Search

PURPOSE With the availability of multigene panel testing (MGPT) for hereditary cancer risk assessment, clinicians need to assess the likelihood of pathogenic germline variants (PGVs) across numerous genes in parallel. This study's aim was to develop and validate a clinical prediction model (PREMMplus) for MGPT risk assessment. MATERIALS AND METHODS PREMMplus was developed in a single-institution cohort of 7,280 individuals who had undergone MGPT. Logistic regression models with Least Absolute Shrinkage and Selection Operator regularization were used to examine candidate predictors (age, sex, ethnicity, and personal/family history of 18 cancers/neoplasms) to estimate one's likelihood of carrying PGVs in 19 genes (broadly categorized by phenotypic overlap and/or relative penetrance: 11 category A [ APC, BRCA1/ 2, CDH1, EPCAM, MLH1, MSH2, MSH6, biallelic MUTYH, PMS2, and TP53] and eight category B genes [ ATM, BRIP1, CDKN2A, CHEK2, PALB2, PTEN, RAD51C, and RAD51D]). Model performance was validated in nonoverlapping data sets of 8,691 and 14,849 individuals with prior MGPT ascertained from clinic- and laboratory-based settings, respectively. RESULTS PREMMplus (score ≥ 2.5%) had 93.9%, 91.7%, and 89.3% sensitivity and 98.3%, 97.5%, and 97.8% negative-predictive value (NPV) for identifying category A gene PGV carriers in the development and validation cohorts, respectively. PREMMplus assessment (score ≥ 2.5%) had 89.9%, 85.6%, and 84.2% sensitivity and 95.0%, 93.5%, and 93.5% NPV, respectively, for identifying category A/B gene PGV carriers. Decision curve analyses support MGPT for individuals predicted to have ≥ 2.5% probability of a PGV. CONCLUSION PREMMplus accurately identifies individuals with PGVs in a diverse spectrum of cancer susceptibility genes with high sensitivity/NPV. Individuals with PREMMplus scores ≥ 2.5% should be considered for MGPT.

PURPOSE Physician workforce diversity can be a driver of institutional excellence, improving innovation and reducing health disparities. However, the current diversity of the hematology/oncology (HO) workforce does not reflect that of the US population. METHODS We conducted a cross-sectional online survey of current trainees and faculty within 5 years of completing terminal training in oncology specialties. RESULTS Of the 306 respondents, 64 (21%) were under-represented in medicine (URiM) and 161 (53%) identified as male. URiM participants were less likely to have a primary mentor (66%) than non-URiM participants (80%; P = .015). Among those who had a primary mentor, URiMs met less frequently (once every 3-6 months or less) with their mentor (19% v 7% non-URiM; P = .003). Furthermore, URiMs were more likely to report having mentors outside their own institution (47% v 40% non-URiM; P = .002) and making compromises to gain access to mentorship (36% v 23% non-URiM; P ≤ 0.001). URiMs were also less likely to apply for grants (34% v 42% non-URiM; P = .035) and awards (28% v 43% non-URiM; P = .019). In multivariable models, URiM individuals were more likely to make compromises to gain access to mentors (odds ratio [OR], 1.96; 95% CI, 1.01 to 3.82) and this remained significant for females (OR, 2.17; 95% CI, 1.26 to 3.75). CONCLUSION URiM individuals may be less likely to have effective mentorship and apply for awards and grant support. Understanding the challenges of URiM trainees can help shape training environments in academic medicine to ensure that they are grounded in diversity, inclusion, and retention.

