Risk of hepatitis B virus reactivation in patients treated with ibrutinib

TO THE EDITOR:

Certain therapies for hematologic malignancies, including anti-CD20 monoclonal antibodies and allogeneic hematopoietic cell transplantation (allo-HCT), are associated with increased risk of hepatitis B virus (HBV) reactivation in patients previously infected with HBV.^1,2  Past HBV infection is defined by concurrent detectable HBV core immunoglobulin G antibodies (HBcAb), undetectable HBV surface antigen (HBsAg), and undetectable HBV DNA. HBV reactivation in patients with hematologic malignancies with past HBV infection being treated with new, targeted cancer therapies has been reported, but its incidence is unknown.

One such agent, ibrutinib, inhibits Bruton tyrosine kinase and thereby interrupts B-cell receptor signaling. Ibrutinib gained US Food and Drug Administration approval for treatment of mantle cell lymphoma in 2013,³  chronic lymphocytic leukemia (CLL) in 2014,⁴  Waldenstrom macroglobulinemia in 2015,⁵  and marginal zone lymphoma⁶  and chronic graft-versus-host disease in allo-HCT⁷  in 2017. Patients with past or chronic HBV were excluded from clinical trials of ibrutinib. With widespread use in a real-world population, infectious diseases complications that were not commonly seen in clinical trials have been reported. These include invasive aspergillosis involving the central nervous system⁸  and atypical Pneumocystis jirovecii pneumonia.⁹  Two case reports have described fulminant HBV reactivation in patients with serologic evidence of previous HBV infection treated with ibrutinib for CLL.^10,11  In contrast, a case series reported no reactivation among 7 HBcAb-positive patients treated with ibrutinib, with median follow up of 24 months.¹²  Thus, it remains unclear whether ibrutinib treatment is associated with increased risk of HBV reactivation in patients with past HBV infection.

After observing HBV reactivation in a patient treated with ibrutinib at our institution, we designed this retrospective study to systematically assess the incidence of HBV reactivation among patients with past HBV infection and hematologic malignancy during and after ibrutinib therapy. We identified patients treated with ibrutinib at Dana-Farber Cancer Institute (DFCI) between 1 January 2010 and 31 December 2016 with serologic evidence of past HBV infection. Clinical characteristics were collected including previous therapy with anti-CD20 antibodies (rituximab, ofatumumab, or obinutuzumab) and allo-HCT. HBV monitoring is routinely performed on patients receiving anti-CD20 therapy or undergoing allo-HCT at DFCI; liver function test results are routinely monitored in all patients with hematologic malignancy.

HBV reactivation was defined as development of HBV DNA >100 IU/mL on 2 consecutive measurements with or without reappearance of HBsAg in patients with evidence of past HBV infection. Data were censored on 31 March 2017. This research was approved by the Dana-Farber/Harvard Cancer Center Office for Human Research Studies.

During the study period, 412 patients were treated with ibrutinib at DFCI, of whom 21 (5.1%) had evidence of past HBV infection and were thus at risk for reactivation. Median duration of ibrutinib therapy among those at risk for HBV reactivation was 9.5 months (range, 1-49 months), and median duration of follow up from the time ibrutinib started was 18.3 months (range, 2-69 months). Two patients developed HBV reactivation; thus, the cumulative incidence of HBV reactivation was 9.5% among those at risk (95% confidence interval, 1.2% to 30.4%); risk in the entire cohort including those without past HBV infection was 0.5% (95% confidence interval, 0.06% to 1.7%). Demographic, oncologic, and HBV-related characteristics of the at-risk cohort are described in Table 1.

The index patient is a 57-year-old man diagnosed with CLL 12 years before reactivation. At baseline, he was HBcAb positive, and HBsAb was >10 IU/L. Six years before HBV reactivation, he was treated with fludarabine and rituximab for 6 cycles, the last 2 of which included cyclophosphamide. Approximately 19 months later, he started therapy with ibrutinib for CLL progression. HBsAb had declined to <5 IU/L at this time. Forty-two months after starting ibrutinib, he developed detectable HBV DNA at 120 IU/mL that persists between 120 and 920 IU/mL with undetectable HBsAg. The patient has been monitored closely without antiviral therapy, and alanine and aspartate aminotransferase levels have remained normal on continued ibrutinib 47 months after starting therapy.

The second patient is a 75-year-old man diagnosed with CLL 9 years before HBV reactivation. At baseline, he was HBcAb positive and HBsAb was 51.7 IU/L. He was treated with 2 cycles of rituximab and lenalidomide 7 years before reactivation, followed by observation. Four and a half years before reactivation, he was treated with rituximab and bendamustine for 1 cycle, which was complicated by a severe allergic reaction to rituximab, followed by 1 cycle of bendamustine, after which he was again observed. Due to increased bulky lymphadenopathy 3 years before reactivation, he was treated with 2 cycles of bendamustine. He was then switched to ibrutinib, which was stopped 9 months later due to neurologic toxicity. He has been observed without CLL therapy since. Twenty-two months after stopping ibrutinib he was noted on routine labs to have alanine and aspartate aminotransferase levels ∼2× normal. Further testing revealed HBV reactivation, with positive HBsAg and HBV DNA >170 000 000 IU/mL. He was asymptomatic. Treatment with entecavir resulted in a prompt reduction in HBV DNA.

In this retrospective assessment of patients with evidence of past HBV infection treated with ibrutinib, we have observed 2 cases of HBV reactivation to date, both of which occurred years after starting ibrutinib. Our retrospective observations are limited by interpatient variability in HBV monitoring. Nonetheless, this is the largest study assessing HBV reactivation in at-risk patients treated with ibrutinib. HBV reactivation in patients treated with ibrutinib has been described only in 2 case reports,^10,11  and the only case series to assess this population included 7 patients who had no reactivation events.¹²  The cumulative incidence of reactivation in our larger cohort was 9.5%. Given that our median follow-up was relatively short (18 months) and both cases occurred after this, it is possible that with longer follow-up, the reactivation rate in our cohort will be higher.

Many questions remain about the impact of ibrutinib on HBV reactivation in this population. For example, what is the mechanism underlying this increased risk of reactivation? For how long after discontinuing ibrutinib does increased risk persist? Rituximab is known to increase risk for HBV reactivation in patients with past HBV infection for 1 year or more after therapy stops. Patient 2 in this cohort and another patient reported by Herishanu et al¹⁰  developed reactivation several months after ibrutinib stopped, which suggests that, as with rituximab, the reactivation risk may linger for months to years after ibrutinib ends. Another important question is: how important is HBV viremia without HBsAg reverse seroconversion in a patient on ibrutinib? The index cohort patient had fluctuating levels of HBV DNA for 6 months without developing abnormalities in aminotransferases or reverse seroconversion. Is it possible that low-level viremia during ibrutinib is of no clinical importance and may resolve without antiviral therapy? Another unanswered question is whether the cumulative probability of HBV reactivation is high enough to merit routine HBV DNA monitoring or HBV prophylaxis during or after therapy.

In conclusion, we demonstrate that HBV reactivation may be a risk during and after ibrutinib therapy in those with past HBV infection. We recommend that practitioners systematically screen patients for HBV infection before starting ibrutinib therapy. Further studies are needed to understand the mechanism, timing of risk, impact of underlying hematologic malignancy and pretreatment, importance of low-level viremia during ibrutinib therapy, and optimal monitoring strategies for patients at risk for HBV reactivation on this effective, widely used novel therapy.