Biochemical progression free and overall survival among Black men with stage IV prostate cancer in South Africa: Results from a prospective cohort study.

BACKGROUND: Men of African descent are disproportionately affected by prostate cancer (PCa), and many have metastatic disease at presentation. In South Africa (SA), androgen deprivation therapy (ADT) is the first-line treatment for stage IV PCa. OBJECTIVE: To identify predictors of overall survival (OS) in Black South African men with stage IV PCa treated with ADT. DESIGN, SETTING, AND PARTICIPANTS: Men diagnosed with prostate cancer (3/22/2016-10/30/2020) at Chris Hani Baragwanath Academic Hospital in Soweto, Johannesburg, were recruited for the Men of African Descent with Cancer of the Prostate study. We included men with newly diagnosed stage IV PCa treated with ADT who had a prostate-specific antigen (PSA) level drawn prior to initiation of ADT and had ≥1 PSA drawn ≥12 weeks after ADT start. OUTCOMES MEASURES AND STATISTICAL ANALYSIS: We used Kaplan-Meier statistics to estimate OS and Cox regression models to identify predictors of OS. RESULTS AND LIMITATIONS: Of the 1097 men diagnosed with prostate cancer, we included 153 men with stage IV PCa who received ADT and met PSA requirements. The median age was 68.0 years (interquartile range 64-73 years). Median OS from time of ADT initiation was 3.39 years (95% confidence interval (CI): 3.14%-noncalculable), while biochemical progression-free survival was 2.36 years (95% CI: 2.03%-3.73%). Biochemical progression (HR 3.52, 95% CI: 1.85%-6.70%), PSA nadir level >4 ng/mL (HR 3.77, 95% CI: 1.86%-7.62%), alkaline phosphatase level at diagnosis >150 IU/dL (HR 3.09, 95% CI: 1.64%-5.83%), and hemoglobin at diagnosis <13.5 g/dL (HR 2.90, 95% CI: 1.28%-6.56%) were associated with worse OS. CONCLUSIONS: In this study, we identified factors associated with poor OS among Black South African men with stage IV PCa treated with ADT. These factors may be useful in identifying patients for upfront treatment escalation, including the use of docetaxel chemotherapy or escalation of therapy at the time of biochemical progression. PATIENT SUMMARY: In this study, we found that high alkaline phosphatase level, anemia at diagnosis, and high PSA nadir after initiation of androgen deprivation therapy are associated with worse overall survival among Black South African men treated with androgen deprivation therapy for metastatic prostate cancer.

Cost-Effectiveness of Liquid Biopsy for Colorectal Cancer Screening in Patients Who Are Unscreened.

Importance: Despite recommendations for universal screening, adherence to colorectal cancer screening in the US is approximately 60%. Liquid biopsy tests are in development for cancer early detection, but it is unclear whether they are cost-effective for colorectal cancer screening. Objective: To estimate the cost-effectiveness of liquid biopsy for colorectal cancer screening in the US. Design, Setting, and Participants: In this economic evaluation, a Markov model was developed to compare no screening and 5 colorectal cancer screening strategies: colonoscopy, liquid biopsy, liquid biopsy following nonadherence to colonoscopy, stool DNA, and fecal immunochemical test. Adherence to first-line screening with colonoscopy, stool DNA, or fecal immunochemical test was assumed to be 60.6%, and adherence for liquid biopsy was assumed to be 100%. For colonoscopy, stool DNA, and fecal immunochemical test, patients who did not adhere to testing were not offered other screening. In colonoscopy-liquid biopsy hybrid, liquid biopsy was second-line screening for those who deferred colonoscopy. Scenario analyses were performed to include the possibility of polyp detection for liquid biopsy. Exposures: No screening, colonoscopy, fecal immunochemical test, stool DNA, liquid biopsy, and colonoscopy-liquid biopsy hybrid screening. Main Outcomes and Measures: Model outcomes included life expectancy, total cost, and incremental cost-effectiveness ratios. A strategy was considered cost-effective if it had an incremental cost-effectiveness ratio less than the US willingness-to-pay threshold of $100 000 per life-year gained. Results: This study used a simulated cohort of patients aged 45 years with average risk of colorectal cancer. In the base case, colonoscopy was the preferred, or cost-effective, strategy with an incremental cost-effectiveness ratio of $28 071 per life-year gained. Colonoscopy-liquid biopsy hybrid had the greatest gain in life-years gained but had an incremental cost-effectiveness ratio of $377 538. Colonoscopy-liquid biopsy hybrid had a greater gain in life-years if liquid biopsy could detect polyps but remained too costly. Conclusions and Relevance: In this economic evaluation of liquid biopsy for colorectal cancer screening, colonoscopy was a cost-effective strategy for colorectal cancer screening in the general population, and the inclusion of liquid biopsy as a first- or second-line screening strategy was not cost-effective at its current cost and screening performance. Liquid biopsy tests for colorectal cancer screening may become cost-effective if their cost is substantially lowered.

The Impact of HIV Infection on Neoadjuvant and Adjuvant Chemotherapy Relative Dose Intensity in South African Patients with Breast Cancer.

