Akanksha Dua
Research Subjects
Akanksha Dua has a total of 37 co-authors. Here are the top co-authors by number of shared publications:
- Jorge Gomez
University of Illinois College of Medicine at Peoria, Peoria, IL, USA
2 shared publications Kellie Ryan
AstraZeneca, Gaithersburg, Maryland, United States of America
2 shared publicationsAdias Marcelin
GHESKIO, Port-au-Prince, Haiti
1 shared publicationAncilla W Fernandes
US Medical Affairs, AstraZeneca, Gaithersburg, MD 20878, USA
1 shared publicationAtul J Butte
University of California, San Francisco
1 shared publicationChristopher Yee
Analysis Group, Inc., Boston, Massachusetts, United States of America
1 shared publication
Latest Publications
- Open Access
Real-World Dose Modifications for FOLFIRINOX in Pancreatic Cancer: Evaluating the Feasibility of a Machine-Learning Framework
Background: FOLFIRINOX is a cornerstone regimen for eligible patients with pancreatic ductal adenocarcinoma (PDAC), but its clinical benefit is limited by substantial toxicity and frequent dose modification. In real-world practice, dose modifications are often individualized, and the clinical factors associated with these decisions remain incompletely characterized. Objective: To develop and evaluate an electronic medical record (EMR)-based machine-learning framework for modeling cycle-specific FOLFIRINOX dose modification decisions in patients with PDAC. Methods: We included patients with PDAC who received FOLFIRINOX at UCSF oncology clinics between November 2011 and December 2023. Predictors included demographic, clinical, laboratory, and treatment variables derived from the EMR. Logistic regression, random forest, and XGBoost models were trained using group-based 5-fold cross-validation to predict cycle-specific dose modifications for 5-fluorouracil, irinotecan, and oxaliplatin. Model performance was evaluated using area under the receiver operating characteristic curve. Results: The cohort included 514 patients receiving FOLFIRINOX across 5,041 treatment cycles. The mean age was 59 years, 60% of patients were White, 41% had a history of smoking, and patients received a median of 6 chemotherapy cycles. More than 60% of patients required at least one dose modification during treatment. XGBoost demonstrated the highest performance across component drugs, with AUCs ranging from 0.53 to 0.70. Clinically plausible predictors of irinotecan and oxaliplatin dose modification included hepatic and renal function markers, cumulative drug exposure, treatment-related symptoms, and demographic or behavioral characteristics. Conclusion: We developed an EMR-based machine-learning framework to model real-world FOLFIRINOX dose modification and identified clinically plausible, routinely available predictors, particularly for irinotecan and oxaliplatin. Variable model performance suggests that dosing decisions are only partially captured by structured EMR data, highlighting both the limitations of current data-driven approaches and clinical domains where ML-based models may support individualized dosing and toxicity surveillance. Future informatics efforts should incorporate dose-modification rationale, patient-reported and functional outcomes, and validation across diverse practice settings.
