Third time’s a charm
As its name implies, the AGO DESKTOP III trial is the third in a series. AGO DESKTOP I developed the hypothesis that a positive AGO score – consisting of an Eastern Cooperative Oncology Group performance status score of 0, complete resection during first-line therapy, and ascites less than 500 mL – could be predictive of favorable outcomes with debulking surgery (Ann Surg Oncol. 2006 Dec;13[12]:1702-10).
AGO DESKTOP II was a prospective, multicenter trial testing the score in patients with platinum-free intervals of more than 6 months (Int J Gynecol Cancer. 2011 Feb;21[2]:289-95). In this trial, 51% of patients had a positive AGO score, and the score was shown to predict, with 95% probability, complete resectability in two-thirds of these patients, Dr. Du Bois said.
In AGO DESKTOP III, 407 patients were prospectively randomized to second-line chemotherapy alone (n = 201) or to cytoreductive surgery (n = 206) followed by the same chemotherapy, with platinum-containing regimens highly recommended.
Patient characteristics were well balanced between the arms. Nearly all patients (99%) in each arm had prior platinum exposure, and 75% had a platinum-free interval of more than 12 months (a median of 21.1 months in the surgery arm versus 18.7 months in the no-surgery arm).
Complete resections extend OS
There were 8 patients (4%) in the chemotherapy-only arm and 14 (6.8%) in the surgery arm who were noncompliant with randomization. The complete resection rate was 74.2%.
Following randomization, 88.8% of patients in the surgery arm and 90% in the no-surgery arm received platinum-containing chemotherapy; 22.8% and 23.4%, respectively, received bevacizumab; and 3.9% and 6.0% received a poly (ADP-ribose) polymerase inhibitor.
The median overall survival in the intention-to-treat population was 53.7 months in the surgery arm and 46 months in the no-surgery arm, an absolute difference of 7.7 months (hazard ratio, 0.75; P = .02).
The median progression-free survival, assessed in the intention-to-treat population after database closure in January 2020, was 18.4 months with surgery and 14 months without (HR, 0.66; P < .001).
A post hoc analysis showed the importance of complete versus partial resection. The median overall survival was 61.9 months in patients with complete resections and 28.8 months for patients with residual tumor after cytoreductive surgery, an absolute difference of nearly 3 years (HR, 0.40; P < .001).
Comparing only those patients with complete resections with patients who did not undergo surgery, the respective median overall survival was 61.9 months and 46 months (HR, 0.57; P < .001).
AGO DESKTOP III was sponsored by the AGO study group in collaboration with other oncology groups, Medac, and GlaxoSmithKline. Dr. du Bois and Dr. Coleman disclosed relationships with many companies.
SOURCE: du Bois A et al. ASCO 2020, Abstract 6000.