Tocilizumab in patients with symptomatic Kaposi sarcoma herpesvirus–associated multicentric Castleman disease

TO THE EDITOR:

Kaposi sarcoma (KS) herpesvirus (KSHV), also known as human herpesvirus-8, is the causative agent for KS and a plasmablastic form of multicentric Castleman disease (MCD; KSHV-MCD).¹  KSHV-MCD is a relapsing and remitting B-cell lymphoproliferative disorder that occurs primarily in people living with HIV.^2,3  Patients present with anemia, thrombocytopenia, and inflammatory symptoms associated with elevated cytokines including interleukin-6 (IL-6) and IL-10, elevated levels of C-reactive protein (CRP), and elevated KSHV viral loads (KSHV-VLs). KSHV encodes for viral IL-6 (vIL-6), an analog of human IL-6 (hIL-6), which contributes to KSHV-MCD pathogenesis, in part by driving production of hIL-6.^4-6  Treatment options include rituximab (a humanized monoclonal antibody against CD20), rituximab with liposomal doxorubicin or etoposide, or virus-activated cytotoxic therapy with high-dose zidovudine (AZT) and valganciclovir (VGC).^7-13  However, some patients do not respond to rituximab, and, when used alone, it is associated with worsening or development of KS in 35% to 67%.^8,9 

The IL-6 coreceptor (glycoprotein 80 [gp80]) is required for hIL-6 signaling and is a rational B-cell–sparing target for KSHV-MCD. Tocilizumab is a humanized anti-gp80 antibody that inhibits membrane-bound and soluble gp80 signal transduction. It is approved for rheumatoid arthritis and has demonstrated clinical benefit in KSHV⁻ MCD.^14,15  Although KSHV-encoded vIL-6 can signal through the gp130 receptor without requiring gp80,^16,17  mouse models demonstrated that IL-6 is required for development of clinical features in the setting of vIL-6 excess.⁶  We hypothesized that blocking hIL-6 signaling would be sufficient to control KSHV-MCD flares. We considered siltuximab, an anti–hIL-6 monoclonal antibody, but found that it does not bind to vIL-6 (supplemental Figure 1, available on the Blood Web site). Additionally, promising results were seen in a tocilizumab trial that included 2 HIV⁻ patients with KSHV-MCD.¹⁵  With this background, we initiated an open-label, single-center pilot study to evaluate the clinical benefit of tocilizumab in patients with symptomatic KSHV-MCD and included the option of adding AZT/VGC if tocilizumab alone did not lead to adequate responses.

Adults with pathologically confirmed KSHV-MCD and at least 1 clinical symptom (such as persistent fever, fatigue, gastrointestinal symptoms) and laboratory abnormality (anemia, hypoalbuminemia, thrombocytopenia, or elevated CRP) attributable to KSHV-MCD were eligible. Patients with HIV infection were required to use effective combination antiretroviral therapy. Tocilizumab (8 mg/kg; Roche) was administered IV on day 1 of 14-day cycles for a maximum of 6 cycles (12 weeks). Oral AZT (600 mg every 6 hours) and VGC (900 mg every 12 hours) were administered during days 1 to 5 of a 14-day cycle in combination with tocilizumab where patients had an inadequate response or relapse on tocilizumab monotherapy. Responses were evaluated using the KSHV-MCD Clinical Benefit Response Criteria (CBR), which uses 8 indicator abnormalities (4 symptoms and 4 laboratory findings) closely associated with KSHV-MCD activity and adapted from previous National Cancer Institute (NCI)-MCD response criteria published by our group (supplemental Table 1).^7,10  The criteria for the addition of AZT/VGC was elevated CRP (>3 mg/L) and progressive disease by CBR from entry, or failure to achieve partial response (PR) by cycle 4 of tocilizumab monotherapy. The addition of AZT/VGC did not alter the number of tocilizumab doses. Patients with concurrent KS were assessed using the previously described modified AIDS Clinical Trial Group criteria.¹⁸  The primary objective was to estimate the clinical benefit of tocilizumab in patients with symptomatic KSHV-MCD using CBR. We also evaluated the benefit and tolerability of the addition of AZT/VGC in cases in which tocilizumab monotherapy was inadequate in achieving or sustaining a response. Exploratory analyses of differences in laboratory findings and serum cytokines between baseline and 48 hours, 2 cycles, and end of treatment were evaluated using the Wilcoxon signed rank test. Progression-free survival (PFS) for all patients was determined using Kaplan-Meier methods from baseline to a subsequent KSHV-MCD flare necessitating further treatment. The protocol (NCT01441063) and accompanying KSHV-MCD natural history protocol (NCT00092222) were approved by the NCI Institutional Review Board. All patients provided written informed consent in accordance with the Declaration of Helsinki.

