Using Machine Learning to Leverage Biomarker Change and Predict Colorectal Cancer Recurrence.

PURPOSE: The risk of colorectal cancer (CRC) recurrence after primary treatment varies across individuals and over time. Using patients' most up-to-date information, including carcinoembryonic antigen (CEA) biomarker profiles, to predict risk could improve personalized decision making. METHODS: We used electronic health record data from an integrated health system on a cohort of patients diagnosed with American Joint Committee on Cancer stage I-III CRC between 2008 and 2013 (N = 3,970) and monitored until recurrence or end of follow-up. We addressed missingness in recurrence outcomes and longitudinal CEA measures, and engineered CEA features using current and past biomarker values for inclusion in a risk prediction model. We used a discrete time Superlearner model to evaluate various algorithms for predicting recurrence. We evaluated the time-varying discrimination and calibration of the algorithms and assessed the role of individual predictors. RESULTS: Recurrence was documented in 448 (11.3%) patients. XGBoost with depth = 1 (XGB-D1) predicted recurrence substantially better than all other algorithms at all time points, with AUC ranging from 0.87 (95% CI, 0.86 to 0.88) at 6 months to 0.94 (95% CI, 0.92 to 0.96) at 54 months. The only variable used by XGB-D1 was 6-month change in log CEA. Predicted 1-year risk of recurrence was nearly zero for patients whose log CEA did not increase in the last 6 months, between 12.2% and 34.1% for patients whose log CEA increased between 0.10 and 0.40, and 43.6% for those with a log CEA increase >0.40. Compared with XGB, penalized regression approaches (lasso, ridge, and elastic net) performed poorly, with AUCs ranging from 0.58 to 0.69. CONCLUSION: A flexible, machine learning approach that incorporated longitudinal CEA information yielded a simple and high-performing model for predicting recurrence on the basis of 6-month change in log CEA.

Underutilization or appropriate care? Assessing adjuvant chemotherapy use and survival in 3 heterogenous subpopulations with stage II/III colorectal cancer within a large integrated health system.

BACKGROUND: Clinical guidelines have recommended adjuvant chemotherapy (ACT) for patients with high-risk stage II colon cancer, although the survival benefit is unclear. ACT is also recommended for patients with stage III colon cancer to reduce the risk of recurrence and mortality. For stage II/III rectal cancer, however, the role of perioperative chemotherapy (PCT, adjuvant or neoadjuvant) remains controversial, resulting in substantial variation in its use in clinical practice. OBJECTIVES: To understand real-world use and predictors of ACT or PCT use and survival outcomes in 3 heterogeneous patient groups with colorectal cancer (CRC), and to inform the evidence gap between guideline-based care and clinical practice. METHODS: This retrospective cohort study included patients with an initial stage II/III CRC diagnosis between 2008 and 2013 identified from Kaiser Permanente Southern California electronic health record databases. Patients were eligible if they were aged 18-74 years at diagnosis and received primary curative surgery. We fitted mixed effects logistic regression models to evaluate predictors of ACT receipt and Cox proportional hazards models on propensity score-matched (PS-matched) samples to assess the association between ACT/PCT receipt and survival. RESULTS: We included 1,690 patients with colon cancer (stage II: 820 and stage III: 870) and 587 patients with rectal cancer (stage II: 241 and stage III: 346). We found that 65% of patients with high-risk stage II colon cancer, 15% of those with stage III colon, and 15% of those with stage II/III rectal cancer did not receive ACT/PCT. Patients with stage II colon cancer with T4 stage (odds ratio [OR] = 5.79, 95% CI = 3.33 - 10.06) and a lower comorbidity score were more likely to receive ACT (high vs low Charlson score: OR = 0.69, 95% CI = 0.55 - 0.87). Patients with stage III rectal cancer were more likely to receive PCT than those with stage II disease (OR = 7.85, 95% CI = 2.07 - 29.74). Patients with another cancer diagnosis prior to CRC diagnosis were less likely to receive PCT (OR = 0.37, 95% CI = 0.16 - 0.85). ACT/PCT use was associated with improved overall survival among patients with high-risk stage II colon cancer (PS-matched hazard ratio [HR] = 0.42, 95% CI = 0.25 - 0.70) and those with stage III CRC (stage III colon: PS-matched HR = 0.3, 95% CI = 0.25 - 0.36; stage III rectal: PS-matched HR = 0.2, 95% CI = 0.13 - 0.31). CONCLUSIONS: We found potential underuse of appropriate chemotherapy treatment in patients with high-risk stage II colon cancer and stage III CRC. Clinicians' and providers' decisions on ACT administration may not be fully guided by the risk of recurrence and 5-year survival benefits in stage II colon cancer. DISCLOSURES: This research was supported by the National Cancer Institute of the National Institutes of Health (NIH) (under R37-CA218413). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

