Factors associated with grade 3 or 4 treatment-related toxicity in women with advanced or recurrent cervical cancer: An exploratory analysis of NRG oncology/gynecologic oncology group trials 179 and 204

Abstract

Objective: This study aimed to describe pretreatment patient characteristics and baseline quality-of-life scores as they relate to the development of grade 3 or 4 toxicity in patients receiving chemotherapy for advanced/recurrent cervical cancer. Methods: The study sample was drawn from Gynecologic Oncology Group protocols 179 and 204. Grade 3 or 4 toxicities were considered in 4 specified categories as follows: peripheral neuropathy, fatigue, hematological, and gastrointestinal (GI). The data variables explored included age, stage, pretreatment radiation, performance status (PS) at treatment initiation, and baseline Functional Assessment of Cancer Therapy-Cervix (FACT-Cx) score.A logistic regression model was developed with various adverse events as binary (0/1) outcomes. Results: Six hundred seventy-three patient-reported questionnaires were used in the analyses. At baseline, pain was the most severe patient-reported symptom. Baseline line-item patient concerns did demonstrate specific correlations with the development of individual toxicities. In 401 patients who were enrolled on Gynecologic Oncology Group 204 (fatigue not measured on 179), a worse PS predicted the development of grade 3 or 4 fatigue (odds ratio, 2.78; 95%confidence interval, 1.66-4.68). Exposure to previous radiation, treatment regimen, and aworse FACT-Cx scorewere associated with the reporting of both grade 3 or 4 leukopenia (P < 0.05) and anemia (P < 0.0005). Performance status and treatment regimen (P < 0.05) were associated with the development of grade 3 or 4 thrombocytopenia. Age and treatment regimen (P < 0.05) were associated with the development of grade 3 or 4 neutropenia. The FACT-Cx score (P = 0.0016) predicted grade 3 or 4 GI toxicity. Conclusions: The development of fatigue, hematological, and GI toxicity might be predictable based on factors other than treatment assignment such as age, PS, and patient-reported quality-of-life measurement.