Objective: In clinical practice, chemotherapy agents demonstrating modest second-line activity against platinum-refractory epithelial ovarian cancer (PROC) are frequently used in patients who have received multiple prior chemotherapy agents. Whether the response rates reported in selected patients can be expected in heavily pretreated patients is not known. Similarly, the costs of palliative chemotherapy are not known. We sought to determine the response, survival, and predictors of response in an unselected cohort of PROC patients receiving liposomal doxorubicin (LD) for relapsed disease, and the overall costs of delivering liposomal doxorubicin in this setting.
Methods: In a cohort of 62 consecutive patients who initiated LD as second- or greater-line therapy for PROC, the following variables were examined: age, number of prior regimens for relapse disease, duration of first clinical remission, time from last prior treatment, dose intensity of LD received, response/clinical benefit, time to progression, toxic effects, and survival. Multivariate analyses were used to identify predictors of clinical benefit and overall survival. Direct medical charges were calculated and converted to costs, and major cost drivers determined.
Results: Sixty-two patients received a total of 174 cycles of LD. The mean number of cycles per patient was 2 (range, 1-8); the median number of prior regimens for recurrent PROC was 2 (range, 0-8); and the median duration of the first clinical remission was 6 months. Median dose intensity of LD delivered was 11.4 mg/m(2)/week (range, 2.8-16.7 mg/m(2)/week). Nine of sixty-two patients (14.5%) had an objective clinical response by CA-125 and/or CT scan (95% confidence interval, 6-23%). Grade 3/4 toxicity occurred in 11% of patients. In the full cohort, median time to progression was 2.2 months, and median overall survival, 9.6 months (range, 0.2-26 months). Dose intensity was the only independent predictor of overall response. Duration of first clinical remission and number of prior salvage regimens were associated with longer overall survival. The mean total direct medical cost per cycle of LD was $5763, and the major cost drivers were hospitalizations and drug acquisition/delivery costs.
Conclusion: LD is an active agent in PROC, even when used as greater-than-second-line therapy. Among heavily pretreated patients, delivering a dose intensity of at least 9.0 mg/m(2)/week was associated with a higher probability of response. The cost per cycle of LD is driven by hospitalizations and drug acquisition/delivery costs.
Copyright 2001 Academic Press.