co., LTD (Shanghai, China). An ACQUITY UPLC I-Class plus(Waters Corporation, Milford, USA) fitted with Q-Exactive mass spectrometer equipped with heated electrospray ionization (ESI) source (Thermo Fisher Scientific, Waltham, MA, USA) was used to analyze the metabolic profiling in both ESI positive and ESI negative ion modes. An ACQUITY UPLC HSS T3 column (1.8 \u0026micro;m, 2.1 \u0026times; 100 mm) were employed in both positive and negative modes. The binary gradient elution system consisted of (A) water (containing 0.1% formic acid, v/v) and (B) acetonitrile (containing 0.1% formic acid, v/v) and separation was achieved using the following gradient: 0.01 min, 5% B; 2min, 5% B; 4min, 30% B; 8min, 50% B; 10min, 80% B; 14min, 100% B; 15 min, 100% B; 15.1 min, 5% and 16 min, 5%B. The flow rate was 0.35 mL/min and column temperature were 45℃. All the samples were kept at 10℃ during the analysis.\u003c/p\u003e \u003cp\u003eThe mass range was from m/z 100 to 1,000. The resolution was set at 70,000 for the full MS scans and 17500 for HCD MS/MS scans. The Collision energy was set at 10, 20 and 40 eV. The mass spectrometer operated as follows: spray voltage, 3800 V (+) and 3200 V (\u0026minus;); sheath gas flow rate, 35 arbitrary units; auxiliary gas flow rate, 8 arbitrary units; capillary temperature, 320\u0026deg;C; Aux gas heater temperature, 350\u0026deg;C; S-lens RF level, 50.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eData Preprocessing and Statistical Analysis\u003c/h2\u003e \u003cp\u003eThe original LC-MS data were processed by software Progenesis QI V2.3 (Nonlinear, Dynamics, Newcastle, UK) for baseline filtering, peak identification, integral, retention time correction, peak alignment, and normalization. Main parameters of 5 ppm precursor tolerance, 10 ppm product tolerance, and 5% product ion threshold were applied. Compound identifications were based on precise mass-to-charge ratio (M/z), secondary fragments, and isotopic distribution using The Human Metabolome Database (HMDB), Lipidmaps (V2.3), Metlin, and self-built databases. The extracted data were then further processed by removing any peaks with a missing value (ion intensity\u0026thinsp;=\u0026thinsp;0) in more than 50% in groups, by replacing zero value by half of the minimum value, and by screening according to the qualitative results of the compound. Compounds with resulting scores below 36 (out of 60) points were also deemed to be inaccurate and removed. A data matrix was combined from the positive and negative ion data.\u003c/p\u003e \u003cp\u003eThe matrix was imported in R to carry out Principal Component Analysis (PCA) to observe the overall distribution among the samples and the stability of the whole analysis process. Orthogonal Partial Least-Squares-Discriminant Analysis (OPLS-DA) and Partial Least-Squares-Discriminant Analysis (PLS-DA) were utilized to distinguish the metabolites that differ between groups. To prevent overfitting, 7-fold cross-validation and 200 Response Permutation Testing (RPT) were used to evaluate the quality of the model. V variable Importance of Projection (VIP) values obtained from the OPLS-DA model were used to rank the overall contribution of each variable to group discrimination. A two-tailed Student\u0026rsquo;s T-test was further used to verify whether the metabolites of difference between groups were significant. Differential metabolites were selected with VIP values greater than 1.0 and p-values less than 0.05. Differential metabolites were further used to for KEGG pathway (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.genome.jp/kegg/\u003c/span\u003e\u003cspan address=\"http://www.genome.jp/kegg/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) enrichment analysis.\u003c/p\u003e \u003cp\u003e \u003cb\u003eConstruction of Machine learning models.\u003c/b\u003e \u003c/p\u003e \u003cp\u003eMachine learning was conducted to identify the responses. The graphical machine learning model construction pipeline is illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e6\u003c/span\u003eA. The machine learning pipeline was built in Python (version 3.9.15) using the following libraries: scikit-learn (version 1.2.1), numpy (version 1.26.3), scipy (version 1.12.0), and pandas (version 1.5.3).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003ePatient cohorts and treatment regimens\u003c/h2\u003e \u003cp\u003eThe analysed samples were collected from 50 patients with LARC who were receiving neoadjuvant radiochemotherapy plus PDL1 treatment (\u003cb\u003eTable\u0026nbsp;1\u003c/b\u003e). Patients who responded to neoadjuvant therapy (pCR) (specifically, patients with 0% residual viable tumour (RVT) after resection) were included in the analysis. Compared with the pCR cohort, the nonpCR cohort did not have differences in clinical characteristics at baseline \u003cb\u003e(Table\u0026nbsp;1)\u003c/b\u003e.\u003c/p\u003e \u003cp\u003eBlood samples were collected at baseline, postradiotherapy and postimmunotherapy during the therapy cycle for haematological evaluation (\u003cb\u003eFigure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e\u003c/b\u003e and \u003cb\u003eMethods\u003c/b\u003e). Ultimately, 117 plasma samples were collected from the 50 enrolled tumour patients before and during neoadjuvant therapy. Among the 117 samples, 35 samples were collected at the baseline stage, 46 samples were collected after radiotherapy, and 36 samples were collected after the PDL1i stage. Among the 50 patients, a total of 20 patients were pathologically recognized as having CR, reaching a pCR rate of 40%. The study design is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e1\u003c/span\u003eA.