Journal of the National Cancer Institute最新文献

{"title":"Effect of low-dose tamoxifen on benign gynecological and breast conditions in a phase III trial in non-invasive breast cancer","authors":"Alessio Carbone, Martino Oliva, Matteo Puntoni, Aliana Guerrieri-Gonzaga, Irene Maria Briata, Matteo Lazzeroni, Davide Serrano, Livia Giordano, Maria Digennaro, Laura Cortesi, Francesco Millo, Katia Cagossi, Giuseppe Aprile, Patrizia Serra, Elisa Gallerani, Bernardo Bonanni, Andrea DeCensi","doi":"10.1093/jnci/djaf250","DOIUrl":"https://doi.org/10.1093/jnci/djaf250","url":null,"abstract":"Despite the proven efficacy of tamoxifen (T) in breast cancer prevention, its uptake remains limited due to concerns over side effects. A phase III trial showed that low-dose T (5 mg/day for 3 years) significantly reduces recurrence with minimal toxicity. After 10 years of follow-up, we observed no significant differences in benign gynecological or breast events between the tamoxifen and placebo (P) arms. The uterus was the most frequently affected site (30 cases per arm), followed by the breast (18) and ovary (5 vs 3 on T and P, respectively). Endometrial polyps were similarly distributed. Endometrial thickness remained stable in premenopausal women and showed only a mild, non-clinically significant increase in postmenopausal women on T (∼1.5 mm). Compared with data on standard-dose tamoxifen, rates of benign events were lower. These findings reinforce the favorable safety profile of low-dose tamoxifen. ClinicalTrials.gov: NCT01357772.","PeriodicalId":501635,"journal":{"name":"Journal of the National Cancer Institute","volume":"16 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-08-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144919387","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Genetic predisposition to persistent fatigue after a diagnosis of colorectal cancer","authors":"Elham Kazemian, Qianxing Mo, Marco Matejcic, Ya-Yu Tsai, Daniel Sobieski, Xiaoyin Li, Aasha I Hoogland, Sylvia L Crowder, Brian D Gonzalez, Laura B Oswald, Alix G Sleight, Nathalie Nguyen, Nicole C Loroña, Victoria Damerell, Khaled R Komrokji, Kathi Mooney, Mary C Playdon, Cornelia M Ulrich, Christopher I Li, David Shibata, Adetunji T Toriola, Jennifer Ose, Anita R Peoples, Sheetal Hardikar, Christoph Kahlert, Erin M Siegel, Julienne E Bower, Stephanie L Schmit, Biljana Gigic, Heather S L Jim, Jane C Figueiredo","doi":"10.1093/jnci/djaf140","DOIUrl":"https://doi.org/10.1093/jnci/djaf140","url":null,"abstract":"Background Cancer-related fatigue (fatigue) is a common and persistent symptom after cancer treatment, yet the role of genetic susceptibility remains unclear. Methods We leveraged data from a prospective cohort study, ColoCare Study (ie, five U.S. sites and Germany). Fatigue was assessed at five timepoints using the EORTC QLQ-C30 fatigue subscale and analyzed as (1) a binary summary measure of the trajectory from diagnosis into survivorship (defined as severe: yes, no), (2) a mean score across all time points, and (3) the highest (ie, worst) score across all time points. We genotyped samples using Infinium Global Diversity Array with imputation using the TOPMed reference panel to conduct a genome-wide analysis (GWAS). SuSiE was used to identify independent secondary signals. Transcriptome-wide association studies (TWAS) using S-PrediXcan and MultiXcan were conducted to examine genetic regulation of gene expression. COLOC assessed whether variants identified in the GWAS influence gene expression through colocalization analysis. Results Among 1,219 participants, 31.0% experienced severe fatigue over the course of their diagnosis. A locus near LINC02505 on chromosome 4 was associated with severe fatigue (rs6531463, OR = 3.25, p = 3.88 × 10−8). When modeling mean fatigue levels, significantly associated variants were identified in or near NEK10 and SLC4A7. Integrative analyses linked the predicted expression of NEK10 in liver tissue to risk of fatigue (p < 4.36 × 10−6). Colocalization analysis identified genetic loci and gene expression near NEK10 (posterior probabilities > 0.9). Conclusions This study identified novel genetic loci associated with fatigue in CRC patients and may be useful for identifying high-risk individuals for preventative strategies.","PeriodicalId":501635,"journal":{"name":"Journal of the National Cancer Institute","volume":"23 