Lesion Volume as a Predictor for Return to Work After Endovascular Therapy - a Four-year Prospective Cohort Study

Lesion Volume as a Predictor for Return to Work After Endovascular Therapy – a Four-year Prospective Cohort Study | medRxiv /* */ /* */ <!-- <!-- /*! * 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-P4HH5NV'); Skip to main content Home About Submit ALERTS / RSS Search for this keyword Advanced Search Lesion Volume as a Predictor for Return to Work After Endovascular Therapy – a Four-year Prospective Cohort Study View ORCID Profile Gisle Berg Helland , View ORCID Profile Mona Kristiansen Beyer , Brian Anthony B. Enriquez , View ORCID Profile Hege Ihle-Hansen , View ORCID Profile Håkon Ihle-Hansen , Stein Andersson , Esten Høyland Leonardsen , Helle Stangeland , Bettina Ùjhelyi , View ORCID Profile Guri Hagberg , View ORCID Profile Hanne Flinstad Harbo , View ORCID Profile Anne Hege Aamodt , Einar August Høgestøl doi: https://doi.org/10.1101/2024.04.16.24305935 Gisle Berg Helland 1 Department of Neurology, Oslo University Hospital , Oslo, Norway 2 Department of Medicine, Bærum Hospital, Vestre Viken Hospital Trust , Drammen, Norway 3 Institute of Clinical Medicine, University of Oslo , Oslo, Norway MD Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Gisle Berg Helland For correspondence: gishel{at}ous-hf.no Mona Kristiansen Beyer 3 Institute of Clinical Medicine, University of Oslo , Oslo, Norway 4 Division of Radiology and Nuclear Medicine, Oslo University Hospital , Norway MD, PhD Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Mona Kristiansen Beyer Brian Anthony B. Enriquez 1 Department of Neurology, Oslo University Hospital , Oslo, Norway 3 Institute of Clinical Medicine, University of Oslo , Oslo, Norway 5 Institute of Health and Society, University of Oslo , Oslo, Norway MD Find this author on Google Scholar Find this author on PubMed Search for this author on this site Hege Ihle-Hansen 1 Department of Neurology, Oslo University Hospital , Oslo, Norway 2 Department of Medicine, Bærum Hospital, Vestre Viken Hospital Trust , Drammen, Norway MD, PhD Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Hege Ihle-Hansen Håkon Ihle-Hansen 2 Department of Medicine, Bærum Hospital, Vestre Viken Hospital Trust , Drammen, Norway MD, PhD Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Håkon Ihle-Hansen Stein Andersson 6 Department of Psychology, University of Oslo , Oslo, Norway PhD Find this author on Google Scholar Find this author on PubMed Search for this author on this site Esten Høyland Leonardsen 6 Department of Psychology, University of Oslo , Oslo, Norway 7 Norwegian Centre for Mental Disorders Research (NORMENT), Oslo University Hospital & Institute of Clinical Medicine, University of Oslo , Oslo, Norway MSc Find this author on Google Scholar Find this author on PubMed Search for this author on this site Helle Stangeland 8 Department of Research and Innovation, Division of Clinical Neuroscience, Oslo University Hospital , Oslo, Norway MSc Find this author on Google Scholar Find this author on PubMed Search for this author on this site Bettina Ùjhelyi 1 Department of Neurology, Oslo University Hospital , Oslo, Norway MSc Find this author on Google Scholar Find this author on PubMed Search for this author on this site Guri Hagberg 1 Department of Neurology, Oslo University Hospital , Oslo, Norway MD, PhD Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Guri Hagberg Hanne Flinstad Harbo 1 Department of Neurology, Oslo University Hospital , Oslo, Norway 3 Institute of Clinical Medicine, University of Oslo , Oslo, Norway MD, PhD Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Hanne Flinstad Harbo Anne Hege Aamodt 1 Department of Neurology, Oslo University Hospital , Oslo, Norway 9 Department of Neuromedicine and Movement Science, The Norwegian University of Science and Technology , Trondheim, Norway MD, PhD Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Anne Hege Aamodt Einar August Høgestøl 1 Department of Neurology, Oslo University Hospital , Oslo, Norway 3 Institute of Clinical Medicine, University of Oslo , Oslo, Norway 6 Department of Psychology, University of Oslo , Oslo, Norway MD, PhD Find this author on Google Scholar Find this author on PubMed Search for this author on this site Abstract Full Text Info/History Metrics Preview PDF Abstract Background There are no known objective biomarkers for predicting Return to Work (RTW) in ischemic stroke survivors. This study aims to explore the predictive utility and define a cut-off value of lesion volume on RTW after endovascular treatment (EVT). Methods We included patients <65 years undergoing EVT at Oslo University Hospital (OUS) between January 2017 and May 2019. Employment status was obtained at both baseline and a four-year follow-up. Stroke lesion volumes were segmented using magnetic resonance imaging (MRI) scans 24 hours post EVT. Logistic regression models were conducted to assess the impact of lesion volume on RTW-status at follow-up, adjusted for patients’ characteristics, stroke related factors and treatment. We calculated the Receiver Operating Characteristic