Evidence suggests that a large proportion of melanoma patients who are being treated with systemic therapy would have been excluded from the supporting registrational phase III trials. As a result, efficacy and survival outcomes are unknown in these patients. Real-world evidence shows that although these patients generally have worse survival outcomes than patients originally included in those trials, outcomes have improved in recent years as more systemic therapies become available.
A significant proportion of advanced melanoma patients are not represented in the pivotal phase III trials of the systemic therapy that they are prescribed.1 This includes patients with an ECOG score ≥ 2, and patients with auto-immune diseases and immunosuppression, amongst other exclusion criteria. As such, these patients are actually being treated without firm clinical evidence. Recently, a study enrolled 3,009 unresectable stage IIIc-IV melanoma patients from the Dutch Melanoma Treatment Registry, which was established in 2012 to gather patient outcome data for patients with melanoma. This real world data study aimed to determine the outcomes of patients who were not eligible for the original immunotherapy clinical trials, to identify prognostic factors for survival, predict the survival of prognostic subgroups and to produce a decision tree that incorporated prognostic factors to guide clinical decision-making.
Of the 3,009 patients, 44% (N= 1,331) were considered ineligible for previous phase III immunotherapy trials, despite 75% (N= 1,004) being treated with systemic therapy. Similarly, 56% (N= 1,678) were considered eligible, with 91% (N= 1,532) of these patients being treated with systemic therapy. Some baseline characteristics were significantly more common in ineligible patients compared to eligible patients, including elevated LDH levels (p< 0.001), liver metastases (p= 0.001) and metastasis in ≥ 3 organ sites (p< 0.001). BRAF mutations were also more common in ineligible patients (67% vs. 54%). Furthermore, treatment heterogeneity was evident in ineligible patients based on BRAF status. Whilst 50.8% of patients with BRAF wild-type received anti-PD-1 therapy, the most common treatment for patients who were BRAF-mutant was BRAF plus MEK inhibitors (36.7%). Eligible patients had better survival outcomes than ineligible patients, with a median overall survival (OS) of 23 months vs. 8.8 months. Similarly, the 36-month OS probability was 41% vs. 22%, respectively. However, ineligible patients did appear to have better survival outcomes as more systemic treatments have become available in recent years. Ineligible patients treated in 2013 had a median OS of 5.7 months, with a 36-month OS probability of 7.5%. Conversely, ineligible patients treated between 2014 and 2017 had a median OS of 8.8 months and a 36-month OS probability of 22%. Combining prognostic indicators (LDH level, presence of brain metastasis and ECOG PS), the study was able to identify prognostic subgroups in ineligible patients. Patients with an ECOG PS of 0-1, absent brain metastasis and normal LDH levels had the highest 3-year survival probability, at 44.5%. Meanwhile, the group with the lowest OS probability, of just 0.3% at 3 years, were patients who had an ECOG PS ≥2, symptomatic brain metastasis and an elevated LDH level. Incorporating the patient’s age, gender, ECOG PS, LDH level, distant metastasis, brain and liver metastasis and BRAF mutational status, a conditional survival tree was produced to guide decision making that provided a Kaplan-Meier curve for overall survival, specific to the patient characteristics.2
A large proportion of patients with unresectable stage III-IV melanoma treated in a real-world setting would not have been eligible for pivotal trials. Three-year survival rates in these ineligible patients were lower compared to eligible patients (22%). Interestingly, the survival rates for these ineligible patients increased with a growing number of available therapies. Prognostic models based on LDH, ECOG, brain metastases, number of metastatic sites and age could support clinical decision-making.
References
Top