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-08-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144927905","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Censoring in trials testing immunotherapy in advanced cancers. A systematic review and a meta-research study","authors":"Filippo Vitale, Fabio Salomone, Massimo Di Maio, Simeone D’Ambrosio, Annarita Avanzo, Fabiana Napolitano, Angela Viggiano, Luigi Liguori, Anna Russo, Maria Carmela Isernia, Lucia Longo, Antonio Santaniello, Luigi Formisano, Roberto Bianco, Alberto Servetto","doi":"10.1093/jnci/djaf237","DOIUrl":"https://doi.org/10.1093/jnci/djaf237","url":null,"abstract":"Background Informative censoring affects interpretation of trials results. We investigated censoring rates in randomized controlled trials (RCTs) of immune checkpoint inhibitors (ICIs). Methods We searched articles of RCTs testing ICIs in advanced cancers, published up to 12/2023. For both progression free (PFS) and overall survival (OS) Kaplan-Meier (K-M) curves, we collected the rates of censored patients at the first (T1), median PFS/OS (TmPFS/OS) and last (T2) study intervals. We calculated the unweighted difference in censoring rates (ΔC-E) and the weighted difference adjusted for enrolment size (wΔC-E), in control (C) versus experimental (E) arm at T1. Tm and T2. Results Of the selected 140 trials, censoring data at T1, Tm and T2 were available for 53/140 (37.8%) and 55/140 (39.2%) trials for PFS and OS K-M curves, respectively. Rates of censoring in C and E were: at T1, 8.19% and 4.92%, for PFS; TmPFS, 15.5% and 12.5%; T1, 2.33% and 1.16%, for OS; TmOS, 20.1% and 21.3%; T2, 23.29% and 26.34%, for PFS; T2, 33.3% and 39.49%, for OS. Analysis of wΔC-E revealed more censoring in C at T1 (PFS: 1.32; OS: 0.40) and in E at T2 (PFS: -2.61; OS: -5.23). Finally, at T1, we found larger rates of censoring in C of open-label compared to double-blinded RCTs. Conclusions Multiple RCTs of ICIs did not report censoring data. The rate of censoring is higher in C at the start and increases in E over the course of the trial. Further studies might elucidate the role of censoring on survival outcomes.","PeriodicalId":501635,"journal":{"name":"Journal of the National Cancer Institute","volume":"32 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144906483","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Industry promotion of oncology drugs with accelerated approval that failed confirmatory trials","authors":"Maryam Mooghali, Reshma Ramachandran, Ayman Mohammad, Aaron P Mitchell","doi":"10.1093/jnci/djaf241","DOIUrl":"https://doi.org/10.1093/jnci/djaf241","url":null,"abstract":"Industry payments to physicians influence prescribing, raising concern for drugs granted accelerated approval that failed confirmatory studies. We measured industry payments for oncology drugs granted accelerated approval before and after negative confirmatory study results. From 2009-2021, of 73 drugs granted accelerated approval, 6 (8.2%) had negative confirmatory studies and available OpenPayments data. These were withdrawn a median of 0.3 (IQR, 0.3-0.6) years after announcement of negative results. Two drugs had almost no payments after negative result announcement. Among other 4, average monthly payments in the year preceding announcement of negative results to the period between results and market withdrawal went from $138 to $183 for vinCRIStine sulfate, $12,317 to $4,606 for Panobinostat, $152,417 to $119,066 for belantamab mafodotin, and $23,139 to $13,588 for umbralisib. While payments for oncology drugs with accelerated approvals mostly decreased, industry promotion continued for certain drugs after confirmatory studies failed until drugs were withdrawn from market.","PeriodicalId":501635,"journal":{"name":"Journal of the National Cancer Institute","volume":"26 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144901592","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"International neuroblastoma risk group consortium: a model of a networking for rare cancers","authors":"Susan L Cohn, Wendy B London, Gudrun Schleiermacher, Lucas Moreno, Inge M Ambros, Peter F Ambros, Rochelle Bagatell, Maja Beck Popovicmd, Klaus Hermann Beiske, Frank Berthold, Suzi Birz, Hervé J Brisse, Garrett M Brodeur, Penelope R Brock, Susan Burchill, Angelika Eggert, Sara M Federico, Matthias Fischer, Brian T Furner, Barbara