Complement activation profiles in patients with immune checkpoint inhibitor-associated neuromuscular immune-related adverse events

Abstract

Background: Immune-related neuropathy (irNeuropathy) and myositis (irMyositis) are the most common neurologic adverse events (irAE-n) associated with immune checkpoint inhibitors. Although case reports suggest benefits of complement inhibitors, the role of complement activation in irAE-n is understudied.

Methods: In a retrospective multicenter study, we enrolled patients with irNeuropathy or irMyositis, cancer controls (CCs), and healthy controls (HCs). Serum levels of 11 complement components were measured using multiplex enzyme-linked immunosorbent assays. Associations with irAE-n severity and outcomes were assessed by Spearman's correlation. C5b-9-positive complement deposition was analyzed in muscle and nerve specimens from a subset of patients.

Results: Thirty-one irMyositis patients, 25 irNeuropathy patients, 25 CCs, and 17 HCs were included. Complement component levels were elevated in irNeuropathy (C3a, C5a, sC5b-9, C3, Ba, C4a), irMyositis (C3a, Ba), and CCs (C3a, C5a, sC5b-9, Bb, Ba, C4a), compared to HCs. In irMyositis, higher levels of C5a and complement regulators Factor H and I correlated with lower irAE-n severity (p = 0.02, rho = -0.45; p = < 0.01, rho = -0.56; p = < 0.001, rho = -0.67, respectively), and improved outcomes (p = 0.03, rho = -0.42; p = 0.05, rho = -0.40; p = < 0.001, rho = -0.64, respectively). Subtle C5b-9 deposition was detected in all tissue samples but showed non-specific patterns.

Discussion: Systemic complement activation is detectable in cancer patients regardless of irAE-n status, and tissue complement deposition is unspecific. Our findings suggest that complement activation is not a major driver of irAE-n, leaving the therapeutic potential of complement inhibitors uncertain.

Keywords: Complement activation; Immune checkpoint inhibitor; Immune-related adverse events; Myositis; Neuropathy; Neurotoxicity.

Fig. 1

Complement profiles in patients with irAE-n, controls with cancer (CC), and healthy controls (HC). Serum levels of central complement components C3/C3a and components of the terminal complement pathway (a), the alternative complement pathway (b), the classical and lectin complement pathway (c), and the complement regulator Factor H (d) were compared between patients with irMyositis (MYO), patients with irNeuropathy (NEU), CCs and HCs using the Kruskal–Wallis test and the Dunn’s test for post-hoc comparison. Tukey’s box plots display interquartile range (lower and upper edge of the box), range of 1.5 × IQR (whiskers), median (central line), mean (cross), and outliers (dots). * = p < 0.05, ** = p < 0.01, *** = p < 0.001, **** = p < 0.0001

Fig. 2

Associations between complement components, irAE-n severity, and outcomes. Serum levels of complement regulators Factor H (upper left) and Factor I (upper middle) as well as serum levels of complement activation product C5a (upper right) show moderate to strong negative correlations with Common Terminology Criteria for Adverse Events (CTCAE) grade, a measure of toxicity severity, in patients with irMyositis (Spearman’s correlation, n = 25, n = 24, n = 26, respectively) (a). Serum levels of complement regulators Factor H (middle left) and Factor I (middle middle) as well as serum levels of complement activation product C5a (middle right) show moderate to strong negative correlations with irAE-n outcomes, defined as complete remission (CR), a relapsing–remitting course (RR), recovery with residual symptoms (RS), or death, in patients with irMyositis (Spearman’s correlation, n = 25, n = 24, n = 26, respectively) (b). Serum levels of complement activation product Ba show a weak positive correlation with tumor outcomes, defined as complete remission (CR), partial remission (PR), stable disease (SD), or progressive disease (PD), in all cancer patients (lower left). In contrast, no statistically significant correlations were observed in patients with irNeuropathy alone (lower middle) and irMyositis alone (lower right, p = > 0.05) (Spearman’s correlation, n = 69, n = 19, n = 26, respectively) (c). Each point represents an individual patient. The dark blue line represents the best-fit linear regression line, the grey area depicts the 95% confidence interval for the regression estimate

Fig. 3

Staining patterns of C5b-9 in patients with irAE-n. Mild sarcolemmal staining of terminal complement complex C5b-9 on non-necrotic myofibers without perifascicular accumulation (a-c). Mild C5b-9-positive deposition on the sarcolemma of myofibers and on capillaries (d). Sarcoplasmic staining for C5b-9 in necrotic myofibers with complement deposition on thickened capillary walls (e). In one patient with irMyositis, sarcolemmal C5b-9 deposition clustered in the perifascicular region (f). C5b-9-positive complement deposition on capillaries (g, h). C5b-9 deposition on endoneurial capillaries of two patients with irMyositis (i, note the physiological perineurial complement positivity, and j) and one patient with irNeuropathy (k). Myocardial biopsy from a patient with irMyocarditis showing C5b-9 deposition in the sarcoplasm of necrotic cardiomyocytes, while adjacent non-necrotic fibers remain negative (l)