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eDeep proteomics dynamics of plasma in LARC patients receiving neoadjuvant chemotherapy plus PDL1i\u003c/h2\u003e \u003cp\u003eTo detect changes in the plasma proteome as a result of PDL1 treatment, we analysed serial pretrm and trm plasma samples. An EasyPept DeeP low-abundance protein enrichment and pretreatment reagent kit (Omicron, Shanghai, China) was used to enrich low-abundance proteins in the plasma samples\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. For quality control of the performance of deep proteomics, mixtures of all of the plasma samples were measured every twenty samples, and this protocol was adopted in the proteomic studies. Pearson\u0026rsquo;s correlation coefficient was calculated for all of the quality-control runs, and the results are shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e1\u003c/span\u003eB. Proteomic analysis revealed 2023\u0026ndash;5035 gene products (GPs) in each sample (median number of 4282 in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e1\u003c/span\u003eB). A total of 5611 gene products (GPs) were identified in all of the plasma samples of the recruited cohort, of which 83250 peptide GPs were identified (\u003cb\u003eFigure S2A\u003c/b\u003e and \u003cb\u003eS2B\u003c/b\u003e; \u003cb\u003eSupplementary data 1 and data 2\u003c/b\u003e). The molecular weights, peptide lengths and numbers are also shown in \u003cb\u003eFigure S2C\u003c/b\u003e to \u003cb\u003eS2E\u003c/b\u003e. To explore the molecular differences between CRC patients and healthy controls, we performed a comparative proteomic analysis of response samples and nonresponse samples. Heatmap analysis revealed the top 50 proteins between the samples from the two cohorts (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e1\u003c/span\u003eC and \u003cb\u003eFigure S3A\u003c/b\u003e). We observed that 204 proteins were significantly upregulated and that 68 were significantly downregulated (adjusted P value\u0026thinsp;\u0026lt;\u0026thinsp;0.05, fold change\u0026thinsp;\u0026gt;\u0026thinsp;2 in \u003cb\u003eFigure S3B\u003c/b\u003e and \u003cb\u003eSupplementary data 3\u003c/b\u003e) according to the protein classes in the Human Protein Atlas (HPA) database\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. Thus far, our study has presented a comprehensive overview of the plasma proteomic landscape of the CRC cohort. We then performed pathway enrichment analysis on the differentially expressed proteins (DEPs) via Gene Ontology (GO)\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e, Interpro\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e, Reactome\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e and Wikipathway\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e analyses. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e1\u003c/span\u003eD, \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e1\u003c/span\u003eE, \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e1\u003c/span\u003eF and \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e1\u003c/span\u003eG, metabolism, fatty acid and lipoprotein transport mechanisms, phosphatidylinositol 3-kinase regulator activity and oxidative phosphorylation were enriched in the plasma samples of responsive patients. In addition, the KEGG analysis revealed genes related to neutrophil degranulation and the innate immune system, among other activities (adjusted P value\u0026thinsp;\u0026lt;\u0026thinsp;0.05) (\u003cb\u003eFigure S3C\u003c/b\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eAlterations in plasma metabolites in association with ICB response in LARC patients\u003c/h2\u003e \u003cp\u003eTo investigate the metabolomic profile before and during neo-CRT plus PDL1 blockade, we also collected 117 plasma samples and performed untargeted liquid chromatography‒mass spectrometry (LC‒MS) metabolomics analysis (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e1\u003c/span\u003eA). Four levels and a total of 4,791 endogenous metabolites were identified (\u003cb\u003eFigure S4A\u003c/b\u003e). A box plot of the metabolite intensity of each sample is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e2\u003c/span\u003eA, which indicates that the skewness of the distribution was greater for further analysis. The results of the quality control analysis are shown in \u003cb\u003eFigure S4B\u003c/b\u003e. The subclasses of the detected metabolites are shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e2\u003c/span\u003eB. Most of the classes included amino acids, peptides, and analogues (10.48%), followed by fatty acids and fatty acid conjugates (8.50%), carbohydrates and carbohydrate conjugates (3.99%), and bile acids (2.44%). PCA and OPLS-DA analyses revealed that the plasma metabolome was significantly different between the response and nonresponse cohorts (both p\u0026thinsp;\u0026lt;\u0026thinsp;0.05; Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e2\u003c/span\u003eC and \u003cb\u003eFigure S4C\u003c/b\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo characterize the altered plasma metabolites according to response status, we conducted a heatmap comparison to identify metabolites with significantly differential peaks between different groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e2\u003c/span\u003eD, \u003cb\u003eFigure S4D\u003c/b\u003e; \u003cb\u003eSupplementary data 4\u003c/b\u003e). Differential plasma metabolites with consistent shifts across the response and nonresponse cohorts were revealed, with 31 sequentially enriched metabolites, including glycoursodeoxycholic acid, (9E)-10-nitrophenyloctadec-9-enoylcarnitine, trans-3-octenedioic acid and multiple fatty acids (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e2\u003c/span\u003eE and \u003cb\u003eFigure S4E\u003c/b\u003e). We also observed 67 sequentially depleted metabolites, including