New results, presented at the 2021 ASCO annual meeting, show the benefit of adding durvalumab to neoadjuvant anthracycline and taxane–based chemotherapy for patients with early triple-negative breast cancer.
Durvalumab (Imfinzi) added to neoadjuvant anthracycline and taxane–based chemotherapy was found to significantly improve survival in patients with early triple-negative breast cancer, according to the results of the phase 2 GeparNUEVO trial (NCT02685059), presented during the 2021 ASCO Annual Meeting.1
Results from the trial indicate that patients in the durvalumab arm achieved a pathologic complete response (pCR) of 53.4% vs 44.2% in the placebo arm (Adjusted overall response, 1.53; 95% CI, 0.82-2.84; P = 0.182).
“The pCR was increased by a small amount of 9%, which was not statistically significantly different,” lead study author Sibylle Loibl, MD, PhD, chief executive officer and chair of the German Breast Group, said during an oral presentation of the data. “In the subgroup analysis, the patients overexposed to durvalumab alone seemed to derive a larger benefit in terms of an increased pCR. Otherwise, [patients who were] stage IIa and higher, and younger patients, had a larger benefit from adding durvalumab.”
The addition of PD-1 and PD-L1 inhibitors to single-agent chemotherapy has been shown to improve progression-free survival in patients with PD-L1–positive metastatic TNBC.2,3 Moreover, rates of pCR have been observed to improve with the addition of PD-1 and PD-L1 inhibitors to neoadjuvant chemotherapy.4,5 Responses to checkpoint inhibitors were also found to occur, independent of PD-L1 status, in patients with early-stage disease.
Several large phase 2/3 studies have examined the use of neoadjuvant chemotherapy plus anti–PD-L1/PD-1 inhibitors, Loibl noted.4-6
“The [phase 1/2 NeoTRIPaPDL1 Michelangelo; NCT002620280] trial investigated the addition of atezolizumab [Tecentriq] to an anthracycline chemotherapy backbone,” she said. “The pCR did not improve long-term results, which was the primary end point. [The phase 3] IMpassion031 trial [NCT03197935] was also investigating the addition of atezolizumab to carboplatinum and paclitaxel–containing chemotherapy. The phase 3 KEYNOTE-522 trial [NCT03036488] used a carboplatinum and anthracycline–based chemotherapy backbone. The last trials both increased the pCR to 57% and 65%, respectively.”
The IMpassion031 trial, which examined the neoadjuvant combination in patients with early-stage disease, and the KEYNOTE-522 trial, which examined the addition of pembrolizumab (Keytruda) to neoadjuvant chemotherapy in patients with early TNBC, found that there were no differences in response between patients who were PD-L1–positive and PD-L1–negative, according to Loibl. Additionally, both trials had a relatively short follow-up time of 20.6 months and 15.5 months, respectively.
“The delta of the events is about 4%, leading to a hazard ratio [HR] of 0.63 [95% CI, 0.43-0.93] in the KEYNOTE-522 study, and 0.76 [95% CI, 0.4-1.44] in the IMpassion031 study, which is the smaller of the 2 [trials],” Loibl said.
The GeparNUEVO trial enrolled 174 patients with early TNBC, who were stratified based on tumor-infiltrating lymphocytes (TILs) status of low, medium, and high to receive either durvalumab or placebo for 2 weeks. Patients in the durvalumab arm received a dose of (0.75 g) 1.5 g on day 1 of every 28-day cycle before receiving a core biopsy and moving on to receive 125 mg/m2 of nab paclitaxel weekly, for 12 weeks. Those who achieved a clinical response then went on to receive 90 mg/m2 of epirubicin, and 600 mg/m2 of cyclophosphamide on day 1 of every 14-day cycle, for 8 weeks, plus durvalumab followed by surgery.
The primary end point of the study was pCR, with key secondary end points including invasive disease-free survival (iDFS), distant DFS (DDFS), and overall survival (OS).
