Biomonitoring of Exposures to Solvents and Metals in Electronics Manufacturing Facilities in Batam, Indonesia

Concerns about chemical exposure in the electronics manufacturing industry have long been recognized, but data are lacking in Southeast Asia. We conducted a study in Batam, Indonesia, to evaluate chemical exposures in electronics facilities, using participatory research and biological monitoring approaches. A convenience sample of 36 workers (28 exposed, 8 controls) was recruited, and urine samples were collected before and after shifts. Five solvents (acetone, methyl ethyl ketone, toluene, benzene, and xylenes) were found in 46%-97% of samples, and seven metals (arsenic, cadmium, cobalt, tin, antimony, lead, and vanadium) were detected in 60%-100% of samples. Biological monitoring and participatory research appeared to be useful in assessing workers' exposure when workplace air monitoring is not feasible due to a lack of cooperation from the employer. Several logistical challenges need to be addressed in future biomonitoring studies of electronics workers in Asia in factories where employers are reluctant to track workers' exposure and health.

Increasing Use of Conservative Management of Low-Risk Prostate Cancer in the Veterans Affairs System from 2012 to 2021

PURPOSE/OBJECTIVE(S)

Conservative management (CM), including either active surveillance or watchful waiting, is the preferred management strategy for men with low-risk prostate cancer. We sought to describe contemporary trends in CM among men with localized prostate cancer across the equal-access Veterans Affairs (VA) system.

MATERIALS/METHODS

Using VA tumor registry data, we identified all men with a new diagnosis of biopsy-proven, clinically node-negative, non-metastatic prostatic adenocarcinoma from 2012 to 2021, excluding men with missing National Comprehensive Cancer Network risk group. We quantified the yearly proportion of patients with low and favorable intermediate-risk (FIR) prostate cancer undergoing conservative management and examined age-specific rates. Multivariable logistic regression was used to identify independent clinical predictors of CM among men with low-risk and FIR disease.

RESULTS

The cohort included 86,415 patients with localized prostate cancer, of whom 20,290 (23.5%) had low risk and 25,447 (29.5%) had FIR disease. The proportion of men diagnosed with low-risk disease decreased from 27.7% in 2012 to 20.1% in 2021. The proportion of men with localized prostate cancer who were 70 years or older at diagnosis increased from 19.5% in 2012 to 46.4% in 2021, with similar increases seen across risk groups. The proportion of men with low-risk disease undergoing CM increased from 63.1% in 2012 to 86.6% in 2021. CM rates in 2021 among low-risk patients were largely consistent across age groups (50-59: 85.1%; 60-69: 87.4%; 70+: 86.3%). For FIR patients, CM rates increased among the subset with Gleason 6 disease (49.5% in 2012 to 80.7% in 2021). In multivariable models, independent predictors of CM in low-risk patients included older age, more recent year, lower PSA at diagnosis, North Atlantic or Pacific region, and presence of another cancer at diagnosis; similar predictors were found in FIR patients. Self-reported race (Black, White, or Other) was not associated with CM. While there was wide geographic variation in CM rates among low-risk patients early in the study period (2012-2014: 56.1% in Southeast region vs 74.1% in Pacific), these disparities resolved in recent years (2019-2021: 83.2% in Southeast vs 83.8% in Pacific).

CONCLUSION

Rates of conservative management for localized prostate cancer increased dramatically over time in the VA, with over 85% of low-risk patients managed conservatively in 2021. Recent years showed no differences in CM rates by self-reported race, geographic region, or age group. Despite this progress, we observed a concomitant increase in the proportion of men 70 years or older diagnosed with localized prostate cancer, including low risk disease. This raises a need to consider strategies to reduce the diagnosis of low-risk disease in elderly Veterans.

Association of Proton Pump Inhibitor Use and Severe Pneumonitis in Stage III Non-Small-Cell Lung Cancer Treated with Primary Chemoradiation and Adjuvant Durvalumab

PURPOSE/OBJECTIVE(S)

Prior studies have suggested gut microbiome changes induced by long-term acid-reducing medication use could modulate immunotherapy efficacy and toxicity. We assessed the relationship between baseline acid-reducing medication use (proton pump inhibitors [PPI] or H2 antagonists [H2A]) on treatment-related toxicity and efficacy in a large cohort of stage III non-small-cell lung cancer (NSCLC) patients treated with or without immunotherapy.