curve to determine the optimal lesion volume cut-off. Machine learning (ML) regression models were employed to assess the predictive abilities of baseline clinical and imaging variables for RTW. Results Of the 109 individuals treated, 81 (74%) were employed at baseline. Among these, 60 completed four-year follow-up with MRI available for stroke lesion segmentation and were included in the analyses. Mean age at stroke onset was 51.8 years (range 23.5–64.9), 50% were female. Median lesion volume was 18 ml (IQR 45-7). After four years, 34 (57%) had successfully RTW. The odds for not RTW increased by 5% for every 1 ml increase in lesion volume (adjusted odds ratio [aOR], 95% CI, 1.02–1.11], P=0.02). A lesion volume cut-off value of 29 cm 3 yielded a sensitivity of 0.91 and specificity of 0.65 for predicting RTW. Notably, the most influential feature in the ML model for predicting RTW was lesion volume. Conclusion Lesion volume was the most robust predictor of RTW four years after EVT. Our findings suggest that a cut-off of 29 cm 3 is suitable to distinguish between those with high and low chance of RTW. Introduction Return to work (RTW) is associated with enhanced well-being and improved quality of life for stroke-survivors. 1 Stroke imposes a substantial burden on society, and work-related issues account for a considerable share of those costs. 2 , 3 In heterogenous stroke populations, the reported rates of RTW in working-age adults vary, ranging from 7% to 81%. 4 , 5 Studies focusing on patients receiving endovascular therapy (EVT) report that approximately one out of three achieve RTW within the first year after discharge. 6 , 7 Several factors have been identified as determinants for successful RTW, with male sex and younger age being the most consistent protective factors. 8 , 9 Time since the acute incident is also important, as there is a growing number of patients achieving RTW in the years following a stroke. 8 Other potential prognostic factors for RTW in patients undergoing EVT include the addition of intravenous thrombolysis (IVT), successful recanalization (Thrombolysis in Cerebral Ischemia (TICI) 2b – 3), medical complications and length of hospital stay. 7 The modified Rankin Scale (mRS) and the National Institutes of Health Stroke Scale (NIHSS) are key measures of functional post-stroke outcomes and stroke severity, and studies indicate that lower mRS and NIHSS scores during the index stay increases the likelihood of RTW. 6 , 8 , 10 None of the factors listed above, individually, are sufficient predictors of RTW. 11 , 12 In recent years, lesion volume has emerged as a potential biomarker for functional and cognitive outcome, 13 – 15 but its capacity to predict RTW has not been investigated. We aimed to explore the predictive value of lesion volume for RTW in patients undergoing EVT and establish a lesion volume cut-off to guide prognostic expectations. Secondary, we aimed to assess the predictive capabilities and influence of clinical and imaging variables on employment status four years post-EVT, employing an interpretable machine learning model. Methods Study sample and design This study utilizes data from "The Oslo Acute Reperfusion Stroke Study" (OSCAR, study-ID: NCT06220981 ), a prospective cohort study of consecutive acute ischemic stroke patients treated with EVT at Oslo University Hospital (OUH). The OUH is a comprehensive stroke center receiving EVT candidates from local stroke units. In this sub-study, all patients 65 years and younger who underwent EVT between January 1, 2017, and May 9, 2019, were included. Follow-up was performed from September 2021 to June 2022. Baseline data were collected from the OSCAR study and medical records. We adhered to the STROBE guidelines when reporting our findings. 16 The first author had access to all data and assumes responsibility for its integrity and accuracy of the data analysis. Imaging acquisition and imaging variables Magnetic Resonance Imaging (MRI) scans were acquired 24 hours post EVT, at four different Siemens scanners at OUH (Table S1). Lesions were identified as areas with diffusion restriction on Diffusion Weighted Images (DWI), indicative of irreversible brain infarction. 17 In two patients without 24-hour imaging, initial admission DWIs were used. Lesion segmentation was manually executed using the ’Insight Segmentation and Registration Toolkit-Snap’ (IKT-Snap) software (v. 4.0.0) ( Figure 1 ). 18 Segmentations were performed by trained researcher G.B.H and overseen by experienced neuroradiologist M.K.B, who was blinded to the outcome and clinical data. Download figure Open in new tab Figure 1. Examples of diffusion weighted images together with its corresponding lesion segmentation and categorization. A) Left Middle Cerebral Artery infarct, small lesion (volume 6.1 cm 3 ). B) Posterior Cerebral Artery infarct, medium lesion (volume 22.0 cm 3 ). C) Right Middle Cerebral Artery infarct, large lesion (volume 94.7 cm 3 ). Lesion location is grouped to left or right anterior- or posterior circulation, as described on the diagnostic CT angiography acquired at admission. Reperfusion grade was assessed using the revised modified Treatment In Cerebral Ischemia (mTICI) scoring system (score 0-3) by the interventional neuroradiologist during the final Digital Subtraction Angiography procedure. 