Cholest-5-en-3-ol (3. beta.) butanoate, N-palmitoyl taurine and maracin A (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e2\u003c/span\u003eE). To gain insight into the functional significance of the altered metabolites, we conducted a correlation network analysis, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e2\u003c/span\u003eF, and the associated metabolites mostly included bile acids and lipid acids. Among the plasma metabolites, the top 3 differential pathways between the two response cohorts were (1) primary bile acid biosynthesis, (2) cholesterol metabolism, and (3) taurine and hypotaurine metabolism (KEGG analysis in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e2\u003c/span\u003eG). The top 3 differential pathways from the reactome database were (1) fructose biosynthesis, (2) digestion of dietary carbohydrates, and (3) fructose metabolism (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e2\u003c/span\u003eH). Collectively, metabolomics analysis demonstrated that fatty acid, bile acid, and lipid metabolism were involved in ICB resistance in LARC plasma.\u003c/p\u003e \u003cp\u003e \u003cb\u003eAnalysis of the plasma proteome and metabolites revealed a significant relationship between cholesterol metabolism and response in LARC patients after nCRT plus PDL1i\u003c/b\u003e \u003c/p\u003e \u003cp\u003eAnother aim of this study was to investigate the effects of PDL1i on the plasma proteome and metabolites of patients. Therefore, we compared both proteomic and metabolomic characteristics among the baseline cohort, post-radiotherapy cohort, and post-immunotherapy cohort. Differential expression protein (DEP) analysis was performed among the different treatment stages, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e3\u003c/span\u003eA, \u003cb\u003eFigure S5\u003c/b\u003e and \u003cb\u003eSupplementary Data 5 to 8\u003c/b\u003e. A Venn diagram was drawn, and 92 response-specific proteins were screened out while 88 nonresponse-related proteins were revealed (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e3\u003c/span\u003eB). These genes include multiple cholesterol-related markers, such as ABCA13, RAB3IP, GBA2, BCS1L, FAH, ABHD17A and SREBF2. Protein N-linked glycosylation and peptidyl-asparagine modification were significantly enriched in nonresponsive proteins, whereas KEGG analysis revealed significant functional alterations in N-glycan biosynthesis and protein processing in the endoplasmic reticulum (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e3\u003c/span\u003eB). As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e3\u003c/span\u003eC, protein-to-protein interaction (PPI) network analysis revealed dramatic alterations in proteome interactions, including cholesterol metabolism markers (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05). Pathway alterations in the response cohort after PDL1 blockade were revealed at baseline and post immunotherapy, and Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e3\u003c/span\u003eD shows alterations in the drug metabolism signature. In addition, differential expression metabolite (DEM) analysis was performed to explore the significant alterations in LARC patients after nCRT plus PDL1i. As demonstrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e3\u003c/span\u003eE and \u003cb\u003eFigure S6\u003c/b\u003e, in a similar manner, the response in rectal cancer patients to nCRT plus PDL1i exhibited increased expression of many cholesterol-related markers, including 1,4-cholestadienone, 7-methyloctanoylcarnitine and 3-hydroxybutyrylcarnitine. Functional alterations from metabolomics were also compared between different response cohorts, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e3\u003c/span\u003eF. The biosynthesis of unsaturated fatty acids, as well as arginine and proline metabolism, were significantly enriched in the response cohort, whereas ether lipid metabolism was enriched in the nonresponse cohort. Metabolites in the response cluster, which suggested the recovery of cholesterol metabolism markers after nCRT plus PDL1 blockade, included 1,4-cholestadienone, 7-oxo-cholesterol and PA (18:0/0:0) (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e3\u003c/span\u003eG and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e3\u003c/span\u003eH). Collectively, these results demonstrated the strong impacts of PDL1i on cholesterol processes and oncogenic signalling.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe correlation between paired plasma proteins and metabolites in response to ICBs in LARC patients was examined using Spearman correlation analysis. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e4\u003c/span\u003eA and \u003cb\u003eFigure S7\u003c/b\u003e, we identified the top 20 significantly correlated metabolites and proteins between multi-omics (correlation coefficient\u0026thinsp;\u0026gt;\u0026thinsp;0.3, FDR\u0026thinsp;\u0026lt;\u0026thinsp;0.05, p\u0026thinsp;\u0026lt;\u0026thinsp;0.05). These included \u003cem\u003eglycochenodeoxycholic acid\u003c/em\u003e, \u003cem\u003e14(R)-hydroxyretrovitamin A\u003c/em\u003e, \u003cem\u003ePG (20:4(5Z,8Z,10E,14Z)-OH(12S)/i-24:0)\u003c/em\u003e and \u003cem\u003echolest-5-en-23-yn-3beta-ol\u003c/em\u003e. Accordingly, integrated proteomics and metabolomics analyses were performed at baseline, postradiotherapy, and posti-mmunotherapy(Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e4\u003c/span\u003eB to \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e4\u003c/span\u003eE). Notably, sedoheptulose and the splicing factor C9orf78 were regarded as being common altered factors between the treatments (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e4\u003c/span\u003eB). Functionally, we also noted that there were very few overlapping pathways between the proteomic and metabolomic data (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e4\u003c/span\u003eF) from the KEGG database, which included those related to cholesterol metabolism, bile secretion and galactose metabolism (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e4\u003c/span\u003eG). Collectively, these results demonstrate the potential role of the cholesterol-related pathway as a biomarker for the PDL1 inhibitor response in LARC.