The hazard ratio (HR) has been the most popular measure to quantify the magnitude of treatment effect on time-to-event outcomes in clinical research. However, the traditional Cox's HR approach has several drawbacks. One major issue is that there is no clear interpretation when the proportional hazards (PH) assumption does not hold, because the estimated HR is affected by study-specific censoring time distribution in non-PH cases. Another major issue is that the lack of a group-specific absolute hazard value in each group obscures the clinical significance of the magnitude of the treatment effect. Given these, we propose average hazard with survival weight (AH-SW) as a summary metric of event time distribution and will use difference in AH-SW (DAH-SW) or ratio of AH-SW (RAH-SW) to quantify the treatment effect magnitude. The AH-SW is interpreted as a person-time incidence rate that does not depend on random censoring. It is defined as the ratio of cumulative incidence probability and restricted mean survival time (RMST), which can be estimated non-parametrically. Numerical studies demonstrate that DAH-SW and RAH-SW offer almost identical power to Cox's HR-based tests under PH scenarios and can be more powerful for delayed-difference patterns often seen in immunotherapy trials. Like median and RMST differences, the proposed approach is a good model-free alternative to the HR-based approach for evaluating the treatment effect magnitude. Such a model-free measure will increase the likelihood that results from clinical studies are correctly interpreted and generalized to future populations.

PURPOSE Tumor testing for microsatellite instability and/or mismatch repair-deficiency (MSI/IHC) and clinical prediction models effectively screen for Lynch syndrome (LS)–associated colorectal cancer (CRC) and endometrial cancer (EC), but they have not been assessed for their ability to identify non-LS forms of inherited risk. The aim of this study was to compare MSI/IHC and the PREMM5 prediction model to identify carriers of LS and non-LS pathogenic variants (PVs) among patients with CRC and EC. PATIENTS AND METHODS Data were retrospectively analyzed from two single-institution cohorts: 706 patients with CRC and/or EC referred for genetic evaluation/testing (high-risk cohort) and 1,058 consecutively ascertained patients with CRC (oncology clinic cohort), unselected for familial risk. All participants underwent germline multigene panel testing. PREMM5 scores were calculated from personal/family cancer history. The primary outcome was the proportion of individuals with germline PVs (LS PVs, high-penetrance PVs, and any PVs) who had abnormal MSI/IHC testing and/or PREMM5 score ≥ 2.5%. RESULTS MSI/IHC and PREMM5 had comparable sensitivity for identifying LS carriers in high-risk (89.3% v 85.7%; P = .712) and oncology clinic patients (96.6% v 96.6%; P = 1.000), although MSI/IHC had significantly superior specificity for LS (81.3% v 20.1%; P < .001; 92.3% v 24.3%; P < .001). In both cohorts, PREMM5 had superior sensitivity to MSI/IHC at identifying patients with any high-penetrance PVs and any low-, moderate-, and high-penetrance PVs. Among patients with normal MSI/IHC, PREMM5 identified 84.2% and 83.3% of high-risk patients with CRC/EC and oncology clinic CRC patients with high-penetrance PVs, respectively. CONCLUSION MSI/IHC and PREMM5 effectively identify patients with CRC and/or EC with LS, although MSI/IHC has better specificity for LS. Because PREMM5 identifies non-LS, high-penetrance germline PVs, patients with CRC and/or EC with PREMM5 score ≥ 2.5%, including those with normal MSI/IHC, should be offered multigene panel testing.