19 , 20 Missing reperfusion scores were retrospectively completed by an interventional neuroradiologist blinded for clinical data including outcomes. For analysis, mTICI scores were dichotomized, with 3 defined as an “excellent“ outcome. 21 Follow-up All patients included at baseline were invited to a four-year follow-up examination. Those unable to attend in person were interviewed by telephone to collect basic demographic and clinical information. The patients were provided two self-report questionnaires related to their mental health and quality of life, including the Hospital Anxiety and Depression Scale (HADS) and the EuroQol-5Dimensions-5Levels (EQ-5D-5L). 22 , 23 Fatigue was assessed using a nine-point fatigue severity scale (FSS). 24 Information regarding adherence to work was obtained and “RTW” was defined as part or full-time paid work, or being enlisted as a student. Participants who stated that they retired early due to stroke impairments, were considered as Not Returned to Work (NRTW). Statistical comparisons between those who RTW, did NRTW and “Retired” can be found in Table S2. Information concerning education level was collected retrospectively at follow-up. For the analyses, ’primary school’ and ’high school’ were categorized into ’high school’, and ’college’ and ’university’ into ’higher education’. Statistical analysis Traditional statistical analyses were performed using the R (v. 4.3.2) statistical software. 25 Baseline characteristics are presented as number (%), medians and interquartile ranges (IQR), means and standard deviations (SD), as well as ranges, as appropriate. We conducted two-sided t-tests to analyze continuous variables that followed a normal distribution, while the Mann-Whitney U test was used for data with non-normal distribution. For categorical variables, Chi-square test and Fischer’s Exact test was used based on observed frequencies in the contingency tables. To explore the association between lesion volume and RTW, we performed a logistic regression analysis with lesion volume as a continuous variable. Results were reported for both univariable and multivariable models, with odds ratios (OR) and 95% confidence intervals (CI). Variables selected for the univariable logistic regression analysis were based on clinical expertise and insights from previous studies. Variables included sex, age, education, affected vascular territory, NIHSS on admission and at discharge, mTICI score and whether IVT was administered or not. 8 In the multivariable logistic regression model, sex, age, and education were predefined variables. Other variables were incorporated if they had a p-value below 0.2 in univariable analyses and variation inflation factor among variables was below 5. 26 An interaction analysis between NIHSS and lesion volume was performed. To further validate our findings, we conducted the aforementioned logistic regressions, normalizing lesion volume to the estimated total intracranial volume, as determined by FreeSurfer, version 7.3.2. 27 T1 weighted images taken 24 hours post-EVT were used for this analysis. To establish a lesion volume cut-off, we made a Receiver Operating Characteristic (ROC) curve and determined the lesion volume to optimally balance sensitivity and specificity using Youden’s index. To enhance the interpretability further, we performed a univariable logistic regression analysis, categorizing lesion volume and NIHSS scores at discharge. Lesion volumes were classified into small, medium and large (70 cm 3 ), based on previously published cut-offs from the ESCAPE-NA1 trial, which looked at stroke lesion volume and functional outcome (mRS). 14 NIHSS scores were categorized as minor (0-4), moderate (5-15) and severe (16-42), in line with international guidelines. 28 Categorization of admission NIHSS scores was not performed due to the minimal number of patients falling into the reference category. Machine learning model To explore the predictive properties of a larger set of baseline data using a data-driven approach, a machine learning model was developed using python with building blocks from scikit-learn. 29 A dataset was assembled using domain expertise, comprising 28 baseline features (Table S3). While inspecting the dataset we identified some missing values (35/1740 = 2%), which were imputed with the mean of their respective features. To establish the model, we performed a 10-fold cross-validation stratified for outcome and sex to increase the robustness of the results. Evaluation metrics included ROC and area under the curve (AUC), sensitivity, specificity, and accuracy. We utilized a LASSO regularization to perform automatic feature selection. 30 Furthermore, we conducted a permutation test to assess the statistical significance of the predictive performance of our machine learning model compared to chance levels. In this test, we repeatedly shuffled the employment outcome variable while keeping the predictors intact. This process was repeated 100 times to establish a distribution of baseline performances of models fitted to noise, against which the performance of the true model was compared. 