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eCirculating proteomic and metabolic signatures associated with nCRT plus PDL1i-related side effects in LARC patients\u003c/h2\u003e \u003cp\u003eNext, we identified circulating proteomic and metabolic differences between adverse events (AEs) and Nonadverse events (NAEs) and demonstrated more differential proteins or metabolites associated with AEs, which were highly biologically distinct. PCA between AE samples and NAE samples at metabolic and protein level were shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003eA. A broad overview of expression heatmap was also displayed in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003eB, volcano map for significantly expressed proteins was shown in \u003cb\u003eFigure S8A-S8B\u003c/b\u003e and \u003cb\u003eSupplementary data 9\u003c/b\u003e. Protein to protein interaction between AE and NAE samples was shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003eC, which revealed that various kinds of interactions, including MYD88, SF3A3, and SNW1 were statistically correlated with AEs. Functional alteration at protein level was also shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003eD. In addition, metabolite alterations between the AE and NAE cohorts, including alterations in \u003cem\u003eN-Palmitoyl Histidine\u003c/em\u003e, \u003cem\u003eChenodeoxycholylphenylalanine\u003c/em\u003e, \u003cem\u003e4,4'-Thiobis-2-butanone\u003c/em\u003e, \u003cem\u003eNaratriptan\u003c/em\u003e, \u003cem\u003eChenodeoxyglycocholic acid\u003c/em\u003e and \u003cem\u003e11-deoxy-PGF1a\u003c/em\u003e were shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003eE, \u003cb\u003eFigure S8C\u003c/b\u003e and \u003cb\u003eSupplementary data 10\u003c/b\u003e. Besides, functional alterations from KEGG database demonstrated Ferroptosis, cytochrome P450, Pyrimidine metabolism pathway were enriched in AE cohorts (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003eF).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eAlso, correlation between paired plasma proteins and metabolites in AE to ICBs in LARC patients was examined using Spearman correlation analysis. The top 20 significantly correlated metabolites and proteins were identified between multi-omics (correlation coefficient\u0026thinsp;\u0026gt;\u0026thinsp;0.3, FDR\u0026thinsp;\u0026lt;\u0026thinsp;0.05, p\u0026thinsp;\u0026lt;\u0026thinsp;0.05) in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003eG. Functionally, integrated analysis noted that there were very few overlapping pathways between the proteomic and metabolomic data from the KEGG database, which included those related to cytochrome P450 metabolism and Pyrimidine metabolism (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003eH). Collectively, our results revealed unique proteomic and metabolic profiles in different AE groups, which may be associated with the mechanisms underlying immunotherapy-related AEs in LARC patients.\u003c/p\u003e \u003cp\u003e \u003cb\u003eIntegrated proteome-metabolic feature-based machine learning model provides an accurate measure for the nCRT plus PDL1i response\u003c/b\u003e \u003c/p\u003e \u003cp\u003eIn this study, we aimed to construct an integrated proteome‒metabolic predictive signature for responses. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e6\u003c/span\u003eA, we identified significant metabolites and proteomes via a machine learning framework. The models were based on a multistep processing pipeline (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e6\u003c/span\u003eA). Inside of the pipeline, for each feature or classifier combination, we compared different data preprocessing strategies along with machine learning models and selected the top N (such as N\u0026thinsp;=\u0026thinsp;5) best models\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e. We screened the rfcv function, performed k-fold cross validation, repeated it 1000 times, and calculated the average error cv value 1000 times. This process was repeated N times, and a line graph was drawn on the basis of the average value of N times in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e6\u003c/span\u003eB. The top 30 metabolites and proteomes are shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e6\u003c/span\u003eC and were ranked by using the mean decrease Gini score. Via feature-based biomarker construction, the proteomes (APBB1IP, OLFM4, and DNAJC19) demonstrated better predictive power for PDL1 responses, whereas the proteome panels (APBB1IP, OLFM4, MRPL58, ERGIC3 and DNAJC19) were determined to have better predictive power via classifier-based statistics (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e6\u003c/span\u003eD and \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e6\u003c/span\u003eE). Predictive metabolites were also screened out, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e6\u003c/span\u003eF.