Nivolumab plus ipilimumab versus nivolumab in individuals with treatment-naive programmed death-ligand 1 positive metastatic soft tissue sarcomas: a multicentre retrospective study

Soft tissue sarcoma (STS) is a heterogeneous malignant tumour derived from mesenchymal cells that displays a heterogeneous mix of clinical and pathologic characteristics and is largely resistant to conventional therapies [1‐3]. Evidence-based statistics [4, 5] have indicated that the majority of individuals with STS tend to develop recurrent or metastatic disease and are associated with poor outcomes. Few, if any, chemotherapy regimens, either alone or in combination, can reverse this situation [6]. For individuals with treatment-naive metastatic STS (MSTS), several approved chemotherapy regimens (i.e., doxorubicin, either alone or in combination) seem to have similar effects, with a low response rate, progression-free survival (PFS) of nearly 0.5 years and overall survival (OS) of 1–1.5 years [7, 8]. Except for first-line chemotherapeutics, all other chemotherapeutics that have been approved, to a certain extent, have improved OS in the absence of progression or metastasis of the STS [9, 10]. However, for MSTS, the median PFS tends to be approximately 4 months, and OS from a diagnosis of MSTS is less than 14 months [7, 11]. Management of such individuals is still a challenge, and a poor prognosis seems to be inevitable [5].

Nivolumab, a fully human immunoglobulin G4PD-1 immune checkpoint-blocking antibody, explicitly binds to programmed death 1(PD-1) and interrupts negative signalling to restore T-cell anti-tumour function, which leads to improved survival and a promising safety profile in individuals with specific progressed solid tumours involving STS [7, 11, 12]. Findings from a recent randomized clinical trial [7] demonstrated that nivolumab, alone or combined with ipilimumab (a cytotoxic T-lymphocyte antigen-4 checkpoint inhibitor), had promising efficacy for specified sarcoma subtypes, with a controllable safety profile consistent with current confirmed alternatives. However, there is a paucity of published information regarding the utilization of nivolumab and/or ipilimumab in treatment-naive programmed death-ligand 1 (PD-L1) positive MSTS individuals [11]. We report herein a retrospective study assessing the efficacy and safety of nivolumab alone or combined with ipilimumab in this setting.

This study shows that for treatment-naive PD-L1 positive MSTS, the superiority of NPI over NIV in terms of survival benefit tends to be positive, which is in line with previous reports involving individuals with untreated MSTS [7, 11]. Safety profiles were consistent with those of other solid tumours (i.e., melanoma) [15].

Our findings might provide a confirmation that NPI improves survival for individuals with untreated MSTS. In a multicentre, open-label, non-comparative, randomized phase 2 study [7], 85 eligible patients with metastatic sarcoma who were treated using NPI (42 cases) or NIV (43 cases) showed a median PFS of 4.1 months (2.6–4.7) and 1.7 months (95% CI 1.4–4.3), respectively; the median OS was 14.3 months (9.6–not reached) with NPI and 10.7 months (5.5–15.4) with NIV. These findings may be instructive when placed in the context of presently accessible treatment options for individuals with untreated MSTS [16]. The classic treatment for MSTS tends to be based on cytotoxic chemotherapy, with first-line therapy predictably accomplishing objective responses in 15–18% of individuals, with a median PFS of 4–6 months [7, 17]. Activity beyond the first-line options tends to decline, with less than 10% of individuals reaching objective responses and a median PFS of 1–4 months [18]. In the current review, the choice of NPI or NIV as a monotherapy, regardless of its combination with cytotoxic chemotherapy, may have a negative impact on survival. However, a key challenge with MSTS is that well-established protocols for management tend to be lacking, and in the absence of distinguishable signs or symptoms identifiable by the clinicians, diagnosis tends to be difficult; indeed, once diagnosed with STS, the patient is generally in the late stage of the disease, ultimately leading to reduced survival [16, 19].