31 Statistical significance was assessed by evaluating whether the mean AUC achieved by the true model was in the 95 th percentile of the AUC’s in the baseline distribution, equivalent to a significance threshold of 0.05. The ML modelling methodology and reporting of results in this study align with published recommendations. 32 Ethical approval The study was carried out in accordance with the Helsinki Declaration and received approval from the regional ethics committee (Regional Committee for Medical Ethics Southeast Norway, (REC ID: 152864). Informed consent was obtained from all eligible study participants, and for those unable to provide consent, their next-of-kin were approached to seek consent on their behalf. Data availability Supporting data for the findings in this study will be made available from the corresponding author upon reasonable request. To facilitate replicability, the pre-trained machine learning model is made publicly available at https://github.com/gislebh/Return_to_Work_after_Stroke . Results A total of 416 patients underwent EVT, of whom 133 were below 65 years of age at stroke onset. Among those employed at baseline, 60 provided consent and were followed up after four years, as seen in the flow chart in Figure 2 . Download figure Open in new tab Figure 2. Flow chart of patient selection with an illustration of enrollment, inclusion, exclusion and loss to follow up. Mean age at baseline was 52 years (SD 9 years, range 23.5 – 64.9). 50% were female. Median NIHSS score on admission was 12 (IQR 19-7) and decreased to 3 (IQR 7-1) upon discharge. The median lesion volume within the first 24 hours was 18 ml (IQR 45-7). The patient who RTW and had the largest lesion volume showed a volume of 79 cm 3 , whereas the patient who NRTW with the smallest lesion volume had a volume of 6.1 cm 3 . Excellent recanalization was achieved in 21 (35%) of the participants. At the four-year follow-up, 34 (57%) participants had RTW and 26 (43%) were NRTW. Characteristics of the study population according to RTW are presented in Table 1 . Figure 3 shows a box plot of lesion volume (cm 3 ), including individual data, based on employment status at follow-up. Descriptive statistics of patients lost to follow-up is provided in Table S4. Download figure Open in new tab Figure 3. Employment status at follow up (≈ 4 years) after stroke and lesion volume at baseline, here visualized through a box plot with individual data, including median and interquartile range. Green indicate subjects who returned to work, and orange those who did not return to work. View this table: View inline View popup Table 1: Descriptive characteristics at baseline and follow-up. N = 60 Prognostic accuracy of Lesion volume on NRTW Results from the univariable and multivariable regression analyses are presented in table 2 . The unadjusted estimates found a substantial and significant increase in the odds of NRTW with higher NIHSS at admission (p=0.04) and discharge (p=0.04), as well as higher lesion volume, both as a continuous (p<0.01) value and when entered as a categorical variable (p<0.01). Higher education (p=0.03) and posterior infarcts (p=0.05) were found as protective factors against NRTW. View this table: View inline View popup Download powerpoint Table 2: Logistic regression, association between baseline variables and RTW. Lesion volume and NIHSS was treated as continuous variables in the multivariable model. N = 60 After adjusting the model for age, sex, stroke territory, education, IVT, mTICI and NIHSS at discharge, lesion volume remained significant (OR 1.05, 95%CI 1.02 to 1.11, p= 0.02). No significant interaction was observed between NIHSS at discharge and lesion volume (p≍0.9). Furthermore, adjusting for intracranial volume did not alter the results for lesion volume. The results are shown in Table S5. Establishing a prognostic cut-off value for Lesion volume In the univariable model, lesion volume alone showed an AUC-ROC of 0.86. The optimal cut-off for lesion volume was 29 cm 3 , yielding a sensitivity of 0.91 and a specificity of 0.65 for RTW. The positive predictive value for RTW was 78% and the negative predictive value was 85%. Machine learning model The logistic regression machine learning model, as illustrated in Figure 4 , yielded a mean AUC of 0.86 (SD = 0.13), demonstrating an accuracy of 75%. The model achieved a sensitivity of 0.77 and specificity of 0.72 for identifying subjects who RTW and NRTW, respectively. The permutation test saw none of its 100 iterations yield AUCs above 0.86, instead resulting in a mean AUC of 0.53 (SD = 0.10). This substantial difference in AUC’s highly suggests that the model’s predictive ability significantly outperforms random models. Download figure Open in new tab Figure 4. This figure provides a representation of the ten-fold cross-validation process used to validate the predictive performance of the machine learning model on