An annotated biobank of triple negative breast cancer patient-derived xenografts featuring treatment-naïve and longitudinal samples throughout neoadjuvant chemotherapy

An annotated biobank of triple negative breast cancer patient-derived xenografts featuring treatment-naïve and longitudinal samples throughout neoadjuvant chemotherapy | bioRxiv /* */ /* */ <!-- <!-- /*! * yepnope1.5.4 * (c) WTFPL, GPLv2 */ (function(a,b,c){function d(a){return"[object Function]"==o.call(a)}function e(a){return"string"==typeof a}function f(){}function g(a){return!a||"loaded"==a||"complete"==a||"uninitialized"==a}function h(){var a=p.shift();q=1,a?a.t?m(function(){("c"==a.t?B.injectCss:B.injectJs)(a.s,0,a.a,a.x,a.e,1)},0):(a(),h()):q=0}function i(a,c,d,e,f,i,j){function k(b){if(!o&&g(l.readyState)&&(u.r=o=1,!q&&h(),l.onload=l.onreadystatechange=null,b)){"img"!=a&&m(function(){t.removeChild(l)},50);for(var d in y[c])y[c].hasOwnProperty(d)&&y[c][d].onload()}}var j=j||B.errorTimeout,l=b.createElement(a),o=0,r=0,u={t:d,s:c,e:f,a:i,x:j};1===y[c]&&(r=1,y[c]=[]),"object"==a?l.data=c:(l.src=c,l.type=a),l.width=l.height="0",l.onerror=l.onload=l.onreadystatechange=function(){k.call(this,r)},p.splice(e,0,u),"img"!=a&&(r||2===y[c]?(t.insertBefore(l,s?null:n),m(k,j)):y[c].push(l))}function j(a,b,c,d,f){return q=0,b=b||"j",e(a)?i("c"==b?v:u,a,b,this.i++,c,d,f):(p.splice(this.i++,0,a),1==p.length&&h()),this}function k(){var a=B;return a.loader={load:j,i:0},a}var l=b.documentElement,m=a.setTimeout,n=b.getElementsByTagName("script")[0],o={}.toString,p=[],q=0,r="MozAppearance"in l.style,s=r&&!!b.createRange().compareNode,t=s?l:n.parentNode,l=a.opera&&"[object Opera]"==o.call(a.opera),l=!!b.attachEvent&&!l,u=r?"object":l?"script":"img",v=l?"script":u,w=Array.isArray||function(a){return"[object Array]"==o.call(a)},x=[],y={},z={timeout:function(a,b){return b.length&&(a.timeout=b[0]),a}},A,B;B=function(a){function b(a){var a=a.split("!"),b=x.length,c=a.pop(),d=a.length,c={url:c,origUrl:c,prefixes:a},e,f,g;for(f=0;f<d;f++)g=a[f].split("="),(e=z[g.shift()])&&(c=e(c,g));for(f=0;f<b;f++)c=x[f](c);return c}function g(a,e,f,g,h){var i=b(a),j=i.autoCallback;i.url.split(".").pop().split("?").shift(),i.bypass||(e&&(e=d(e)?e:e[a]||e[g]||e[a.split("/").pop().split("?")[0]]),i.instead?i.instead(a,e,f,g,h):(y[i.url]?i.noexec=!0:y[i.url]=1,f.load(i.url,i.forceCSS||!i.forceJS&&"css"==i.url.split(".").pop().split("?").shift()?"c":c,i.noexec,i.attrs,i.timeout),(d(e)||d(j))&&f.load(function(){k(),e&&e(i.origUrl,h,g),j&&j(i.origUrl,h,g),y[i.url]=2})))}function h(a,b){function c(a,c){if(a){if(e(a))c||(j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}),g(a,j,b,0,h);else if(Object(a)===a)for(n in m=function(){var b=0,c;for(c in a)a.hasOwnProperty(c)&&b++;return b}(),a)a.hasOwnProperty(n)&&(!c&&!--m&&(d(j)?j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}:j[n]=function(a){return function(){var b=[].slice.call(arguments);a&&a.apply(this,b),l()}}(k[n])),g(a[n],j,b,n,h))}else!c&&l()}var h=!!a.test,i=a.load||a.both,j=a.callback||f,k=j,l=a.complete||f,m,n;c(h?a.yep:a.nope,!!i),i&&c(i)}var i,j,l=this.yepnope.loader;if(e(a))g(a,0,l,0);else if(w(a))for(i=0;i (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];var j=d.createElement(s);var dl=l!='dataLayer'?'&l='+l:'';j.src='//www.googletagmanager.com/gtm.js?id='+i+dl;j.type='text/javascript';j.async=true;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-M677548'); Skip to main content Home About Submit ALERTS / RSS Search for this keyword Advanced Search New Results An annotated biobank of triple negative breast cancer patient-derived xenografts featuring treatment-naïve and longitudinal samples throughout neoadjuvant chemotherapy View ORCID Profile Amanda L. Rinkenbaugh , Yuan Qi , Shirong Cai , Jiansu Shao , Faiza Hancock , Sabrina Jeter-Jones , Xiaomei Zhang , Emily Powell , Lei Huo , Rosanna Lau , Chunxiao Fu , Rebekah Gould , Petra den Hollander , Elizabeth E. Ravenberg , Jason B. White , Gaiane M. Rauch , Banu Arun , Clinton Yam , Alastair M. Thompson , View ORCID Profile Gloria V. Echeverria , Stacy L. Moulder , W. Fraser Symmans , View ORCID Profile Jeffrey T. Chang , View ORCID Profile Helen Piwnica-Worms doi: https://doi.org/10.1101/2024.11.25.625287 Amanda L. Rinkenbaugh 1 Department of Experimental Radiation Oncology, University of Texas MD Anderson Cancer Center , Houston, TX USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Amanda L. Rinkenbaugh Yuan Qi 1 Department of Experimental Radiation Oncology, University of Texas MD Anderson Cancer Center , Houston, TX USA 2 Department of Bioinformatics and Computational Biology, University of Texas MD Anderson Cancer Center , Houston, TX USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Shirong Cai 1 Department of Experimental Radiation Oncology, University of Texas MD Anderson Cancer Center , Houston, TX USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Jiansu Shao 1 Department of Experimental Radiation Oncology, University of Texas MD Anderson Cancer Center , Houston, TX USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Faiza Hancock 1 Department of Experimental Radiation Oncology, University of Texas MD Anderson Cancer Center , Houston, TX USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Sabrina Jeter-Jones 1 Department of Experimental Radiation Oncology, University of Texas MD Anderson Cancer Center , Houston, TX USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Xiaomei Zhang 1 Department of Experimental Radiation Oncology, University of Texas MD Anderson Cancer Center , Houston, TX USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Emily Powell 1 Department of Experimental Radiation Oncology, University of Texas MD Anderson Cancer Center , Houston, TX USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Lei Huo 3 Department of Pathology, University of Texas MD Anderson Cancer Center , Houston, TX USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Rosanna Lau 3 Department of Pathology, University of Texas MD Anderson Cancer Center , Houston, TX USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Chunxiao Fu 3 Department of Pathology, University of Texas MD Anderson Cancer Center , Houston, TX USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Rebekah Gould 3 Department of Pathology, University of Texas MD Anderson Cancer Center , Houston, TX USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Petra den Hollander 4 Department of Translational Molecular Pathology, University of Texas MD Anderson Cancer Center , Houston, TX USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Elizabeth E. Ravenberg 5 Department of Breast Medical Oncology, University of Texas MD Anderson Cancer Center , Houston, TX USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Jason B. White 5 Department of Breast Medical Oncology, University of Texas MD Anderson Cancer Center , Houston, TX USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Gaiane M. Rauch 6 Department of Abdominal Imaging, University of Texas MD Anderson Cancer Center , Houston, TX USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Banu Arun 5 Department of Breast Medical Oncology, University of Texas MD Anderson Cancer Center , Houston, TX USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Clinton Yam 5 Department of Breast Medical Oncology, University of Texas MD Anderson Cancer Center , Houston, TX USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Alastair M. Thompson 7 Department of Breast Surgical Oncology, University of Texas MD Anderson Cancer Center , Houston, TX USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Gloria V. Echeverria 1 Department of Experimental Radiation Oncology, University of Texas MD Anderson Cancer Center , Houston, TX USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Gloria V. Echeverria Stacy L. Moulder 5 Department of Breast Medical Oncology, University of Texas MD Anderson Cancer Center , Houston, TX USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site W. Fraser Symmans 3 Department of Pathology, University of Texas MD Anderson Cancer Center , Houston, TX USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Jeffrey T. Chang 2 Department of Bioinformatics and Computational Biology, University of Texas MD Anderson Cancer Center , Houston, TX USA 8 Department of Integrative Biology and Pharmacology, University of Texas Health Science Center , Houston, TX USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Jeffrey T. Chang For correspondence: hpiwnica-worms{at}mdanderson.org jeffrey.t.chang{at}uth.tmc.edu Helen Piwnica-Worms 1 Department of Experimental Radiation Oncology, University of Texas MD Anderson Cancer Center , Houston, TX USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Helen Piwnica-Worms For correspondence: hpiwnica-worms{at}mdanderson.org jeffrey.t.chang{at}uth.tmc.edu Abstract Full Text Info/History Metrics Supplementary material Preview PDF SUMMARY Triple negative breast cancer (TNBC) that fails to respond to neoadjuvant chemotherapy (NACT) can be lethal. Developing effective strategies to eradicate chemoresistant disease requires experimental models that recapitulate the heterogeneity characteristic of TNBC. To that end, we established a biobank of 92 orthotopic patient-derived xenograft (PDX) models of TNBC from the tumors of 75 patients enrolled in the ARTEMIS clinical trial ( NCT02276443 ) at MD Anderson Cancer Center, including 12 longitudinal sets generated from serial patient biopsies collected throughout NACT and from metastatic disease. Models were established from both chemosensitive and chemoresistant tumors, and nearly 30% of PDX models were capable of lung metastasis. Comprehensive molecular profiling demonstrated conservation of genomes and transcriptomes between patient and corresponding PDX tumors, with representation of all major transcriptional subtypes. Transcriptional changes observed in the longitudinal PDX models highlight dysregulation in pathways associated with DNA integrity, extracellular matrix interactions, the ubiquitin-proteasome system, epigenetics, and inflammatory signaling. These alterations reveal a complex network of adaptations associated with chemoresistance. This PDX biobank provides a valuable resource for tackling the most pressing issues facing the clinical management of TNBC, namely chemoresistance and metastasis. INTRODUCTION Triple negative breast cancer (TNBC) is a subtype of breast cancer characterized by the absence of estrogen and progesterone hormone receptor expression and the lack of amplification or overexpression of ERBB2 , encoding human epidermal growth factor receptor 2. The lack of these receptors precludes the use of some of the most effective targeted breast cancer treatments for these patients. Currently, systemic neoadjuvant chemotherapy (NACT) is the established treatment approach for early stage TNBC 1 , with recent approval for immunotherapy in the neoadjuvant setting 2 . However, up 40-50% of patients with TNBC are found to have residual disease post-NACT, even with the addition of immunotherapy, leading to a heightened risk of early recurrence after surgical resection 1 , 3 . Identifying upfront predictors of response to NACT remains a critical clinical challenge, as it could help tailor treatments to those that would be most effective with the least toxicity. To understand response to NACT, the ARTEMIS trial (A Robust TNBC Evaluation FraMework to Improve Survival, NCT02276443 ) evaluated whether a gene expression signature 4 could be used to predict patients with chemoresistant disease and direct them to phase II clinical trials of targeted therapies based on biomarkers from their pre-treatment tumors. Tumor tissue was collected at multiple clinical timepoints: before NACT initiation, from tumors showing a poor response after four cycles of doxorubicin and cyclophosphamide (AC), at surgical resection following completion of NACT, and from metastatic disease. Patient tumors were profiled by whole exome sequencing (WES) and transcriptome sequencing (RNA-seq). It is important to note that the ARTEMIS trial preceded KEYNOTE-522 2 , 5 , 6 , and therefore none of the patients received immunotherapy as part of the standard of care treatment, although some patients were enrolled in a phase II trial combining atezolizumab with nab-paclitaxel 7 . Patient-derived xenograft (PDX) models have been established as generally faithful recapitulations of their source tumors, providing a platform for both experimental and translational research 8 – 13 . To support translational research in TNBC and mitigate the challenges posed by limited patient tumor tissue, we established a biobank of orthotopic PDX models from tumors collected in the ARTEMIS trial. The biobank includes 92 PDX models derived from the tumors of 75 patients, including 12 longitudinal sets established from serially sampled tumors across trial timepoints or metastatic lesions. The models encompassed a range of TNBC subtypes. A high degree of conservation in genomes, transcriptomes, and molecular subtypes was observed between PDX tumors and their respective patient tumors based on WES and RNA-Seq. The genomic profiles of the PDX tumors accurately reflected the landscape of TNBC, with TP53 being the most frequently observed genetic alteration. In addition, lung metastases were detected in a significant fraction of PDX models. Thus, this annotated PDX biobank comprehensively captures the genomic and transcriptional diversity of TNBC, including representation of each molecular subtype, and promises to be a valuable resource for addressing critical challenges facing