Timing isn't everything: an analysis of when to start salvage chemotherapy in ovarian cancer

Results from GOG 178 showed a prolongation of progression-free survival (PFS) with the immediate use of chemotherapy (CT) following a complete clinical response (CR) in patients with stage III-IV ovarian cancer. We wanted to evaluate our strategy of reserving second line (2nd line) chemotherapy to the time of clinical recurrence by determining PFS intervals following first, second, and third line agents and to compare these finding to results of GOG 178.

We conducted a retrospective parallel study to GOG 178 using identical criteria for PFS definitions. Patients (pts) with stage III-IV cancer achieving a CR following surgery and five to eight cycles of platinum-based CT were identified. Patients not obtaining a CR and those with a CR who underwent second look surgery were excluded. Rather than immediately beginning consolidation CT after CR, second line agents were started at recurrence and were followed by a third line when pts progressed. Clinical-pathologic characteristics were abstracted, and time intervals including time to recurrence, time to use of second line CT, time to use of third line (3rd line) CT, and survival were recorded. Time intervals were studied by Kaplan-Meier method.

Of 217 reviewed pts (1991-2001), 59 eligible pts were identified. Forty-nine patients had stage III disease and ten had stage IV. At completion of surgery, 44 were optimally debulked. With a median follow-up of 51 months, the median PFS (from CR) of all patients was 20 months. At 5 years, 36% of pts remain disease-free, and 66% of pts are alive. Twenty-three pts have not received second line agents, and thirty-six have received them. For all pts, the median time from CR to start of second line chemotherapy was 21 months, and the median time to start of third line agents was 43 months. Recurrences occurred after 6 months from completion of first line (1st line) therapy in 87% of cases and after 12 months in 50%.

Nearly 70% of pts achieving a CR after primary therapy eventually recurred. Most recurrences occurred greater than 6 months from completion of primary chemotherapy, and the use of second line agents at the time of recurrence was effective. In this study, the median time from CR to start of third line agent at 43 months compares favorably with the median PFS of 28 months following 12 months of Taxol reported in GOG 178 and challenges the concept of consolidation chemotherapy in ovarian cancer. A randomized trial to evaluate when to institute second line agents should be performed.