not returning to work. It delineates the division into ten unique folds, each consisting of 6 subjects serving as the test set in one iteration of the validation process. The performance of the model is quantified through area under the curve, calculated as mean of folds. Among the features considered in the model, lesion volume (mean OR=4.98) measured as a continuous variable, emerged as the most influential factor in predicting the outcome. For a comprehensive overview, we have presented the top eight most influential factors in Table S6. Discussion In this study of working-age adults undergoing EVT for acute ischemic stroke, including machine learning, we found lesion volume as an imaging biomarker to be the most robust predictor of RTW. Our findings suggest that a cut-off of 29 cm 3 is suitable to distinguish between those with high and low chance of RTW. To our knowledge, the importance of lesion volume and the use of AI to assess clinical variables on RTW have not been previously explored. The relationship between lesion volume and RTW underscores the critical impact of stroke severity on vocational outcomes and highlights its role as a promising biomarker of RTW. Furthermore, our analysis of ML modelling demonstrated the potential to anticipate RTW outcomes based on lesion volume and other baseline variables, although this warrants further validation. The proportion of patients in this study who RTW is consistent with findings from other stroke studies. 6 – 8 , 10 , 33 , 34 Education emerged as a protective factor for RTW. However, its significance diminished when considered alongside other variables. Contrary to previous studies, we identified an association between lesion location and RTW, 35 , 36 although significance was not retained in the multivariable model. Nonetheless, a more precise exploration of localization could be valuable, as previously demonstrated by the concept of strategic locations. 37 Previous research have uncovered the imprecise predictions of clinicians and randomness in the selection of individuals for rehabilitation. 38 – 40 ML models offer a potential pathway towards a more objective, equitable and personalized stroke follow-up approach, hence reducing the undue influence of individual clinician biases. As a contribution to the growing field of stroke prediction models, we have made our pre-trained model available for download online. An obstacle to utilizing our study’s result in clinical practice is the absence of routine measurement of lesion volume. Looking ahead, the integration of AI-powered tools for image analysis holds promise for enhancing lesion volume assessments, also when it comes to regular clinical scans, such as those used in our study. 41 Strengths of our study were the well described cohort with extensive clinical and radiological measures both at baseline and four-year follow-up. The main limitation of this study is the relatively modest sample size. Although the inclusion of diverse patients enhances generalizability, particularly to young stroke patients undergoing EVT, caution is necessary when extending our findings to broader ischemic stroke populations. Furthermore, the applicability of our results to countries with different social security systems and rehabilitation facilities than Norway should be approached with care. Additionally, several important variables were not addressed in our study, such as motivation, job type, income, rehabilitation, and the perception of social support, all of which have been established as influential factors in RTW outcomes. 8 , 42 Future research could potentially benefit from obtaining the perspective of employers, as their role in facilitating RTW is a reasonable hypothesis and deserves investigation. As with other stroke prognostic machine learning models so far, another limitation is the lack of a separate test set. This diminishes certainty about the model’s generalizability, posing challenges in comparing different models and predicting their real-world performance, and is one of the reasons for making our pre-trained model available to others. In conclusion, our study underscores the role of lesion volume as a predictor of RTW after stroke. To our knowledge, this is the first objective imaging biomarker found to be associated with RTW. Sources of funding This paper was funded by a grant from South-Eastern Norway Regional Health Authority (Grant No. 2020078). The OSCAR registry was funded by Oslo University Hospital. Additional funding for the project was obtained from Bærum Vestre Viken Hospital Trust, and The National Association for Stroke Victims, Norway. Disclosures The other authors report no conflicts of interest related to this manuscript. Supplemental material list Supplemental methods Tables S1-S6 Reference #43 (Leonardsen EH, Peng H, Kaufmann T, Agartz I, Andreassen OA, Celius EG, Espeseth T, Harbo HF, Hogestol EA, Lange AM, et al. Deep neural networks learn general and clinically relevant representations of the ageing brain. Neuroimage . 