INTRODUCTION: In the South African Breast Cancer and HIV Outcomes (SABCHO) study, we previously found that breast cancer patients living with HIV and treated with neoadjuvant chemotherapy achieve lower rates of complete pathologic response than patients without HIV. We now assess the impact of comorbid HIV on receipt of timely and complete neoadjuvant and adjuvant chemotherapy. MATERIALS AND METHODS: Since June 2015, the SABCHO study has collected data on women diagnosed with breast cancer at 6 South African hospitals. We selected a sample of participants with stages I-III cancer who received ≥2 doses of neoadjuvant or adjuvant chemotherapy. Data on chemotherapies prescribed and received, filgrastim receipt, and laboratory values measured during treatment were captured from patients' medical records. We calculated the mean relative dose intensity (RDI) for all prescribed chemotherapies. We tested for association between full regimen RDI and HIV status, using linear regression to control for demographic and clinical covariates, and for association of HIV with laboratory abnormalities. RESULTS: The 166 participants living with HIV and 159 without HIV did not differ in median chemotherapy RDI: 0.89 (interquartile range (IQR) 0.77-0.95) among those living with HIV and 0.87 (IQR 0.77-0.94) among women without HIV. Patients living with HIV experienced more grade 3+ anemia and leukopenia than those without HIV (anemia: 10.8% vs. 1.9%, P = .001; leukopenia: 8.4% vs. 1.9%, P = .008) and were more likely to receive filgrastim (24.7% vs. 10.7%, P = .001). CONCLUSIONS: HIV status did not impact neoadjuvant or adjuvant chemotherapy RDI, although patients with breast cancer living with HIV experienced more myelotoxicity during treatment.

Deep learning on time series laboratory test results from electronic health records for early detection of pancreatic cancer.

The multi-modal and unstructured nature of observational data in Electronic Health Records (EHR) is currently a significant obstacle for the application of machine learning towards risk stratification. In this study, we develop a deep learning framework for incorporating longitudinal clinical data from EHR to infer risk for pancreatic cancer (PC). This framework includes a novel training protocol, which enforces an emphasis on early detection by applying an independent Poisson-random mask on proximal-time measurements for each variable. Data fusion for irregular multivariate time-series features is enabled by a "grouped" neural network (GrpNN) architecture, which uses representation learning to generate a dimensionally reduced vector for each measurement set before making a final prediction. These models were evaluated using EHR data from Columbia University Irving Medical Center-New York Presbyterian Hospital. Our framework demonstrated better performance on early detection (AUROC 0.671, CI 95% 0.667 - 0.675, p < 0.001) at 12 months prior to diagnosis compared to a logistic regression, xgboost, and a feedforward neural network baseline. We demonstrate that our masking strategy results greater improvements at distal times prior to diagnosis, and that our GrpNN model improves generalizability by reducing overfitting relative to the feedforward baseline. The results were consistent across reported race. Our proposed algorithm is potentially generalizable to other diseases including but not limited to cancer where early detection can improve survival.

The Impact of Breast Cancer Treatment Delays on Survival Among South African Women.

BACKGROUND: In high-income settings, delays from breast cancer (BC) diagnosis to initial treatment worsen overall survival (OS). We examined how time to BC treatment initiation (TTI) impacts OS in South Africa (SA). METHODS: We evaluated women enrolled in the South African BC and HIV Outcomes study between July 1, 2015 and June 30, 2019, selecting women with stages I-III BC who received surgery and chemotherapy. We constructed a linear regression model estimating the impact of sociodemographic and clinical factors on TTI and separate multivariable Cox proportional hazard models by first treatment (surgery and neoadjuvant chemotherapy (NAC)) assessing the effect of TTI (in 30-day increments) on OS. RESULTS: Of 1260 women, 45.6% had upfront surgery, 54.4% had NAC, and 19.5% initiated treatment >90 days after BC diagnosis. Compared to the surgery group, more women in the NAC group had stage III BC (34.8% vs 81.5%). Living further away from a hospital and having hormone receptor positive (vs negative) BC was associated with longer TTI (8 additional days per 100 km, P = .003 and 8 additional days, P = .01, respectively), while Ki67 proliferation index >20 and upfront surgery (vs NAC) was associated with shorter TTI (12 and 9 days earlier; P = .0001 and.007, respectively). Treatment initiation also differed among treating hospitals (P < .0001). Additional 30-day treatment delays were associated with worse survival in the surgery group (HR 1.11 [95%CI 1.003-1.22]), but not in the NAC group. CONCLUSIONS: Delays in BC treatment initiation are common in SA public hospitals and are associated with worse survival among women treated with upfront surgery.

Impact of HIV infection on survival among women with stage I-III breast cancer: Results from the South African breast cancer and HIV outcomes study.