Open AccessReal-World Dose Modifications for FOLFIRINOX in Pancreatic Cancer: Evaluating the Feasibility of a Machine-Learning Framework
Background: FOLFIRINOX is a cornerstone regimen for eligible patients with pancreatic ductal adenocarcinoma (PDAC), but its clinical benefit is limited by substantial toxicity and frequent dose modification. In real-world practice, dose modifications are often individualized, and the clinical factors associated with these decisions remain incompletely characterized. Objective: To develop and evaluate an electronic medical record (EMR)-based machine-learning framework for modeling cycle-specific FOLFIRINOX dose modification decisions in patients with PDAC. Methods: We included patients with PDAC who received FOLFIRINOX at UCSF oncology clinics between November 2011 and December 2023. Predictors included demographic, clinical, laboratory, and treatment variables derived from the EMR. Logistic regression, random forest, and XGBoost models were trained using group-based 5-fold cross-validation to predict cycle-specific dose modifications for 5-fluorouracil, irinotecan, and oxaliplatin. Model performance was evaluated using area under the receiver operating characteristic curve. Results: The cohort included 514 patients receiving FOLFIRINOX across 5,041 treatment cycles. The mean age was 59 years, 60% of patients were White, 41% had a history of smoking, and patients received a median of 6 chemotherapy cycles. More than 60% of patients required at least one dose modification during treatment. XGBoost demonstrated the highest performance across component drugs, with AUCs ranging from 0.53 to 0.70. Clinically plausible predictors of irinotecan and oxaliplatin dose modification included hepatic and renal function markers, cumulative drug exposure, treatment-related symptoms, and demographic or behavioral characteristics. Conclusion: We developed an EMR-based machine-learning framework to model real-world FOLFIRINOX dose modification and identified clinically plausible, routinely available predictors, particularly for irinotecan and oxaliplatin. Variable model performance suggests that dosing decisions are only partially captured by structured EMR data, highlighting both the limitations of current data-driven approaches and clinical domains where ML-based models may support individualized dosing and toxicity surveillance. Future informatics efforts should incorporate dose-modification rationale, patient-reported and functional outcomes, and validation across diverse practice settings.
- Open Access
- 12 August 2024
- Published by Elsevier in The Lancet Regional Health - Americas
- Vol. 37, 100847
Continuity of care during severe civil unrest with a model of community-based HIV care: a retrospective cohort study from Haiti
- 12 August 2024
- Published by Elsevier in The Lancet Regional Health - Americas
- Vol. 37, 100847
Open AccessContinuity of care during severe civil unrest with a model of community-based HIV care: a retrospective cohort study from Haiti
- Open Access
- 31 March 2023
- Published by Springer Nature in Neurology and Therapy
- Vol. 12 (3) , 849-861
Cost Per Relapse Avoided for Ozanimod Versus Other Selected Disease-Modifying Therapies for Relapsing–Remitting Multiple Sclerosis in the United States
Introduction This study assessed the cost-effectiveness of ozanimod compared with commonly used disease-modifying therapies (DMTs) for relapsing–remitting multiple sclerosis (RRMS). Methods Annualized relapse rate (ARR) and safety data were obtained from a network meta-analysis (NMA) of clinical trials of RRMS treatments including ozanimod, fingolimod, dimethyl fumarate, teriflunomide, interferon beta-1a, interferon beta-1b, and glatiramer acetate. ARR-related number needed to treat (NNT) relative to placebo and annual total MS-related healthcare costs was used to estimate the incremental annual cost per relapse avoided with ozanimod vs each DMT. ARR and adverse event (AE) data were combined with drug costs and healthcare costs to manage relapses and AEs in order to estimate annual cost savings with ozanimod vs other DMTs, assuming a 1 million USD fixed treatment budget. Results Treatment with ozanimod was associated with lower incremental annual healthcare costs to avoid a relapse, ranging from $843,684 vs interferon beta-1a (30 μg; 95% confidence interval [CI] − $1,431,619, − $255,749) to $72,847 (95% CI − $153,444, $7750) vs fingolimod. Compared with all other DMTs, ozanimod was associated with overall healthcare cost savings ranging from $8257 vs interferon beta-1a (30 μg) to $2178 vs fingolimod. Compared with oral DMTs, ozanimod was associated with annual cost savings of $6199 with teriflunomide 7 mg, $4737 with teriflunomide 14 mg, $2178 with fingolimod, and $2793 with dimethyl fumarate. Conclusion Treatment with ozanimod was associated with substantial reductions in annual drug costs and total MS-related healthcare costs to avoid relapses compared with other DMTs. In the fixed-budget analysis, ozanimod demonstrated a favorable cost-effective profile relative to other DMTs.