TNBC patients with chemoresistant and metastatic disease. RESULTS Generation of a TNBC PDX biobank in alignment with the ARTEMIS trial Patients with TNBC enrolled in the prospective neoadjuvant ARTEMIS trial ( NCT02276443 ) provided tumor material for WES, RNA-seq, histology, and establishment of PDX models via implantation into mice. Longitudinal patient samples were collected at different stages of neoadjuvant treatment: prior to treatment (pre-NACT), after four cycles of doxorubicin and cyclophosphamide (AC; mid-NACT), following treatment with paclitaxel, alone or in combination with an experimental targeted therapy (post-NACT), and from metastatic disease ( Figure 1A ). For each patient sample, we pooled cells from three independent fine needle aspirations (FNAs) and engrafted them into the pre-humanized fourth mammary fat pads (MFPs) of non-obese diabetic/severe combined immunodeficient (NOD/SCID) mice as described previously 8 , 14 , 15 . We classified PDX models as established once they reached the third passage and consistently passed quality control assays. These tests included short-tandem repeat (STR) DNA fingerprinting to verify the unique tumor profile and corresponding patient match of each tumor, quantitative PCR (qPCR) using human and mouse-specific DNA probes to confirm the human origin of tumor cells, and PCR to confirm the absence of GFP, which was present in immortalized human stromal fibroblasts used to humanize the mouse MFPs before the initial engraftment of patient tumor cells. Download figure Open in new tab Figure 1: Generation of a TNBC PDX biobank in alignment with the ARTEMIS trial (A) Schematic of ARTEMIS trial design where all treatment-naïve patients received four cycles of AC, followed by a response assessment via ultrasound or MRI. Patients with ≥70% tumor volume reduction remained on standard of care treatment, while patients with <70% tumor volume reduction were offered one of four clinical trials. The included table describes the associated phase II clinical trials that offered targeted therapies to patients with suboptimal AC response. Samples were collected from patients at key timepoints (pre-, mid-, and post-NACT and metastasis) to be analyzed by WES, RNA-seq, and histology, as well as implanted into mice for PDX model generation. (B) Visualization of 12 longitudinal sets encompassing pre-, mid-, and post-NACT, and metastatic samples. Patient identifier is shown above each column of corresponding samples. Biopsy timepoint is indicated to the left of each row. AC: doxorubicin and cyclophosphamide, RCB: residual cancer burden, WES: whole exome sequencing, Out of the 92 established PDX models derived from tumors of 75 patients, 71 originated from pre-NACT biopsies, 11 from mid-NACT biopsies, eight from post-NACT biopsies and two from metastases (PIM356: chest wall, PIM411: brain). Additionally, the collection includes 12 longitudinal sets, each featuring a pre-NACT model along with at least one additional model from mid-NACT, post-NACT, or metastasis ( Figure 1B ). As part of ARTEMIS, four phase II therapeutic trials ( Figure 1A ) were available to explore targeted therapies in combination with paclitaxel for those patients with suboptimal responses to AC, based on tumor volume reduction by ultrasound or MRI and clinical assessment. Eligibility for the phase II trials followed a hierarchical selection process based on specific tumor characteristics determined at diagnosis. Patients with tumors exhibiting ≥20% tumor infiltrating lymphocytes (TILs) were eligible for treatment with atezolizumab and nab-paclitaxel ( NCT02530489 ) 7 , then tumors with metaplastic histology or vimentin IHC expression >50% were eligible for liposomal doxorubicin and bevacizumab combined with temsirolimus or everolimus ( NCT02456857 ) 16 , then tumors with androgen receptor IHC expression ≥10% were eligible for enzalutamide and paclitaxel ( NCT02689427 ) 17 . If none of these biomarkers were present, patients were eligible for panitumumab in combination with carboplatin and paclitaxel ( NCT02593175 ) 18 . Of the eight PDX models established from post-NACT tumors, two were from patients who received paclitaxel as standard of care (PIM114, PIM311); three were from patients enrolled in NCT02530489 (PIM051, PIM321, PIM328); one was from a patient enrolled in NCT02689427 (PIM407); and two were from patients enrolled in NCT02593175 (PIM291 and PIM404) ( Figure 1A ). The PDX collection was established from a diverse patient population with 21% identifying as black, 16% as Hispanic or Latino, and 7% as Asian (Table S1). Histological analysis demonstrated that 96% of the PDX tumors were invasive ductal carcinoma, while the remaining four models were classified as metaplastic tumors. Longitudinal PDX models generally retained the same histology across timepoints and shared the histology of the corresponding patient samples (Figure S1) 19 . Detailed clinical annotation and characterization of each PDX model can be found in Table S1. We collected 392 patient tumors to engraft into mice, which produced the final collection of 92 models resulting in a take rate of 23%, in agreement with our previously reported interim analysis 14 . In that study, we also found that therapy-resistant tumors, defined either as RCB-II/III after NACT or with metastatic relapse within two years of surgery, were positively correlated with successful PDX engraftment 14 . Assessing the entire biobank, we confirmed that both RCB-II/RCB-III tumors and tumors with metastatic relapse more frequently led to successful establishment of a PDX model. The recent prospective clinical trial TOWARDS further underscored the relationship between successful PDX engraftment and metastatic relapse 20 . The patient biopsy timepoint did not correlate with take rate (Figure S2A-2SF). Finally, we investigated the time required for each model to reach passage three and observed that higher RCB status correlated with a faster time to PDX establishment, independent of molecular subtype (Figure S2G-S2I). Given that TNBC is known for its aggressive metastatic behavior, we investigated our PDX models for the presence of lung metastases at endpoints following primary tumor resection or declining health of the mice. We analyzed hematoxylin and eosin-stained lung sections, and in some cases, stained for cytokeratins and human mitochondria ( Figure 2A ). Our findings revealed that 28% of the examined models (21/75) could colonize the lungs of engrafted mice. Metastasis rates between transcriptional subtypes did not vary significantly perhaps due to sample size limitations, though BL1 and LAR trended towards higher rates ( Figure 2B ). There was no significant association between RCB status and the propensity for lung metastasis ( Figure 2C ). It is possible that we would have detected metastases in additional models if primary tumors were given more time in vivo or if more sensitive detection methods were employed. Based on rapid and overt lung metastases detected in PIM056, a pre-treatment model, we engineered it to express click beetle red-luciferase and used bioluminescence imaging to assess metastasis in vivo. We identified metastatic lesions in the lung, liver, brain, and bone through this approach ( Figure 2D ). These results agreed in part with the medical history of the patient who exhibited both lung and liver metastases. Multi-organ metastasis has previously been reported in other PDX models 11 , 21 , 22 , and it will be important to assess the extent to which additional PDX models in our collection are capable of multi-organ metastasis. Download figure Open in new tab Figure 2: Lung metastasis is a common feature of TNBC PDX models (A) H&E and IHC staining of pan-cytokeratin and human-specific mitochondria in the primary tumors and lung metastases from PIM056 and PIM084. Scale bars are 100 µm. (B) Quantification of metastasis-competent PDX models based on transcriptional subtype (C) Quantification of metastasis-competent PDX models based on RCB status (D) Bioluminescence imaging of PIM056 labeled with CBR-luciferase identifies metastases in the lung, liver, brain, and bone. Genomic and transcriptomic profiles of PDX tumors reflect the diversity of human TNBC We profiled all 92 PDX models, representing the entire PDX collection, by WES and RNA-seq. TP53 was the most frequently altered gene, with 50 tumors exhibiting a somatic mutation, including 70% missense, 22% stopgain, and 10% splicing mutations ( Figure 3A ), in line with other studies of TNBC 23 – 25 . To evaluate the distribution of somatic mutations in our PDX dataset, we compared mutation frequencies with those from the METABRIC TNBC cohort 26 , 27 , focusing on the top 20 most frequently mutated genes in that dataset. Fisher’s exact tests showed no significant difference in mutation frequencies between our PDX dataset and the METABRIC TNBC cohort, except for a lower prevalence of TP53 mutations in our PDX models ( Figure 3B ), and their corresponding patient tumors. Download figure Open in new tab Figure 3: Genomic and transcriptomic profiles of PDX tumors reflect the diversity of human TNBC (A) Oncoprint showing mutations of commonly mutated cancer genes from the COSMIC database. Models derived from the same patient are indicated by bars below the PDX ID. (B) Comparison of mutation frequency of the top 20 most frequently mutated genes in METABRIC TNBC samples and TNBC PDX biobank. (C) Stacked barplot shows the relative contributions of mutational signatures in each PDX model. Models are in order from highest to lowest overall mutation burden, reflected in bar graph at the top. RCB status is annotated at the bottom. (D) Heatmap of the 250 most variable genes across the RNAseq dataset of all 92 PDX models. Annotations for each PDX model include patient ID, transcriptional subtype, RCB status, and biopsy timepoint. (E) Summary of the number of PDX models assigned to each transcriptional subtype for a 6-class or 4-class system We also found multiple pathogenic alterations in the PI3K pathway, including four PIK3CA missense mutations (E542K in PIM049, PIM165, and PIM285; H1047R in PIM053) and three nonsense PTEN mutations. None of the patients had germline mutations in BRCA1/2 as those patients were recruited to a different clinical trial at MD Anderson Cancer Center 28 , 29 . However, five models had somatic BRCA1 mutations, and nine models had somatic BRCA2 mutations. While most of these mutations were classified as benign or variants of unknown significance, PIM371 harbored a stopgain BRCA1 mutation (R1443X) that is reported to be deleterious 30 . Comparing the longitudinal models, we found that 96 out of 213 non-silent mutations in COSMIC genes were consistently present across all timepoints. Notably, PIM190 and PIM231, established from the pre- and mid-NACT samples of patient TNBC283, exhibited the greatest discordance. This discrepancy aligns with the variability observed in STR profiling across multiple passages of these models, where the number of microsatellite repeats fluctuated within the acceptable range of the assay, allowing both models to still be assigned the same profile. We assigned the PDX models into one of three categories based on their tumor mutation burden: high (>15 non-synonymous somatic mutations/Mb), medium (5-15 mutations/Mb), and low (<5 mutations/Mb). 71% of tumors fell in the low category, 25% in the medium category, and 4% in the high category ( Figure 3C , top). Analysis of the mutational signatures showed that the high and medium mutation burden tumors were dominated by signature SBS5, while the low mutation burden tumors were characterized by signature SBS3 ( Figure 3C , bottom) 31 . Signature SBS5 is associated with a clock-like etiology, while signature SBS3 is linked with defects in homologous recombination repair. Neither signature aligned with readily discernible characteristics of the patients (such as age) or their corresponding PDX models (such as BRCA status). Additionally, we used WES data to generate genome-wide copy number profiles for each PDX model. We observed frequent gains of chromosomes 1q and 8q, as well as losses on chromosome 5, which is consistent with previous findings in human TNBC (Figure S3). Comparing longitudinal changes across the matched sets, we noted a gain of 16p in five of the mid- or post-NACT models. The gain of 16p has been linked to increased immune evasion 32 , which may indicate an adaptive response of these tumors to treatment. Based on transcriptional profiling, we clustered the PDX models using the top 250 most variable genes. We found that 11 out of 12 longitudinal sets were closely clustered together, suggesting that patient-level heterogeneity is a more significant factor influencing transcriptional state than treatment timepoint ( Figure 3D ). To assign transcriptional subtypes to each tumor, we utilized TNBCtype, opting for the 4-type 33 over the 6-type 34 classifier. The 4-type classifier better reflects cancer cell properties and minimizes stromal influences, which is better aligned with our gene expression profiles since we discarded the host stromal reads when processing the data from PDX tumors. All four molecular subtypes of TNBC, basal-like 1 (BL1), basal-like 2 (BL2), luminal androgen receptor (LAR), and mesenchymal (M) 33 , 34 , are represented in our PDX collection ( Figure 3E ). Additionally, our analysis revealed significant intratumoral heterogeneity when examining subtype coefficients as continuous variables. Specifically, 34 out of 92 (37%) of the PDX models scored significantly for more than one