2022;256:119210. doi: 10.1016/j.neuroimage.2022.119210) Acknowledgements We express our gratitude to all the patients included in this study for their time, effort, and contribution. ChatGPT was used for improving the quality of the written English to ensure readability of the article and participated in writing the scripts for the machine learning model. Non-standard abbreviations and acronyms AI Artificial Intelligence DWI Diffusion Weighted Imaging EVT Endovascular treatment IVT Intravenous Thrombolysis MCA Middle Cerebral Artery ML Machine Learning MRI Magnetic Resonance Imaging mRS Modified Rankin Scale NIHSS National Institute of Health Stroke Scale NRTW Not Return to Work OR Odds Ratio OSCAR The Oslo Acute Reperfusion Stroke Study RTW Return to Work mTICI modified Thrombolysis in Cerebral Ischemia References 1. ↵ Vestling M , Tufvesson B , Iwarsson S . Indicators for return to work after stroke and the importance of work for subjective well-being and life satisfaction . J Rehabil Med . 2003 ; 35 : 127 – 131 . doi: 10.1080/16501970310010475 OpenUrl CrossRef PubMed Web of Science 2. ↵ Terent A , Marke LA , Asplund K , Norrving B , Jonsson E , Wester PO . Costs of stroke in Sweden. A national perspective . Stroke . 1994 ; 25 : 2363 – 2369 . doi: 10.1161/01.str.25.12.2363 OpenUrl Abstract / FREE Full Text 3. ↵ Gerstl JVE , Blitz SE , Qu QR , Yearley AG , Lassaren P , Lindberg R , Gupta S , Kappel AD , Vicenty-Padilla JC , Gaude E , et al. Global, Regional, and National Economic Consequences of Stroke . Stroke . 2023 . doi: 10.1161/STROKEAHA.123.043131 OpenUrl CrossRef 4. ↵ Hofgren C , Bjorkdahl A , Esbjornsson E , Sunnerhagen KS . Recovery after stroke: cognition, ADL function and return to work . Acta Neurol Scand . 2007 ; 115 : 73 – 80 . doi: 10.1111/j.1600-0404.2006.00768.x OpenUrl CrossRef PubMed Web of Science 5. ↵ Vlachos G , Ihle-Hansen H , Wyller TB , Braekhus A , Mangset M , Hamre C , Fure B . Factors Determining Not Returning to Full-Time Work 12 Months After Mild Ischemic Stroke . Arch Rehabil Res Clin Transl . 2023 ; 5 : 100245 . doi: 10.1016/j.arrct.2022.100245 OpenUrl CrossRef 6. ↵ Parameshwaran B , Cordato D , Parsons M , Cheung A , Manning N , Wenderoth J , Cappelen-Smith C . The Benefit of Endovascular Thrombectomy for Stroke on Functional Outcome Is Sustained at 12 Months . Cerebrovasc Dis Extra . 2021 ; 11 : 81 – 86 . doi: 10.1159/000517929 OpenUrl CrossRef 7. ↵ Hahn M , Groschel S , Hayani E , Brockmann MA , Muthuraman M , Groschel K , Uphaus T , German Stroke Registry-Endovascular Treatment Study G. Sex Disparities in Re-Employment in Stroke Patients With Large Vessel Occlusion Undergoing Mechanical Thrombectomy . Stroke . 2022 ; 53 : 2528 – 2537 . doi: 10.1161/STROKEAHA.121.037386 OpenUrl CrossRef 8. ↵ Edwards JD , Kapoor A , Linkewich E , Swartz RH . Return to work after young stroke: A systematic review . Int J Stroke . 2018 ; 13 : 243 – 256 . doi: 10.1177/1747493017743059 OpenUrl CrossRef PubMed 9. ↵ Orange C , Lanhers C , Coll G , Coste N , Dutheil F , Hauret I , Pereira B , Coudeyre E . Determinants of Return to Work After a Stroke: A Systematic Review and Meta-analysis . Arch Phys Med Rehabil . 2024 ; 105 : 359 – 368 . doi: 10.1016/j.apmr.2023.08.027 OpenUrl CrossRef 10. ↵ Aarnio K , Rodriguez-Pardo J , Siegerink B , Hardt J , Broman J , Tulkki L , Haapaniemi E , Kaste M , Tatlisumak T , Putaala J . Return to work after ischemic stroke in young adults: A registry-based follow-up study . Neurology . 2018 ; 91 : e1909 – e1917 . doi: 10.1212/WNL.0000000000006510 OpenUrl CrossRef 11. ↵ Boyd LA , Hayward KS , Ward NS , Stinear CM , Rosso C , Fisher RJ , Carter AR , Leff AP , Copland DA , Carey LM , et al. Biomarkers of Stroke Recovery: Consensus-Based Core Recommendations from the Stroke Recovery and Rehabilitation Roundtable . Neurorehabil Neural Repair . 2017 ; 31 : 864 – 876 . doi: 10.1177/1545968317732680 OpenUrl CrossRef 12. ↵ Zhang JJ , Sanchez Vidana DI , Chan JN , Hui ESK , Lau KK , Wang X , Lau BWM , Fong KNK . Biomarkers for prognostic functional recovery poststroke: A narrative review . Front Cell Dev Biol . 2022 ; 10 : 1062807 . doi: 10.3389/fcell.2022.1062807 OpenUrl CrossRef 13. ↵ Vogt G , Laage R , Shuaib A , Schneider A , Collaboration V . Initial lesion volume is an independent predictor of clinical stroke outcome at day 90: an analysis of the Virtual International Stroke Trials Archive (VISTA) database . Stroke . 2012 ; 43 : 1266 – 1272 . doi: 10.1161/STROKEAHA.111.646570 OpenUrl Abstract / FREE Full Text 14. ↵ Ospel JM , Hill MD , Menon BK , Demchuk A , McTaggart R , Nogueira R , Poppe A , Haussen D , Qiu W , Mayank A , et al. Strength of Association between Infarct Volume and Clinical Outcome Depends on the Magnitude of Infarct Size: Results from the ESCAPE-NA1 Trial . AJNR Am J Neuroradiol . 2021 ; 42 : 1375 – 1379 . doi: 10.3174/ajnr.A7183 OpenUrl Abstract / FREE Full Text 15. ↵ Aamodt EB , Lydersen S , Alnaes D , Schellhorn T , Saltvedt I , Beyer MK , Haberg A . Longitudinal Brain Changes After Stroke and the Association With Cognitive Decline . Front Neurol . 2022 ; 13 : 856919 . doi: 10.3389/fneur.2022.856919 OpenUrl CrossRef 16. ↵ von Elm E , Altman DG , Egger M , Pocock SJ , Gotzsche PC , Vandenbroucke JP , Initiative S . The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies . Lancet . 