Neoadjuvant Nivolumab/Ipilimumab Shows Superior EFS in Resectable Melanoma

“Combining IFN-γ and TMB or PD-L1 and TMB further discriminates favorable and unfavorable subgroups in the neoadjuvant and adjuvant arm,” Minke W. Lucas, MD, said.

The combination of nivolumab (Opdivo) and ipilimumab (Yervoy) as neoadjuvant therapy displayed a sustained event-free survival (EFS) and distant metastasis-free survival (DMFS) advantage vs adjuvant nivolumab alone in patients with resectable stage III melanoma, according to 2-year follow-up findings from the phase 3 NADINA trial (NCT04949113), presented at the 2025 European Society for Medical Oncology (ESMO) Congress.1 The trial also previewed the potential for using baseline markers such as IFN-γ signature, tumor mutational burden (TMB), and PD-L1 expression as biomarkers of response.

At a median follow-up of 25.2 months, updated EFS results from the trial indicated that the 12- and 24-month EFS rates were 85.2% and 77.3% with neoadjuvant (n = 212) therapy, respectively, vs 61.7% and 55.7% with adjuvant therapy (n = 211), respectively (HR, 0.40; 95% CI, 0.28-0.57; P <.001). The updated 12- and 24-month DMFS rates were 88.5% and 82.8% with neoadjuvant therapy, respectively, vs 71.9% and 63.9% with adjuvant therapy, respectively (HR, 0.43; 95% CI, 0.29-0.64; P <.001).

What is the history of PD-1 checkpoint blockade with or without CTLA4 inhibition in NADINA?

Neoadjuvant PD-1 checkpoint blockade with or without CTLA4 inhibition has been shown to improve EFS vs adjuvant PD-1 inhibition in patients with resectable, macroscopic stage III melanoma.2 At a median follow-up of 9.9 months, the NADINA trial previously showed estimated 12-month EFS [rates] of 83.7% (99.9% CI, 73.8%-94.8%) and 57.2% (99.9% CI, 45.1%-72.7%) with neoadjuvant nivolumab plus ipilimumab and adjuvant nivolumab, respectively.2 Moreover, TMB, the 10-gene IFN-γ signature, and PD-L1 expression has demonstrated predictive capability in phase 2 neoadjuvant trials, Lukas explained.

The NADINA trial enrolled patients with stage III de novo or recurrent pathologically proven resectable melanoma with at least 1 lymph node metastasis, 3 or fewer in-transit metastases, and no prior exposure to PD-1, CTLA4, LAG3, or BRAF and MEK inhibition.

Patients were randomly assigned to 1 of 2 arms. In the experimental arm, patients received 2 courses of 80 mg of ipilimumab plus 240 mg of nivolumab every 3 weeks, followed by therapeutic lymph node dissection (TLND); patients with major pathologic response (MPR) entered follow-up and those without received 11 courses of nivolumab at a dose of 480 mg every 4 weeks if BRAF wild-type or 46 weeks of dabrafenib (Tafinlar) at 150 mg twice daily plus trametinib (Mekinist) at 2 mg once daily with or without adjuvant radiation if BRAF V600E/K mutant within 12 weeks of surgery. Patients in the control arm received TLND followed by 480 mg of nivolumab within 12 weeks of surgery every 4 weeks with or without adjuvant radiotherapy.

Patients were stratified for BRAF mutational status, their continent, and the presence of in-transit metastases.

The primary end point was EFS.

What subgroup data were presented during ESMO 2025?

Additional data were presented with respect to recurrence-free survival (RFS) and DMFS according to pathologic response in the neoadjuvant arm. The 24-month RFS rates were 95.1%, 81.2%, 69.4%, and 48.4% in patients with pathologic complete response (pCR), near pCR, pathologic partial response (pPR), and pathologic non-response (pNR), respectively (P = .001). The 24-month DMFS rates were 98.0%, 81.2%, 78.5%, and 55.0%, respectively (P < .001).

RFS was further elucidated according to adjuvant treatment after pPR or pNR in the neoadjuvant arm. A total of 212 patients received neoadjuvant therapy: 129 achieved MPR, 74 achieved pPR/pNR, 6 experienced progression, and 3 were non-evaluable. Within the three respective arms, 43 received adjuvant dabrafenib/trametinib, 21 received adjuvant nivolumab, and 10 received no adjuvant therapy.

The 12-month RFS rates in the dabrafenib/trametinib, nivolumab, and no treatment arms were 85.1%, 75.0%, and 52.4%, respectively. The respective 24-month RSF rates were 67.2%, 58.3%, and 37.7%. “Patients with pCR have very favorable EFS and DMFS at 24 months,” Lukas said.

Pathologic response was also evaluated in the neoadjuvant arm according to baseline TMB, IFN-γ signature, and PD-L1 expression. pCR, near pCR, pPR, pNR, and progression occurred in 61.0%, 6.8%, 6.8%, 23.7%, and 1.7% of patients with high TMB (n = 59), respectively, vs 35.9%, 17.9%, 7.7%, and 32.1%, and 5.1% of those with low TMB (n = 78); 1.3% of patients were not evaluable in the low TMB cohort (P = .006).

pCR, near pCR, pPR, pNR, and progression occurred in 58.1%, 16.2%, 5.4%, 17.6%, and 1.4% of patients with high IFN-γ (n = 74), respectively, vs 30.5%, 10.2%, 10.2%, 42.4%, and 6.8% of those with high IFN-γ (n = 59); 1.4% of patients were not evaluable in the high IFN-γ cohort (P = .002).

pCR, near pCR, pPR, pNR, and progression occurred in 58.6%, 14.3%, 7.1%, 17.1%, and 0% of patients with positive (≥ 1%) PD-L1 expression (n = 70), respectively, vs 38.3%, 11.7%, 8.3%, 36.7%, and 5.0% of those with negative PD-L1 expression (< 1%; n = 60); 2.9% of patients were not evaluable in the positive PD-L1 expression cohort (P = .045).