subtype (Figures S4, S7). For tumors with multiple significant subtype coefficients, we found that the BL1 and M subtypes significantly co-occurred (p = 0.005), while the BL2 and LAR subtypes were the next most frequently co-occurring but did not reach statistical significance (p = 0.13). PDX models conserve genotype and phenotype of patient samples After analyzing the genomic and transcriptional landscape of our PDX biobank, we next assessed how well the PDX models preserved the features of the corresponding patient tumors. WES data was available for 71 out of 75 patients, accounting for 87 PDX tumors and 11 of the 12 longitudinal sets, while RNA-seq data was available from 70 patients, accounting for 84 PDX tumors and 10 of 12 longitudinal sets (Table S2). Our analysis of conserved mutations between patient tumor and corresponding PDX tumor revealed a median conservation rate of 0.55, and no correlation with tumors from non-corresponding patients ( Figure 4A ). As TP53 was the most frequently mutated gene, we specifically examined if TP53 mutations were conserved between patient and PDX tumors. There was a 94% concordance in TP53 alterations between patient and PDX tumors. Exceptions included PIM166 and PIM263 with pathogenic Y220C and G245D mutations respectively, not found in the corresponding patient tumors, and PIM056, which lacked the pathogenic R342X mutation detected in the patient tumor. Additionally, the P72R mutation present in the pre-NACT patient tumor from TNBC249 was not detected in the mid-NACT patient tumor or the associated PDX models, PIM137 (pre-NACT) or PIM172 (mid-NACT). P72R is a common TP53 polymorphism, which is characterized as benign in ClinVar but has been associated with increased metastasis and poorer prognosis in p53-mutant breast cancer 35 , 36 . Download figure Open in new tab Figure 4: PDX models conserve genotype and phenotype of patient samples (A) Heatmap of consistency rates for mutations between patient samples and PDX models. Longitudinal sets of samples are indicated with patient color scheme. (B) Scatter plot showing the correlation between mutation burden in patient and PDX samples, with color scheme representing high, medium, and low mutation burden categories. (C) Heatmap of the Spearman correlation coefficients for the 1000 most variable genes between patient samples and PDX models. Longitudinal sets of samples are indicated with patient color scheme. (D) Differential ssGSEA pathway scores for the Hallmark pathways between patient samples and PDX models (E) Confusion matrix of subtype assignments between patient samples and PDX models. We also assessed the overall mutation burden and discovered that 90 out of 92 PDX models (98%) were classified into the same category of mutation burden as their corresponding patient tumor ( Figure 4B ). The two exceptions, PIM056 and PIM137, may be attributed to variations in the subclonal populations sampled for sequencing and PDX model establishment, or the possibility that certain subclones, while detected in patient tumors, were unable to proliferate in mice. Transcriptomic analysis via bulk RNA sequencing showed a median Spearman correlation of 0.59 for the 1000 most variable genes between patient and PDX tumors, indicating a notable preservation of expression patterns between patient tumors and their corresponding PDX tumors ( Figure 4C ). Gene set enrichment analysis (GSEA) of the 50 Hallmarks pathways 37 revealed distinct patterns of pathway expression: patient samples exhibited higher expression of angiogenesis and inflammation-related pathways, whereas PDX tumors showed increased proliferation and DNA repair pathways ( Figure 4D ). These differences may be due, in part, to the inclusion of stromal transcriptomes in the patient samples that were removed from the PDX samples by discarding transcripts mapped to a mouse reference genome. Additionally, the predominance of the BL1 molecular subtype in the PDX collection ( Figure 3E ) likely contributes to the observed enrichment in proliferation and DNA repair pathways noted in the PDX samples. We also assessed conservation of transcriptional subtypes between patient and PDX tumors using the 4-type classification system. 54% of the most significant subtypes in the patient tumors were also the most significant in the corresponding PDX models, with the highest concordance in BL1 tumors ( Figure 4E ). As previously noted, the subtype assignments can be ambiguous, where tumors can obtain statistically significant scores for multiple subtypes. When comparing all statistically significant subtype assignments, we found a 71% concordance between patient and PDX tumor subtypes (Table S1). While the PDX models generally recapitulated properties of their corresponding patient samples, some discrepancies were noted. To quantify their overall similarity, we calculated conservation scores across five categories: mutations, mutation burden, subclonal architecture, gene expression, and transcriptional subtype. Each PDX model was scored as 0 (weak conservation), 0.5 (moderate conservation), or 1 (strong conservation) in each category, with the total conservation score derived from the sum across all categories. There was no significant correlation between conservation scores and various biological variables, including treatment timepoint, mutation burden, RCB status, patient metastasis, or patient or PDX transcriptional subtype (Supplemental Figure 5A-F). Among the longitudinal sets, we found that seven out of nine PDX sets either maintained or improved their conservation score across treatment timepoints (Supplemental Figure 5G). Tumors within longitudinal PDX sets exhibit AC resistance We conducted preclinical studies to evaluate the response of several longitudinal PDX tumor sets to AC ( Figure 5 ). These sets included four pre-/mid-NACT pairs, two pre-/post-NACT pairs, and two pre-/mid-/post-NACT triplets. Mice engrafted with these tumors were treated with either vehicle or a single dose of AC at our previously determined maximum tolerated dose for NOD/SCID mice 38 , and tumor volumes were measured twice weekly for up to 90 days. As anticipated, all tumors exhibited resistance to AC, given that seven out of the eight sets were derived from resistant patient tumors (RCB-II/RCB-III). Although the patient whose tumor was used to establish the PIM137/PIM172 pair achieved a pCR on the phase II immunotherapy trial ( NCT02530489 ) 7 , her tumor volume increased during AC treatment as measured by ultrasound, predicting it was AC-resistant. Download figure Open in new tab Figure 5: Tumors within longitudinal PDX sets exhibit AC resistance Tumor response to one dose of AC on Day 0, showing mean tumor volume relative to starting tumor volume. Longitudinal sets are indicated by color scheme and model timepoint in indicated to the left of each row. Growth curves are based on n=2-5 mice/group, as indicated in the graph legends. Data are presented as the mean ± SEM. Statistical significance is reported based on unpaired two-tailed t-tests comparing the area under the curve values for vehicle vs. AC-treated tumors from treatment start until vehicle tumors reached endpoint. The increase in tumor volume observed in most AC-treated mice was either similar or slightly slowed relative to the vehicle-treated controls ( Figure 5 ), suggesting intrinsic resistance. In one case, PIM254, a pre-NACT tumor, initially regressed during AC treatment, but then ceased responding and resumed growth ( Figure 5 , bottom left). In contrast its corresponding post-NACT tumor, PIM311, grew continuously on AC treatment. This finding aligns with the patient’s treatment history, as the tumor showed a 95% decrease in volume at the mid-treatment timepoint but was classified as RCB-II at the time of surgery. Analysis of the subclonal architectures, transcriptomes, and proteomes of longitudinal PDX tumor sets We analyzed both patient and PDX tumors before and during treatment to determine how chemotherapy impacted subclonal architecture. Using PyClone 39 , we estimated cancer cell frequency after accounting for tumor purity and local copy number status. This analysis did not reveal subclone enrichment after AC treatment in any of the patient or PDX tumors when comparing pre-NACT tumors to mid-NACT tumors (Figure S6). By contrast, the analysis of one patient tumor (TNBC139) showed subclone expansion in both the post-NACT tumor and its corresponding PDX tumor, PIM311 (Figure S6, bottom left). Notably, these samples correspond to the PIM254/PIM311 PDX pair with distinct AC responses. Additionally, two longitudinal sets included metastatic lesions: one (PIM411) paired with a pre-NACT tumor (PIM384), and the other (PIM356) paired with both pre- (PIM077) and mid-NACT (PIM110) tumors. In both cases, the metastatic lesions exhibited subclone selection relative to the earlier timepoints, consistent in both patient and PDX samples. We analyzed the transcriptomes of tumors within each longitudinal PDX set to identify therapy-induced changes in the Hallmark pathways 37 . Table S3 summarizes data available for each model among the longitudinal sets. Each set of models exhibited a distinct set of altered pathways, reflecting the substantial patient-level heterogeneity seen in TNBC datasets 40 , 41 . In the mid-NACT models, the most commonly upregulated pathways relative to pre-NACT models could be grouped into three major categories: proliferation (E2F targets, MYC targets V1, G2M checkpoint), inflammatory signaling (interferon alpha and gamma responses, and inflammatory response), and epithelial-mesenchymal transition ( Figure 6A ). These pathways were upregulated in an overlapping, but not identical, subset of models with upregulation observed in 4-5 mid-/pre-NACT comparisons. Conversely, the most commonly downregulated pathways in mid-NACT compared to pre-NACT tumors were hypoxia and TNFɑ signaling via NF-κB. The dichotomy between upregulated interferon signaling and downregulated TNFɑ signaling may reflect two distinct types of chemotherapy response, as these signatures typically change in concert. Notably, only one mid-/pre-NACT comparison (TNBC231) showed these pathways significantly altered in opposite directions; in all other comparisons, either only one pathway was significantly changed, or both pathways moved in the same direction. To gain deeper insights into transcriptional changes over time, we also analyzed the samples using the MSigDB C2 pathways 42 . We identified 81 pathways that were significantly altered in at least 5 out of 9 instances, which fell into four main categories: extracellular matrix (ECM), ubiquitin-proteasome system, DNA integrity, and epigenetics ( Figure 6B ). Download figure Open in new tab Figure 6: Analysis of the transcriptomes of longitudinal PDX tumor sets (A) Pairwise comparisons of differentially regulated Hallmark pathways within the longitudinal sets. Models are indicated by patient ID with color scheme. Dot size represents significance, while color indicates the direction of change. (B) Number of models that up- or down-regulated each pathway when comparing mid-NACT to pre-NACT models. (C) Number of models that up- or down-regulated each pathway when comparing post-NACT to pre-NACT models. Color scheme indicates biological category for each pathway. Comparing post-NACT models to their pre-NACT or mid-NACT counterparts, we observed downregulation of multiple Hallmark pathways, many of which were related to inflammatory signaling (interferon alpha and gamma responses, inflammatory response, TNFɑ signaling via NF-κB). Other commonly downregulated pathways included epithelial-mesenchymal transition and UV response down ( Figure 6A ). A caveat to comparisons with the post-NACT models is that they were derived after different treatment courses, including targeted therapies. Due to the limited representation of post-NACT models from each of the associated phase II trials, we cannot draw conclusions about distinct responses to each therapy. As a result, we treated these models as a single group in our analyses. Similar changes were observed in metastasis-derived models relative to their pre-NACT models. Using the MSigDB C2 pathway set for post-versus pre-NACT tumor comparisons, we identified 27 pathways significantly altered in at least 3 out of the 6 comparisons, primarily associated with ECM or inflammatory signaling ( Figure 6C ). We examined how these transcriptional changes impacted subtype correlation, both as categorical ( Table 1 ) or continuous variables (Figure S7), finding instances of both subtype maintenance and subtype switching with treatment. View this table: View inline View popup Download powerpoint Table 1: Molecular subtyping of longitudinal patient and PDX tumors Additionally, we used reverse-phase protein arrays (RPPA) to analyze proteomic changes in a subset of the longitudinal PDX sets. Clustering of the RPPA profiles showed that all but one of the matched sets clustered together, reinforcing the significant patient-level heterogeneity we consistently observed across these models and in TNBC generally ( Figure 6A , Table S4). We also investigated protein markers associated with subtype coefficients and found the BL1 and BL2 subtypes were inversely correlated with many markers ( Figure 6B ). We observed associations with growth factor signaling in BL2 tumors and PI3K signaling in LAR tumors, consistent with other studies 41 . DISCUSSION Here, we report the establishment of an annotated biobank of TNBC PDX models, generated from tumor tissue obtained in the context of the ARTEMIS neoadjuvant