\u003c/p\u003e \u003cp\u003eBy overlapping the significant biomarkers, we identified metabolite (phosphatidylcholine, cholest-5-en-23-yn-3beta-ol) and proteome (APBB1IP, OLFM4, DNAJC19) features that were present in the longitudinal cohort after nCRT plus PDL1i. This result led to the use of a machine learning framework (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e6\u003c/span\u003eA) to integrate features into a predictive model of PDL1 responses to predict disease progression in rectal cancer patients. There are five optionally different feature combinations based on metabolic and proteome features: (1) phosphatidylcholine, (2) cholest-5-en-23-yn-3beta-ol, (3) APBB1IP, (4) OLFM4, (5) DNAJC19, and (6) all of the integrated features. To examine the robustness and efficacy of the diagnostic strategy, we conducted repeated group-to-group validation analyses of the baseline cohort and postradiotherapy cohort and calculated a model predictive ability of 0.954; in contrast, postimmunotherapy validation revealed a predictive ability of 0.877 (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e6\u003c/span\u003eG). Overall, the selected features from each group showed excellent potential for identifying LARC patients with pathological responses to neoadjuvant immunotherapy, thus indicating the great potential of this screening strategy for people at pretreatment stages.\u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eNeoadjuvant immunotherapy is designed to stimulate the immune system prior to surgical intervention, thus potentially improving outcomes by targeting the tumour microenvironment (TME) and enhancing systemic antitumour responses\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e. Recent studies have indicated that immune checkpoint inhibitors, such as PD-1/PD-L1 and CTLA-4 inhibitors, can lead to significant tumour regression in patients with dMMR or MSI-H rectal cancers\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. For MSS rectal cancer, with the stimulated TMEs elicited by radiotherapy, neoadjuvant immunotherapy has emerged as a promising approach for rectal cancer treatment, particularly in cases where traditional modalities have shown limited efficacy\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eHowever, most published studies have indicated that 40% of MSS rectal cancer patients achieve pathological complete response (pCR), which demonstrates that another 60% of patients have a limited response to nCRT plus PDL1 blockade. Responses can vary widely among individuals, thus highlighting the need for reliable biomarkers to predict which patients will benefit most from such therapies. In this context, the integration of plasma proteomics and metabolomics offers a novel framework for predicting therapeutic responses and enhancing treatment strategies\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. Plasma proteomics focuses on the comprehensive analysis of proteins that are present in the blood plasma, which can reflect tumour biology and host immune responses. Recent advances in mass spectrometry and bioinformatics have enabled researchers to identify specific protein signatures associated with tumour progression and treatment response\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e. For example, certain cytokines and immune-related proteins have been linked to favourable outcomes in immunotherapy\u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e. In addition, metabolites can provide insights into the metabolic states of tumours and their microenvironments. In rectal cancer, alterations in metabolic pathways have been associated with tumour aggressiveness and treatment resistance\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e. For example, specific metabolic signatures may indicate a greater likelihood of immune evasion or lower responsiveness to therapy\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. The use of metabolomic profiling in conjunction with proteomic analysis can enhance our understanding of the biochemical changes associated with rectal cancer and improve the ability of these methods to predict therapeutic efficacy.\u003c/p\u003e \u003cp\u003eIn this work, we aimed to systematically describe comprehensive alterations in the plasma proteome and metabolites and explore potential biomarkers of PDL1 blockade in MSS rectal cancer patients. We collected longitudinal samples from rectal cancer patients treated with PDL1 blockade and performed advanced sample preparation, in-depth MS detection, MS data preprocessing, LC-MS sequencing and bioinformatic analysis, thus resulting in a comprehensive proteome and metabolite workflow\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. This longitudinal study of PDL1 blockade revealed proteome and metabolite fluctuations, which may facilitate biomarker exploration and further speculation/research on the biological mechanisms. In this study, we described the link between the PDL1 blockade response and cholesterol metabolism. Furthermore, we proposed a metabolite (phosphatidylcholine, cholest-5-en-23-yn-3beta-ol) and proteome (APBB1IP, OLFM4, DNAJC19) as being potential response biomarkers for PDL1 blockade.\u003c/p\u003e \u003cp\u003eIn our cohort, we observed enrichment of cholesterol metabolism-related processes in nonresponsive samples after radiotherapy, whereas functional alterations in the cholesterol transport process were also observed via both proteomics and metabolomics. These results implied that, for nonresponsive cohorts, increased cholesterol was detected in the blood, which is consistent with the findings of a previously published study\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e. On the basis of these results, the increase in blood cholesterol was regarded as being the result of a nonresponse to nCRT plus PDL1 blockade, which demonstrated that anti-cholesterol metabolism treatment may be a drug target for overcoming PDL1 resistance; consequently, the specific biological mechanism needs to be further investigated.