- 31 March 2023
- Published by Springer Nature in Neurology and Therapy
- Vol. 12 (3) , 849-861
Open AccessCost Per Relapse Avoided for Ozanimod Versus Other Selected Disease-Modifying Therapies for Relapsing–Remitting Multiple Sclerosis in the United States
Introduction This study assessed the cost-effectiveness of ozanimod compared with commonly used disease-modifying therapies (DMTs) for relapsing–remitting multiple sclerosis (RRMS). Methods Annualized relapse rate (ARR) and safety data were obtained from a network meta-analysis (NMA) of clinical trials of RRMS treatments including ozanimod, fingolimod, dimethyl fumarate, teriflunomide, interferon beta-1a, interferon beta-1b, and glatiramer acetate. ARR-related number needed to treat (NNT) relative to placebo and annual total MS-related healthcare costs was used to estimate the incremental annual cost per relapse avoided with ozanimod vs each DMT. ARR and adverse event (AE) data were combined with drug costs and healthcare costs to manage relapses and AEs in order to estimate annual cost savings with ozanimod vs other DMTs, assuming a 1 million USD fixed treatment budget. Results Treatment with ozanimod was associated with lower incremental annual healthcare costs to avoid a relapse, ranging from $843,684 vs interferon beta-1a (30 μg; 95% confidence interval [CI] − $1,431,619, − $255,749) to $72,847 (95% CI − $153,444, $7750) vs fingolimod. Compared with all other DMTs, ozanimod was associated with overall healthcare cost savings ranging from $8257 vs interferon beta-1a (30 μg) to $2178 vs fingolimod. Compared with oral DMTs, ozanimod was associated with annual cost savings of $6199 with teriflunomide 7 mg, $4737 with teriflunomide 14 mg, $2178 with fingolimod, and $2793 with dimethyl fumarate. Conclusion Treatment with ozanimod was associated with substantial reductions in annual drug costs and total MS-related healthcare costs to avoid relapses compared with other DMTs. In the fixed-budget analysis, ozanimod demonstrated a favorable cost-effective profile relative to other DMTs.
- Open Access
- 18 March 2020
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 15 (3) , e0230444
Predictors of chemoradiotherapy versus single modality therapy and overall survival among patients with unresectable, stage III non-small cell lung cancer
Concurrent chemoradiotherapy (cCRT) was the standard of care for patients with unresectable stage III non-small cell lung cancer (NSCLC) prior to the PACIFIC trial, however, patients also received single modality therapy. This study identified predictors of therapy and differences in overall survival (OS). This retrospective study included stage III NSCLC patients aged ≥65 years, with ≥1 claim for systemic therapy (ST) or radiotherapy (RT) within 90 days of diagnosis, identified in SEER-Medicare data (2009–2014). Patients who had overlapping claims for chemotherapy and RT ≤90 days from start of therapy were classified as having received cCRT. Patients who received sequential CRT or surgical resection of tumor were excluded. Predictors of cCRT were analyzed using logistic regression. OS was compared between therapies using adjusted Cox proportional hazards models. Of 3,799 patients identified, 21.7% received ST; 26.3% received RT; and 52.0% received cCRT. cCRT patients tended to be younger (p <0.001), White (p = 0.002), and have a good predicted performance status (p<0.001). Patients who saw all three specialist types (medical oncologist, radiation oncologist, and surgeon) had increased odds of receiving cCRT (p<0.001). ST and RT patients had higher mortality risk versus cCRT patients (hazard ratio [95% CI]: ST: 1.38 [1.26–1.51]; RT: 1.75 [1.61, 1.91]); p<0.001). Several factors contributed to treatment selection, including patient age and health status, and whether the patient received multidisciplinary care. Given the survival benefit of receiving cCRT over single-modality therapy, physicians should discuss treatment within a multidisciplinary team, and be encouraged to pursue cCRT for patients with unresectable stage III NSCLC.