MATERIALS/METHODS

Patients with unresectable stage III NSCLC treated with primary concurrent chemoradiation with or without adjuvant durvalumab from 2015 to 2021 were identified in the Veterans Affairs (VA) system. We defined baseline acid-reducing medication use with VA and non-VA pharmacy records in the year prior to radiation start; the number of prescriptions and the cumulative duration of therapy were quantified. Using multivariable Cox models adjusting for potential baseline confounders and stratified by adjuvant durvalumab use, we estimated the association between PPI or H2A use and subsequent severe pneumonitis, progression-free survival, and overall survival. Pneumonitis was determined and graded by manual chart review.

RESULTS

We included 1994 patients with stage III NSCLC treated with primary chemoradiation, of whom 1005 (50%) received adjuvant durvalumab, 1064 (53%) received any PPI and 1030 (52%) received any H2A. In the overall sample, baseline use of any PPI was associated with increased risk of grade 3-5 pneumonitis (adjusted hazard ratio [aHR] 1.53, 95% CI 1.12-2.10, p = 0.008) and was found to be significant only in durvalumab-treated patients (aHR 1.67, 95% CI 1.10-2.54, p = 0.016), but not for patients treated without durvalumab (aHR 1.34, 95% CI 0.82-2.20, p = 0.2). Higher number of PPI prescriptions were associated with increased risk of severe pneumonitis (aHR 1.38 per 5 prescriptions, 95% CI 1.03-1.85, p = 0.03) and longer duration of PPIs trended toward significance (aHR 1.04 per 90 days, 95% CI 1.00-1.09, p = 0.066). Any PPI use was associated with worse OS in durvalumab-treated patients (aHR 1.30, 95% CI 1.05-1.61, p = 0.016) but not for patients without durvalumab use (aHR 0.99, 95% CI 0.86-1.14, p = 0.9). PPI use had no association with progression-free survival in either cohort. No significant associations between baseline H2A use and pneumonitis, PFS, or OS in either cohort were seen.

CONCLUSION

PPIs, but not H2As, are associated with increased risk of treatment-related pneumonitis and inferior OS in stage III NSCLC patients treated with chemoradiation and immunotherapy. This association was not observed among patients treated with chemoradiation alone. No association was found between PPI use and cancer progression. Further work is warranted to confirm these findings in other immunotherapy-treated cohorts.

Symptomatic Pneumonitis Rates Among Patients with Stage III Non-Small Cell Lung Cancer Receiving Adjuvant Durvalumab

PURPOSE/OBJECTIVE(S)

Adjuvant durvalumab after definitive chemoradiation for stage III non-small cell lung cancer (NSCLC) has improved overall survival with low reported pneumonitis rates in clinical trials. However, real-world rates of treatment-associated pneumonitis remain poorly defined. We sought to describe pneumonitis rates among a largest-to-date real-world cohort of stage III NSCLC patients receiving adjuvant durvalumab.

MATERIALS/METHODS

We performed a retrospective cohort study of patients with stage III NSCLC in the national Veterans Health Administration who received concurrent chemoradiation alone from 2015-2016 or concurrent chemoradiation followed by at least one dose of adjuvant durvalumab from 2017-2021. Incident pneumonitis was defined as new or worsening symptoms, radiographic changes on CT chest, at least one steroid prescription, and a likely diagnosis of pneumonitis by the treating physician within 2 years of the final radiation treatment; severity was graded by CTCAE v4.03 criteria. Pneumonitis events were ascertained by manual chart review assisted by keyword searches and steroid prescription data. Cox regression was used to identify baseline clinical factors associated with Grade 2+ pneumonitis.