\u003c/p\u003e \u003cp\u003eAnother key strength of this study was the use of longitudinal blood samples from MSS rectal cancer patients, which allowed us to evaluate fluctuations in protein and metabolite levels by setting the treatment time as a covariate. Notably, the use of these samples in combination with proteomic and metabolomic features outperformed other methods for predicting the response of patients to nCRT plus PDL1 blockade. The combination of proteomics and metabolomics represents a powerful approach for developing predictive models for neoadjuvant immunotherapy. This integrative strategy allows for a more comprehensive understanding of tumour biology and its interactions with the host immune system. By correlating protein expression profiles with metabolic alterations, such models could significantly improve the precision of immunotherapy, thus enabling tailored approaches that consider both the proteomic landscape and the metabolic context of the tumour.\u003c/p\u003e \u003cp\u003eThis study has several limitations. First, although our plasma-based analysis provided valuable insights, the heterogeneity of patient responses could not be fully identified due to the inherent complexity of the TME. Second, another limitation of the study includes the relatively small sample size, which may not fully represent the broader patient population; however, the longitudinal study design can partly offer a more comprehensive understanding of treatment response over time. Another limitation is the potential for bias in patient selection, as our study cohort may not encompass the full spectrum of rectal cancer patients. This could affect the representativeness of the proteomic and metabolomic profiles that we identified.\u003c/p\u003e \u003cp\u003eIn conclusion, our findings suggest several avenues for future research. First, we introduced the proteomic and metabolomic landscape of the plasma and designed a longitudinal study with MSS rectal cancer patient cohorts receiving nCRT plus PDL1 blockade. Additionally, explorations of the mechanisms underlying the observed alterations in cholesterol metabolism could provide deeper insights into how immunotherapy exerts its effects and can help in identifying novel therapeutic targets. Moreover, this integrated model may ultimately lead to more effective and personalized treatment strategies for RC.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis manuscript has not been submitted to any other journal and is not currently being considered for publication by another journal. We Thank the doctors and nurses who performed the treatment.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWritten informed consent was obtained from all patients. The study protocol followed the ethical guidelines of the Declaration of Helsinki and was approved by the Ethics Committee of Zhongshan Hospital of Fudan University. The ethics approval ID was ZS2022-019.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe have obtained consent to publish from the participant to report individual patient data.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and material\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets used and/or analyzed during the current study are available from the corresponding authors upon reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflicts of interest\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026apos; contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eProf. Xu JianMin and Prof. Ye LeChi contributed to the design of this study. Dr. Lv Yang, Zhu ZheHui, Tang WenTao, Ji MeiLing and Zheng Peng performed the research, and Dr. Lv Yang, Tang Wentao, Zhu DeXiang and Zheng Peng analyzed and interpreted the patient data. Dr. Lv Yang was a major contributor to the writing of the manuscript. Dr. Ji MeiLing provided research background and perspectives. Prof. Xu JianMin, Dr. Tang WenTao, and Prof. Ye LeChi were the corresponding authors and approved the final version of this manuscript for publication.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by the Natural Science Foundation of China (82303889, 82172816), the Natural Science Foundation of Shanghai (23ZR1410400) and Shanghai Sailing Program (23YF1406100). The funding bodies had no role in the design of the study; collection, analysis, and interpretation of data; or in the writing of the manuscript.\u003c/p\u003e\n"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eLi Y\u003cem\u003e, et al.\u003c/em\u003e A Review of Neoadjuvant Chemoradiotherapy for Locally Advanced Rectal Cancer. \u003cem\u003eInt J Biol Sci\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 1022-1031 (2016).\u003c/li\u003e\n \u003cli\u003eNIH consensus conference. Adjuvant therapy for patients with colon and rectal cancer. \u003cem\u003eJAMA\u003c/em\u003e \u003cstrong\u003e264\u003c/strong\u003e, 1444-1450 (1990).\u003c/li\u003e\n \u003cli\u003eSauer R\u003cem\u003e, et al.\u003c/em\u003e Preoperative versus postoperative chemoradiotherapy for rectal cancer. \u003cem\u003eN Engl J Med\u003c/em\u003e \u003cstrong\u003e351\u003c/strong\u003e, 1731-1740 (2004).\u003c/li\u003e\n \u003cli\u003eKasi A\u003cem\u003e, et al.\u003c/em\u003e Total Neoadjuvant Therapy vs Standard Therapy in Locally Advanced Rectal Cancer: A Systematic Review and Meta-analysis. \u003cem\u003eJAMA Netw Open\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, e2030097 (2020).\u003c/li\u003e\n \u003cli\u003ePetrelli F\u003cem\u003e, et al.\u003c/em\u003e Total Neoadjuvant Therapy in Rectal Cancer: A Systematic Review and Meta-analysis of Treatment Outcomes. \u003cem\u003eAnn Surg\u003c/em\u003e \u003cstrong\u003e271\u003c/strong\u003e, 440-448 (2020).\u003c/li\u003e\n \u003cli\u003eMalmberg R\u003cem\u003e, et al.\u003c/em\u003e Alternative dosing strategies for immune checkpoint inhibitors to improve cost-effectiveness: a special focus on nivolumab and pembrolizumab. \u003cem\u003eLancet Oncol\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e, e552-e561 (2022).