Background: Uptake of genetic testing for cancer susceptibility in family members of cancer patients is suboptimal. The GENetic Education, Risk Assessment, and TEsting (GENERATE) study evaluated two strategies of remote genetic education and testing in relatives of pancreatic ductal adenocarcinoma (PDAC) patients. Methods: Eligible participants had: a first-degree relative with PDAC or had a known pathogenic germline variant (PGV) in one of thirteen PDAC predisposition genes (APC, ATM, BRCA1, BRCA2, CDKN2A, EPCAM, MLH1, MSH2, MSH6, PALB2, PMS2, STK11, or TP53) and a first- or second-degree relative with PDAC. Participants were cluster-randomized by family into one of two arms. Arm 1 included an interactive telemedicine session with a genetic counselor, followed by genetic testing at a commercial laboratory. Arm 2 involved remote online genetic education and testing at the commercial laboratory without the interactive session. The primary outcome was uptake of genetic testing across study arms, which was compared by permutation tests and mixed-effects logistic regression models. Results: Between 5/8/2019 and 6/1/2021, 424 families were randomized, including 601 participants (n=296 Arm 1; n=305 Arm 2). The uptake of genetic testing was 87% (257/296) in Arm 1 and 93% (284/305) in Arm 2 (p=0.014). Participants in Arm 1 were significantly less likely to obtain genetic testing compared to Arm 2 (Adjusted ratio [Arm1/Arm2] 0.90, 95% confidence interval 0.78-0.98). BRCA2, ATM, CDKN2A and PALB2 were the most common PDAC susceptibility genes in which PGVs were identified. Conclusions: Remote methods of genetic education and testing are successful alternatives to traditional germline susceptibility testing. Citation Format: Nicolette J. Rodriguez, C. Sloane Furniss, Matthew B. Yurgelun, Chinedu Ukaegbu, Pamela E. Constantinou, Ileana Fortes, Alyson Caruso, Alison N. Schwartz, Jill E. Stopfer, Meghan Underhill-Blazey, Barbara Kenner, Scott H. Nelson, Sydney Okumura, Alicia Y. Zhou, Tara B. Coffin, Hajime Uno, Miki Horiguchi, Allyson J. Ocean, Florencia McAllister, Andrew M. Lowy, Scott M. Lippman, Alison P. Klein, Lisa Madlensky, Gloria M. Petersen, Judy E. Garber, Michael G. Goggins, Anirban Maitra, Sapna Syngal. A randomized study of two Strategies of remote Genetic Education, Risk Assessment, and Testing (GENERATE) for family members of patients with pancreatic cancer [abstract]. In: Proceedings of the AACR Special Conference on Pancreatic Cancer; 2022 Sep 13-16; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2022;82(22 Suppl):Abstract nr A029.

11047 Background: Physician workforce diversity can be a driver of institutional excellence, improving innovation and reducing health disparities. The current diversity of the hematology/oncology (HO) workforce does not reflect that of the US population. The role of mentorship in increasing HO fellows’ interest in pursuing careers in HO has been described previously. However, the mentorship experiences of fellows and early career faculty from backgrounds underrepresented in medicine (UIM) in HO has not been fully characterized. Therefore, we compared the mentorship experiences of UIM and non-UIM trainees and early career faculty in HO subspecialties. Methods: We conducted cross-sectional online survey of HO subspecialists including current fellows and faculty within 5 years of end of training. The anonymous survey was distributed via email and social media channels in April 2020. Fisher’s exact test and multivariable logistic regression models were used to conduct comparisons between study groups. Results: Of the 306 respondents, 64 (21%) were UIM and 161 (53%) identified as male. UIM participants were less likely to have a primary mentor (66%) than non-UIM participants (80%, p = 0.015). Among those who had a primary mentor, UIMs were more likely to meet infrequently (greater than every 3 months, p = 0.007). Furthermore, UIMs were more likely to report having mentors outside their own institution (47% vs 40% non-UIM, p = 0.002) and making compromises to gain access to mentorship (36% vs 23% non-UIM, p = < 0.001). In addition, UIM participants were less likely to have an advisor (38% vs 54% non-UIM, p = 0.017), a coach (13% vs 20% non-UIM, p = 0.054), or a sponsor (19% vs 26% non-UIM, p = 0.046). UIMs were also less likely to apply for grants (34% vs 42% non-UIM, p = 0.035) and awards (28% vs 43%, non-UIM p = 0.019). In multivariable models, UIM individuals were more likely to make compromises to gain access to mentors (OR = 1.96, p = 0.047) and this remained significant for females (OR = 2.17, p = 0.005). Lastly, US born individuals had higher odds of having a primary mentor than non-US-born individuals (OR = 2.43, p = 0.004). Conclusions: In the largest study to date characterizing the mentorship experience of HO trainees and junior faculty, we found that UIM individuals were significantly less likely to find effective mentorship and were less likely to apply for awards and grant support. Understanding the challenges of UIM trainees can help shape training environments in academic medicine to ensure these are grounded in diversity and inclusion. Given the importance of workforce diversity, training programs in HO must consider structured programs and other innovations to improve mentoring experiences of UIM trainees and junior faculty.