Hero, David Machin, Takehiko Kamijo, Katherine K Matthay, Akira Nakagawara, Arlene Naranjo, Ulrike Pötschger, Dominique Valteau-Couanet, Michael T Watkins, Meredith S Irwin, Samuel L Volchenboum, Julie R Park, Andrew D J Pearson","doi":"10.1093/jnci/djaf242","DOIUrl":"https://doi.org/10.1093/jnci/djaf242","url":null,"abstract":"It is critical to share knowledge and harmonize approaches to optimize progress in rare cancers. The International Neuroblastoma Risk Group (INRG) Task Force was formed by the four major neuroblastoma cooperative groups in 2004 to achieve this goal. Strategies developed for neuroblastoma are an exemplar for other rare malignancies. Data from an initial cohort of 8,800 patients were transferred to the INRG Data Commons, and a data-sharing model was developed. Currently, information on > 25,000 patients are available to the research community. The INRG staging and risk classification systems have led to harmonized approaches for therapeutic groupings. INRG consensus manuscripts have led to uniform criteria for classifying biological data, evaluating the extent of disease, and defining treatment response. More than forty INRG research studies have been performed by investigators from around the world, including analyses of rare patients which would not otherwise be possible. The success of this approach for neuroblastoma has been leveraged to create the Pediatric Cancer Data Commons and the Data for the Common Good. Efforts to enrich the INRG Commons with additional genomic and biomarker data, extracted electronic health records, and digital medical images are ongoing. The international networking model developed by the INRG Task Force has led to new research discoveries and progress in neuroblastoma. The approach has now been applied to sixteen other cancers and conditions, including rhabdomyosarcoma, germ cell tumor, Lynch Syndrome and cancer predisposition. This framework of international collaboration and data sharing serves as a model for advancing rare adult malignancies.","PeriodicalId":501635,"journal":{"name":"Journal of the National Cancer Institute","volume":"17 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144899120","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Increase of early-onset colorectal cancer: a cohort effect","authors":"Laura Downham, Mathieu Laversanne, Sandra Perdomo, Adalberto M Filho, Freddie Bray, Paul Brennan","doi":"10.1093/jnci/djaf238","DOIUrl":"https://doi.org/10.1093/jnci/djaf238","url":null,"abstract":"Increasing incidence rates of early-onset colorectal cancer (eoCRC, <50 years) have been reported across multiple countries. We investigated long-term cancer incidence data (from 1995 or earlier) from Australia, Canada, England, and the U.S, separately by sex. Estimated annual percentage change (EAPC) and age-period-cohort (APC) models were used to assess trends by country and sex. All countries showed increasing eoCRC incidence in successive birth cohorts since 1960, with those born in the 1990s facing at least five-fold higher risks than those born in the 1960s. Cohort effects were observed across all countries, with sharper increases at younger ages. Over the most recent decade, EAPC ranged from 3.7% in Canada to 6.0% in England, with steep rises before age of 40. The emergence of these trends from ages 20–29 suggests contributing factors may originate early in life and may reflect exposures whose effect begin in youth and accumulate throughout the lifespan.","PeriodicalId":501635,"journal":{"name":"Journal of the National Cancer Institute","volume":"22 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144899128","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Risk of lymphomas among people with HIV in the United States during 2001-2019","authors":"Jun Tao, Qianlai Luo, Cameron B Haas, Sam M Mbulaiteye, Jennifer Hayes, Colby Cohen, Karen S Pawlish, Sai Cherala, Eric A Engels, Lindsay M Morton, Meredith S Shiels","doi":"10.1093/jnci/djaf244","DOIUrl":"https://doi.org/10.1093/jnci/djaf244","url":null,"abstract":"Introduction Lymphomas are among the most common malignancies in people with HIV (PWH). Although the widespread use of combination antiretroviral therapy has significantly reduced the risk of lymphoma among PWH, data on specific lymphoma subtypes remain limited and outdated. Methods Using data from the HIV/AIDS Cancer Match Study, a