“Baseline IFN-γ signature, PD-L1 expression and TMB were associated with EFS in the neoadjuvant arm and the IFN-γ signature and PD-L1 expression also in the adjuvant arm. The neoadjuvant arm was superior [to] the adjuvant arm across all biomarker subgroups,” Minke W. Lucas, MD, of the Department of Medical Oncology at the Netherlands Cancer Institute in Amsterdam, Netherlands, said.

The 12- and 24-month EFS rates were 93.2% and 87.7%, 65.4% and 74.0%, 66.8% and 72.1%, and 36.6% and 43.9%, respectively, in the neoadjuvant/IFN-γ-high (n = 74), neoadjuvant IFN-γ-low (n = 59), adjuvant IFN-γ high (n = 85), and adjuvant IFN-γ low (n = 67) populations, respectively (P <.0001).

The 12- and 24-month EFS rates were 92.9% and 87.6%, 79.6% and 73.3%, 74.7% and 67.2%, and 32.0% and 32.0%, respectively, in the neoadjuvant/PD-L1–positive (n = 70), neoadjuvant/PD-L1–negative (n = 60), adjuvant PD-L1–positive (n = 61), and adjuvant/PD-L1–negative (n = 51) populations, respectively (P <.0001).

The 12- and 24-month EFS rates were 91.5% and 91.5%, 76.5% and 62.0%, 66.3% and 59.5%, and 56.5% and 49.3%, respectively, in the neoadjuvant/TMB high (n = 59), neoadjuvant/TMB low (n = 78), adjuvant TMB high (n = 58), and adjuvant/TMB low (n = 94) populations, respectively (P <.0001).

The combination of baseline biomarkers in the adjuvant arm also provided predictive information.

In the neoadjuvant setting, the 12- and 24-month EFS rates were 96.4% and 96.4%, 89.3% and 78.5%, 88.5% and 88.5%, and 61.1% and 45.7%, respectively, in the IFN-γ-high/TMB high (n = 28), IFN-γ-high/TMB low (n = 38), IFN-γ-low/TMB high (n = 26), and IFN-γ-low/TMB low (n = 29) populations, respectively (P < .0001).

The 12- and 24-month EFS rates were 100.0% and 100.0%, 89.5% and 80.5%, 84.6% and 84.6%, and 61.5% and 43.1%, respectively, in the PD-L1–positive/TMB high (n = 21), PD-L1–positive/TMB low (n = 17), PD-L1–negative/TMB high (n = 31), and PD-L1–negative/TMB low (n = 25) populations, respectively (P = .0017).

In the adjuvant setting, the 12- and 24-month EFS rates were 76.2% and 72.7%, 70.8% and 63.8%, 52.2% and 45.7%, and 36.8% and 28.0%, respectively, in the IFN-γ-high/TMB high (n = 31), IFN-γ-high/TMB low (n = 49), IFN-γ-low/TMB high (n = 23), and IFN-γ-low/TMB low (n = 39) populations, respectively (P = .00036).

The 12- and 24-month EFS rates were 80.0% and 80.0%, 70.0% and 54.4%, 50.0% and 50.0%, and 20.0% and 20.0%, respectively, in the PD-L1–positive/TMB high (n = 21), PD-L1–positive/TMB low (n = 17), PD-L1–negative/TMB high (n = 31), and PD-L1–negative/TMB low (n = 25) populations, respectively (P < .0001).

With respect to updated safety data, 97.6% of patients in the neoadjuvant arm (n = 212) experienced an AE (grade ≥3, 49.5%) vs 97.6% (grade ≥3, 39.1%) of those in the adjuvant arm (n = 207). Surgery-related AEs also occurred in the neoadjuvant (any grade, 75.9%; grade ≥3, 14.8%) and adjuvant (any grade, 81.2%; grade ≥3, 16.4%) arms.

Systemic treatment-related AEs occurred in 87.7% (grade ≥3, 31.1%) and 84.1% (grade ≥3, 15.9%) of patients in the neoadjuvant and adjuvant arms, respectively. AEs leading to discontinuation occurred in 9.4% and 18.4% of patients in the neoadjuvant and adjuvant arms, respectively.

One death due to a treatment-related AE occurred in the adjuvant arm.

“Combining IFN-γ and TMB or PD-L1 and TMB further discriminates favorable and unfavorable subgroups in the neoadjuvant and adjuvant arm,” Lucas said in conclusion.

Disclosures: Lucas had no disclosures to report.

References

1. Lucas MW, Menzies AM, Dimitriadis P, et al. Two-year clinical update and first biomarker analyses of the phase 3 NADINA trial comparing neoadjuvant nivolumab plus ipilimumab versus adjuvant nivolumab in resectable stage III melanoma. Presented at: 2025 ESMO Congress; October 17-25, 2025; Berlin, Germany. Abstract LBA57.
2. Blank CU, Lucas MW, Scolyer RA, et al. Neoadjuvant nivolumab and ipilimumab in resectable stage III melanoma. N Engl J Med. 2024;391(18):1696-1708. doi:10.1056/NEJMoa2402604