PURPOSE: Tumor testing for microsatellite instability and/or mismatch repair-deficiency (MSI/IHC) and clinical prediction models effectively screen for Lynch syndrome (LS)–associated colorectal cancer (CRC) and endometrial cancer (EC), but they have not been assessed for their ability to identify non-LS forms of inherited risk. The aim of this study was to compare MSI/IHC and the PREMM(5) prediction model to identify carriers of LS and non-LS pathogenic variants (PVs) among patients with CRC and EC. PATIENTS AND METHODS: Data were retrospectively analyzed from two single-institution cohorts: 706 patients with CRC and/or EC referred for genetic evaluation/testing (high-risk cohort) and 1,058 consecutively ascertained patients with CRC (oncology clinic cohort), unselected for familial risk. All participants underwent germline multigene panel testing. PREMM(5) scores were calculated from personal/family cancer history. The primary outcome was the proportion of individuals with germline PVs (LS PVs, high-penetrance PVs, and any PVs) who had abnormal MSI/IHC testing and/or PREMM(5) score ≥ 2.5%. RESULTS: MSI/IHC and PREMM(5) had comparable sensitivity for identifying LS carriers in high-risk (89.3% v 85.7%; P = .712) and oncology clinic patients (96.6% v 96.6%; P = 1.000), although MSI/IHC had significantly superior specificity for LS (81.3% v 20.1%; P < .001; 92.3% v 24.3%; P < .001). In both cohorts, PREMM(5) had superior sensitivity to MSI/IHC at identifying patients with any high-penetrance PVs and any low-, moderate-, and high-penetrance PVs. Among patients with normal MSI/IHC, PREMM(5) identified 84.2% and 83.3% of high-risk patients with CRC/EC and oncology clinic CRC patients with high-penetrance PVs, respectively. CONCLUSION: MSI/IHC and PREMM(5) effectively identify patients with CRC and/or EC with LS, although MSI/IHC has better specificity for LS. Because PREMM(5) identifies non-LS, high-penetrance germline PVs, patients with CRC and/or EC with PREMM(5) score ≥ 2.5%, including those with normal MSI/IHC, should be offered multigene panel testing.

Background: More than 95% of recent cancer randomized controlled trials used the log-rank test to detect a treatment difference making it the predominant tool for comparing two survival functions. As with other tests, the log-rank test has both advantages and disadvantages. One advantage is that it offers the highest power against proportional hazards differences, which may be a major reason why alternative methods have rarely been employed in practice. The performance of statistical tests has traditionally been investigated both theoretically and numerically for several patterns of difference between two survival functions. However, to the best of our knowledge, there has been no attempt to compare the performance of various statistical tests using empirical data from past oncology randomized controlled trials. So, it is unknown whether the log-rank test offers a meaningful power advantage over alternative testing methods in contemporary cancer randomized controlled trials. Focusing on recently reported phase III cancer randomized controlled trials, we assessed whether the log-rank test gave meaningfully greater power when compared with five alternative testing methods: generalized Wilcoxon, test based on maximum of test statistics from multiple weighted log-rank tests, difference in t-year event rate, and difference in restricted mean survival time with fixed and adaptive [Formula: see text]. Methods: Using manuscripts from cancer randomized controlled trials recently published in high-tier clinical journals, we reconstructed patient-level data for overall survival (69 trials) and progression-free survival (54 trials). For each trial endpoint, we estimated the empirical power of each test. Empirical power was measured as the proportion of trials for which a test would have identified a significant result ( p value Results: For overall survival, t-year event rate offered the lowest (30.4%) empirical power and restricted mean survival time with fixed [Formula: see text] offered the highest (43.5%). The empirical power of the other types of tests was almost identical (36.2%–37.7%). For progression-free survival, the tests we investigated offered numerically equivalent empirical power (55.6%–61.1%). No single test consistently outperformed any other test. Conclusion: The empirical power assessment with the past cancer randomized controlled trials provided new insights on the performance of statistical tests. Although the log-rank test has been used in almost all trials, our study suggests that the log-rank test is not the only option from an empirical power perspective. Near universal use of the log-rank test is not supported by a meaningful difference in empirical power. Clinical trial investigators could consider alternative methods, beyond the log-rank test, for their primary analysis when designing a cancer randomized controlled trial. Factors other than power (e.g. interpretability of the estimated treatment effect) should garner greater consideration when selecting statistical tests for cancer randomized controlled trials.

