Neo-Adjuvant Treatment of Melanoma and Directed Therapy from the PRADO Trial

Immunotherapy and targeted treatment of melanoma have drastically improved prognosis. Well established in the metastatic melanoma setting, studies have proven PD1 inhibition in the adjuvant setting for high-risk resectable melanoma, and this is now a PBS-funded treatment. High-risk resectable melanoma is classified as stage 3B to 3D according to the AJCC TNM staging system for melanoma. AJCC TNM measures thickness of primary tumour, ulceration, mitotic rate, and lymph nodal involvement. Other risk factors that influence prognosis and treatment include BRAF mutation status and patient characteristics such as age, comorbidities, and treatment preferences.

There is evidence for improved survival with neo-adjuvant treatment of high-risk resectable melanoma, and the suggested mechanism behind this is that tumour load has an impact with regard to immunotherapy response. Hence if you remove the tumour prior to therapy (adjuvant), it has been proposed to be less effective than if you treat the tumour up front and then undergo surgery (neo-adjuvant).

The primary endpoint of the SWOG 1801 study was event-free survival (EFS), a study that compared the PD1 inhibitor pembrolizumab in both the adjuvant (18 doses post-surgery) and neo-adjuvant (3 doses pre-surgery and then 15 doses post-surgery) settings. Three hundred thirteen participants with resectable stage IIIB to IV melanomas were included. The neo-adjuvant group experienced better control than the adjuvant group (HR 0.58, P=0.004). At 2 years, the neo-adjuvant group EFS was 72% compared to 49% for the adjuvant group, representing a significant reduction. All participants underwent recommended surgery to remove the primary tumour regardless of response to the neoadjuvant treatment. It should be noted that the control group was statistically planned to have a 64% reduction in EFS compared with the 49% reduction that was observed during this study and therefore the results may be magnified.

The PRADO trial assessed stage IIIB to IIID melanoma neoadjuvant treatment with Nivolumab 3mg/kg and Ipilimumab 1mg/kg for 6 weeks prior to surgery. Personalised response directed therapy was also assessed by measurement of response in an index node (largest lymph node metastasis at baseline) after neo-adjuvant treatment. In previous studies, it has been shown that using an index node to measure response was a reliable indicator of response in the total tumour bed.

Therapeutic lymph node dissection (TLND) usually is indicated for high-risk resectable melanoma and is associated with significant morbidity. There is also a risk of side effects when treating a patient with adjuvant immunotherapy or MEK/BRAF inhibition. The risk that when a patient no longer responds to adjuvant therapy, it can’t be used in the “arsenal” as part of treatment if relapse occurs is also a factor. Therefore, it would be optimal to only treat those patients with additional surgery and adjuvant treatment when necessary, and it has been suggested in preliminary studies that those patients who respond strongly to neoadjuvant therapy can forgo TLND and subsequent adjuvant treatment.

In the PRADO study, patients who had a major response (≤ 10% viable tumour in index node after 2 cycles) to neoadjuvant Nivolumab and Ipilimumab did not undergo major TLND and did not receive additional adjuvant treatment. Patients who had a partial response (>10% to ≤50% viable tumour after 2 cycles) received TLND without adjuvant treatment. Patients with minimal response (≥50% viable tumour after 2 cycles) received both a TLND and adjuvant immunotherapy or BRAF/MEK inhibition if BRAF^V600E/K-mutated tumours present for 12 months. Primary endpoints included pathological Response Rates, 24-month relapse-free survival and 24-month relapse-free survival in patients with minimal response after neoadjuvant therapy.

Pathological responses were recorded in 72% of patients who successfully underwent ILN resection and analysis. Major pathological response occurred in 61% of patients, and partial responses in 11% of patients ILN successfully analysed. PD1 expression was the only tumour or patient characteristic that showed definitive predictive value in response, with the higher PD1 expression associated with higher neoadjuvant response rates.

The relapse-free survival rate after 2 years for patients with major pathological response was 93%, with all 4 patients who had relapsed harbouring a BRAF^V600E/K-mutation and multiple lymph node involvement. Patients with partial pathological response who underwent TLND was 64%, and patients with minimal pathological response who received both TLND and adjuvant therapy was 71%. Grade 3-4 immune-related adverse effects occurred in 22% of patients, with the most common grade 3-4 adverse effects being increased transaminases and diarrhoea/colitis. 10% of patients received only one dose of neoadjuvant therapy due to adverse effects. A flowchart of the outcomes and interventions in the study is below here.

This PRADO trial investigated an exciting development in personalised treatment of melanoma that is tailored to the response from 6 weeks of neo-adjuvant therapy. Although the outcomes achieved for the major pathological response cohort are consistent with previous investigational studies, the small sample sizes, particularly in the partial and non-responsive subgroups, limit the certainty of any outcomes. It should also be noted that PRADO did not randomise TLND versus lymph node dissection, only allowing for indirect comparisons with historical studies. Also of note is that although previous studies have shown that the Index node response can be translated to the entire tumour bed, in two of the patients with multiple node involvement, there was a non-response in a small non-index node that did not alter the response subgroup in the entire TLND tumour bed. The authors speculate that less responsive sub-clones in a minority of patients may be the reason for the increased risk seen with patients with multiple nodal involvement.

The NADINA trial is currently investigating if neoadjuvant with standard TLND versus adjuvant will add to the body of work being pursued in the neoadjuvant treatment of advanced stage 3 melanoma. As TLND is associated with morbidity and poor quality of life outcomes and the burden of yearly adjuvant immunotherapy or targeted therapy, those with major pathological response to neoadjuvant therapy could forgo both TLND and adjuvant therapy according to the results from the PRADO trial.  Those in the partial response group should definitely be considered for yearly adjuvant therapy as well due to the unfavourable results in this subgroup (2 year relapse-free survival 64%). A larger randomised trial comparing TLND and index lymph node dissection is necessary to draw any conclusive evidence.