RESULTS

Our cohort included 1,994 patients with stage IIII NSCLC, 989 of whom received concurrent chemoradiation alone and 1005 of whom received adjuvant durvalumab. The overall incidence of symptomatic pneumonitis among patients receiving concurrent chemoradiation alone vs those receiving adjuvant immunotherapy was 6.8% vs 11% (Grade 2, p = 0.0005), 6% vs 8.4% (Grade 3, p = 0.007) and 1.1% vs 1.6% (Grade 5, p = 0.292), respectively. The cumulative incidence of Grade 2-5 pneumonitis at 2 years was 13.8% among patients treated with chemoradiation alone versus 22.1% for patients treated with chemoradiation followed by durvalumab (p ≤0.0001). Of those with Grade 2 pneumonitis, pneumonitis occurred in 85% of patients within one year of receiving radiation, and 84% of those with Grade 3-5 pneumonitis developed pneumonitis within one year. Patients prescribed outpatient steroids for pneumonitis had active prescriptions lasting a mean of 153 days, and among those hospitalized for pneumonitis, ¾ needed supplemental oxygen support. On multivariate analysis, the use of durvalumab is associated with higher risk of Grade 2-5 pneumonitis (HR 1.4, p = 0.004). Current smokers have a 50% risk reduction for Grade 2+ pneumonitis relative to never smokers (p≤0.001).

CONCLUSION

In this largest-to-date cohort, we found that adjuvant immunotherapy is associated with higher risk of Grade 2+ pneumonitis. The increase in symptomatic pneumonitis associated with durvalumab should spur future work to refine patient selection, balance the risk/benefit of adjuvant immunotherapy, and potentially modify radiation dosimetric constraints to decrease pneumonitis rates.

Use of Explainable AI Algorithm Revealing Longitudinal Changes in Practice Patterns and Toxicity Models

PURPOSE/OBJECTIVE(S)

Dosimetric constraints evolve as clinicians implement practice changes, requiring modeling approaches to be dynamic. We applied a semi-automated explainable artificial intelligence (eAI) algorithm and dashboard visualizations to model dysphagia and xerostomia for head and neck cancer patients. We coupled a large, comprehensive, "real-world" database to the eAI for discovery of features with the strongest combined statistical and machine learning based evidence and to identify clinically actionable thresholds.

MATERIALS/METHODS

Cohort included 758 patients treated 2017-2021 for HN cancer with conventional fractionation. Features included age, sex, diagnosis, staging, chemotherapy, smoking and alcohol status, BMI, weight loss, re-simulation, DVH curves, PTV and OAR volumes. Patients were scored for toxicity within 2 yrs of RT for dysphagia grade ≥ 3 and xerostomia grade ≥ 2. Bootstrap resampling of thresholds, ROC-AUC, PR-ROC, SN, SP, F1 and diagnostic odds ratio was used to statistically profile strength of evidence for candidate features. XGBoost models with 10-fold cross validation were repeated (n = 20) to identify mean and CIs for statistical measures of predictions. DVH metrics included standard template values and those with highest statistical evidence and low cross correlation with other features. Backward feature selection was used to identify the most relevant feature subset, where the least informative feature is iteratively removed from the model. This workflow was repeated by year and overall.

RESULTS

Annual incidence of dysphagia averaged 0.13 ± 0.02 overall years. Xerostomia incidence decreased from 0.32 to 0.12 (2017-2021). Box-whisker plots by year showed consistent reductions in standard practice toxicity linked DVH metric values. Median dose to superior constrictors (PCM), contralateral parotid and contralateral submandibular gland (SMG) declined from 2017 to 2021 by 48 to 33 Gy, 17 to 10 Gy, and 28 to 22 Gy respectively. Statistics of XGBoost models of dysphagia for all years were ROC-AUC = 0.72 ± 0.05. Strongest overall years predictors were Oral Cavity (OC) D50%[Gy] < 32, inferior PCM Max [Gy] < 60, contralateral SMG D10%[Gy] < 53 and use of Paclitaxel. Xerostomia models were less predictive with ROC-AUC = 0.65 ± 0.05. Strongest predictors over each year were ipsilateral parotid D30%[Gy] < 35, contralateral SMG D96%[Gy] < 18.4, and overall staging < II. Predictive features varied substantially by year for both, showing the most consistency for SMG doses. For example, OC D50%[Gy] < 27 and contralateral SMG D96%[Gy] < 18 dominated xerostomia model in 2017 but not in 2021 when practice norms shifted to lower doses.

CONCLUSION

As OAR doses were systematically reduced, statistical and AI models evidence highlighted contralateral SMG dose as important to both dysphagia and xerostomia for clinical practice change. The "real-world" database + eAI + visualization dashboards provided a method for continuous learning as clinical practice changes.