In randomized clinical trials, the magnitude of the treatment effect is often reported using the hazard ratio (HR) even when the proportional hazards (PH) assumption is not met. Conducting numerical studies, this commentary illustrates how/why the HR estimate via the standard Cox's procedure is difficult to interpret even as an “average” treatment effect for non-PH cases.

In randomized clinical trials where time-to-event is the primary outcome, almost routinely, the logrank test is prespecified as the primary test and the hazard ratio is used to quantify treatment effect. If the ratio of 2 hazard functions is not constant, the logrank test is not optimal and the interpretation of hazard ratio is not obvious. When such a nonproportional hazards case is expected at the design stage, the conventional practice is to prespecify another member of weighted logrank tests, eg, Peto-Prentice-Wilcoxon test. Alternatively, one may specify a robust test as the primary test, which can capture various patterns of difference between 2 event time distributions. However, most of those tests do not have companion procedures to quantify the treatment difference, and investigators have fallen back on reporting treatment effect estimates not associated with the primary test. Such incoherence in the "test/estimation" procedure may potentially mislead clinicians/patients who have to balance risk-benefit for treatment decision. To address this, we propose a flexible and coherent test/estimation procedure based on restricted mean survival time, where the truncation time τ is selected data dependently. The proposed procedure is composed of a prespecified test and an estimation of corresponding robust and interpretable quantitative treatment effect. The utility of the new procedure is demonstrated by numerical studies based on 2 randomized cancer clinical trials; the test is dramatically more powerful than the logrank, Wilcoxon tests, and the restricted mean survival time-based test with a fixed τ, for the patterns of difference seen in these cancer clinical trials.

There exist a wide variety of statistical methods for clinical trials. Despite decades of clinical trial research, little is known about which methods are used in data analysis practice of the real world. This review describes the evidence of existing practices on survival data analysis and identifies potential opportunities to improve them.

Background & Aims Lynch syndrome (LS) is associated with increased risks of various gastrointestinal, gynecologic, genitourinary, and other cancers. Many clinical practice guidelines recommend that LS carriers' screening strategies be devised based on their family history of various cancers, in addition to age-, sex-, and gene-specific considerations. The aim of this study was to examine the association between family history and other clinical factors with LS carriers' histories of various cancers. Methods Two cohorts of LS carriers were analyzed: a laboratory-based cohort of consecutively ascertained individuals undergoing germline LS testing and a clinic-based cohort of LS carriers undergoing clinical care at an academic medical center. Multivariable logistic regression was performed to assess clinical factors associated with LS carriers' histories of various cancers/neoplasms. Familial burden was defined as LS carriers' aggregate number of first-/second-degree relatives with a history of a given malignancy. Results Multivariable analysis of the laboratory-based cohort (3828 LS carriers) identified familial burden as being incrementally associated with LS carriers' personal history of endometrial (odds ratio [OR], 1.37 per affected first-/second-degree relative; 95% confidence interval [CI], 1.21–1.56), urinary tract (OR, 2.72; 95% CI, 2.02–3.67), small bowel (OR, 3.17; 95% CI, 1.65–6.12), gastric (OR, 1.93; 95% CI, 1.24–3.02), and pancreaticobiliary cancers (OR, 2.10; 95% CI, 1.21–3.65) and sebaceous neoplasms (OR, 7.39; 95% CI, 2.71–20.15). Multivariable analysis of the clinic-based cohort (607 LS carriers) confirmed a significant association of familial burden of endometrial and urinary tract cancers. Conclusions Familial burden – in addition to age, sex, and specific LS gene – should be used to assess LS carriers' risks of specific cancers and guide decision-making about organ-specific surveillance.