2007 ; 370 : 1453 – 1457 . doi: 10.1016/S0140-6736(07)61602-X OpenUrl CrossRef PubMed Web of Science 17. ↵ van Everdingen KJ , van der Grond J , Kappelle LJ , Ramos LM , Mali WP . Diffusion-weighted magnetic resonance imaging in acute stroke . Stroke . 1998 ; 29 : 1783 – 1790 . doi: 10.1161/01.str.29.9.1783 OpenUrl Abstract / FREE Full Text 18. ↵ Yushkevich PA , Piven J , Hazlett HC , Smith RG , Ho S , Gee JC , Gerig G . User-guided 3D active contour segmentation of anatomical structures: significantly improved efficiency and reliability . Neuroimage . 2006 ; 31 : 1116 – 1128 . doi: 10.1016/j.neuroimage.2006.01.015 OpenUrl CrossRef PubMed Web of Science 19. ↵ Dargazanli C , Fahed R , Blanc R , Gory B , Labreuche J , Duhamel A , Marnat G , Saleme S , Costalat V , Bracard S , et al. Modified Thrombolysis in Cerebral Infarction 2C/Thrombolysis in Cerebral Infarction 3 Reperfusion Should Be the Aim of Mechanical Thrombectomy: Insights From the ASTER Trial (Contact Aspiration Versus Stent Retriever for Successful Revascularization) . Stroke . 2018 ; 49 : 1189 – 1196 . doi: 10.1161/STROKEAHA.118.020700 OpenUrl Abstract / FREE Full Text 20. ↵ Jang KM , Nam TK , Ko MJ , Choi HH , Kwon JT , Park SW , Byun JS . Thrombolysis in Cerebral Infarction Grade 2C or 3 Represents a Better Outcome than 2B for Endovascular Thrombectomy in Acute Ischemic Stroke: A Network Meta-Analysis . World Neurosurg . 2020 ; 136 : e419 – e439 . doi: 10.1016/j.wneu.2020.01.020 OpenUrl CrossRef PubMed 21. ↵ LeCouffe NE , Kappelhof M , Treurniet KM , Lingsma HF , Zhang G , van den Wijngaard IR , van Es A , Emmer BJ , Majoie C , Roos Y , et al. 2B, 2C, or 3: What Should Be the Angiographic Target for Endovascular Treatment in Ischemic Stroke? Stroke . 2020 ; 51 : 1790 – 1796 . doi: 10.1161/STROKEAHA.119.028891 OpenUrl CrossRef PubMed 22. ↵ Golicki D , Niewada M , Buczek J , Karlinska A , Kobayashi A , Janssen MF , Pickard AS . Validity of EQ-5D-5L in stroke . Qual Life Res . 2015 ; 24 : 845 – 850 . doi: 10.1007/s11136-014-0834-1 OpenUrl CrossRef PubMed 23. ↵ Kari Ann Leiknes TKD, Johan Siqveland . Psychometric assessment of the Norwegian version of the Hospital Anxiety and Depression Scale (HADS). In : Internet : Norwegian Institute of Public Health ; 2016 :32. 24. ↵ Lerdal A , Wahl A , Rustoen T , Hanestad BR , Moum T . Fatigue in the general population: a translation and test of the psychometric properties of the Norwegian version of the fatigue severity scale . Scand J Public Health . 2005 ; 33 : 123 – 130 . doi: 10.1080/14034940410028406 OpenUrl CrossRef PubMed Web of Science 25. ↵ R Core Team . A Language and Environment for Statistical Computing; 4.3.0 . Vienna, Austria : R Foundation for Statistical Computing ; 2023 . 26. ↵ Akinwande MO DH . Variance Inflation Factor: As a Condition for the Inclusion of Suppressor Variable(s) in Regression Analysis . Open Journal of Statistics . 2015 ; 5 . doi: 10.4236/ojs.2015.57075 OpenUrl CrossRef 27. ↵ Fischl. FreeSurfer . NeuroImage . 2012 ; 62 . doi: 10.1016/j.neuroimage.2012.01.021 OpenUrl CrossRef PubMed Web of Science 28. ↵ Berge E , Whiteley W , Audebert H , De Marchis GM , Fonseca AC , Padiglioni C , de la Ossa NP , Strbian D , Tsivgoulis G , Turc G . European Stroke Organisation (ESO) guidelines on intravenous thrombolysis for acute ischaemic stroke . Eur Stroke J . 2021 ; 6 :I-LXII. doi: 10.1177/2396987321989865 OpenUrl CrossRef 29. ↵ Fabian Pedregosa GV , Alexandre Gramfort , Vincent Michel , Bertrand Thirion , Olivier Grisel , Mathieu Blondel , Peter Prettenhofer , Ron Weiss , Vincent Dubourg , Jake Vanderplas , Alexandre Passos , David Cournapeau , Matthieu Brucher , Matthieu Perrot , Édouard Duchesnay . Scikit-learn: Machine Learning in Python . Journal of Machine Learning Research . 2011 : 2825 – 2830 . 30. ↵ Tibshirani R . Regression Shrinkage and Selection via The Lasso . Journal of the Royal Statistical Society . 1996 ; 58 : 267 – 288 . OpenUrl CrossRef Web of Science 31. ↵ Bonnini S , Bonnini S , ProQuest. Nonparametric hypothesis testing : rank and permutation methods with applications in R. 1st edition ed . Chichester, England : Wiley ; 2014 . 32. ↵ Stevens LM , Mortazavi BJ , Deo RC , Curtis L , Kao DP . Recommendations for Reporting Machine Learning Analyses in Clinical Research . Circ Cardiovasc Qual Outcomes . 2020 ; 13 : e006556 . doi: 10.1161/CIRCOUTCOMES.120.006556 OpenUrl CrossRef PubMed 33. ↵ Daniel K , Wolfe CD , Busch MA , McKevitt C . What are the social consequences of stroke for working-aged adults? A systematic review . Stroke . 2009 ; 40 : e431 – 440 . doi: 10.1161/STROKEAHA.108.534487 OpenUrl Abstract / FREE Full Text 34. ↵ Maaijwee NA , Rutten-Jacobs LC , Arntz RM , Schaapsmeerders P , Schoonderwaldt HC , van Dijk EJ , de Leeuw FE . Long-term increased risk of unemployment after young stroke: a long-term follow-up study . Neurology . 2014 ; 83 : 1132 – 1138 . doi: 10.1212/WNL.0000000000000817 OpenUrl Abstract / FREE Full Text 35. ↵ Saeki S , Hachisuka K . The association between stroke location and return to work after first stroke . J Stroke Cerebrovasc Dis . 