In some countries of sub-Saharan Africa, the prevalence of HIV exceeds 20%; in South Africa, 20.4% of people are living with HIV. We examined the impact of HIV infection on the overall survival (OS) of women with nonmetastatic breast cancer (BC) enrolled in the South African Breast Cancer and HIV Outcomes (SABCHO) study. We recruited women with newly diagnosed BC at six public hospitals from 1 July 2015 to 30 June 2019. Among women with stages I-III BC, we compared those with and without HIV infection on sociodemographic, clinical, and treatment factors. We analyzed the impact of HIV on OS using multivariable Cox proportional hazard models. Of 2367 women with stages I-III BC, 499 (21.1%) had HIV and 1868 (78.9%) did not. With a median follow-up of 29 months, 2-year OS was poorer among women living with HIV (WLWH) than among HIV-uninfected women (72.4% vs 80.1%, P < .001; adjusted hazard ratio (aHR) 1.49, 95% confidence interval (CI) = 1.22-1.83). This finding was consistent across age groups ≥45 years and <45 years, stage I-II BC and stage III BC, and ER/PR status (all P < .03). Both WLWH with <50 viral load copies/mL and WLWH with ≥50 viral load copies/mL had poorer survival than HIV-uninfected BC patients [aHR: 1.35 (1.09-1.66) and 1.54 (1.20-2.00), respectively], as did WLWH who had ≥200 CD4+ cells/mL at diagnosis [aHR: 1.39 (1.15-1.67)]. Because receipt of antiretroviral therapy has become widespread, WLWH is surviving long enough to develop BC; more research is needed on the causes of their poor survival.

The Impact of HIV on Non-AIDS defining gastrointestinal malignancies: A review.

BACKGROUND: Cancer is the leading cause of morbidity and mortality among people living with HIV (PLWH). Although gastrointestinal (GI) cancers are not associated with HIV, their incidence is rising among PLWH, and yet little is known about how HIV affects their presentation, treatment and outcomes. METHODS: We searched PubMed using "HIV" and "cancer", "esophageal cancer", "gastric cancer", "stomach cancer", "gastroesophageal cancer", "colorectal cancer", "colon cancer", or "rectal cancer". We included studies comparing an HIV-positive group (n ≥ 4) to an HIV-negative group, with respect to clinical presentation, treatment, or mortality of GI cancers. RESULTS: Of 18 articles that met inclusion criteria, 17 were retrospective, and 13 described patients in the United States. At diagnosis with colorectal, but not pancreatic, gastric, or esophageal cancer, PLWH were younger than patients who were HIV-negative. PLWH did not present with more advanced stage GI cancers than patients who were HIV-negative. Compared to HIV-negative controls, PLWH with colorectal cancer had a higher proportion of right-sided versus left-sided colon cancers and a higher proportion of rectal versus colon cancers. Among patients diagnosed with colorectal or pancreatic cancer, PLWH were less likely to receive cancer treatment than other patients; no studies examined the association of HIV status with treatment for esophageal or gastric cancer. PLWH with GI malignancies had higher all-cause mortality compared to patients who were HIV-negative, but evidence for cancer-specific mortality was limited and mixed. CONCLUSION: PLWH with GI malignancies were less likely to receive cancer treatment and had higher all-cause mortality than patients who were HIV-negative. Most of the studies focused on colorectal cancer; more studies are needed in pancreatic, gastric and esophageal cancer. Future studies should investigate the effects of HIV on cancer-specific mortality, especially among patients in low- and middle-income countries, including those with high HIV prevalence.

Impact of HIV infection on overall survival among women with stage IV breast cancer in South Africa.

PURPOSE: Advanced breast cancer (BC) at diagnosis is common in sub-Saharan Africa (SSA), including among women living with HIV (WLWH). In public hospitals across South Africa (SA), 10-15% of women present with stage IV BC, compared to < 5% in the United States (US); 20% of new BC diagnoses in SA are in WLWH. We evaluated the impact of HIV on overall survival (OS) among women with stage IV BC. METHODS: We conducted a prospective cohort study of women diagnosed with stage IV BC between February 2, 2015 and September 18, 2019 at six public hospitals in SA. Multivariate Cox regression models were used to estimate the association between HIV status and OS. RESULTS: Among 550 eligible women, 147 (26.7%) were WLWH. Compared to HIV-negative BC patients, WLWH were younger (median age 45 vs. 60 years, p < 0.001), predominantly black (95.9% vs. 77.9%, p < 0.001), and more likely to have hormone receptor-negative (hormone-negative) BC (32.7% vs. 22.6%, p = 0.016). Most women received systemic cancer-directed therapy (80.1%). HIV status was not associated with treatment or OS (Hazard Ratio (HR) 1.13 [95%CI 0.89-1.44]). On exploratory subgroup analysis, WLWH and hormone-negative BC had shorter OS compared to HIV-uninfected women (1-year OS: 27.1% vs. 48.8%, p = 0.003; HR 1.94 [95%CI 1.27-2.94]; p = 0.002), which was not observed for hormone receptor-positive BC. CONCLUSION: HIV status was not associated with worse OS in women with stage IV BC in SA and cannot account for the poor survival in this cohort. Subgroup analysis revealed that WLWH with hormone-negative BC had worse OS, which warrants further investigation.