RECIST/CA-125 progression-free survival and the role of CA-125 surveillance in the phase III PAOLA-1/ENGOT-ov25 trial evaluating maintenance olaparib plus bevacizumab in patients with newly diagnosed advanced high-grade ovarian carcinoma

Abstract

<h3>Objectives:</h3> In the PAOLA-1/ENGOT-ov25 primary analysis (NCT02477644), adding the PARP inhibitor olaparib to maintenance bevacizumab (bev) after first-line platinum-based chemotherapy with bev led to a significant progression-free survival (PFS) benefit vs placebo (pbo) + bev in patients (pts) with advanced high-grade ovarian cancer (HGOC), using modified RECIST v1.1 criteria (HR 0.59; 95% CI 0.49-0.72; <i>P</i><0.001; median PFS 22.1 vs 16.6 months) (Ray-Coquard <i>et al. NEJM</i> 2019). Limited evidence supports use of serum CA-125, a surrogate biomarker for progression, to detect relapse in pts receiving PARP inhibitor maintenance therapy. We evaluated RECIST/CA-125 progression and CA-125 surveillance in PAOLA-1. <h3>Methods:</h3> Pts with newly diagnosed, FIGO stage III-IV HGOC in response after platinum-based chemotherapy plus bev received bev (15 mg/kg q3w for 15 months) + olaparib (300 mg bid for 24 months) or pbo. Scans were performed every 24 weeks, or every 12 weeks if there was evidence of clinical or CA-125 progression. CA-125 levels were assessed every 12 weeks. Time to RECIST or CA-125 progression or death was a secondary endpoint; RECIST/CA-125 progression by biomarker status, response to initial therapy and CA-125 level at baseline were explored. <h3>Results:</h3> 537 pts were randomized to olaparib + bev and 269 to pbo + bev with median follow-up for PFS by RECIST/CA-125 of 24.2 vs 24.7 months (DCO 22 March 2019). In the ITT population, the benefit provided by olaparib + bev vs pbo + bev for PFS by RECIST/CA-125 (HR 0.58; 95% CI 0.48-0.70; <i>P</i><0.0001; median PFS 22.1 vs 15.4 months) was consistent with the primary analysis of PFS by RECIST. Of 462 pts with RECIST progression (ITT), 39% had RECIST progression alone and 45% had CA-125 progression more than 7 days before RECIST progression, with a median 2.4 months between CA-125 and RECIST progression (Table). In subgroup analyses, 49% of t<i>BRCA</i>m pts and 45% of HRD-positive pts had RECIST progression alone compared with only 37% of HRD-negative pts. 42% of pts with NED or CR had RECIST progression alone compared with 33% of pts with PR. 44% of pts with normal baseline CA-125 levels had RECIST progression alone compared with only 17% of pts with abnormal baseline CA-125 levels (<i>P</i><0.001). The rate of PFS events by RECIST/CA-125 was higher for pbo + bev vs olaparib + bev (74% vs 53%) (ITT). Of pts with RECIST progression, a similar proportion in each treatment arm had RECIST progression alone, including in subgroup analyses by biomarker status and response to initial therapy. <h3>Conclusions:</h3> Median PFS and HRs were consistent when progression was assessed by either RECIST or RECIST/CA-125 in PAOLA-1. Scans appear particularly important for detecting progression in HRD-positive pts, t<i>BRCA</i>m pts or pts with normal CA-125 levels at baseline as CA-125 levels did not seem to predict relapse in these pts. Results suggest that CA-125 surveillance alone may be sufficient to detect progression in most pts with abnormal baseline CA-125 levels when starting maintenance therapy.