clinical trial, where patient response to therapy is known. The biobank consists of 92 PDX models derived from the tumors of 75 patients, including 12 longitudinal sets established from serially sampled tumors across trial timepoints and metastatic lesions. Most published PDX collections have been established using single biopsies obtained from the primary and metastatic tumors of patients who had undergone therapy 9 , 10 , 12 , 43 – 46 . In contrast, our collection was established from pooled fine needle aspirates (FNAs) and features 71 treatment-naïve models. Combining FNAs from different regions of the tumor is expected to capture the heterogeneity of a TNBC tumor more fully than a single biopsy can. Additionally, FNAs are less painful to patients than repeated core biopsies, increasing the likelihood that patients will consent to serial sampling. This approach enabled us to generate 12 longitudinal PDX models from serial patient sampling throughout NACT, as well as from metastatic disease. Another distinctive feature of our biobank is the acquisition of tumor tissue alongside a single neoadjuvant clinical trial, where detailed information on patient characteristics, tumor features, and pathologic response outcomes are known. PDX models were established from patients with long-term follow-up, providing valuable clinical insight into tumor aggressiveness and sensitivity to various systemic treatments. Furthermore, nearly all patient tumors have been characterized using WES and RNA-seq, allowing for direct comparisons to their corresponding PDX tumors. Comparative analyses demonstrated that PDX tumors largely recapitulated the genomic and transcriptomic characteristics of their corresponding patient tumors and reflect the molecular and phenotypic heterogeneity of TNBC. The greatest differences in the transcriptomes between patient and PDX tumors were those related to immune and stroma interactions, which is expected when adapting human tumors for growth in immune-deficient murine hosts. Nonetheless, we saw a strong recapitulation of gene expression patterns related to signaling pathways, suggesting that these PDX models effectively capture many of the key biological features of the corresponding patient tumors. The collection includes representation from each transcriptional subtype (BL1, BL2, LAR, and M), demonstrating that the PDX biobank represents the extensive inter-patient heterogeneity characteristic of TNBC. While we successfully established PDX models from patients who achieved pCR or RCB-I following NACT, these models did not respond with full tumor elimination in AC treatment preclinical studies in vivo 38 . This could be due in part to the inability to reach human-equivalent doses of chemotherapy in mice due to toxicity. Additionally, it is possible that an intact immune system, which is absent in NOD/SCID mice, is necessary to achieve this level of response 47 . Indeed, analysis of the ARTEMIS patient samples found that nearly all immune cold tumors were chemoresistant, underscoring the interplay between the immune system and chemotherapy response (Seth et al ., in preparation ). With the increasing utilization of immune checkpoint inhibitors in the neoadjuvant setting, it will be important to complement PDX studies with immunocompetent models and primary patient tumors whenever possible. Nonetheless, given that most immune cold and a portion of immune hot tumors do not respond to immunotherapy, PDX models will be particularly useful in identifying intrinsic mechanisms of therapy resistance in TNBC. Models derived from treatment-naïve tumors provide the opportunity to investigate tumors that have not undergone therapy-induced selection, providing a valuable platform to identify potential therapies for targeting resistance in primary, treatment-naïve TNBC. A high-throughput drug screen of 16 treatment-naïve models from our collection revealed that a kinesin spindle protein inhibitor (KSPi) was broadly active across all molecular subtypes 48 . The KSPi targets the mitotic spindle through mechanisms that are independent of microtubule stability and showed efficacy in PDX models that were resistant to paclitaxel, a microtubule inhibitor used in standard TNBC treatment. In addition, subtype selectivity was observed with PRIMA-1 MET which exhibited enhanced efficacy in mesenchymal subtype tumors 48 . The longitudinal models are another unique feature of our PDX collection. These models facilitate studies on how NACT impacts TNBC heterogeneity. The subclonal architecture of tumors within the longitudinal sets remained largely unchanged in response to AC, which is not unexpected given that these tumors came from patients whose tumors were largely resistant to AC (RCB-II and RCB-III). These results contrast with some post-NACT tumors and metastases where changes in subclonal composition were observed and are consistent with barcoding studies investigating chemoresistance and metastasis conducted in PDX tumors 22 , 38 . Throughout our studies, we found that patient identity had a greater impact on transcriptional state than treatment timepoint or resistance status in the longitudinal PDX models. A recent study of isogenic carboplatin-sensitive and -resistant PDX models reached a similar conclusion 49 . Analysis of transcriptional changes in longitudinal tumors identified several pathways that were dysregulated in response to treatment. Inflammatory signaling was commonly downregulated in post-NACT and metastatic tumors relative to pre-and mid-NACT tumors. This finding is consistent with therapy-resistant tumors being able to evade immune recognition, perhaps by suppressing the immune system. Accordingly, immunotherapy has been a recent addition to TNBC treatment paradigms, aiming to engage the immune system in tumor elimination 2 , 5 . The ubiquitin-proteasome pathway was dysregulated in mid-NACT models relative to treatment-naïve models. A previously reported genome-wide siRNA screen identified a dependence on proteasome activity in TNBC 50 . Although bortezomib, a proteasome inhibitor, did not show efficacy as a single agent in patients with metastatic breast cancer 51 , 52 , some responses were observed when bortezomib was combined with capecitabine 53 . We conducted a high throughput drug screen on tumors in our longitudinal collection and identified increased sensitivity to inhibitors targeting both the proteasome and neddylation pathways in mid-/post-NACT tumors compared with pre-NACT tumors 54 . Neddylation is essential for full activity of the Cullin-RING ligase complexes and thus is directly tied to the function of the ubiquitin-proteasome system 55 . Pevonedistat, a neddylation inhibitor, showed efficacy in several of the longitudinal PDX models 54 . There continues to be broad interest in investigating additional inhibitors that impact the ubiquitin-proteasome system in breast cancer, including those that target NRF1, UBA1, and PIM kinase 56 – 58 . Other pathways commonly dysregulated in mid-NACT models relative to pre-NACT models included genome integrity and epigenetics. Drugs targeting proteins in these pathways are also under investigation for the treatment of chemoresistant TNBC 59 – 66 . In summary, our PDX collection provides a valuable resource for the TNBC community, as it recapitulates the heterogeneity characteristic of TNBC and can be used as an experimental model to address the most pressing issues facing the clinical management of TNBC, namely chemoresistance and metastasis. Resource Availability Lead Contact Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Helen Piwnica-Worms ( hpiwnica-worms{at}mdanderson.org ) Materials Availability All PDX models generated in this study are available from the authors upon request. Data and Code Availability Processed RNA-seq data are available at GEO #######. Additionally, clinical and genomics data characterizing the PDX biobank is available at pdxportal.research.bcm.edu. Code used to parse, analyze, and visualize the genomics data is available at ####. Author contributions Conceptualization: ALR, JTC, HPW Methodology: ALR, EP, PdH, GVE, WFS, AMT, SLM, JTC, HPW Formal Analysis: ALR, YQ, JTC Investigation: SC, JSS, SJJ, XZ, CF, RL Resources: EER, JBW, LH, GMR, BA, CY, SLM Data Curation: FH, EER, JBW, RG, CY, YQ, JTC Writing – Original Draft: ALR, YQ, JTC, HPW Writing – Review and Editing: ALR, GVE, AMT, WFS, LH, SLM, JTC, HPW Visualization: ALR, YQ Supervision: WFS, SLM, JTC, HPW Project Administration: ALR, FH, EER, JBW, RG, GMR, BA, CY, WFS, SLM, JTC, HPW Funding Acquisition: WFS, AMT, SLM, JTC, HPW Declaration of interests S.L.M. and E.E.R. are employees of and hold stock in Eli Lilly and Company. The spouse of A.M.T. works for Eli Lilly and Company. W.F.S. owns founder stock in Delphi Diagnostics and publicly traded stock in IONIS Pharmaceuticals and Eiger BioPharmaceuticals; is a consultant/advisor to Merck, Astra Zeneca. W.F.S. and C.F. are co-inventors of pending patent “Targeted Measure of Transcriptional Activity Related to Hormone Receptors”, United States, Provisional Patent Application Serial No. 62/329,774. G.V.E. receives sponsored research funding from Chimerix, Inc. and experimental compounds from the Lead Discovery Center of Germany. C.Y. has received research funding (to the institution) from Genentech, Gilead, BostonGene, Sanofi, Amgen, Pfizer, Astellas, and Novartis and has served on advisory boards for Gilead and Cytodyn. The remaining authors declare no competing interests. Supplemental Information Document S1. Figures S1-S8 and Tables S2 and S4 Table S1. Excel file containing characteristic information for each PDX model which is too large to fit in PDF, related to Figure 1 . Table S4. Excel file containing RPPA data which is too large to fit in PDF, related to Figure S8. METHODS Collection of patient-derived materials All research conducted in human patients followed national guidelines including the Health Insurance Portability and Accountability Act (HIPAA) privacy and security rules 67 and the Common Rule ( http://www.hhs.gov/ohrp/humansubjects/commonrule/ ). Informed consent was obtained from all patients who provided samples for PDX model establishment. Patients were enrolled in the ARTEMIS trial ( NCT02276443 ), an MD Anderson Cancer Center IRB-approved protocol (2014-0185). Establishment of PDX collection All experimental procedures were approved by the Institutional Animal Care and Use Committee (IACUC) at MD Anderson Cancer Center under IACUC protocol 00000978-RN01. Ends for animal experiments were selected in accordance with IACUC approved criteria. Female NOD/SCID mice (NOD.CB17- Prkdc scid /NcrCrl) were obtained from Charles River, National Cancer Institute colony. PDX models were established according to published protocols 8 , 14 . Three to four weeks prior to tumor cell engraftment, the fourth mammary fat pads (MFPs) of 4- to 5-week-old NOD/SCID female mice were pre-humanized with green fluorescent protein (GFP)–labeled immortalized human mammary stromal fibroblasts (EG cells). FNA biopsies obtained in the clinic were immediately placed in DMEM/F12 media (HyClone Cat# SH30023.01) with 1x antibiotic/antimycotic (Corning cat # MT30004CI) and maintained on ice during transport to the laboratory. Enzymatic digestion of the samples used complete DMEM/F12 media supplemented with 3mg/mL collagenase (Roche cat# 1088793) and 250 U/mL of hyaluronidase (Sigma cat# H-3506) with incubation at 37°C in a rotator for 30 to 90 minutes. Following digestion, samples were washed with media and incubated with red blood cell lysis buffer (Sigma Cat# R7757). Cell pellets (passage 0, P0 cells) were washed with media and resuspended in DMEM/F12 supplemented in 5% bovine calf serum (BCS) and Matrigel (50/50; Corning) before injection into the right fourth MFPs of mice. Tumor cells were injected in a total volume of 30-50 µL. On average, 4,000-100,000 cells were implanted into two mice for the first PDX generation, depending on the tumor cell yield. Typically, cells were implanted into mice within two hours of the FNA biopsy reaching the laboratory. Engrafted mice were monitored for up to 6 months for tumor growth, at which point mice without palpable tumors were euthanized. When PDX tumors reached approximately 1000 mm 3 , they were collected and stored as snap-frozen tumor pieces, dissociated single cell suspensions, or formalin-fixed, paraffin-embedded tissue blocks. The dissociation of PDX tumors was the same as for FNA biopsy processing, with 3 to 5 h digestion times. P1 tumor cells were immediately implanted into the next generation of mice (n=2), which were not pre-humanized with stromal fibroblasts. Tumors were passaged through a total of three generations of mice, with 6 mice used for each subsequent generation. PDX models that reached P3 (the 3rd generation of mice) and were validated as described below were considered established. Quality control of PDX models Genomic DNA (gDNA) was extracted from PDX tumor cells using the DNeasy Blood & Tissue Kit (Qiagen). We used three approaches to validate the identity of each PDX model throughout passaging in mice 14 , 22 . First, we determined the human origin of each PDX tumor by quantifying the ratio of human-to-mouse gDNA in PDX tumors via qPCR with a human RNaseP gene probe (20x human RNaseP copy number assay, FAM-TAMRA, Life Technologies) and a mouse Trfc gene probe (20x mouse Trfc copy number assay, VIC-TAMRA, Life Technologies). The absolute calibrators were gDNA from TaqMan™ Control Human Genomic DNA (ThermoFisher) and from Mouse Genomic DNA (Promega). qPCR was conducted