It’s a BRAF new world for hairy cell leukemia

In this issue of Blood, Handa et al¹ describe the long-term outcomes in patients with relapsed or refractory hairy cell leukemia (HCL) treated with vemurafenib monotherapy, focusing on retreating relapsed patients with vemurafenib and confirming that the treatment is a safe and effective strategy.

Since the first description of HCL by Bouroncle et al in 1958,² there have been significant advances in understanding the biology of the disease and improvements in treatment. In the 1980s, the expected median survival was only 4 years.³ The advent of successful therapy with purine analogues (PAs), which induce durable remissions for most patients, changed this gloomy prognosis, delivering near normal life expectancy for patients.⁴ However, approximately half of the patients treated with PAs relapse and some eventually become refractory to, or unable to tolerate, the agents. Until recently, treatment options for patients with PA-refractory disease have been limited. Discovery of the almost universal occurrence of BRAF V600E mutation in classic HCL and the importance of BRAF-MEK-extracellular signal-regulated kinase signaling in the pathogenesis has been a defining moment in the understanding of this disease.⁵ Inhibition of BRAF in vitro reverses the classic morphological and genetic features of the hairy cell and induces the death of the cell. In 2011, the same year the high frequency of BRAF V600E mutations was first reported in HCL, a highly selective inhibitor of BRAF V600E, vemurafenib, received US Food and Drug Administration approval for treatment of patients with melanoma harboring the mutation. Subsequent studies, including Handa et al, have confirmed in vivo the therapeutic efficacy of vemurafenib for HCL.⁶ 

Handa et al present the results from the US arm of a previously reported prospective trial⁶ of vemurafenib treatment of relapsed/refractory (R/R) HCL, now including 36 patients with a median of 40 months of follow-up. There is considerable interest in following clinical trial data for an extended period, especially in a disease with prolonged survival such as HCL. It is particularly important to establish the rate of response to further therapy at subsequent relapse. The high objective response rate (ORR) of 86% reported previously was unchanged after longer follow-up of this larger cohort and was independent of other variables. However, two-thirds of the patients relapsed after a median relapse-free survival of 19 months. Of those patients, two-thirds were retreated with vemurafenib with the same high ORR and a remission duration of approximately a year. This finding confirms that although relapse is frequent, retreatment with vemurafenib is possible in most cases, although remission durations have been reported to be shorter with each subsequent relapse. This “law of diminishing returns” is also seen with PA monotherapy, but can be addressed by changing therapy, particularly delivering a combination regimen such as PA+rituximab.⁷ Only 2 patients were shown to have acquired resistance to vemurafenib associated with identifiable mutations (CDKN2 loss and activating KRAS mutation), although not all patients underwent genetic analysis. Importantly, overall survival for the 36 patients in the study remained good with 82% at 4 years. However, patients whose initial relapse <1 year after treatment with vemurafenib had a much less favorable outcome (see figure). Again, this is a pattern seen with other therapies in HCL. Patients with poor response and short remissions merit further investigation. There is increasing knowledge of poor prognostic features such as IGHV4-34 rearrangements, TP53 mutations, and unmutated IGHV genes that may help identify such patients earlier and allow for a more tailored treatment strategy.⁸ In addition, sequencing data may identify newly acquired mutations that predict therapeutic resistance.

The safety profile of vemurafenib is good, with no evidence of cumulative toxicity during retreatment. Handa et al make an interesting observation about dose and duration of therapy. The design of the study provided for patients to receive dose escalation in cases of suboptimal response. On analysis of the data, “more was not better” and had an inverse correlation with response. This result is consistent with those in other reports in the literature showing that lower dose vemurafenib (240 mg twice daily) achieves equivalent outcomes compared with the “melanoma” dose (960 mg twice daily). The lower dose was sufficient to block BRAF V600E signaling and induce loss of cyclin D1 expression while maintaining similar overall and complete response rates.⁹ The exquisite sensitivity of hairy cells to inhibition of BRAF translates into rapid clinical responses with relatively small doses of vemurafenib. This knowledge may help to reduce both clinical toxicity and financial burden in future study design. Furthermore, identification of patients who are sensitive to BRAF inhibition and are therefore most likely to benefit from vemurafenib, as well as monitoring for emergence of resistant mutations, may aid rationalization of therapy. The data on resistant mutations were limited in this study, and this area merits further work. Although much can be learned from studying other BRAF-driven diseases, such as melanoma, the uniqueness of HCL means that observations in other diseases may not always be applicable, and mechanisms of resistance may be different.

Most patients with HCL can expect long remissions with PA monotherapy, but there is no sign of a plateau in progression-free survival curves, and alternative therapies are needed to ensure good long-term outcomes for all patients. It is also important to note that PAs, although well tolerated, are highly immunosuppressive drugs. The past 2 years of the COVID pandemic has driven changes in practice to maintain patient safety.¹⁰ More patients with HCL have received a BRAF inhibitor in the frontline setting, and it is important to collect those data. The big question is whether BRAF inhibitors can come close to delivering the exceptionally long remissions seen after frontline PA, an unlikely possibility, given the relatively short remission (12-18 months) reported by Handa et al, albeit in the R/R setting. It is not yet clear which patients would benefit from combination therapies with novel agents at relapse or in the frontline setting, although several trials are in progress to address this question. The focus is now and will be on optimizing dose schedules and developing rational combination regimens to deliver deeper and more durable remissions and to overcome the mechanisms of disease resistance.

Conflict-of-interest disclosure: The author declares no competing financial interests.