Patients across both study arms had a median age of 49.5 years (range, 23-76), and 8% of patients in the durvalumab arm had T3 tumors. Additionally, 30.7% of patients in the experimental arm had nodal involvement at baseline, and over half of the cohort (63.6%) had stage IIA or higher disease. Most patients in the durvalumab arm had grade 3 tumors (84.1%), but a notably small proportion of patients had high TILs (13.6%). Moreover, 38.6% of patients in the experimental arm had low TILs, and 47.7% had intermediate TILs.
In an evaluation of pCR according to stromal TILs, intratumoral TILs (iTILs), and PD-L1 changes, investigators reported that neither TILs nor PD-L1 status predicted the effect of durvalumab, or the pCR rate, Loibl said. However, the change from the iTILs from pre- to post-exposure predicted responses to durvalumab, she explained.
After a median follow-up of 43.7 months (range, 4.9-56.1), 34 iDFS events were reported, 12 of which were in the durvalumab arm. Distant relapse occurred in 6 patients vs 13 in the durvalumab and placebo arms, respectively, with an additional 4 patients experiencing invasive locoregional relapse in the durvalumab arm, and 5 in the placebo arm. While 2 patients in the experimental arm developed invasive contralateral breast cancer, 3 patients in the placebo arm developed a second malignancy, and 1 experienced death as a first event.
Additionally, after a median follow-up of 43.7 months (range, 4.9-56.1), patients treated with durvalumab had a 3-year iDFS rate of 85.6% vs 77.2% in the placebo arm (HR, 0.48; 95% CI, 0.24-0.97; P = .0398). Moreover, patients in the durvalumab arm achieved a 3-year DDFS rate of 91.7% vs 78.4% with placebo (HR, 0.31; 95% CI, 0.13-0.74; P = .0078). Three-year OS rates of 95.2% and 83.5% were reported in the durvalumab and placebo arms, respectively (HR, 0.24; 95% CI, 0.08-0.72; P = .0108).
An iDFS subgroup analysis was also included on the study.
“Looking at subgroups for the iDFS, we did not observe any differences,” Loibl said. “There are small hypothesis-generating results for the PD-L1 group, but this year we used the 2.263 with a cutoff at 1%. Those seemed to have a better iDFS with durvalumab and the pCR patients seemed to have a better outcome with durvalumab.”
When examining iDFS by pCR, investigators reported that patients with a pCR who were treated with durvalumab achieve a 3-year iDFS rate of 95.5%, while non-pCR patients who received the same treatment had a rate of 76.3%. Among the placebo cohort, patients who achieved a pCR had a 3-year iDFS of 86.1% vs 69.7% in non-pCR patients (HR [pCR vs non-pCR], 0.34; 95% CI, 0.16-0.73; P = .004).
Additionally, patients in the durvalumab arm who experienced a pCR had a 3-year DDFS of 100% (95% CI, 100%-100%) vs 84.3% (95% CI, 68.3%-92.6%) in non-pCR patients. Patients in the placebo arm had a 3-year DDFS rate of 86.1% (95% CI, 69.8%-94.0%) and 71.9% (95% CI, 55.8%-83.0%) in pCR and non-pCR patients, respectively. Lastly, patients who had a pCR in the durvalumab group had a 3-year OS rate of 100% (95% CI, 100%-100%) vs 92.0% (95% CI, 77.1%-97.3%) in non-pCR patients, while in the placebo cohort, patients had a pCR of 88.9% (95% CI, 73.1%-95.7%) and 78.8% (95% CI, 63.2%-88.4%) in the pCR and non-pCR patients, respectively.
“Similar results were seen in the iDFS, DDFS, and OS analysis,” according to Loibl. “There were absolutely no DDFS events in the durvalumab arm, [as well as] no death. This was overall statistically significantly different. The HR for all of those time-to-event end points between pCR and non-pCR was also statistically significantly different.”