Between-group comparison based on the restricted mean survival time (RMST) is getting attention as an alternative to the conventional logrank/hazard ratio approach for time-to-event outcomes in randomized controlled trials (RCTs). The validity of the commonly used nonparametric inference procedure for RMST has been well supported by large sample theories. However, we sometimes encounter cases with a small sample size in practice, where we cannot rely on the large sample properties. Generally, the permutation approach can be useful to handle these situations in RCTs. However, a numerical issue arises when implementing permutation tests for difference or ratio of RMST from two groups. In this article, we discuss the numerical issue and consider six permutation methods for comparing survival time distributions between two groups using RMST in RCTs setting. We conducted extensive numerical studies and assessed type I error rates of these methods. Our numerical studies demonstrated that the inflation of the type I error rate of the asymptotic methods is not negligible when sample size is small, and that all of the six permutation methods are workable solutions. Although some permutation methods became a little conservative, no remarkable inflation of the type I error rates were observed. We recommend using permutation tests instead of the asymptotic tests, especially when the sample size is less than 50 per arm.

1579 Background: Current clinical prediction models provide syndrome-specific numeric estimates of an individual's likelihood of having a specific hereditary cancer syndrome (e.g., PREMM5 for Lynch syndrome; BRCAPRO for BRCA1/2). With the emergence of multigene panel testing (MGPT), there is a need to evaluate individuals' risk of carrying a pathogenic variant in a diverse array of cancer susceptibility genes in parallel. This study’s aim was to develop and validate the PREMMplus clinical prediction model for multigene cancer risk assessment. Methods: PREMMplus was developed in a cohort of 7296 individuals who had undergone germline MGPT at a single center. Logistic regression models were used to examine candidate predictive variables – including age, sex, ethnicity, and personal/family history of cancer – to provide a numeric estimate of an individual’s likelihood of carrying a pathogenic/likely pathogenic germline variant in one of 18 cancer susceptibility genes (11 high- [ APC, BRCA1/2, CDH1, EPCAM, MLH1, MSH2, MSH6, biallelic MUTYH, PMS2, and TP53] and 7 moderate-penetrance [ ATM, CDKN2A, CHEK2, PALB2, PTEN, RAD51C, and RAD51D]). Model performance was validated in an independent dataset of 14845 individuals who had undergone MGPT at a commercial laboratory. Results: Using clinical characteristics, including personal/family history of 18 cancers plus colorectal adenoma burden, PREMMplus demonstrated an excellent ability to predict pathogenic variants in high penetrance genes at 90% sensitivity. PREMMplus had acceptable performance with the addition of 7 moderate penetrance genes. PREMMplus was well-calibrated and demonstrated comparable performance in the external validation dataset. Conclusions: PREMMplus is the first validated risk assessment model to quantify an individual’s likelihood of carrying pathogenic variants in a wide diversity of cancer risk genes, and can be used to select individuals who should undergo MGPT. As expected, PREMMplus’s discriminatory capacity was reduced with the inclusion of moderate penetrance cancer risk genes. [Table: see text]

This study compares conventional hazard ratios and an alternative, clinically interpretable approach to quantify the long-term survival benefit in patients receiving chemotherapy with immunotherapy.