The RaDaR assay successfully identified the patients with locoregionally advanced urothelial cancer most likely to benefit from treatment with ipilimumab and nivolumab in findings from the second cohort of the phase 1b/2a NABUCCO trial.
The RaDaR assay successfully identified the patients with locoregionally advanced urothelial cancer (UC) most likely to benefit from treatment with ipilimumab (Yervoy) and nivolumab (Opdivo) in findings from the second cohort of the phase 1b/2a NABUCCO trial (NCT03387761). Investigators concluded that the absence of circulating tumor DNA in plasma, as detected by RaDaR(R), is a predictor for progression-free survival and pathologic complete response (pCR).1
Thirty patients received 2 cycles of 3 mg/kg ipilimumab plus 1 mg/kg nivolumab (n = 15; cohort A) or 1 mg/kg ipilimumab plus 3 mg/kg nivolumab (n = 15; cohort B). All patients received a third cycle of 3 mg/kg nivolumab.
Six (43%) patients in cohort A achieved pCR (95% CI, 21%-67%) as did 1 (7%) in cohort B (95% CI, 0%-34%; P = .008). In comparison, the pCR rate was 46% in phase 1 (n = 11/24).
Combining data from both phases of the trial, the absence of ctDNA in blood plasma was highly associated with both pCR (odds ratio, 45.0; 95% CI, 4.9-416.5) and PFS (HR, 10.4; 95% CI, 2.9-37.5). Investigators found that the absence of ctDNA in urine was linked to pCR, but not PFS.
“Although cisplatin-based chemotherapy plus radical cystectomy remains the currently recommended treatment for muscle-invasive bladder cancer [MIBC], there are some patients who are not eligible to receive chemotherapy and the prognosis for this patient population is poor,” Shashikant Kulkarni, chief scientific officer, NeoGenomics, said in a news release.2 “The findings from the NABUCCO trial offer initial evidence in hopefully improving the outlook for these patients. Importantly, the research shows that the RaDaR(R) assay can be used successfully to guide decision-making and help oncologists personalize patient care based on their risk of recurrence.”
The primary treatment for MIBC is radical cystectomy, a surgical procedure known to cause significant morbidity and mortality. Resection may not be necessary however, given the high rate of pCR following treatment with ipilimumab/nivolumab, a bladder-sparing approach could be feasible in a subset of patients. However, current diagnostic tools cannot accurately predict a pathological response upon neoadjuvant therapy.
Previous results have shown that ctDNA in plasma correlated with clinical outcome in patients with MIBC who received preoperative and adjuvant atezolizumab (Tecentriq). Investigators in NABUCCO used RaDaR(R), a personalized sequencing test to detect minimal residual disease and risk for recurrence, to identify which patients may not have to undergo surgery.
“Most patients with muscle-invasive urothelial cancer are at a high risk for recurrence and most will undergo cystectomy that results in the removal of the bladder,” Vishal Sikri, president, Advanced Diagnostics, NeoGenomics, said in a news release. “If possible, preserving a patient's bladder is clearly an important outcome if new approaches to neoadjuvant treatments are successful. By determining if a patient has responded well to neoadjuvant treatment, RaDaR(R) provides necessary insights to help oncologists decide whether a patient may be a candidate for bladder-sparing treatment strategies.”
Investigators assessed ctDNA in urine and plasma before, during, and after treatment to evaluate the effect of preoperative treatment with ipilimumab and nivolumab in patients with locoregionally advanced MIBC. Based on whole-exome sequencing (WES) of tumor and germline material, they identified tumor-specific genetic variants for all patients in phase 1 (n = 24) and isolated cell-free DNA from plasma and urine samples.
Twenty-six (87%) patients received all treatment cycles. Four (13%) missed at least 1 cycle due to immune-related adverse effects (irAEs). Twenty-six patients underwent radical surgery. Three refused surgery and 1 patient progressed while on treatment and was no longer eligible.
In plasma, the mean estimated variant allele frequency before surgery was significantly lower in responders compared with nonresponders (P <.01). ctDNA was not detectable in the latest available plasma sample before surgery in 13 (93%) responders and in 4 (40%) nonresponders (P <.01).
Investigators found no correlation between pCR and the absence of ctDNA in urine before surgery (P = .39). When comparing patients with and without a pCR in the bladder, there was a correlation with the absence of urinary ctDNA before surgery (P < .01). There was no correlation between absence of urinary ctDNA and pCR (P = 0.08) or clinical outcome, making urinary ctDNA a suboptimal biomarker for clinical decision-making.
All patients in cohort A experienced any-grade irAEs compared with 73% in cohort B. Five (33%) patients in cohort A had grade 3 or higher irAEs compared with 3 (20%) in cohort B. All patients in phase 1 experienced any-grade irAEs and 13 (55%) had grade 3 or higher irAEs.