Between 2011 and 2018, 8 cisgender patients (7 male and 1 female) with symptomatic KSHV-MCD and HIV were enrolled (Table 1). All were on antiretroviral therapy. Four patients (50%) had no prior KSHV-MCD therapy. Two patients had active concurrent KS (25%); 1 patient received prior KS therapy. Five of 8 patients (63%; 95% confidence interval, 25% to 92%) experienced a response as assessed by the modified KSHV-MCD CBR with tocilizumab alone. Best responses on tocilizumab alone were PRs in 4 patients and complete response (CR) in 1 patient (Table 1). Among the 5 tocilizumab-monotherapy responders, beneficial CBR responses were seen within the first 2 cycles (4 weeks) of treatment. Patient 4 (case report previously published [Manion et al¹⁹ ]), who had a concurrent diagnosis and treatment of mycobacteria immune reconstitution inflammatory syndrome, had rapid improvement in symptoms and normalization of KSHV-VL, experiencing a CR after cycle 5. She had no further recurrence of KSHV-MCD.

Three patients had AZT/VGC added to tocilizumab, resulting in either PR (2 patients) or CR (1 patient) within 4 weeks of combination treatment (Table 1). Clinical benefit was noted in all 3 patients during the first cycle with AZT/VGC, with decreases in KSHV-VL, IL-6, and CRP levels (supplemental Figure 2). Responses were sustained among 2 of 3 patients at the end-of-treatment assessment, whereas 1 patient (patient 2) had progressive KSHV-MCD flare symptoms prior to cycle 6. Three patients did not receive AZT/VGC following incomplete tocilizumab responses due to worsening pulmonary KS (patient 6), thrombocytopenia prohibiting the initiation of AZT/VGC (patient 7), or rapidly progressive KSHV-MCD symptoms (patient 8).

There were decreases in CRP at 48 hours (median, −10.3 mg/L [interquartile range, −49.2 to −5.4 mg/L]; P = .02) in all 7 assessed patients; however, these changes did not consistently persist (Figure 1; supplemental Table 2). Although there was an expected increase in soluble IL-6 and h-IL-6 due to blocking of receptor binding, other cytokines and laboratory parameters associated with KSHV-MCD activity did not change significantly as compared with baseline. vIL-6 messenger RNA levels measured in peripheral blood mononuclear cells, assessed at baseline, cycles 2 and 6, highlighted varying levels of expression among all patients. The messenger RNA vIL-6 transcripts appeared to mirror the KSHV-VL levels in 6 of the 8 patients (supplemental Figure 3).

Two patients with concurrent KS had KS progression, and 1 patient developed cutaneous KS with tocilizumab and AZT/VGC treatment on study. The median PFS was 3.2 months for all study participants (95% confidence interval, 1-8.3 months; supplemental Figure 4). Seven patients received rituximab-based regimens or pomalidomide and liposomal doxorubicin for subsequent flares (Table 1). Treatment in this study was generally well tolerated; the most common grade 2 and 3 adverse events possibly, probably, or definitely attributed to treatment were thrombocytopenia and neutropenia for patients treated with tocilizumab monotherapy or in combination with AZT/VGC (supplemental Table 3).

In conclusion, tocilizumab is safe and has activity among patients with a diagnosis of KSHV-MCD and HIV. However, unlike treatment with rituximab, which can yield sustained responses,¹²  the PFS in this study was short, and tocilizumab did not alter KSHV-VL, IL-10, or IL-1β levels. Incomplete responses may occur because tocilizumab does not block signaling by vIL-6, which binds directly to the gp130 subunit without requiring gp80. Nonetheless, tocilizumab may be useful in the management of KSHV-MCD as part of a multiagent regimen.