Remission after rituximab for HHV8⁺ MCD: what next?

In this issue of Blood Advances, Rasmussen et al evaluate the outcomes and prognostic factors of 99 patients treated for human herpesvirus 8 (HHV8)–associated multicentric Castleman disease (MCD).¹ HHV8 is also the etiologic agent of Kaposi sarcoma (KS) and primary effusion lymphoma; these conditions, including MCD, are frequently observed in people with HIV (PWH). The unique cohort presented by Rasmussen et al included 73 PWH and 26 HIV seronegative (HIV⁻) patients who were uniformly treated with rituximab, which led to clinical resolution of the initial "MCD attack" in 96% of patients. Rituximab is effective in HHV8⁺ MCD flares, eliminating the B-cell reservoir of HHV8 infection and resulting in decreases in disease-associated cytokines and HHV8 viremia.² Studies from resource-rich settings have shown that rituximab, which can also be used in relapsed MCD, improves survival, transforming this once fatal lymphoproliferative disorder into a manageable condition.^3-5 

The authors build on treatment advances by evaluating baseline and time variable factors associated with relapse. Although most progression was because of MCD, 20% of patients progressed with KS or primary effusion lymphoma, highlighting the risk for concurrent HHV8-associated malignancies in this population. A previous study from the United Kingdom evaluating factors associated with MCD relapse in 80 PWH did not identify demographic or HIV-related risk factors for relapse after first-line rituximab.⁵ Here, low baseline hemoglobin and HIV⁻ status are identified as novel prognostic factors. After rituximab, covariates such as increase in CD19 B cells, HHV8 viral load (VL), and C-reactive protein (CRP) levels were also investigated for their impact on progression. In these analyses, increases of ≥3 log of blood HHV8, CRP increases of ≥20 mg/L, and HIV^– status were independently associated with progression. Similar to findings of the UK study, CD19 B-cell restoration was observed before MCD recurrence but was not an independent risk factor.

Understanding measures of HHV8 in recurrence may guide decisions about additional treatment. Baseline anemia was notable as a predictor of recurrence and may reflect HHV8 burden, lytic activation, and interleukin-6 (IL-6)–related inflammatory signs and symptoms. A previous study demonstrated that detectable plasma HHV8 of >50 copies/mL was a risk factor for MCD recurrence, although levels in the setting of an MCD flare were generally 4 to 5 log.⁶ In this study, the authors evaluated a cutoff blood HHV8 of ≥3 log during follow-up as a risk factor for progression based on the fact that, in their experience, patients with KS or asymptomatic MCD may have HHV8 levels ≤3 log. HHV8 detection above this threshold was associated with recurrence but with a latency of >2 years. Together with our observation that patients who received virus-activated cytotoxic therapy with zidovudine and ganciclovir after rituximab-based therapy had a lower rate of relapse,⁴ data on HHV8 VL as a predictor of response raise the possibility that depth of initial remission (eg, HHV8 VL at the end of therapy or residual adenopathy or splenomegaly) may affect relapse risk. However, given differences in assays and long latency between having a threshold >3 log and development of clinical symptoms leaves open the question whether there is a specific measure of HHV8 that warrants intervention in the absence of symptoms. This study supports watchful waiting for those with blood HHV8 of ≤3 log rather than additional rituximab, which carries the risk of new or worsening KS. Likewise, circulating cell–associated HHV8 correlates with several host cytokines including IL-10 and IL-6, as well as a viral homolog of host IL-6, vIL-6,⁷ during MCD flares. CRP may be a surrogate marker for IL-6 increases and, although not specific, it may be useful to follow CRP levels in tandem with fever or inflammatory symptoms to monitor for subsequent flares.

To our knowledge, this is the first study to include the outcomes of patients who are HIV⁻ with HHV8-associated MCD and highlight HIV^– status as a distinct prognostic factor associated with MCD progression. Individuals who are HIV^– were older, and 54% were from North and West Africa, areas with endemic Kaposi sarcoma–associated herpesvirus (KSHV) transmission. These individuals who are HIV^– with normal CD4 T-cell counts likely have unmeasured immune defects that contribute to KSHV persistence. Interestingly, PWH in this cohort may also provide insights into why patients who are HIV^– may have more MCD progression. All PWH were on antiretroviral therapy and had a median CD4 T-cell count of ≥200 cells/μL at baseline MCD flare. Only 25% of the cohort of PWH had undetectable HIV. For the majority, MCD was diagnosed at the time of HIV diagnosis. The authors observed that 53% of those with an undetectable HIV VL had MCD progression, whereas this was the case in only 28% of those with initially uncontrolled HIV. Immune reconstitution on antiretroviral therapy likely has a role in controlling HIV and HHV8, resulting in longer duration of remission in these patients.

Patients who are HIV^– and those with controlled HIV who develop HHV8⁺ MCD may have inadequate anti-HHV8–specific T-cell responses or other innate immune deficits. Invariant natural killer T (iNKT) cells are innate-like T cells that may have a role in the control of viral replication and that are decreased in aging, HIV infection, and active MCD.⁸ Although the initial study of patients with HHV8 disorders showed that iNKT cell numbers were not different by HIV status, it is possible that older patients who are HIV⁻ may have decreased iNKT cell numbers, contributing to recurrent MCD. Immune defects observed in KS may also be relevant to HHV8⁺ MCD, because individuals who are HIV^– with KS have a higher proportion of senescent (CD57⁺ CD8⁺) T cells as compared with PWH with KS.⁹ The course and outcomes between those without HIV and those with well-controlled HIV are often similar because these groups often experience recurrent disease, albeit more indolent disease, than those have HIV viremia and KS.¹⁰ 

Findings by Rasmussen et al advance our understanding of the natural history of HHV8⁺ MCD after initial treatment and identify differences in outcomes by HIV status. They raise new questions about what modifiable immune factors may exist to improve outcomes across all groups.

Conflict-of-interest disclosure: R.R. is an investigator on studies supported by Cooperative Research and Development Agreements between the institution and the following: Bristol Myers Squibb-Celgene, Merck, PDS Biotech, Janssen, and Eli-Lilly. T.S.U. is a full-time employee of Regeneron Pharmaceuticals, and is a coinventor on US Patent 10 001 483 entitled “Methods for the treatment of Kaposi’s sarcoma or KSHV-induced lymphoma using immunomodulatory compounds and uses of biomarkers” and patent application.