U.S. cohort linking cancer and HIV registries and including 822,702 PWH, we assessed the relative risks of 14 lymphoma subtypes by calendar period and age group during 2001-2019. We estimated the risk of lymphoma subtypes using standardized incidence ratios (SIRs) adjusted for age, sex, race, ethnicity, and registry. Trends of SIRs for lymphoma subtypes across calendar periods were estimated using Poisson regression overall and by age groups. Results There were 6,577 non-Hodgkin lymphomas (NHL) and 1,783 Hodgkin lymphomas (HL) during 2001-2019. SIRs declined significantly over time for most NHL subtypes including diffuse large B-cell lymphoma (DLBCL), Burkitt lymphoma (BL), peripheral T-cell lymphoma (PTCL), central nervous system (CNS) NHL, and follicular lymphoma. During 2015-2019, lymphoma subtypes with elevated risk among PWH included DLBCL (SIR = 5.38, 95%CI = 5.02-5.76), BL (15.1, 12.6-17.9), PTCL (1.68, 1.26-2.20), CNS lymphoma (8.85, 7.05-11.0), nodular sclerosis HL (4.04, 3.22-5.02), and other classical HL (8.99, 8.00-10.1). For most lymphoma subtypes, SIRs for the 20-39-year-old-group were higher than 40-59-year-old and 60-84-year-old groups. Conclusion The risk of lymphomas among PWH in the US declined during 2001-2019, with variations across histological subtypes and age groups. Despite overall declines, several NHL subtypes and HL risk remained elevated recently, underscoring the continued lymphoma burden among PWH.","PeriodicalId":501635,"journal":{"name":"Journal of the National Cancer Institute","volume":"19 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144905985","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The impact of eligibility criteria on KRASG12C inhibitor trials in patients with NSCLC","authors":"Margaux Wooster, Michael May, Prashasti Agrawal, Jonathan Lee, Benjamin May, Xin Ma, Stephanie Bogdan, Catherine A Shu, Brian S Henick, Anjali Saqi, Mahesh Mansukhani, Gregory Riely, Dawn L Hershman, Christine Garcia, Kathryn C Arbour, Benjamin O Herzberg","doi":"10.1093/jnci/djaf236","DOIUrl":"https://doi.org/10.1093/jnci/djaf236","url":null,"abstract":"Background 20% of cancer patients are estimated to be ineligible for phase III trials due to restrictive eligibility criteria. In response, several groups, including the FDA, have advocated for more inclusive study designs. We examined KRAS G12C inhibitor trials to determine if inclusivity has shifted in the development of molecularly-targeted therapies. Methods We evaluated Phase I–III studies of KRAS G12C inhibitors in non–small cell lung cancer (NSCLC) by applying criteria from 15 US trials to a multi-institutional real-world cohort of patients with metastatic NSCLC and universal KRAS testing (N = 2383). Eligibility analysis, multivariate logistic regression for ineligibility, and a Cox proportional hazards model were used on patient with KRAS G12C–mutated NSCLC (N = 185) to compare trial enrollment and overall survival under various eligibility modifications. Results 60-70% of patients with metastatic KRAS G12C-mutated NSCLC were ineligible for any KRAS inhibitor clinical trial, including studies aiming to establish first-line standard of care. Eligibility criteria remained unchanged from Phase I to Phase III. Performance status, renal function, and active brain metastases were the main causes of trial ineligibility. Liberalizing criteria for renal function and brain metastases increased enrollment by 25% without affecting overall survival (p = .49), whereas allowing worse performance status reduced study effect sizes (p = .001 in second line and p = .04 in first line). Conclusions Most patients with metastatic KRAS G12C-mutated NSCLC are excluded from trials. There is significant potential to refine trial entry criteria to better balance generalizability, safety, speed, and success.","PeriodicalId":501635,"journal":{"name":"Journal of the National Cancer Institute","volume":"14 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144899146","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Real-world generalizability of clinical trial cytomolecular risk in pediatric acute myeloid leukemia: a report from the REAL-AML cohort","authors":"Daniel J Zheng, Gary Hettinger, Catherine Aftandilian, Kira Bona, Emi H Caywood, Anderson B Collier, Caitlin W Elgarten, Cody Gathers, Taumoha Ghosh, M Monica Gramatges, Meret Henry, Yuan-Shung V