Krithika Suresh

Research Assistant Professor

University of Michigan
Country flagUnited States
ApplyKite TwinLearns as you apply

Make your next move.

Prepare your application

Review drafts before you send them.

Explore related searches

Articles (2)

Purpose: We conducted a phase II multicenter clinical trial to test the hypothesis that 18 F-fluorodeoxyglucose positron emission tomography (FDG-PET)–based chemoradiation (CRT) dose deescalation would provide noninferior locoregional control compared with historical controls among patients with early-stage p16+ oropharyngeal cancer. We also hypothesized that human papillomavirus (HPV) circulating tumor DNA (ctDNA) changes during treatment predict locoregional recurrence (LRR). Patients and Methods: Patients with stage I/II p16+ oropharyngeal squamous cell carcinoma were planned to receive radiation 70 Gy in 35 fractions with concurrent weekly carboplatin and paclitaxel. All patients underwent FDG-PET at baseline and at radiation therapy (RT) fraction 10. Patients with ≥50% decrease from baseline to mid-treatment metabolic tumor volume had treatment deescalated to 54 Gy in 27 fractions. The primary endpoint was LRR. Plasma HPV ctDNA was evaluated weekly and in surveillance. Results: Of 84 evaluable patients, 43% met deescalation criteria. With a median follow-up of 37.8 months, 24-month LRR for the entire cohort was 7.8% [90% confidence interval (CI), 2.6%–12.6%], which was less than the 25% rate specified for assessing noninferiority, thereby meeting the primary endpoint. At 1 month after RT, the mean of multiple quality of life measures between the two groups was improved in the 54 Gy cohort, exceeding the minimal clinically important difference threshold. During CRT, week 1 percentage increase in ctDNA relative to baseline was significantly associated with worse loco-regional control (LRC; HR = 1.052 per 10 percentage point increase in ctDNA; 95% CI, 1.007–1.099; P = 0.023) and loco-regional progression-free survival (HR = 1.038; 95% CI, 1.002–1.076; P = 0.035). Conclusions: FDG-PET–based RT dose personalization resulted in promising LRR outcomes in early-stage oropharynx cancer with improved short-term patient-reported outcomes. Furthermore, HPV ctDNA changes early in treatment may predict LRC.

Year:

2026

Advanced machine learning methods capable of capturing complex and nonlinear relationships can be used in biomedical research to accurately predict time‐to‐event outcomes. However, these methods have been criticized as “black boxes” that are not interpretable and thus are difficult to trust in making important clinical decisions. Explainable machine learning proposes the use of model‐agnostic explainers that can be applied to predictions from any complex model. These explainers describe how a patient's characteristics are contributing to their prediction, and thus provide insight into how the model is arriving at that prediction. The specific application of these explainers to survival prediction models can be used to obtain explanations for (i) survival predictions at particular follow‐up times, and (ii) a patient's overall predicted survival curve. Here, we present a model‐agnostic approach for obtaining these explanations from any survival prediction model. We extend the local interpretable model‐agnostic explainer framework for classification outcomes to survival prediction models. Using simulated data, we assess the performance of the proposed approaches under various settings. We illustrate application of the new methodology using prostate cancer data.

Year:

2024

Collaborators (8)

Steven B. Chinn

Assistant Professor

University of Michigan

UNITED STATES
View details

Marisa Buchakjian

Assistant Professor

University of Iowa Hospitals and Clinics

UNITED STATES
View details

Keith Casper

-

UNITED STATES
View details

Muneesh Tewari

Associate Professor

University of Michigan

UNITED STATES
View details

Paul Swiecicki

University of Michigan

UNITED STATES
View details

Yue Cao

University of Michigan

UNITED STATES
View details

THEODORE S. LAWRENCE

Chair, Professor

University of Michigan

UNITED STATES
View details

J. Chad Brenner

Director Head and Neck Oncology Program

University of Michigan

UNITED STATES
View details