Racial and Ethnic Bias in Risk Prediction Models for Colorectal Cancer Recurrence When Race and Ethnicity Are Omitted as Predictors.

Importance: Including race and ethnicity as a predictor in clinical risk prediction algorithms has received increased scrutiny, but there continues to be a lack of empirical studies addressing whether simply omitting race and ethnicity from the algorithms will ultimately affect decision-making for patients of minoritized racial and ethnic groups. Objective: To examine whether including race and ethnicity as a predictor in a colorectal cancer recurrence risk algorithm is associated with racial bias, defined as racial and ethnic differences in model accuracy that could potentially lead to unequal treatment. Design, Setting, and Participants: This retrospective prognostic study was conducted using data from a large integrated health care system in Southern California for patients with colorectal cancer who received primary treatment between 2008 and 2013 and follow-up until December 31, 2018. Data were analyzed from January 2021 to June 2022. Main Outcomes and Measures: Four Cox proportional hazards regression prediction models were fitted to predict time from surveillance start to cancer recurrence: (1) a race-neutral model that explicitly excluded race and ethnicity as a predictor, (2) a race-sensitive model that included race and ethnicity, (3) a model with 2-way interactions between clinical predictors and race and ethnicity, and (4) separate models by race and ethnicity. Algorithmic fairness was assessed using model calibration, discriminative ability, false-positive and false-negative rates, positive predictive value (PPV), and negative predictive value (NPV). Results: The study cohort included 4230 patients (mean [SD] age, 65.3 [12.5] years; 2034 [48.1%] female; 490 [11.6%] Asian, Hawaiian, or Pacific Islander; 554 [13.1%] Black or African American; 937 [22.1%] Hispanic; and 2249 [53.1%] non-Hispanic White). The race-neutral model had worse calibration, NPV, and false-negative rates among racial and ethnic minority subgroups than non-Hispanic White individuals (eg, false-negative rate for Hispanic patients: 12.0% [95% CI, 6.0%-18.6%]; for non-Hispanic White patients: 3.1% [95% CI, 0.8%-6.2%]). Adding race and ethnicity as a predictor improved algorithmic fairness in calibration slope, discriminative ability, PPV, and false-negative rates (eg, false-negative rate for Hispanic patients: 9.2% [95% CI, 3.9%-14.9%]; for non-Hispanic White patients: 7.9% [95% CI, 4.3%-11.9%]). Inclusion of race interaction terms or using race-stratified models did not improve model fairness, likely due to small sample sizes in subgroups. Conclusions and Relevance: In this prognostic study of the racial bias in a cancer recurrence risk algorithm, removing race and ethnicity as a predictor worsened algorithmic fairness in multiple measures, which could lead to inappropriate care recommendations for patients who belong to minoritized racial and ethnic groups. Clinical algorithm development should include evaluation of fairness criteria to understand the potential consequences of removing race and ethnicity for health inequities.

Racial Disparities in the Ascertainment of Cancer Recurrence in Electronic Health Records.