Huang, Yimei Li, Craig Lotterman, Kelly Maloney, Amir Mian, Tamara P Miller, Arunkumar Modi, Rajen Mody, Elaine Morgan, Regina Myers, Haley Newman, Jose Ortiz, Alix E Seif, Caroline Smith, Jamie Stokke, Xin Wang, Naomi Winick, Jennifer J Wilkes, Victor Wong, Richard Aplenc, Kelly D Getz","doi":"10.1093/jnci/djaf234","DOIUrl":"https://doi.org/10.1093/jnci/djaf234","url":null,"abstract":"Cytomolecular features critical for risk-stratified treatment determination in pediatric acute myeloid leukemia (AML) were expanded in Children’s Oncology Group (COG) Phase III trial AAML1831 based on previous trials. It remains unknown whether the cytomolecular risk profiles are generalizable to the real-world. We addressed this knowledge gap using a nationally representative real-world cohort of 913 pediatric AML patients. Distributions of cytomolecular risk profiles and individual markers were comparable for trial-enrolled and non-enrolled patients, as well as across social drivers of trial enrollment (race/ethnicity, language, insurance, acuity). Compared to patients with only favorable cytomolecular markers (4-year OS 89.48%; 95% CI: 84.46%-92.95%), patients with both favorable and unfavorable (hazards ratio [HR] = 2.49, 95% CI : 1.18-5.23), neutral (HR = 4.33, 95% CI : 2.75-6.82), and only unfavorable (HR = 5.80, 95% CI: 3.70-9.11) markers all had increased hazards of death. Cytomolecular risk informed by trial data appears to be generalizable to the real-world setting in pediatric AML.","PeriodicalId":501635,"journal":{"name":"Journal of the National Cancer Institute","volume":"195 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144899204","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Basal cell carcinoma risk prediction in survivors of childhood cancer","authors":"Cindy Im, Christina Boull, Zhe Lu, Kenneth Liao, Hasibul Hasan, Linwan Xu, Yadav Sapkota, Rebecca M Howell, Michael A Arnold, Miriam R Conces, Ashley J Housten, Judith Gebauer, Thorsten Langer, Jop C Teepen, Leontien C M Kremer, Louis S Constine, Yutaka Yasui, Melissa M Hudson, Kirsten K Ness, Gregory T Armstrong, Joseph P Neglia, Yan Yuan, Lucie M Turcotte","doi":"10.1093/jnci/djaf228","DOIUrl":"https://doi.org/10.1093/jnci/djaf228","url":null,"abstract":"Background Survivors of childhood cancer face excess risk of developing basal cell carcinoma (BCC). Age-specific BCC risk prediction models for survivors may support targeted screening recommendations. Methods We developed models predicting BCC risk by ages 40 and 50 years featuring detailed cancer treatment predictors, utilizing statistical/machine learning algorithms and data from 23,166 five-year survivors in the Childhood Cancer Survivor Study (CCSS), a multi-institutional retrospective cohort study. Selected models were externally validated in 5,314 survivors in the St Jude Lifetime Cohort (SJLIFE). Model discrimination and precision were evaluated using the areas under the receiver operating characteristic curve (AUROC) and precision-recall curve (AUPRC), and benchmarked against the current Children’s Oncology Group Long-Term Follow-Up Guidelines (COG LTFU, v6.0) for skin cancer screening. Results By ages 40 and 50 years, BCC cumulative incidence was 5% and 15% in CCSS and 7% and 21% in SJLIFE. The XGBoost algorithm-based models with treatment dose-specific predictors performed best, showing good external discrimination (AUROC40y=0.75; AUROC50y=0.76) and precision (AUPRC40y=0.20; AUPRC50y=0.52), outperforming COG LTFU Guideline-directed risk stratification (AUROC40y=0.65, AUROC50y=0.62; AUPRC40y=0.09, AUPRC50y=0.26; P < .01). These novel models reclassified 37% of survivors with COG-recommended skin cancer screening as low risk by age 40 and 29% of survivors without COG-recommended screening as moderate/high risk by age 50, suggesting these recommendations overestimate risk in younger survivors and miss relevant predictors (eg, attained age, chemotherapy). Conclusions In this study, we present validated BCC risk prediction models for childhood cancer survivors that outperform current practice guidelines. The associated online risk calculator can inform risk-/age-based screening recommendations.","PeriodicalId":501635,"journal":{"name":"Journal of the National Cancer Institute","volume":"70 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144899192","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}