\u003c/li\u003e\n \u003cli\u003eMatsumoto K\u003cem\u003e, et al.\u003c/em\u003e Real-world outcomes of nivolumab plus ipilimumab and pembrolizumab with platinum-based chemotherapy in advanced non-small cell lung cancer: a multicenter retrospective comparative study. \u003cem\u003eCancer Immunol Immunother\u003c/em\u003e \u003cstrong\u003e73\u003c/strong\u003e, 4 (2024).\u003c/li\u003e\n \u003cli\u003eYu JH\u003cem\u003e, et al.\u003c/em\u003e Long-Term Outcomes of dMMR/MSI-H Rectal Cancer Treated With Anti-PD-1-Based Immunotherapy as Curative-Intent Treatment. \u003cem\u003eJ Natl Compr Canc Netw\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, (2024).\u003c/li\u003e\n \u003cli\u003eXiao WW\u003cem\u003e, et al.\u003c/em\u003e Effect of neoadjuvant chemoradiotherapy with or without PD-1 antibody sintilimab in pMMR locally advanced rectal cancer: A randomized clinical trial. \u003cem\u003eCancer Cell\u003c/em\u003e, (2024).\u003c/li\u003e\n \u003cli\u003eXia F\u003cem\u003e, et al.\u003c/em\u003e Randomized Phase II Trial of Immunotherapy-Based Total Neoadjuvant Therapy for Proficient Mismatch Repair or Microsatellite Stable Locally Advanced Rectal Cancer (TORCH). \u003cem\u003eJ Clin Oncol\u003c/em\u003e, JCO2302261 (2024).\u003c/li\u003e\n \u003cli\u003eLin ZY\u003cem\u003e, et al.\u003c/em\u003e Neoadjuvant short-course radiotherapy followed by camrelizumab and chemotherapy in locally advanced rectal cancer (UNION): early outcomes of a multicenter randomized phase III trial. \u003cem\u003eAnn Oncol\u003c/em\u003e, (2024).\u003c/li\u003e\n \u003cli\u003eBortolomeazzi M\u003cem\u003e, et al.\u003c/em\u003e Immunogenomics of Colorectal Cancer Response to Checkpoint Blockade: Analysis of the KEYNOTE 177 Trial and Validation Cohorts. \u003cem\u003eGastroenterology\u003c/em\u003e \u003cstrong\u003e161\u003c/strong\u003e, 1179-1193 (2021).\u003c/li\u003e\n \u003cli\u003eShan J, Han D, Shen C, Lei Q, Zhang Y. Mechanism and strategies of immunotherapy resistance in colorectal cancer. \u003cem\u003eFront Immunol\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 1016646 (2022).\u003c/li\u003e\n \u003cli\u003eMeric-Bernstam F, Larkin J, Tabernero J, Bonini C. Enhancing anti-tumour efficacy with immunotherapy combinations. \u003cem\u003eLancet\u003c/em\u003e \u003cstrong\u003e397\u003c/strong\u003e, 1010-1022 (2021).\u003c/li\u003e\n \u003cli\u003eDi Meo F\u003cem\u003e, et al.\u003c/em\u003e A target discovery pipeline identified ILT3 as a target for immunotherapy of multiple myeloma. \u003cem\u003eCell Rep Med\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, 101110 (2023).\u003c/li\u003e\n \u003cli\u003eLoriot Y\u003cem\u003e, et al.\u003c/em\u003e Plasma proteomics identifies leukemia inhibitory factor (LIF) as a novel predictive biomarker of immune-checkpoint blockade resistance. \u003cem\u003eAnn Oncol\u003c/em\u003e \u003cstrong\u003e32\u003c/strong\u003e, 1381-1390 (2021).\u003c/li\u003e\n \u003cli\u003eLyu J\u003cem\u003e, et al.\u003c/em\u003e Plasma proteome profiling reveals dynamic of cholesterol marker after dual blocker therapy. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 3860 (2024).\u003c/li\u003e\n \u003cli\u003eBenson MD\u003cem\u003e, et al.\u003c/em\u003e Protein-metabolite association studies identify novel proteomic determinants of metabolite levels in human plasma. \u003cem\u003eCell Metab\u003c/em\u003e \u003cstrong\u003e35\u003c/strong\u003e, 1646-1660 e1643 (2023).\u003c/li\u003e\n \u003cli\u003eNusinow DP\u003cem\u003e, et al.\u003c/em\u003e Quantitative Proteomics of the Cancer Cell Line Encyclopedia. \u003cem\u003eCell\u003c/em\u003e \u003cstrong\u003e180\u003c/strong\u003e, 387-402 e316 (2020).\u003c/li\u003e\n \u003cli\u003eColwill K, Renewable Protein Binder Working G, Graslund S. A roadmap to generate renewable protein binders to the human proteome. \u003cem\u003eNat Methods\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 551-558 (2011).\u003c/li\u003e\n \u003cli\u003eGene Ontology C. Gene Ontology Consortium: going forward. \u003cem\u003eNucleic Acids Res\u003c/em\u003e \u003cstrong\u003e43\u003c/strong\u003e, D1049-1056 (2015).\u003c/li\u003e\n \u003cli\u003ePaysan-Lafosse T\u003cem\u003e, et al.\u003c/em\u003e InterPro in 2022. \u003cem\u003eNucleic Acids Res\u003c/em\u003e \u003cstrong\u003e51\u003c/strong\u003e, D418-D427 (2023).\u003c/li\u003e\n \u003cli\u003eMilacic M\u003cem\u003e, et al.\u003c/em\u003e The Reactome Pathway Knowledgebase 2024. \u003cem\u003eNucleic Acids Res\u003c/em\u003e \u003cstrong\u003e52\u003c/strong\u003e, D672-D678 (2024).\u003c/li\u003e\n \u003cli\u003eFrohlich K, Furrer R, Schori C, Handschin C, Schmidt A. Robust, Precise, and Deep Proteome Profiling Using a Small Mass Range and Narrow Window Data-Independent-Acquisition Scheme. \u003cem\u003eJ Proteome Res\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e, 1028-1038 (2024).\u003c/li\u003e\n \u003cli\u003eRueda F\u003cem\u003e, et al.\u003c/em\u003e Protein-based cardiogenic shock patient classifier. \u003cem\u003eEur Heart J\u003c/em\u003e \u003cstrong\u003e40\u003c/strong\u003e, 2684-2694 (2019).\u003c/li\u003e\n \u003cli\u003eShen B\u003cem\u003e, et al.\u003c/em\u003e Proteomic and Metabolomic Characterization of COVID-19 Patient Sera. \u003cem\u003eCell\u003c/em\u003e \u003cstrong\u003e182\u003c/strong\u003e, 59-72 e15 (2020).\u003c/li\u003e\n \u003cli\u003eGanesh K\u003cem\u003e, et al.\u003c/em\u003e Immunotherapy in colorectal cancer: rationale, challenges and potential. \u003cem\u003eNat Rev Gastroenterol Hepatol\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 361-375 (2019).\u003c/li\u003e\n \u003cli\u003eChalabi M\u003cem\u003e, et al.