2004 ; 13 : 160 – 163 . doi: 10.1016/j.jstrokecerebrovasdis.2004.06.006 OpenUrl CrossRef PubMed 36. ↵ Wozniak MA , Kittner SJ , Price TR , Hebel JR , Sloan MA , Gardner JF . Stroke location is not associated with return to work after first ischemic stroke . Stroke . 1999 ; 30 : 2568 – 2573 . doi: 10.1161/01.str.30.12.2568 OpenUrl Abstract / FREE Full Text 37. ↵ Weaver NA , Kuijf HJ , Aben HP , Abrigo J , Bae HJ , Barbay M , Best JG , Bordet R , Chappell FM , Chen C , et al. Strategic infarct locations for post-stroke cognitive impairment: a pooled analysis of individual patient data from 12 acute ischaemic stroke cohorts . Lancet Neurol . 2021 ; 20 : 448 – 459 . doi: 10.1016/S1474-4422(21)00060-0 OpenUrl CrossRef 38. ↵ Stinear CM , Smith MC , Byblow WD . Prediction Tools for Stroke Rehabilitation . Stroke . 2019 ; 50 : 3314 – 3322 . doi: 10.1161/STROKEAHA.119.025696 OpenUrl CrossRef PubMed 39. Reid JM , Dai D , Delmonte S , Counsell C , Phillips SJ , MacLeod MJ . Simple prediction scores predict good and devastating outcomes after stroke more accurately than physicians . Age Ageing . 2017 ; 46 : 421 – 426 . doi: 10.1093/ageing/afw197 OpenUrl CrossRef 40. ↵ Xian Y , Thomas L , Liang L , Federspiel JJ , Webb LE , Bushnell CD , Duncan PW , Schwamm LH , Stein J , Fonarow GC , et al. Unexplained Variation for Hospitals’ Use of Inpatient Rehabilitation and Skilled Nursing Facilities After an Acute Ischemic Stroke . Stroke . 2017 ; 48 : 2836 – 2842 . doi: 10.1161/STROKEAHA.117.016904 OpenUrl Abstract / FREE Full Text 41. ↵ Chen L , Bentley P , Rueckert D . Fully automatic acute ischemic lesion segmentation in DWI using convolutional neural networks . Neuroimage Clin . 2017 ; 15 : 633 – 643 . doi: 10.1016/j.nicl.2017.06.016 OpenUrl CrossRef PubMed 42. ↵ Saar K , Tolvanen A , Poutiainen E , Aro T . Returning to Work after Stroke: Associations with Cognitive Performance, Motivation, Perceived Working Ability and Barriers . J Rehabil Med . 2023 ; 55 :jrm00365. doi: 10.2340/jrm.v55.2576 OpenUrl CrossRef View the discussion thread. Back to top Previous Next Posted April 22, 2024. Download PDF Email Thank you for your interest in spreading the word about medRxiv. 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 Lesion Volume as a Predictor for Return to Work After Endovascular Therapy – a Four-year Prospective Cohort Study Message Subject (Your Name) has forwarded a page to you from medRxiv Message Body (Your Name) thought you would like to see this page from the medRxiv 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 Lesion Volume as a Predictor for Return to Work After Endovascular Therapy – a Four-year Prospective Cohort Study Gisle Berg Helland , Mona Kristiansen Beyer , Brian Anthony B. Enriquez , Hege Ihle-Hansen , Håkon Ihle-Hansen , Stein Andersson , Esten Høyland Leonardsen , Helle Stangeland , Bettina Ùjhelyi , Guri Hagberg , Hanne Flinstad Harbo , Anne Hege Aamodt , Einar August Høgestøl medRxiv 2024.04.16.24305935; doi: https://doi.org/10.1101/2024.04.16.24305935 Share This Article: Copy Citation Tools Lesion Volume as a Predictor for Return to Work After Endovascular Therapy – a Four-year Prospective Cohort Study Gisle Berg Helland , Mona Kristiansen Beyer , Brian Anthony B. Enriquez , Hege Ihle-Hansen , Håkon Ihle-Hansen , Stein Andersson , Esten Høyland Leonardsen , Helle Stangeland , Bettina Ùjhelyi , Guri Hagberg , Hanne Flinstad Harbo , Anne Hege Aamodt , Einar August Høgestøl medRxiv 2024.04.16.24305935; doi: https://doi.org/10.1101/2024.04.16.24305935 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 Neurology Subject Areas All Articles Addiction Medicine (574) Allergy and Immunology (865) Anesthesia (304) Cardiovascular Medicine (4460) Dentistry and Oral Medicine (445) Dermatology (383) Emergency Medicine (611) Endocrinology (including Diabetes Mellitus and Metabolic Disease) (1517) Epidemiology (15251) Forensic Medicine (31) Gastroenterology (1132) Genetic and Genomic Medicine (6621) Geriatric Medicine (669) Health Economics (1002) Health Informatics (4564) Health Policy (1372) Health Systems and Quality Improvement (1617) Hematology (544) HIV/AIDS (1272) Infectious Diseases (except HIV/AIDS) (15938) Intensive Care and Critical Care Medicine (1107) Medical Education (624) Medical Ethics (147) Nephrology (670) Neurology (6642) Nursing (346) Nutrition (1001) Obstetrics and Gynecology (1148) Occupational and Environmental Health (957) Oncology (3350) Ophthalmology (981) Orthopedics (369) Otolaryngology (421) Pain Medicine (436) Palliative Medicine (130) Pathology (665) Pediatrics (1698) Pharmacology and Therapeutics (694) Primary Care Research (714) Psychiatry and Clinical Psychology (5464) Public and Global Health (9259) Radiology and Imaging (2212) Rehabilitation Medicine and Physical Therapy (1372) Respiratory Medicine (1198) Rheumatology (598) Sexual and Reproductive Health (716) Sports Medicine (533) Surgery (715) Toxicology (100) Transplantation (289) Urology (265) (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'a038d5a468038650',t:'MTc4MDA5MjY4Mg=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();