PURPOSE: There is growing interest in using computable phenotypes or proxies to identify important clinical outcomes, such as cancer recurrence, in rich electronic health records data. However, the race/ethnicity-specific accuracies of these proxies remain unclear. We examined whether the accuracy of a proxy for colorectal cancer (CRC) recurrence differed by race/ethnicity and the possible mechanisms that drove the differences. METHODS: Using data from a large integrated health care system, we identified a stratified random sample of 282 Black/African American (AA), Hispanic, and non-Hispanic White (NHW) patients with CRC who received primary treatment. Patient 5-year recurrence status was estimated using a utilization-based proxy and evaluated against the true recurrence status obtained using detailed chart review and by race/ethnicity. We used covariate-adjusted probit regression models to estimate the associations between race/ethnicity and misclassification. RESULTS: The recurrence proxy had excellent overall accuracy (positive predictive value [PPV] 89.4%; negative predictive value 96.5%; mean difference in timing 1.96 months); however, accuracy varied by race/ethnicity. Compared with NHW patients, PPV was 14.9% lower (95% CI, 2.53 to 28.6) among Hispanic patients and 4.3% lower (95% CI, -4.8 to 14.8) among Black/AA patients. The proxy disproportionately inflated the 5-year recurrence incidence for Hispanic patients by 10.6% (95% CI, 4.2 to 18.2). Compared with NHW patients, proxy recurrences for Hispanic patients were almost three times as likely to have been misclassified as positive (adjusted risk ratio 2.91 [95% CI, 1.21 to 8.31]). Higher false positives among racial/ethnic minorities may be related to higher prevalence of noncancerous lung-related problems and substantial delays in primary treatment because of insufficient patient-provider communication and abnormal treatment patterns. CONCLUSION: Using a proxy with worse accuracy among racial/ethnic minority patients to estimate population health may misdirect resources and support erroneous conclusions around treatment benefit for these patients.

Symptom Burden Is Lower in Asian and Pacific Islander and Black Men Admitted to Home-Based Palliative Care in an Integrated Health Care System.

Background: Little is known about racial/ethnic differences in symptom severity among patients receiving home-based palliative care (HomePal). Objectives: To determine whether symptom severity differs between White patients and patients of color receiving HomePal and whether gender moderates the difference. Design: This is a cross-sectional exploratory study. Setting/Subjects: Baseline data were from 2090 patients receiving HomePal in Kaiser Permanente Southern California. Measurements: Multivariable median regression analyses were carried out across race/ethnicity groups and stratified by gender to assess differences in Edmonton Symptom Assessment System (ESAS) scores at HomePal admission. Results: Asian/Pacific Islander men and Black men had lower ESAS scores compared than White men (-5 [-7.8, -2.2], p = 0.0005 and -5.4 [-8.7, -2.1], p = 0.001, respectively); there were marginal ESAS differences across race/ethnic groups for women. Conclusion: Patients of color reported lower symptom severity than White patients. More research is needed to understand how the intersection of culture and gender affects symptom experience and reporting in patients living with serious illness. Trial Registration: ClinicalTrials.gov: NCT#03694431.

Perceptions and Utilization of Lung Cancer Screening Among Smokers Enrolled in a Tobacco Cessation Program.

BACKGROUND: Although lung cancer screening (LCS) with low-dose computed tomography (LDCT) reduces lung cancer mortality in high-risk patients, most of those eligible are not referred for screening. Tobacco cessation counseling may be an opportune time to educate people about LCS, but little is known about the utilization and perceptions of LCS among people undergoing tobacco cessation treatment. MATERIALS AND METHODS: We surveyed 185 current smokers, including 122 smokers between the ages of 55 and 80 years, who were attending a tobacco cessation class in a large integrated health care system regarding lung cancer risk perception and perceived benefits, harms, and barriers to LCS. We analyzed results according to whether participants had already undergone LCS and also whether they had undergone colorectal cancer screening. RESULTS: A minority (18.9%) of participants had undergone LCS, and no participant who had not undergone LCS was familiar with LCS. Perceived lung cancer risk was high, and screening was believed to be beneficial. Common barriers included being a current smoker (56.6%), worrying about test results (52.5%), lack of knowledge about the test (50.8%), absence of symptoms of lung cancer (40.2%), costs of the study (35.2%), and worrying about being blamed for having smoked (33.6%). Perceived risk and barriers to LDCT were similar among people who had or had not previously undergone colorectal cancer screening. CONCLUSIONS: Utilization of LCS was low, and few smokers were aware of LDCT for LCS. A number of patient-related barriers to screening exist among smokers. Tobacco cessation counseling may be an opportune time to provide education regarding LCS with LDCT.