\u003c/em\u003e Neoadjuvant Immunotherapy in Locally Advanced Mismatch Repair-Deficient Colon Cancer. \u003cem\u003eN Engl J Med\u003c/em\u003e \u003cstrong\u003e390\u003c/strong\u003e, 1949-1958 (2024).\u003c/li\u003e\n \u003cli\u003eXia F\u003cem\u003e, et al.\u003c/em\u003e Randomized Phase II Trial of Immunotherapy-Based Total Neoadjuvant Therapy for Proficient Mismatch Repair or Microsatellite Stable Locally Advanced Rectal Cancer (TORCH). \u003cem\u003eJ Clin Oncol\u003c/em\u003e \u003cstrong\u003e42\u003c/strong\u003e, 3308-3318 (2024).\u003c/li\u003e\n \u003cli\u003eYang Z\u003cem\u003e, et al.\u003c/em\u003e Efficacy and safety of PD-1 blockade plus long-course chemoradiotherapy in locally advanced rectal cancer (NECTAR): a multi-center phase 2 study. \u003cem\u003eSignal Transduct Target Ther\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 56 (2024).\u003c/li\u003e\n \u003cli\u003eWang F\u003cem\u003e, et al.\u003c/em\u003e Efficacy and safety of combining short-course neoadjuvant chemoradiotherapy with envafolimab in locally advanced rectal cancer patients with microsatellite stability: A phase II PRECAM experimental study. \u003cem\u003eInt J Surg\u003c/em\u003e, (2024).\u003c/li\u003e\n \u003cli\u003eNavas-Carrillo D, Rodriguez JM, Montoro-Garcia S, Orenes-Pinero E. High-resolution proteomics and metabolomics in thyroid cancer: Deciphering novel biomarkers. \u003cem\u003eCrit Rev Clin Lab Sci\u003c/em\u003e \u003cstrong\u003e54\u003c/strong\u003e, 446-457 (2017).\u003c/li\u003e\n \u003cli\u003eGao Y\u003cem\u003e, et al.\u003c/em\u003e Acetylation-dependent regulation of PD-L1 nuclear translocation dictates the efficacy of anti-PD-1 immunotherapy. \u003cem\u003eNat Cell Biol\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 1064-1075 (2020).\u003c/li\u003e\n \u003cli\u003eBabacic H, Lehtio J, Pico de Coana Y, Pernemalm M, Eriksson H. In-depth plasma proteomics reveals increase in circulating PD-1 during anti-PD-1 immunotherapy in patients with metastatic cutaneous melanoma. \u003cem\u003eJ Immunother Cancer\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, (2020).\u003c/li\u003e\n \u003cli\u003eKroemer G, Kepp O. Radiochemotherapy-induced elevations of plasma HMGB1 levels predict therapeutic responses in cancer patients. \u003cem\u003eOncoimmunology\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 2005859 (2021).\u003c/li\u003e\n \u003cli\u003eWang H\u003cem\u003e, et al.\u003c/em\u003e Serum metabolic traits reveal therapeutic toxicities and responses of neoadjuvant chemoradiotherapy in patients with rectal cancer. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 7802 (2022).\u003c/li\u003e\n \u003cli\u003eLiu H\u003cem\u003e, et al.\u003c/em\u003e ADORA1 Inhibition Promotes Tumor Immune Evasion by Regulating the ATF3-PD-L1 Axis. \u003cem\u003eCancer Cell\u003c/em\u003e \u003cstrong\u003e37\u003c/strong\u003e, 324-339 e328 (2020).\u003c/li\u003e\n \u003cli\u003eLiu X\u003cem\u003e, et al.\u003c/em\u003e Inhibition of PCSK9 potentiates immune checkpoint therapy for cancer. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e588\u003c/strong\u003e, 693-698 (2020).\u003c/li\u003e\n \u003cli\u003eYang W\u003cem\u003e, et al.\u003c/em\u003e Potentiating the antitumour response of CD8(+) T cells by modulating cholesterol metabolism. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e531\u003c/strong\u003e, 651-655 (2016).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-5509842/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5509842/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eNeoadjuvant radio-chemotherapy (nCRT) plus immune checkpoint inhibitors (ICIs) have emerged as an effective antitumor regimen for locally advanced rectal cancer. Yet, few effective biomarkers are developed to monitor the therapy response. Herein, we investigate the longitudinal plasma proteome and metabolites profiling including 117 longitudinal samples from 50 patients who received nCRT plus PDL1 blockade therapy. Notably, the cholesterol metabolism is activated in the disease non-response group during the therapy. Correspondingly, the 1,4-cholestadienone, 7-methyloctanoylcarnitine and 3-hydroxybutyrylcarnitine, ABCA13, RAB3IP, GBA2 show significantly positive association with the cholesterol metabolism. Furthermore, by integrating proteome and metabolites approach, we identify a candidate metabolite (phosphatidylcholine, cholest-5-en-23-yn-3beta-ol) and proteome (APBB1IP, OLFM4, DNAJC19) that can reflect nCRT plus PDL1 response. Above, we establish a machine learning model to predict response, and the model performance is validated by repeated group-to-group validation with accuracy is 0.954. Thus, the plasma proteome and metabolites profiling strategy evaluate the alteration of cholesterol metabolism and identifies a panel of biomarkers.\u003c/p\u003e","manuscriptTitle":"Plasma proteome and metabolites profiling reveals dynamics for adverse events and responses after neoadjuvant radiochemotherapy plus PDL1 blockade in microsatellite-stable locally advanced rectal cancer: A prospective longitudinal study","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-07 07:05:57","doi":"10.21203/rs.3.rs-5509842/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"ad14b2d9-b7b6-4087-a7be-16946387f9c3","owner":[],"postedDate":"May 7th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":47862161,"name":"Biological sciences/Immunology"},{"id":47862162,"name":"Health sciences/Biomarkers/Predictive markers"}],"tags":[],"updatedAt":"2025-09-02T13:57:29+00:00","versionOfRecord":[],"versionCreatedAt":"2025-05-07 07:05:57","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-5509842","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5509842","identity":"rs-5509842","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}