A double-blind trial [20] involving 142 patients with treatment-naive MSTS showed that meaningfully longer PFS was detected with NPI than with ipilimumab monotherapy (not reached vs. 4.4 months [95% CI 2.8–5.7]; HR 0.40, 95% CI 0.23–0.68, p< 0.001). The response rate associated with NPI in their study (61%) was higher than with NIV (61% vs. 40%) as first-line therapy in such individuals. The response rate of the combination therapy in our study was also higher than the rate detected in published trials involving anti–PD-1 agent-based monotherapy (i.e., pembrolizumab) [21, 22]. Nevertheless, a comparison of the efficacy of NPI and anti-PD-1 monotherapy may be challenging due to differences in the baseline data of individuals among the studies. The PFS and OS seen with NPI in our review are in accordance with those reported elsewhere [7, 11, 23], with the primary endpoint occurring by the time of the final tumour evaluation and, in a host of patients, OS being prolonged as follow-up continued regardless of termination of treatment, which might be elucidated by the fact that the individuals included in this review were diagnosed with treatment-naive MSTS.

Antibodies against PD-1 or PD-L1 have a positive effect in blocking tumour immune evasion and inducing tumour regression in STS [7, 24]. Previous reports [7, 11, 25, 26] of PD-L1 expression have shown that STS is potentially responsive to PD-1/PD-L1 blockade intervention in STS patients with PD-L1 positivity. The survival benefit of NIV monotherapy is inconsistent with presently existing chemotherapy-based untargeted therapies [7, 11]. Furthermore, in previous trials [8, 23], NIV patients failed to meet the predetermined primary outcome of completing responses in more than 13% of cases to sustain activity in their setting, which could exclude extended trials for heavily treated, unselected patients with MSTS [11, 16]. NPI patients met this predetermined primary outcome among those unselected patients with MSTS [24, 26, 27]. The proportion of NPI patients reaching an objective response appeared to be 16%, approximating that realized via accepted chemotherapy-based management [26, 27]. Additionally, an objective response of approximately 16% in 38 patients is in accordance with FDA-approved chemotherapy regimens, theoretically favouring future trials of NPI not only as second-line management in patients with MSTS but also as a first-line treatment option [7, 26]. Although patients undergoing a treatment regimen approved by the FDA exhibited a median OS of 26.5 months (95% CI, 20.9–31.7), their data might not truly reflect survival in an open-label phase 1b and randomized phase 2 trial [28]. However, the OS seen with NPI in this study is promising and indicates the potential to improve survival in patients with MSTS.

The safety results associated with NPI and NIV were in accordance with prior studies [7, 26]. In this review, NIV tended to be better tolerated, with a lower rate of AEs compared with NPI. The rate of grade 3–5 AEs among individuals experiencing NIV was 27.1%. The safety results contrasted with the results reported in a previous study [7], where the dose of NIV was higher than the recommended dose, and a higher percentage of individuals suffered from grade 3–5 AEs. Adopting a lower dose of NIV could potentially improve the rate of AEs. Remarkably, the proportion of grade 3–5 AEs described in this review for NIV was lower than that of cytotoxic drugs in the current setting.

Several limitations should be recognized in this review. First, this study is a retrospective study, with its inherent shortcomings and some potential confounding variables (i.e., potential comorbidities and complications, some patients who were followed up by telephone), which reduces the reliability of the conclusion. Second, the sample size of this retrospective review is limited, which restricts the generalizability of the results to some extent. Third, gene mutation types are not retested when the disease progresses, and drug resistance mutations during treatment have not been tested for each individual. Therefore, when drug resistance mutations appear in some individuals, the power of this study to reach a reliable conclusion is weakened. Fourth, the current research objects were collected from different tertiary medical centres, and there might be some differences in the diagnosis process of these medical institutions. Nevertheless, these research objects are coded and combined through standardized methods, which guarantees the reliability of the research conclusions.

The results reported in the current review reiterate an increasing body of evidence showing that for individuals with treatment-naive PD-L1 positive MSTS who undergo treatment with NPI or NIV, NPI seems to be less tolerated but has a greater survival advantage as the primary therapy than NIV. Our findings might underline the promise of combined checkpoint inhibition in the current setting. These data should be validated prospectively in subsequent analyses of larger cohorts with treatment-naive MSTS.