Comparison of Anti–Programmed Cell Death Ligand 1 Therapy Combinations vs Sunitinib for Metastatic Renal Cell Carcinoma: A Meta-analysis

A partir d'une revue de la littérature, cette méta-analyse évalue l'efficacité, du point de vue de la survie globale, de la survie sans progression et du taux de réponse globale, et la toxicité d'agents anti-PD-L1 dispensés en combinaison par rapport au sunitinib chez des patients atteints d'un carcinome à cellules rénales de stade métastatique

Résumé en anglais

Introduction: In the past 10 years, immune checkpoint inhibitors (ICIs) have been used to manage metastatic renal cell carcinoma (mRCC). Different combinations of ICIs and tyrosine kinase inhibitors or double ICIs have been approved by regulatory agencies for first-line treatment of mRCC after demonstrating prolonged survival vs sunitinib in randomized clinical trials (RCTs). From a pharmacological perspective, anti–programmed cell death 1 (PD-1) and anti–programmed cell death ligand 1 (PD-L1) agents behave similarly. However, combinations based on anti–PD-1 vs anti–PD-L1 agents currently dominate first-line treatments of mRCC. We aimed to assess the association of anti–PD-L1 treatment combinations with survival and response rate among patients with mRCC.

Methods: The literature search for this meta-analysis was conducted in December 2022 and followed the PRISMA reporting guideline. Randomized clinical trial articles on anti–PD-L1 agents as first-line mRCC treatment were selected (eFigure in Supplement 1). We included 3 RCTs1-3 that assessed overall survival (OS), progression-free survival (PFS), and overall response rate (ORR). For OS and PFS, summary hazard ratios (HRs) were calculated; for ORR, odds ratio (OR) was assessed. Random- or fixed-effects models were used, depending on study heterogeneity. Two-sided P < .05 indicated statistical significance. RevMan 5.4 (Cochrane Collaboration) was used for statistical analyses.

Results: Atezolizumab plus bevacizumab (IMmotion150 and IMmotion151 trials1,3) and avelumab plus axitinib (JAVELIN Renal 101 trial2) were administered in the intervention groups (n = 1100; 784 males [71.3%], 316 females [28.7%], mean [SD] age, 62 [0.45] years), and sunitinib was administered in the control group (n = 1107; 776 males [77.1%], 231 females [22.9%], mean [SD] age, 61 [0.58] years). Overall, the combination of anti–PD-L1 agents was not associated with a risk of death (HR, 0.88; 95% CI, 0.75-1.03; P = .11) but was associated with a reduced risk of progression (HR, 0.78; 95% CI, 0.69-0.88; P < .001) compared with sunitinib. Moreover, the trials did not report a higher ORR with anti–PD-L1 agents than sunitinib (OR, 1.63; 95% CI, 0.79-3.35; P = .19). In patients with PD-L1 expression, no differences in OS were detected (HR, 0.84; 95% CI, 0.67-1.05; P = .12), but PFS (HR, 0.66; 95% CI, 0.56-0.79; P < .001) and ORR (OR, 2.28; 95% CI, 1.17-4.46; P = .02) were favorable with a combination of anti–PD-L1 agents vs sunitinib (Table).

Discussion: Anti–PD-L1 and anti–PD-1 agents prevented PD-L1 from interacting with PD-1; however, the 2 classes slightly differed in pharmacokinetics and pharmacodynamics. Because PD-1 is expressed on the T-cell membrane surface, anti–PD-1 agents are more effective than anti–PD-L1 agents in activating T cells even with lower or no PD-L1 expression, which partially explains why ICIs were active in cases with lower or absent PD-L1 expression. In contrast, PD-L1 is expressed by tumor cells; therefore, anti–PD-L1 agents dysregulate tumor cell signaling rather than simply acting on T cells. In other tumor subtypes, such as non–small cell lung cancer, evidence suggests that anti–PD-1 agents could be more effective than anti–PD-L1 agents, as anti–PD-1 agents simultaneously block PD-1 binding with both PD-L1 and PD-L2.4 However, anti–PD-L1 agents did not influence the PD-1–PD-L2 interaction, which may inhibit T-cell activation. Therefore, in the setting of anti–PD-L1 agent use, the PD-1 or PD-L2 could be used by the tumor to escape the antitumor immune response.
Further evidence of the lack of effectiveness of anti–PD-L1 agents in RCC comes from the IMmotion010 trial,5 in which adjuvant atezolizumab did not affect disease-free survival in intermediate-to-high-risk resected or stage M1 cancer in patients with no evidence of disease. Updated results from the KEYNOTE-564 trial6 support the use of pembrolizumab vs placebo in the adjuvant setting due to a disease-free survival advantage (HR, 0.63).
Study limitations were use of cumulative rather than individual patient data and lack of direct comparisons between anti–PD-1 and anti–PD-L1 agents from RCTs on RCC. Therefore, only speculative analyses were feasible, although head-to-head studies are warranted.