- 18 March 2020
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 15 (3) , e0230444
Open AccessPredictors of chemoradiotherapy versus single modality therapy and overall survival among patients with unresectable, stage III non-small cell lung cancer
Concurrent chemoradiotherapy (cCRT) was the standard of care for patients with unresectable stage III non-small cell lung cancer (NSCLC) prior to the PACIFIC trial, however, patients also received single modality therapy. This study identified predictors of therapy and differences in overall survival (OS). This retrospective study included stage III NSCLC patients aged ≥65 years, with ≥1 claim for systemic therapy (ST) or radiotherapy (RT) within 90 days of diagnosis, identified in SEER-Medicare data (2009–2014). Patients who had overlapping claims for chemotherapy and RT ≤90 days from start of therapy were classified as having received cCRT. Patients who received sequential CRT or surgical resection of tumor were excluded. Predictors of cCRT were analyzed using logistic regression. OS was compared between therapies using adjusted Cox proportional hazards models. Of 3,799 patients identified, 21.7% received ST; 26.3% received RT; and 52.0% received cCRT. cCRT patients tended to be younger (p <0.001), White (p = 0.002), and have a good predicted performance status (p<0.001). Patients who saw all three specialist types (medical oncologist, radiation oncologist, and surgeon) had increased odds of receiving cCRT (p<0.001). ST and RT patients had higher mortality risk versus cCRT patients (hazard ratio [95% CI]: ST: 1.38 [1.26–1.51]; RT: 1.75 [1.61, 1.91]); p<0.001). Several factors contributed to treatment selection, including patient age and health status, and whether the patient received multidisciplinary care. Given the survival benefit of receiving cCRT over single-modality therapy, physicians should discuss treatment within a multidisciplinary team, and be encouraged to pursue cCRT for patients with unresectable stage III NSCLC.
- Open Access
- 23 September 2019
- Published by Taylor & Francis in Future Oncology
- Vol. 15 (29) , 3381-3393
Treatment Patterns and Overall Survival among Patients with Unresectable, Stage III Non-Small-Cell Lung Cancer
Aim: To analyze treatment patterns and overall survival (OS) across time (2009-2014) among patients with unresected, stage III non-small-cell lung cancer (NSCLC). Patients & methods: Stage III NSCLC patients aged ≥65 years who initiated therapy were identified using SEER-Medicare data. Results: Among 4564 patients, 84% received chemotherapy (with or without radiotherapy), and 59% received chemoradiotherapy (CRT). Carboplatin + paclitaxel was the most frequent regimen. Median (interquartile range) OS among chemotherapy patients was 13.2 (6.0-28.9) months, and 14.8 (6.7-33.4) months among CRT patients. Among CRT patients, there was no difference in OS across years of CRT initiation. Conclusion: OS remained static across 2009-2014, indicating stagnancy in clinical outcomes for stage III NSCLC patients and a need for more effective therapeutic options.
- 23 September 2019
- Published by Taylor & Francis in Future Oncology
- Vol. 15 (29) , 3381-3393
Open AccessTreatment Patterns and Overall Survival among Patients with Unresectable, Stage III Non-Small-Cell Lung Cancer
Aim: To analyze treatment patterns and overall survival (OS) across time (2009-2014) among patients with unresected, stage III non-small-cell lung cancer (NSCLC). Patients & methods: Stage III NSCLC patients aged ≥65 years who initiated therapy were identified using SEER-Medicare data. Results: Among 4564 patients, 84% received chemotherapy (with or without radiotherapy), and 59% received chemoradiotherapy (CRT). Carboplatin + paclitaxel was the most frequent regimen. Median (interquartile range) OS among chemotherapy patients was 13.2 (6.0-28.9) months, and 14.8 (6.7-33.4) months among CRT patients. Among CRT patients, there was no difference in OS across years of CRT initiation. Conclusion: OS remained static across 2009-2014, indicating stagnancy in clinical outcomes for stage III NSCLC patients and a need for more effective therapeutic options.