using the 2x TaqMan gene expression master mix (Life Technologies). The relative ratio of human and mouse gDNA in each tumor sample was calculated using the ΔΔCt method as described previously 68 . PDX tumors consistently exhibiting <10% human genomic DNA were discarded and not included in any subsequent analyses. Second, to confirm clearance of the GFP-labeled EG cells used to pre-humanize the first generation of mice, PCR was performed against GFP and human GAPDH. Amplicons were visualized on agarose gel after electrophoresis. Primer pairs for GAPDH were: Fwd 5′-AAGTTCATCTGCACCACCG; Rev 5′-TCCTTGAAGAAGATGGTGCG. Primer pairs for GFP were: Fwd 5′-ACATCATCCCTGCCTCTAC; Rev 5′-TCAAAGGTGGAGGAGTGG. All PDX models herein were confirmed to be GFP negative by this test. Lastly, to establish a DNA ‘fingerprint’ for each PDX model for future validation purposes, short-tandem repeat (STR) DNA fingerprinting was conducted using 50 ng of gDNA through the MDACC Characterized Cell Line Core. The STR profiles from Promega 16 High Sensitivity STR Kit (Catalog# DC2100) analyses were compared to commercial databases (DSMZ/ATCC/JCRB/RIKEN) of approximately 2500 known profiles, as well as the MDACC Characterized Cell Line Core database of approximately 2000 known DNA fingerprint profiles. Each PDX tumor had a unique STR profile, confirming that none were contaminated with human cell lines and that cross-contamination among the models did not occur. All tumors within each longitudinal set shared the same STR profile, indicating they were derived from the same patient. The matching profiles of subsequent PDX passages demonstrated consistent identities. Histology Patient and PDX tumor biopsies stored as formalin-fixed paraffin embedded tissue blocks were cut into 5 μm sections and analyzed by hematoxylin and eosin (H & E) staining. Histologic properties were evaluated by a breast pathologist (Dr. L. Huo). Immunohistochemical staining was performed using standard methods. After deparaffinization of the sections with xylene and rehydration via graded aqueous-ethanol baths, antigen retrieval was performed using Reveal decloaker solution (Biocare Medical, RV1000M) and an EX-Retriever v.3.0 microwave. Endogenous peroxidases were quenched by treatment with Dako Dual Enzyme block (S2003). The human-specific mitochondria antibody (Abcam ab92824) was diluted 1:800 and the pan-cytokeratin antibody was diluted 1:100 (Abcam ab86734). The M.O.M kit (Vector Laboratories MP-2400) was used to develop staining for both antibodies. CBR-luciferase labeling of PIM056 To generate a sub-line of PIM056 expressing fluorescent and bioluminescent markers, we followed previously established protocols 22 , 69 . Briefly, a PIM056 tumor was harvested, dissociated into single cells, and depleted of mouse cells using magnetic-activated cell sorting (Miltenyi, Cat. # 130-104-694). Two million human tumor cells were cultured under mammosphere conditions using 5 mL of complete Mammocult medium (StemCell Technologies) in ultra-low attachment plates (Corning) in the presence of recombinant lentivirus (approximate MOI = 8) encoding Click beetle red luciferase (CBRLuc) and mCherry, followed by adding 10µg/mL polybrene (Sigma) twelve hours later. Twenty-four hours after transduction, the media was refreshed. Four days after lentiviral transduction, cells were pelleted, organoids were dissociated with TrypLE Express dissociation reagent (Gibco), washed, and resuspended in PBS containing 0.5% BSA. Cells were stained with Sytox Blue viability dye (Life Technologies), then sorted on a BD Aria Fusion fluorescence activated cell sorter (FACS) to collect viable, mCherry-positive cells. These cells were washed with media and engrafted into the MFP of one NOD/SCID mouse as described above. Once the tumor reached 1000 mm 3 , it was harvested, dissociated into single cells and immediately transplanted into a second generation of NOD/SCID mice. This process was repeated on second passage CBRLuc-labeled tumor cells. Aliquots of cells were stored in cryopreservation media and frozen after each passage. Bioluminescence imaging (BLI) BLI was performed as previously described 21 . Animals received an intraperitoneal injection of D-luciferin (150 µg/g body weight, GoldBio) in phosphate-buffered saline. Ten minutes after injection, animals were anesthetized via isoflurane and imaged with a charge-coupled device camera-based BLI system (IVIS Spectrum, PerkinElmer). Signals were displayed as photons/s/cm 2 /sr for image representation. To detect distal organ metastases, tissues were assessed with BLI ex vivo at necropsy. Lungs, liver, brain, and bone were subjected to BLI individually. Sequencing patient samples and early-passage PDX tumors Initially, samples were sequenced by WES at the Cancer Genetics Laboratory at MD Anderson Cancer Center. Genomic libraries from approximately 500 ng gDNA were prepared using the standard KAPA paired-end sample preparation kit (KAPA Biosystems) according to the manufacturer’s instructions. SureSelect Human All Exon Kit, version 4 (Agilent Technologies) was used to enrich sequencing libraries for exomes. Samples were pooled 2 per lane and paired end 2 × 75 bp sequencing was performed using the Illumina HiSeq 2000. Later in the study, sequencing was performed by the Sequencing and Microarray Facility at MD Anderson Cancer Center. They followed a similar protocol as described above and also used the Agilent SureSelect Human All Exon Kit, version 4 to enrich for exomes. In contrast, they used the Agilent SureSelectXT Reagent Kit (Agilent Technologies) for library preparation and sequenced on an Illumina HiSeq4000 sequencer. Whole-exome sequencing data analysis We processed the NGS data using the BETSY system 70 . Reads from stromal mouse sequences were identified and discarded using Xenome 71 . We aligned the remaining non-mouse reads to the human reference (hg19) following the GATK best practices 72 , 73 . Bam files were sorted and indexed using samtools 74 . Duplicates were identified using Picard tools 73 and indels were realigned using GATK. Somatic mutation calls For somatic variant callers, we used the patient’s blood sample as the germline reference when available. For unavailable germline samples, we generated a common normal using 213 blood germline samples from the ARTEMIS trial that were each downsampled to 1% of reads and merged. We set minPruning = 3 in MuTect2 to accelerate computation. We also applied a custom filter designed to identify alignment artifacts from repeat regions 22 , 75 . To identify somatic variants, we employed a consensus approach 76 , 77 consisting of three algorithms to call SNPs and short and long indels (MuSE 78 , MuTect2, and Varscan2 79 ). Variants identified by two out of the three algorithms were considered true variant calls. We annotated the variants using Annovar 80 and SnpEff 81 . We further applied a quality filter of a minimal alternative allele of 5. All mutations that passed this filter were used in downstream analysis. Copy number assignments Analysis of copy number alteration for paired tumor-normal whole exome sequencing samples was performed using FACETS 82 . To detect major changes, the critical value for segmentation was set to 100, and minimum number of heterozygous SNPs in a segment used for bivariate t-statistic during clustering was set to 30. To reduce hyper-segmentation, scanning window size was set to 250 bp. Other parameters remain the same as default values. From the segmentation files produced by FACETS, we extracted the estimated total copy number (under column tcn.em) and plotted the values across the human genome as a heatmap. Plots of longitudinal PDX sets derived from the same patient were grouped together. PyClone To select variants for subclone analysis, we discarded those with low (1000) read depth coverage, low alternative allele reads (<5), or frequency (6), and those that occurred in regions where the copy number varied across samples. Using the read count, copy number, and purity prediction for the selected variants, we used PyClone 39 to estimate the cancer cell frequencies (CCFs). We used the runanalysispipeline script in Pyclone (v0.13.1) under the default settings: basemeasureparams (alpha = 1; beta = 1), betabinomialprecisionparams (prior rate = 0.001; shape = 1; proposal precision = 0.01; value = 1000), concentration (prior rate = 0.001; shape = 1; value = 1), density = pyclone_ betabinomial, numiters = 10,000, and errorrates = 0.001, except for prior for method to set the possible genotypes = totalcopynumber, for all samples. Output statistics and plots were generated after 1,000 iterations as burn-in. PyClone analysis identified multiple clusters comprising one or more variants. We noticed that some clusters were unstable and were comprised of variants whose cluster assignments were unstable throughout the MCMC sampling. To quantify the unstable clusters, we calculated a pairwise matrix containing the probabilities that each pair of variants were assigned to the same cluster across MCMC samples. We then calculated a matrix containing the difference of probabilities of a variant with other variants within its cluster and outside its cluster. We accepted a cluster as robust if the number of variants in the cluster is greater than 5 and the median difference of probabilities is greater than 80%. Mutation consistency rate between PDX and patient Consistency rate between corresponding PDX and patient samples was defined as the number of common mutations in both a PDX tumor and corresponding patient sample divided by the number of all mutations in the patient sample. The range of consistency rate is 0 to 1. Frequently mutated genes Mutation frequency of a variant was calculated as patient-based, not sample-based. If a variant occurred in any sample belonging to the same patient, it was considered to occur once in that patient. Silent variants, as well as those falling into intergenic and intronic regions, were removed from this analysis. The ClinVar database 83 was downloaded and processed to obtain a collection of pathogenic mutations. Variants in our dataset that mapped to this collection were considered pathogenic. Tumor mutation burden analysis Mutation burden for each sample was calculated as the number of nonsynonymous mutations divided by the sequenced length. Here, nonsynonymous mutations include all those mutations that lead to a change in protein sequence, e.g., missense, startloss, stopgain, stoploss. Mutation signature analysis Mutation signature analysis was carried out using MuSiCal v1.0 84 by refitting the trinucleotide profile in our dataset to the program’s Breast.AdenoCA signature collection. The resulting exposure scores of each signature were scaled to sum up to 100% for each sample. Signatures with less than 15% exposure in all PDX models were considered as “others” when plotting. RNA sequencing data analysis We processed the data using pipelines implemented in the BETSY system 70 . We aligned the fastq reads to human genome hg19 using STAR 85 . Gene-level read counts were calculated with HTSeq-Count 86 and transcripts per million (TPM) were summarized using RSEM 87 . Batch effect correction of the original TPM was carried out using ComBat 88 . The batch-corrected TPM values were used in downstream analyses, unless stated otherwise. TNBC subtyping Vanderbilt TNBC subtypes 34 , 89 were assigned using the TNBCtype public server: https://cbc.app.vumc.org/tnbc/ . The assignments were carried out by sequencing batch. Original TPM values from samples within the same sequencing batch were submitted to the server for simultaneous assignment, minimizing potential issues related to normalization across different batches. Two of the subtypes assigned by the server, IM and MSL, are heavily impacted by stromal infiltration 33 and were excluded in our PDX tumor analysis. To convert these two subtypes, as well as the unstable (UNS) group, to other valid subtypes (BL1, BL2, M, LAR), we applied the following algorithm, only utilizing p-values and correlation coefficients of the four valid subtypes (BL1, BL2, M, LAR). We assigned the subtype with the largest correlation coefficient among subtypes with p-value ≤ 0.05 and correlation coefficient ≥ 0.1. The difference between this and the next largest coefficient had to be ≥ 0.05. If no subtype met this criteria, “UNS” was assigned 89 . The significance test of subtype co-occurrence was carried out using the statistical model described by Zhou et al 90 . A one-sided p-value was calculated using an implementation of the method in R. Gene expression comparison between corresponding patient and PDX samples Comparisons of the gene expression profiles (nsample=84) were carried out using Spearman correlation of the 1000 most variable genes from PDX models and the corresponding genes from patient tumor samples. Unsupervised consensus clustering We performed unsupervised consensus clustering using Pearson correlation distance and a hierarchical clustering algorithm by using the R package ConsensusClusterPlus 91 . The heatmap was plotted using R package ComplexHeatmap 92 . Single-sample pathway scores We calculated single-sample pathway scores using the ssGSEA algorithm implemented in R package GSVA 93 . Pairwise pathway analyses between timepoints We performed pairwise pathway analyses on the longitudinal sets of PDX models using the pre-ranked GSEA algorithm implemented in R package fgsea 94 with gene expression fold change for ranking. Calculating conservation scores Each patient sample-PDX model pair was scored for degree of conservation (0, 0.5, or 1) in five categories: mutations, mutation burden, subclonal architecture, gene expression, and transcriptional subtype. The total conservation score was the sum across all categories for each pair of samples. For mutations, the median conservation proportion of mutations across the dataset was 0.554. All pairs with a higher conservation proportion scored 1, any pairs between 0.300 and 0.554 scored 0.5, and any pairs below 0.300 scored 0. For mutation burden, we previously classified samples into high (>15 mutations/Mb), medium (5-15 mutations/Mb), and low categories (<5 mutations/Mb). Any pairs that matched scored 1 and pairs that did not match scored 0. For subclonal architecture, pairs with the same number of subclones, with proportions within 25% of each other, scored 1, pairs with the subclones varying >25% in proportion scored 0.5, and pairs with different numbers of subclones scored 0. For gene expression, the median Spearman correlation coefficient for matched patient-PDX samples was 0.588, while the median coefficient between any patient-PDX pair was 0.351. Paired samples with a correlation coefficient greater than the median 0.588 scored 1, pairs between 0.351 and 0.588 scored 0.5, and pairs below 0.351 scored 0. For transcriptional subtype, we previously classified samples using the 4-class TNBCtype, allowing for both a primary subtype assignment and multiple significant subtype correlations. Pairs with the same primary subtype assignment scored 1, pairs with multiple significant subtypes that shared at least one in common scored 0.5, and pairs with no subtypes in common scored 0. Preclinical studies in mice NOD/SCID mice were implanted with 0.5-1.0 x 10 6 cells in the fourth MFP. When tumors reached ∼100-150 mm 3 , mice were randomized into vehicle or chemotherapy groups for dosing. Mice received a single dose of both doxorubicin (0.5 mg/kg) and cyclophosphamide (50 mg/kg) or sterile water (vehicle) via intraperitoneal injection. Tumor volumes were monitored via caliper measurements twice weekly until allowable tumor burden was reached. Reverse-phase protein array (RPPA) Tumor fragments were collected from each PDX model and flash-frozen in liquid nitrogen. Frozen tumor pieces were homogenized using a Next Advance Bullet Blender, followed by sonication in lysis buffer containing 1% Triton X-100, 50mM HEPES pH 7.4, 150mM NaCl, 1.5mM MgCl 2 ,1mM EGTA, 100mM NaF, 10mM Na 4 P 2 O 7 , 1mM Na 3 VO 4 , 10% glycerol, and protease and phosphatase inhibitors. Samples were centrifuged at 13,000rpm for 10 minutes and supernatants were collected. We quantified the protein concentration by Bradford assay, then adjusted each sample to a final concentration of 1.0-1.5 µg/mL with lysis buffer. Protein lysate was mixed with 4X SDS+β-mercaptoethanol buffer and boiled for five minutes, then stored at −80°C. RPPA was then performed as previously described 95 . We used tumor pieces from three individual mice per model whenever possible, otherwise multiple tumor pieces from the same mouse were used as replicates when necessary. Relative protein levels for each sample were determined and normalized for protein loading by algorithm SuperCurve 96 . Log2-transformed normalized values were used in downstream analysis. Samples from longitudinal PDX sets derived from the same patient were measured in the same batch. Outlier samples determined by SuperCurve loading correction factors and principal component analysis were excluded. Unsupervised clustering of RPPA expression was carried out using the top 100 most variable (by MAD) antibody markers out of 331 commonly available markers for all samples, and a heatmap of the median-centered expression values was plotted. Expression values of proteins (CDH2 and VIM) that were missing values (because they were not measured in every batch) were not included in the clustering and added to the heatmap after clustering. Correlation analysis between protein expression and Vanderbilt TNBC subtype coefficients were also performed on these 331 common markers. The top 5 most-significant markers that were positively and negatively correlated for each of the 4 subtypes (BL1, BL2, LAR, M) were selected and shown in a dot plot. Acknowledgements We are grateful to the patients who provided tumors for PDX model establishment. We thank Dr. Sendurai Mani, Dr. Abena Redwood, Dr. Yan Jiang, Aaron McCoy, and Yizheng Tu for contributions in the early stages of PDX establishment. We thank Drs. Naoto Ueno, Jennifer Litton, Bora Lim, and Senthil Damodaran for their leadership as PIs of the post-AC Phase II clinical trial arms for ARTEMIS. Drs. Rosalind Candelaria, Lumarie Santiago, Beatriz E. Adrada, Deanna L. Lane, and Wei T. Yang facilitated tumor acquisition for PDX establishment. PDX models were established through a generous gift from the Cazalot family and from funds from the MD Anderson Cancer Center Breast Cancer Moon Shot Program. Additional funding sources that supported this work include the Cancer Prevention and Research Institute of Texas RP150148 (to H.P.-W.), RP160710 (to W.F.S., H.P-W., and J.T.C.). This study was funded in part by generous philanthropic contributions to The University of Texas MD Anderson Cancer Center Moon Shots Program, a Conquer Cancer Career Development Award supported by Fleur Fairman (to C.Y.), philanthropic support from the Still Water Foundation (to C.Y. and S.L.M.), and the Winterhof Foundation (S.L.M.). Dr. Yam was additionally supported by the 2018 Gianni Bonadonna Breast Cancer Research Fellowship (Conquer Cancer Foundation), the Allison and Brian Grove Endowed Fellowship for Breast Medical Oncology, and the Susan Papizan Dolan Fellowship in Breast Oncology. The MD Anderson Research Histology Core Laboratory, the Advanced Technology Genomics Core, and the Characterized Cell Line Core are supported in part by NCI grant P30CA016672. The Functional Proteomics Reverse Phase Protein Array Core was supported in part by The University of Texas MD Anderson Cancer Center, P30CA016672, and R50CA221675. Footnotes ↵ 13 Lead contact REFERENCES 1. ↵ Liedtke , C. , Mazouni , C. , Hess , K.R. , Andre , F. , Tordai , A. , Mejia , J.A. , Symmans , W.F. , Gonzalez-Angulo , A.M. , Hennessy , B. , Green , M. , et al. ( 2008 ). Response to neoadjuvant therapy and long-term survival in patients with triple-negative breast cancer . J Clin Oncol 26 , 1275 – 1281 . doi: 10.1200/JCO.2007.14.4147 . OpenUrl Abstract / FREE Full Text 2. ↵ Schmid , P. , Cortes , J. , Dent , R. , Pusztai , L. , McArthur , H. , Kummel , S. , Bergh , J. , Denkert , C. , Park , Y.H. , Hui , R. , et al. ( 2022 ). Event-free Survival with Pembrolizumab in Early Triple-Negative Breast Cancer . N Engl J Med 386 , 556 – 567 . doi: 10.1056/NEJMoa2112651 . OpenUrl CrossRef PubMed 3. ↵ Yau , C. , Osdoit , M. , van der Noordaa , M. , Shad , S. , Wei , J. , de Croze , D. , Hamy , A.S. , Lae , M. , Reyal , F. , Sonke , G.S. , et al. ( 2022 ). Residual cancer burden after neoadjuvant chemotherapy and long-term survival outcomes in breast cancer: a multicentre pooled analysis of 5161 patients . Lancet Oncol 23 , 149 – 160 . doi: 10.1016/S1470-2045(21)00589-1 . OpenUrl CrossRef PubMed 4. ↵ Hatzis , C. , Pusztai , L. , Valero , V. , Booser , D.J. , Esserman , L. , Lluch , A. , Vidaurre , T. , Holmes , F. , Souchon , E. , Wang , H. , et al. ( 2011 ). A genomic predictor of response and survival following taxane-anthracycline chemotherapy for invasive breast cancer . JAMA 305 , 1873 – 1881 . doi: 10.1001/jama.2011.593 . OpenUrl CrossRef PubMed Web of Science 5. ↵ Schmid , P. , Cortes , J. , Pusztai , L. , McArthur , H. , Kummel , S. , Bergh , J. , Denkert , C. , Park , Y.H. , Hui , R. , Harbeck , N. , et al. ( 2020 ). Pembrolizumab for Early Triple-Negative Breast Cancer . N Engl J Med 382 , 810 – 821 . doi: 10.1056/NEJMoa1910549 . OpenUrl CrossRef PubMed 6. ↵ Schmid , P. , Cortes , J. , Dent , R. , McArthur , H. , Pusztai , L. , Kummel , S. , Denkert , C. , Park , Y.H. , Hui , R. , Harbeck , N. , et al. ( 2024 ). Overall Survival with Pembrolizumab in Early-Stage Triple-Negative Breast Cancer . N Engl J Med . doi: 10.1056/NEJMoa2409932 . OpenUrl CrossRef 7. ↵ Yam , C. , Mittendorf , E.A. , Garber , H.R. , Sun , R. , Damodaran , S. , Murthy , R.K. , Ramirez , D. , Karuturi , M. , Layman , R.M. , Ibrahim , N. , et al. ( 2023 ). A phase II study of neoadjuvant atezolizumab and nab-paclitaxel in patients with anthracycline-resistant early-stage triple-negative breast cancer . Breast Cancer Res Treat 199 , 457 – 469 . doi: 10.1007/s10549-023-06929-9 . OpenUrl CrossRef PubMed 8. ↵ Dobrolecki , L.E. , Airhart , S.D. , Alferez , D.G. , Aparicio , S. , Behbod , F. , Bentires-Alj , M. , Brisken , C. , Bult , C.J. , Cai , S. , Clarke , R.B. , et al. ( 2016 ). Patient-derived xenograft (PDX) models in basic and translational breast cancer research . Cancer Metastasis Rev 35 , 547 – 573 . doi: 10.1007/s10555-016-9653-x . OpenUrl CrossRef PubMed 9. ↵ Zhang , X. , Claerhout , S. , Prat , A. , Dobrolecki , L.E. , Petrovic , I. , Lai , Q. , Landis , M.D. , Wiechmann , L. , Schiff , R. , Giuliano , M. , et al. ( 2013 ). A renewable tissue resource of phenotypically stable, biologically and ethnically diverse, patient-derived human breast cancer xenograft models . Cancer Res 73 , 4885 – 4897 . doi: 10.1158/0008-5472.CAN-12-4081 . OpenUrl Abstract / FREE Full Text 10. ↵ Bruna , A. , Rueda , O.M. , Greenwood , W. , Batra , A.S. , Callari , M. , Batra , R.N. , Pogrebniak , K. , Sandoval , J. , Cassidy , J.W. , Tufegdzic-Vidakovic , A. , et al. ( 2016 ). A Biobank of Breast Cancer Explants with Preserved Intra-tumor Heterogeneity to Screen Anticancer Compounds . Cell 167 , 260 – 274 e222. doi: 10.1016/j.cell.2016.08.041 . OpenUrl CrossRef PubMed 11. ↵ DeRose , Y.S. , Wang , G. , Lin , Y.C. , Bernard , P.S. , Buys , S.S. , Ebbert , M.T. , Factor , R. , Matsen , C. , Milash , B.A. , Nelson , E. , et al. ( 2011 ). Tumor grafts derived from women with breast cancer authentically reflect tumor pathology, growth, metastasis and disease outcomes . Nat Med 17 , 1514 – 1520 . doi: 10.1038/nm.2454 . OpenUrl CrossRef PubMed 12. ↵ Guillen , K.P. , Fujita , M. , Butterfield , A.J. , Scherer , S.D. , Bailey , M.H. , Chu , Z. , DeRose , Y.S. , Zhao , L. , Cortes-Sanchez , E. , Yang , C.H. , et al. ( 2022 ). A human breast cancer-derived xenograft and organoid platform for drug discovery and precision oncology . Nat Cancer 3 , 232 – 250 . doi: 10.1038/s43018-022-00337-6 . OpenUrl CrossRef PubMed 13. ↵ Lewis , M.T. , and Caldas , C . ( 2024 ). The Power and Promise of Patient-Derived Xenografts of Human Breast Cancer . Cold Spring Harb Perspect Med 14 . doi: 10.1101/cshperspect.a041329 . OpenUrl Abstract / FREE Full Text 14. ↵ Echeverria , G.V. , Cai , S. , Tu , Y. , Shao , J. , Powell , E. , Redwood , A.B. , Jiang , Y. , McCoy , A. , Rinkenbaugh , A.L. , Lau , R. , et al. ( 2023 ). Predictors of success in establishing orthotopic patient-derived xenograft models of triple negative breast cancer . NPJ Breast Cancer 9 , 2 . doi: 10.1038/s41523-022-00502-1 . OpenUrl CrossRef 15. ↵ Kuperwasser , C. , Chavarria , T. , Wu , M. , Magrane , G. , Gray , J.W. , Carey , L. , Richardson , A. , and Weinberg , R.A . ( 2004 ). Reconstruction of functionally normal and malignant human breast tissues in mice . Proc Natl Acad Sci U S A 101 , 4966 – 4971 . doi: 10.1073/pnas.0401064101 . OpenUrl Abstract / FREE Full Text 16. ↵ Abuhadra , N. , Sun , R. , Bassett , R.L. , Jr. , Huo , L. , Chang , J.T. , Teshome , M. , Clayborn , A.R. , White , J.B. , Ravenberg , E.E. , Adrada , B.E. , et al. ( 2023 ). Targeting chemotherapy resistance in mesenchymal triple-negative breast cancer: a phase II trial of neoadjuvant angiogenic and mTOR inhibition with chemotherapy . Invest New Drugs 41 , 391 – 401 . doi: 10.1007/s10637-023-01357-4 . OpenUrl CrossRef PubMed 17. ↵ Lim , B. , Seth , S. , Yam , C. , Huo , L. , Fujii , T. , Lee , J. , Bassett , R. , Nasser , S. , Ravenberg , L. , White , J. , et al. ( 2024 ). Phase 2 study of neoadjuvant enzalutamide and paclitaxel for luminal androgen receptor-enriched TNBC: Trial results and insights into “ARness” . Cell Rep Med , 101595 . doi: 10.1016/j.xcrm.2024.101595 . OpenUrl CrossRef 18. ↵ Yam , C. , Patel , M. , Hill , H.A. , Sun , R. , Bassett , R.L. , Kong , E. , Damodaran , S. , Koenig , K.B. , Abouharb , S. , Saleem , S. , et al. ( 2024 ). Targeting the EGFR pathway in chemotherapy-resistant triple-negative breast cancer - a phase II study . Cancer Res Commun . doi: 10.1158/2767-9764.CRC-24-0255 . OpenUrl CrossRef 19. ↵ Yam , C. , Abuhadra , N. , Sun , R. , Adrada , B.E. , Ding , Q.Q. , White , J.B. , Ravenberg , E.E. , Clayborn , A.R. , Valero , V. , Tripathy , D. , et al. ( 2022 ). Molecular Characterization and Prospective Evaluation of Pathologic Response and Outcomes with Neoadjuvant Therapy in Metaplastic Triple-Negative Breast Cancer . Clin Cancer Res 28 , 2878 – 2889 . doi: 10.1158/1078-0432.CCR-21-3100 . OpenUrl CrossRef PubMed 20. ↵ Vaklavas , C. , Matsen , C.B. , Chu , Z. , Boucher , K.M. , Scherer , S.D. , Pathi , S. , Beck , A. , Brownson , K.E. , Buys , S.S. , Chittoria , N. , et al. ( 2024 ). TOWARDS Study: Patient-Derived Xenograft Engraftment Predicts Poor Survival in Patients With Newly Diagnosed Triple-Negative Breast Cancer . JCO Precis Oncol 8 , e2300724 . doi: 10.1200/PO.23.00724 . OpenUrl CrossRef 21. ↵ Powell , E. , Shao , J. , Yuan , Y. , Chen , H.C. , Cai , S. , Echeverria , G.V. , Mistry , N. , Decker , K.F. , Schlosberg , C. , Do , K.A. , et al. ( 2016 ). p53 deficiency linked to B cell translocation gene 2 (BTG2) loss enhances metastatic potential by promoting tumor growth in primary and metastatic sites in patient-derived xenograft (PDX) models of triple-negative breast cancer . Breast Cancer Res 18 , 13 . doi: 10.1186/s13058-016-0673-9 . OpenUrl CrossRef PubMed 22. ↵ Echeverria , G.V. , Powell , E. , Seth , S. , Ge , Z. , Carugo , A. , Bristow , C. , Peoples , M. , Robinson , F. , Qiu , H. , Shao , J. , et al. ( 2018 ). High-resolution clonal mapping of multi-organ metastasis in triple negative breast cancer . Nat Commun 9 , 5079 . doi: 10.1038/s41467-018-07406-4 . OpenUrl CrossRef PubMed 23. ↵ Cancer Genome Atlas, N. ( 2012 ). Comprehensive molecular portraits of human breast tumours . Nature 490 , 61 – 70 . doi: 10.1038/nature11412 . OpenUrl CrossRef PubMed Web of Science 24. Langerod , A. , Zhao , H. , Borgan , O. , Nesland , J.M. , Bukholm , I.R. , Ikdahl , T. , Karesen , R. , Borresen-Dale , A.L. , and Jeffrey , S.S . ( 2007 ). TP53 mutation status and gene expression profiles are powerful prognostic markers of breast cancer . Breast Cancer Res 9 , R30 . doi: 10.1186/bcr1675 . OpenUrl CrossRef PubMed 25. ↵ Silwal-Pandit , L. , Vollan , H.K. , Chin , S.F. , Rueda , O.M. , McKinney , S. , Osako , T. , Quigley , D.A. , Kristensen , V.N. , Aparicio , S. , Borresen-Dale , A.L. , et al. ( 2014 ). TP53 mutation spectrum in breast cancer is subtype specific and has distinct prognostic relevance . Clin Cancer Res 20 , 3569 – 3580 . doi: 10.1158/1078-0432.CCR-13-2943 . OpenUrl Abstract / FREE Full Text 26. ↵ Curtis , C. , Shah , S.P. , Chin , S.F. , Turashvili , G. , Rueda , O.M. , Dunning , M.J. , Speed , D. , Lynch , A.G. , Samarajiwa , S. , Yuan , Y. , et al. ( 2012 ). The genomic and transcriptomic architecture of 2,000 breast tumours reveals novel subgroups . Nature 486 , 346 – 352 . doi: 10.1038/nature10983 . OpenUrl CrossRef PubMed Web of Science 27. ↵ Pereira , B. , Chin , S.F. , Rueda , O.M. , Vollan , H.K. , Provenzano , E. , Bardwell , H.A. , Pugh , M. , Jones , L. , Russell , R. , Sammut , S.J. , et al. ( 2016 ). The somatic mutation profiles of 2,433 breast cancers refines their genomic and transcriptomic landscapes . Nat Commun 7 , 11479 . doi: 10.1038/ncomms11479 . OpenUrl CrossRef PubMed 28. ↵ Litton , J.K. , Beck , J.T. , Jones , J.M. , Andersen , J. , Blum , J.L. , Mina , L.A. , Brig , R. , Danso , M. , Yuan , Y. , Abbattista , A. , et al. ( 2023 ). Neoadjuvant Talazoparib in Patients With Germline BRCA1/2 Mutation-Positive, Early-Stage Triple-Negative Breast Cancer: Results of a Phase II Study . Oncologist 28 , 845 – 855 . doi: 10.1093/oncolo/oyad139 . OpenUrl CrossRef PubMed 29. ↵ Litton , J.K. , Scoggins , M.E. , Hess , K.R. , Adrada , B.E. , Murthy , R.K. , Damodaran , S. , DeSnyder , S.M. , Brewster , A.M. , Barcenas , C.H. , Valero , V. , et al. ( 2020 ). Neoadjuvant Talazoparib for Patients With Operable Breast Cancer With a Germline BRCA Pathogenic Variant . J Clin Oncol 38 , 388 – 394 . doi: 10.1200/JCO.19.01304 . OpenUrl CrossRef PubMed 30. ↵ Castilla , L.H. , Couch , F.J. , Erdos , M.R. , Hoskins , K.F. , Calzone , K. , Garber , J.E. , Boyd , J. , Lubin , M.B. , Deshano , M.L. , Brody , L.C. , and, et al. ( 1994 ). Mutations in the BRCA1 gene in families with early-onset breast and ovarian cancer . Nat Genet 8 , 387 – 391 . doi: 10.1038/ng1294-387 . OpenUrl CrossRef PubMed Web of Science 31. ↵ Alexandrov , L.B. , Nik-Zainal , S. , Wedge , D.C. , Aparicio , S.A. , Behjati , S. , Biankin , A.V. , Bignell , G.R. , Bolli , N. , Borg , A. , Borresen-Dale , A.L. , et al. ( 2013 ). Signatures of mutational processes in human cancer . Nature 500 , 415 – 421 . doi: 10.1038/nature12477 . OpenUrl CrossRef PubMed Web of Science 32. ↵ Watson , E.V. , Lee , J.J. , Gulhan , D.C. , Melloni , G.E.M. , Venev , S.V. , Magesh , R.Y. , Frederick , A. , Chiba , K. , Wooten , E.C. , Naxerova , K. , et al. ( 2024 ). Chromosome evolution screens recapitulate tissue-specific tumor aneuploidy patterns . Nat Genet 56 , 900 – 912 . doi: 10.1038/s41588-024-01665-2 . OpenUrl CrossRef 33. ↵ Lehmann , B.D. , Jovanovic , B. , Chen , X. , Estrada , M.V. , Johnson , K.N. , Shyr , Y. , Moses , H.L. , Sanders , M.E. , and Pietenpol , J.A . ( 2016 ). Refinement of Triple-Negative Breast Cancer Molecular Subtypes: Implications for Neoadjuvant Chemotherapy Selection . PLoS One 11 , e0157368 . doi: 10.1371/journal.pone.0157368 . OpenUrl CrossRef PubMed 34. ↵ Lehmann , B.D. , Bauer , J.A. , Chen , X. , Sanders , M.E. , Chakravarthy , A.B. , Shyr , Y. , and Pietenpol , J.A . ( 2011 ). Identification of human triple-negative breast cancer subtypes and preclinical models for selection of targeted therapies . J Clin Invest 121 , 2750 – 2767 . doi: 10.1172/JCI45014 . OpenUrl CrossRef PubMed Web of Science 35. ↵ Bonafe , M. , Ceccarelli , C. , Farabegoli , F. , Santini , D. , Taffurelli , M. , Barbi , C. , Marzi , E. , Trapassi , C. , Storci , G. , Olivieri , F. , and Franceschi , C . ( 2003 ). Retention of the p53 codon 72 arginine allele is associated with a reduction of disease-free and overall survival in arginine/proline heterozygous breast cancer patients . Clin Cancer Res 9 , 4860 – 4864 . OpenUrl Abstract / FREE Full Text 36. ↵ Basu , S. , Gnanapradeepan , K. , Barnoud , T. , Kung , C.P. , Tavecchio , M. , Scott , J. , Watters , A. , Chen , Q. , Kossenkov , A.V. , and Murphy , M.E . ( 2018 ). Mutant p53 controls tumor metabolism and metastasis by regulating PGC-1alpha . Genes Dev 32 , 230 – 243 . doi: 10.1101/gad.309062.117 . OpenUrl Abstract / FREE Full Text 37. ↵ Liberzon , A. , Birger , C. , Thorvaldsdottir , H. , Ghandi , M. , Mesirov , J.P. , and Tamayo , P . ( 2015 ). The Molecular Signatures Database (MSigDB) hallmark gene set collection . Cell Syst 1 , 417 – 425 . doi: 10.1016/j.cels.2015.12.004 . OpenUrl CrossRef PubMed 38. ↵ Echeverria , G.V. , Ge , Z. , Seth , S. , Zhang , X. , Jeter-Jones , S. , Zhou , X. , Cai , S. , Tu , Y. , McCoy , A. , Peoples , M. , et al. ( 2019 ). Resistance to neoadjuvant chemotherapy in triple-negative breast cancer mediated by a reversible drug-tolerant state . Sci Transl Med 11 . doi: 10.1126/scitranslmed.aav0936 . OpenUrl Abstract / FREE Full Text 39. ↵ Roth , A. , Khattra , J. , Yap , D. , Wan , A. , Laks , E. , Biele , J. , Ha , G. , Aparicio , S. , Bouchard-Cote , A. , and Shah , S.P . ( 2014 ). PyClone: statistical inference of clonal population structure in cancer . Nat Methods 11 , 396 – 398 . doi: 10.1038/nmeth.2883 . OpenUrl CrossRef PubMed Web of Science 40. ↵ Balko , J.M. , Giltnane , J.M. , Wang , K. , Schwarz , L.J. , Young , C.D. , Cook , R.S. , Owens , P. , Sanders , M.E. , Kuba , M.G. , Sanchez , V. , et al. ( 2014 ). Molecular profiling of the residual disease of triple-negative breast cancers after neoadjuvant chemotherapy identifies actionable therapeutic targets . Cancer Discov 4 , 232 – 245 . doi: 10.1158/2159-8290.CD-13-0286 . OpenUrl Abstract / FREE Full Text 41. ↵ Lehmann , B.D. , Colaprico , A. , Silva , T.C. , Chen , J. , An , H. , Ban , Y. , Huang , H. , Wang , L. , James , J.L. , Balko , J.M. , et al. ( 2021 ). Multi-omics analysis identifies therapeutic vulnerabilities in triple-negative breast cancer subtypes . Nat Commun 12 , 6276 . doi: 10.1038/s41467-021-26502-6 . OpenUrl CrossRef PubMed 42. ↵ Liberzon , A. , Subramanian , A. , Pinchback , R. , Thorvaldsdottir , H. , Tamayo , P. , and Mesirov , J.P . ( 2011 ). Molecular signatures database (MSigDB) 3.0 . Bioinformatics 27 , 1739 – 1740 . doi: 10.1093/bioinformatics/btr260 . OpenUrl CrossRef PubMed Web of Science 43. ↵ Li , S. , Shen , D. , Shao , J. , Crowder , R. , Liu , W. , Prat , A. , He , X. , Liu , S. , Hoog , J. , Lu , C. , et al. ( 2013 ). Endocrine-therapy-resistant ESR1 variants revealed by genomic characterization of breast-cancer-derived xenografts . Cell Rep 4 , 1116 – 1130 . doi: 10.1016/j.celrep.2013.08.022 . OpenUrl CrossRef PubMed Web of Science 44. Evans , K.W. , Yuca , E. , Akcakanat , A. , Scott , S.M. , Arango , N.P. , Zheng , X. , Chen , K. , Tapia , C. , Tarco , E. , Eterovic , A.K. , et al. ( 2017 ). A Population of Heterogeneous Breast Cancer Patient-Derived Xenografts Demonstrate Broad Activity of PARP Inhibitor in BRCA1/2 Wild-Type Tumors . Clin Cancer Res 23 , 6468 – 6477 . doi: 10.1158/1078-0432.CCR-17-0615 . OpenUrl Abstract / FREE Full Text 45. McAuliffe , P.F. , Evans , K.W. , Akcakanat , A. , Chen , K. , Zheng , X. , Zhao , H. , Eterovic , A.K. , Sangai , T. , Holder , A.M. , Sharma , C. , et al. ( 2015 ). Ability to Generate Patient-Derived Breast Cancer Xenografts Is Enhanced in Chemoresistant Disease and Predicts Poor Patient Outcomes . PLoS One 10 , e0136851 . doi: 10.1371/journal.pone.0136851 . OpenUrl CrossRef PubMed 46. ↵ Petrosyan , V. , Dobrolecki , L.E. , Thistlethwaite , L. , Lewis , A.N. , Sallas , C. , Srinivasan , R.R. , Lei , J.T. , Kovacevic , V. , Obradovic , P. , Ellis , M.J. , et al. ( 2023 ). Identifying biomarkers of differential chemotherapy response in TNBC patient-derived xenografts with a CTD/WGCNA approach . iScience 26 , 105799 . doi: 10.1016/j.isci.2022.105799 . OpenUrl CrossRef PubMed 47. ↵ Singh , S. , Lee , N. , Pedroza , D.A. , Bado , I.L. , Hamor , C. , Zhang , L. , Aguirre , S. , Hu , J. , Shen , Y. , Xu , Y. , et al. ( 2022 ). Chemotherapy Coupled to Macrophage Inhibition Induces T-cell and B-cell Infiltration and Durable Regression in Triple-Negative Breast Cancer . Cancer Res 82 , 2281 – 2297 . doi: 10.1158/0008-5472.CAN-21-3714 . OpenUrl CrossRef PubMed 48. ↵ Powell , R.T. , Redwood , A. , Liu , X. , Guo , L. , Cai , S. , Zhou , X. , Tu , Y. , Zhang , X. , Qi , Y. , Jiang , Y. , et al. ( 2020 ). Pharmacologic profiling of patient-derived xenograft models of primary treatment-naive triple-negative breast cancer . Sci Rep 10 , 17899 . doi: 10.1038/s41598-020-74882-4 . OpenUrl CrossRef PubMed 49. ↵ Altman , J.E. , Olex , A.L. , Zboril , E.K. , Walker , C.J. , Boyd , D.C. , Myrick , R.K. , Hairr , N.S. , Koblinski , J.E. , Puchalapalli , M. , Hu , B. , et al. ( 2024 ). Single-cell transcriptional atlas of human breast cancers and model systems . Clin Transl Med 14 , e70044 . doi: 10.1002/ctm2.70044 . OpenUrl CrossRef 50. ↵ Petrocca , F. , Altschuler , G. , Tan , S.M. , Mendillo , M.L. , Yan , H. , Jerry , D.J. , Kung , A.L. , Hide , W. , Ince , T.A. , and Lieberman , J . ( 2013 ). A genome-wide siRNA screen identifies proteasome addiction as a vulnerability of basal-like triple-negative breast cancer cells . Cancer Cell 24 , 182 – 196 . doi: 10.1016/j.ccr.2013.07.008 . OpenUrl CrossRef PubMed Web of Science 51. ↵ Yang , C.H. , Gonzalez-Angulo , A.M. , Reuben , J.M. , Booser , D.J. , Pusztai , L. , Krishnamurthy , S. , Esseltine , D. , Stec , J. , Broglio , K.R. , Islam , R. , et al. ( 2006 ). Bortezomib (VELCADE) in metastatic breast cancer: pharmacodynamics, biological effects, and prediction of clinical benefits . Ann Oncol 17 , 813 – 817 . doi: 10.1093/annonc/mdj131 . OpenUrl CrossRef PubMed Web of Science 52. ↵ Engel , R.H. , Brown , J.A. , Von Roenn , J.H. , O’Regan , R.M. , Bergan , R. , Badve , S. , Rademaker , A. , and Gradishar , W.J. ( 2007 ). A phase II study of single agent bortezomib in patients with metastatic breast cancer: a single institution experience . Cancer Invest 25 , 733 – 737 . doi: 10.1080/07357900701506573 . OpenUrl CrossRef PubMed Web of Science 53. ↵ Schmid , P. , Kuhnhardt , D. , Kiewe , P. , Lehenbauer-Dehm , S. , Schippinger , W. , Greil , R. , Lange , W. , Preiss , J. , Niederle , N. , Brossart , P. , et al. ( 2008 ). A phase I/II study of bortezomib and capecitabine in patients with metastatic breast cancer previously treated with taxanes and/or anthracyclines . Ann Oncol 19 , 871 – 876 . doi: 10.1093/annonc/mdm569 . OpenUrl CrossRef PubMed 54. ↵ Powell , R.T. , Rinkenbaugh , A.L. , Guo , L. , Cai , S. , Shao , J. , Zhou , X. , Zhang , X. , Jeter-Jones , S. , Fu , C. , Qi , Y. , et al. ( 2024 ). Targeting neddylation and sumoylation in chemoresistant triple negative breast cancer . NPJ Breast Cancer 10 , 37 . doi: 10.1038/s41523-024-00644-4 . OpenUrl CrossRef 55. ↵ Ohh , M. , Kim , W.Y. , Moslehi , J.J. , Chen , Y. , Chau , V. , Read , M.A. , and Kaelin , W.G. , Jr . . ( 2002 ). An intact NEDD8 pathway is required for Cullin-dependent ubiquitylation in mammalian cells . EMBO Rep 3 , 177 – 182 . doi: 10.1093/embo-reports/kvf028 . OpenUrl Abstract / FREE Full Text 56. ↵ Byers , H.A. , Brooks , A.N. , Vangala , J.R. , Grible , J.M. , Feygin , A. , Clevenger , C.V. , Harrell , J.C. , and Radhakrishnan , S.K . ( 2023 ). Evaluation of the NRF1-proteasome axis as a therapeutic target in breast cancer . Sci Rep 13 , 15843 . doi: 10.1038/s41598-023-43121-x . OpenUrl CrossRef PubMed 57. Jacob , S. , Turner , T.H. , Cai , J. , Floros , K.V. , Yu , A.K. , Coon , C.M. , Khatri , R. , Alzubi , M.A. , Jakubik , C.T. , Bouck , Y.M. , et al. ( 2022 ). Genomic screening reveals ubiquitin-like modifier activating enzyme 1 as a potent and druggable target in c-MYC-high triple negative breast cancer models . PNAS Nexus 1 , pgac232 . doi: 10.1093/pnasnexus/pgac232 . OpenUrl CrossRef 58. ↵ Kunder , R. , Velyunskiy , M. , Dunne , S.F. , Cho , B.K. , Kanojia , D. , Begg , L. , Orriols , A.M. , Fleming-Trujillo , E. , Vadlamani , P. , Vialichka , A. , et al. ( 2022 ). Synergistic PIM kinase and proteasome inhibition as a therapeutic strategy for MYC-overexpressing triple-negative breast cancer . Cell Chem Biol 29 , 358 – 372 e355. doi: 10.1016/j.chembiol.2021.08.011 . OpenUrl CrossRef 59. ↵ Ma , C.X. , Cai , S. , Li , S. , Ryan , C.E. , Guo , Z. , Schaiff , W.T. , Lin , L. , Hoog , J. , Goiffon , R.J. , Prat , A. , et al. ( 2012 ). Targeting Chk1 in p53-deficient triple-negative breast cancer is therapeutically beneficial in human-in-mouse tumor models . J Clin Invest 122 , 1541 – 1552 . doi: 10.1172/JCI58765 . OpenUrl CrossRef PubMed Web of Science 60. Zheng , H. , Shao , F. , Martin , S. , Xu , X. , and Deng , C.X . ( 2017 ). WEE1 inhibition targets cell cycle checkpoints for triple negative breast cancers to overcome cisplatin resistance . Sci Rep 7 , 43517 . doi: 10.1038/srep43517 . OpenUrl CrossRef PubMed 61. Chopra , N. , Tovey , H. , Pearson , A. , Cutts , R. , Toms , C. , Proszek , P. , Hubank , M. , Dowsett , M. , Dodson , A. , Daley , F. , et al. ( 2020 ). Homologous recombination DNA repair deficiency and PARP inhibition activity in primary triple negative breast cancer . Nat Commun 11 , 2662 . doi: 10.1038/s41467-020-16142-7 . OpenUrl CrossRef PubMed 62. Anand , K. , Patel , T. , Niravath , P. , Rodriguez , A. , Darcourt , J. , Belcheva , A. , Boone , T. , Ensor , J. , and Chang , J . ( 2021 ). Targeting mTOR and DNA repair pathways in residual triple negative breast cancer post neoadjuvant chemotherapy . Sci Rep 11 , 82 . doi: 10.1038/s41598-020-80081-y . OpenUrl CrossRef 63. Li , Y. , Dobrolecki , L.E. , Sallas , C. , Zhang , X. , Kerr , T.D. , Bisht , D. , Wang , Y. , Awasthi , S. , Kaundal , B. , Wu , S. , et al. ( 2023 ). PRMT blockade induces defective DNA replication stress response and synergizes with PARP inhibition . Cell Rep Med 4 , 101326 . doi: 10.1016/j.xcrm.2023.101326 . OpenUrl CrossRef 64. Risom , T. , Langer , E.M. , Chapman , M.P. , Rantala , J. , Fields , A.J. , Boniface , C. , Alvarez , M.J. , Kendsersky , N.D. , Pelz , C.R. , Johnson-Camacho , K. , et al. ( 2018 ). Differentiation-state plasticity is a targetable resistance mechanism in basal-like breast cancer . Nat Commun 9 , 3815 . doi: 10.1038/s41467-018-05729-w . OpenUrl CrossRef PubMed 65. Bagheri , M. , Mohamed , G.A. , Mohamed Saleem , M.A. , Ognjenovic , N.B. , Lu , H. , Kolling , F.W. , Wilkins , O.M. , Das , S. , LaCroix , I.S. , Nagaraj , S.H. , et al. ( 2024 ). Pharmacological induction of chromatin remodeling drives chemosensitization in triple-negative breast cancer . Cell Rep Med 5 , 101504 . doi: 10.1016/j.xcrm.2024.101504 . OpenUrl CrossRef 66. ↵ Zhang , K. , Wei , J. , Zhang , S. , Fei , L. , Guo , L. , Liu , X. , Ji , Y. , Chen , W. , Ciamponi , F.E. , Chen , W. , et al. ( 2024 ). A chemical screen identifies PRMT5 as a therapeutic vulnerability for paclitaxel-resistant triple-negative breast cancer . Cell Chem Biol . doi: 10.1016/j.chembiol.2024.08.003 . OpenUrl CrossRef 67. ↵ U. S. Department of Health and Human Services . Combined Text of All Rules . 68. ↵ Livak , K.J. , and Schmittgen , T.D . ( 2001 ). Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method . Methods 25 , 402 – 408 . doi: 10.1006/meth.2001.1262 . OpenUrl CrossRef PubMed Web of Science 69. ↵ Powell , E. , Shao , J. , Picon , H.M. , Bristow , C. , Ge , Z. , Peoples , M. , Robinson , F. , Jeter-Jones , S.L. , Schlosberg , C. , Grzeskowiak , C.L. , et al. ( 2018 ). A functional genomic screen in vivo identifies CEACAM5 as a clinically relevant driver of breast cancer metastasis . NPJ Breast Cancer 4 , 9 . doi: 10.1038/s41523-018-0062-x . OpenUrl CrossRef 70. ↵ Chen , X. , and Chang , J.T . ( 2017 ). Planning bioinformatics workflows using an expert system . Bioinformatics 33 , 1210 – 1215 . doi: 10.1093/bioinformatics/btw817 . OpenUrl CrossRef PubMed 71. ↵ Conway , T. , Wazny , J. , Bromage , A. , Tymms , M. , Sooraj , D. , Williams , E.D. , and Beresford-Smith , B . ( 2012 ). Xenome--a tool for classifying reads from xenograft samples . Bioinformatics 28 , i172 – 178 . doi: 10.1093/bioinformatics/bts236 . OpenUrl CrossRef PubMed Web of Science 72. ↵ McKenna , A. , Hanna , M. , Banks , E. , Sivachenko , A. , Cibulskis , K. , Kernytsky , A. , Garimella , K. , Altshuler , D. , Gabriel , S. , Daly , M. , and DePristo , M.A . ( 2010 ). The Genome Analysis Toolkit: a MapReduce framework for analyzing next-generation DNA sequencing data . Genome Res 20 , 1297 – 1303 . doi: 10.1101/gr.107524.110 . OpenUrl Abstract / FREE Full Text 73. ↵ DePristo , M.A. , Banks , E. , Poplin , R. , Garimella , K.V. , Maguire , J.R. , Hartl , C. , Philippakis , A.A. , del Angel , G. , Rivas , M.A. , Hanna , M. , et al. ( 2011 ). A framework for variation discovery and genotyping using next-generation DNA sequencing data . Nat Genet 43 , 491 – 498 . doi: 10.1038/ng.806 . OpenUrl CrossRef PubMed Web of Science 74. ↵ Li , H. , Handsaker , B. , Wysoker , A. , Fennell , T. , Ruan , J. , Homer , N. , Marth , G. , Abecasis , G. , Durbin , R. , and Genome Project Data Processing, S. ( 2009 ). The Sequence Alignment/Map format and SAMtools . Bioinformatics 25 , 2078 – 2079 . doi: 10.1093/bioinformatics/btp352 . OpenUrl CrossRef PubMed Web of Science 75. ↵ Jia , P. , Li , F. , Xia , J. , Chen , H. , Ji , H. , Pao , W. , and Zhao , Z . ( 2012 ). Consensus rules in variant detection from next-generation sequencing data . PLoS One 7 , e38470 . doi: 10.1371/journal.pone.0038470 . OpenUrl CrossRef PubMed 76. ↵ Callari , M. , Sammut , S.J. , De Mattos-Arruda , L. , Bruna , A. , Rueda , O.M. , Chin , S.F. , and Caldas , C. ( 2017 ). Intersect-then-combine approach: improving the performance of somatic variant calling in whole exome sequencing data using multiple aligners and callers . Genome Med 9 , 35 . doi: 10.1186/s13073-017-0425-1 . OpenUrl CrossRef PubMed 77. ↵ Goode , D.L. , Hunter , S.M. , Doyle , M.A. , Ma , T. , Rowley , S.M. , Choong , D. , Ryland , G.L. , and Campbell , I.G . ( 2013 ). A simple consensus approach improves somatic mutation prediction accuracy . Genome Med 5 , 90 . doi: 10.1186/gm494 . OpenUrl CrossRef PubMed 78. ↵ Fan , Y. , Xi , L. , Hughes , D.S. , Zhang , J. , Zhang , J. , Futreal , P.A. , Wheeler , D.A. , and Wang , W . ( 2016 ). MuSE: accounting for tumor heterogeneity using a sample-specific error model improves sensitivity and specificity in mutation calling from sequencing data . Genome Biol 17 , 178 . doi: 10.1186/s13059-016-1029-6 . OpenUrl CrossRef PubMed 79. ↵ Koboldt , D.C. , Zhang , Q. , Larson , D.E. , Shen , D. , McLellan , M.D. , Lin , L. , Miller , C.A. , Mardis , E.R. , Ding , L. , and Wilson , R.K . ( 2012 ). VarScan 2: somatic mutation and copy number alteration discovery in cancer by exome sequencing . Genome Res 22 , 568 – 576 . doi: 10.1101/gr.129684.111 . OpenUrl Abstract / FREE Full Text 80. ↵ Wang , K. , Li , M. , and Hakonarson , H . ( 2010 ). ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data . Nucleic Acids Res 38 , e164 . doi: 10.1093/nar/gkq603 . OpenUrl CrossRef PubMed 81. ↵ Cingolani , P. , Platts , A. , Wang le , L. , Coon , M. , Nguyen , T. , Wang , L. , Land , S.J. , Lu , X. , and Ruden , D.M . ( 2012 ). A program for annotating and predicting the effects of single nucleotide polymorphisms, SnpEff: SNPs in the genome of Drosophila melanogaster strain w1118; iso-2; iso-3 . Fly (Austin) 6 , 80 – 92 . doi: 10.4161/fly.19695 . OpenUrl CrossRef PubMed Web of Science 82. ↵ Shen , R. , and Seshan , V.E . ( 2016 ). FACETS: allele-specific copy number and clonal heterogeneity analysis tool for high-throughput DNA sequencing . Nucleic Acids Res 44 , e131 . doi: 10.1093/nar/gkw520 . OpenUrl CrossRef PubMed 83. ↵ Landrum , M.J. , Lee , J.M. , Riley , G.R. , Jang , W. , Rubinstein , W.S. , Church , D.M. , and Maglott , D.R . ( 2014 ). ClinVar: public archive of relationships among sequence variation and human phenotype . Nucleic Acids Res 42 , D980 – 985 . doi: 10.1093/nar/gkt1113 . OpenUrl CrossRef PubMed Web of Science 84. ↵ Jin , H. , Gulhan , D.C. , Geiger , B. , Ben-Isvy , D. , Geng , D. , Ljungstrom , V. , and Park , P.J . ( 2024 ). Accurate and sensitive mutational signature analysis with MuSiCal . Nat Genet 56 , 541 – 552 . doi: 10.1038/s41588-024-01659-0 . OpenUrl CrossRef PubMed 85. ↵ Dobin , A. , Davis , C.A. , Schlesinger , F. , Drenkow , J. , Zaleski , C. , Jha , S. , Batut , P. , Chaisson , M. , and Gingeras , T.R . ( 2013 ). STAR: ultrafast universal RNA-seq aligner . Bioinformatics 29 , 15 – 21 . doi: 10.1093/bioinformatics/bts635 . OpenUrl CrossRef PubMed Web of Science 86. ↵ Anders , S. , Pyl , P.T. , and Huber , W . ( 2015 ). HTSeq--a Python framework to work with high-throughput sequencing data . Bioinformatics 31 , 166 – 169 . doi: 10.1093/bioinformatics/btu638 . OpenUrl CrossRef PubMed Web of Science 87. ↵ Li , B. , and Dewey , C.N . ( 2011 ). RSEM: accurate transcript quantification from RNA-Seq data with or without a reference genome . BMC Bioinformatics 12 , 323 . doi: 10.1186/1471-2105-12-323 . OpenUrl CrossRef PubMed 88. ↵ Johnson , W.E. , Li , C. , and Rabinovic , A . ( 2007 ). Adjusting batch effects in microarray expression data using empirical Bayes methods . Biostatistics 8 , 118 – 127 . doi: 10.1093/biostatistics/kxj037 . OpenUrl CrossRef PubMed Web of Science 89. ↵ Chen , X. , Li , J. , Gray , W.H. , Lehmann , B.D. , Bauer , J.A. , Shyr , Y. , and Pietenpol , J.A . ( 2012 ). TNBCtype: A Subtyping Tool for Triple-Negative Breast Cancer . Cancer Inform 11 , 147 – 156 . doi: 10.4137/CIN.S9983 . OpenUrl CrossRef PubMed 90. ↵ Zhou , X. , Zhou , M. , Huang , D. , and Cui , L . ( 2022 ). A probabilistic model for co-occurrence analysis in bibliometrics . J Biomed Inform 128 , 104047 . doi: 10.1016/j.jbi.2022.104047 . OpenUrl CrossRef PubMed 91. ↵ Wilkerson , M.D. , and Hayes , D.N . ( 2010 ). ConsensusClusterPlus: a class discovery tool with confidence assessments and item tracking . Bioinformatics 26 , 1572 – 1573 . doi: 10.1093/bioinformatics/btq170 . OpenUrl CrossRef PubMed Web of Science 92. ↵ Gu , Z. , Eils , R. , and Schlesner , M . ( 2016 ). Complex heatmaps reveal patterns and correlations in multidimensional genomic data . Bioinformatics 32 , 2847 – 2849 . doi: 10.1093/bioinformatics/btw313 . OpenUrl CrossRef PubMed 93. ↵ Hanzelmann , S. , Castelo , R. , and Guinney , J . ( 2013 ). GSVA: gene set variation analysis for microarray and RNA-seq data . BMC Bioinformatics 14 , 7 . doi: 10.1186/1471-2105-14-7 . OpenUrl CrossRef PubMed 94. ↵ Korotkevich G. , S.V. , Budin N ., Shpak B. , Artyomov M.N. , and Sergushichev A . ( 2021 ). Fast gene set enrichment analysis . bioRxiv . doi: 10.1101/060012 . OpenUrl Abstract / FREE Full Text 95. ↵ Akbani , R. , Ng , P.K. , Werner , H.M. , Shahmoradgoli , M. , Zhang , F. , Ju , Z. , Liu , W. , Yang , J.Y. , Yoshihara , K. , Li , J. , et al. ( 2014 ). A pan-cancer proteomic perspective on The Cancer Genome Atlas . Nat Commun 5 , 3887 . doi: 10.1038/ncomms4887 . OpenUrl CrossRef PubMed 96. ↵ Ju , Z. , Liu , W. , Roebuck , P.L. , Siwak , D.R. , Zhang , N. , Lu , Y. , Davies , M.A. , Akbani , R. , Weinstein , J.N. , Mills , G.B. , and Coombes , K.R . ( 2015 ). Development of a robust classifier for quality control of reverse-phase protein arrays . Bioinformatics 31 , 912 – 918 . doi: 10.1093/bioinformatics/btu736 . OpenUrl CrossRef PubMed View the discussion thread. Back to top Previous Next Posted November 28, 2024. Download PDF Supplementary Material Email Thank you for your interest in spreading the word about bioRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following An annotated biobank of triple negative breast cancer patient-derived xenografts featuring treatment-naïve and longitudinal samples throughout neoadjuvant chemotherapy Message Subject (Your Name) has forwarded a page to you from bioRxiv Message Body (Your Name) thought you would like to see this page from the bioRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share An annotated biobank of triple negative breast cancer patient-derived xenografts featuring treatment-naïve and longitudinal samples throughout neoadjuvant chemotherapy Amanda L. Rinkenbaugh , Yuan Qi , Shirong Cai , Jiansu Shao , Faiza Hancock , Sabrina Jeter-Jones , Xiaomei Zhang , Emily Powell , Lei Huo , Rosanna Lau , Chunxiao Fu , Rebekah Gould , Petra den Hollander , Elizabeth E. Ravenberg , Jason B. White , Gaiane M. Rauch , Banu Arun , Clinton Yam , Alastair M. Thompson , Gloria V. Echeverria , Stacy L. Moulder , W. Fraser Symmans , Jeffrey T. Chang , Helen Piwnica-Worms bioRxiv 2024.11.25.625287; doi: https://doi.org/10.1101/2024.11.25.625287 Share This Article: Copy Citation Tools An annotated biobank of triple negative breast cancer patient-derived xenografts featuring treatment-naïve and longitudinal samples throughout neoadjuvant chemotherapy Amanda L. Rinkenbaugh , Yuan Qi , Shirong Cai , Jiansu Shao , Faiza Hancock , Sabrina Jeter-Jones , Xiaomei Zhang , Emily Powell , Lei Huo , Rosanna Lau , Chunxiao Fu , Rebekah Gould , Petra den Hollander , Elizabeth E. Ravenberg , Jason B. White , Gaiane M. Rauch , Banu Arun , Clinton Yam , Alastair M. Thompson , Gloria V. Echeverria , Stacy L. Moulder , W. Fraser Symmans , Jeffrey T. Chang , Helen Piwnica-Worms bioRxiv 2024.11.25.625287; doi: https://doi.org/10.1101/2024.11.25.625287 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Area Cancer Biology Subject Areas All Articles Animal Behavior and Cognition (7651) Biochemistry (17746) Bioengineering (13928) Bioinformatics (42064) Biophysics (21499) Cancer Biology (18650) Cell Biology (25579) Clinical Trials (138) Developmental Biology (13409) Ecology (19947) Epidemiology (2067) Evolutionary Biology (24373) Genetics (15633) Genomics (22557) Immunology (17774) Microbiology (40504) Molecular Biology (17217) Neuroscience (88793) Paleontology (667) Pathology (2845) Pharmacology and Toxicology (4836) Physiology (7664) Plant Biology (15178) Scientific Communication and Education (2047) Synthetic Biology (4304) Systems Biology (9839) Zoology (2272)