Page 20 - 2019_05-HaematologicaMondo-web

866
Editorials
cell-to-cell contact is necessary for the infection of new T cells, while the expansion of the HTLV-1 proviral load is achieved by proliferation of infected T cells, which leads to a clonally diverse neoplastic population11 (Figure 1).
Extensive molecular aberrations in HTLV-1-infected T cells, often accumulating over decades, lead to the devel- opment of ATL in approximately 3-5% of seropositive carriers. HTLV-1 induced leukemogenesis is a complex, multistep process, driven by Tax and HBZ. Tax-induced upregulation of IL-15, IL-15Ra, and EZH-2 leads to chron- ic inflammation and polycomb repressive complex 2 (PRC2) hyperactivation, with genome-wide H3K27me3 accumulation.12 Expression of HBZ by HTLV-1 infected T cells results in increased proliferation, impaired apoptosis, and disruption of genomic integrity.13 Analysis of the somatic mutation landscape of ATL reveals common mutations at TP53 and IRF4, and copy number alterations at PD-L1 and CDKN2A.14
HTLV-1 seroprevalence rates mean that ATL predomi- nates in endemic regions, accounting for up to 35% of all T-cell lymphomas in endemic areas in Japan and 15-20% in Peru. However, these figures are only 1-2% in North America and Europe.15 ATL can present with four clinical subtypes: acute, lymphomatous, chronic, and smolder- ing. A consensus report highlighting the clinical features and treatment guidelines of these subtypes (including an increasingly appreciated fifth subtype: aggressive extran- odal primary cutaneous) was recently updated.16
Retrospective studies have described significant clinical and biological differences between Japanese ATL and North American ATL, including a slight female predomi- nance, a younger median age at diagnosis (61-67 vs. 50- 54 years), and a higher frequency of aggressive subtypes (acute and lymphomatous) (approx. 75% vs. 88-94%).6,7 There are also differences in the mutational landscape, with significantly higher mutation rates for epigenetic regulators, and fewer T-cell receptor/NF-κB pathway alterations in North American ATL compared to Japanese ATL.17
Despite advances in our understanding of the biology of HTLV-1 and ATL, prognosis remains very poor, with median overall survival (OS) of 8.3 months (acute), 10.6 months (lymphomatous), 31.5 months (chronic), and 55 months (smoldering);18 western ATL patients may have a worse prognosis.6,7 Treatment strategies differ significant- ly between endemic and non-endemic regions. In Japan, the LSG15 regimen was superior to CHOP-14, with high- er complete remission (CR) rate and a trend towards improved 3-year OS (24% vs. 13%).19 However, this regi- men is not routinely used outside Japan, and the most fre- quently used chemotherapies in North American ATL are CHOP-like regimens, with overall response rates (ORR) of approximately 60-75% and CR rates of 13-36%.6,7 Consolidation with allogeneic hematopoietic stem cell transplantation (HSCT) is generally recommended for eli- gible patients with aggressive ATL subtypes, with Japanese studies showing 3-4 year OS ranging between 26% and 36%.20
Unfortunately, most ATL patients relapse, and multia- gent salvage chemotherapy is generally ineffective.18 The discovery that C-C chemokine receptor 4 (CCR4) is expressed in over 90% of ATL cases, led to the clinical
development of mogamulizumab, a glycoengineered anti-CCR4 monoclonal antibody with a defucosylated Fc region that enhances ADCC. In 2012, mogamulizumab was approved for ATL in Japan in the relapsed setting on the basis of a Phase II trial that showed a 50% ORR,4 and in 2014 was approved for chemotherapy-naïve patients, based on a randomized Phase II trial in combination with the mLSG15 regimen.5 Both studies were quite small (28 and 53 patients, respectively) with ORR as the primary end point. Up-dated outcomes analyses appear to show a real, but relatively modest, benefit for mogamulizumab, with median PFS and OS of 5.2 and 14.4 months for the single arm R/R ATL cohort and 1-year progression-free survival (PFS) 47% and 29% for mLSG15 + moga- mulizumab versus mLSG15 in the randomized front-line study.21
In this context, the study by Phillips et al.8 aimed to determine if the incremental, but encouraging, outcome improvements with mogamulizumab in Japanese ATL could be replicated in non-Japanese ATL. This interna- tional Phase II study, conducted at 22 centers, random- ized (2:1 ratio) 71 patients with R/R ATL with at least one prior line of therapy to either mogamulizumab (n=47) or investigator choice chemotherapy (n=24: GemOx=21; pralatrexate=2; DHAP=1). The primary objective of the study was confirmed overall response rate (cORR), defined as a response sustained for ≥8 weeks. In the mogamulizumab arm, cORRs by investiga- tor and independent review were 15% and 11%, respec- tively, notably inferior to that of the Japanese registration study.3 Remarkably, the cORR in the investigator’s choice arm was 0%. Concordant with the Japanese Phase II study, the best responses to mogamulizumab by com- partment were in blood (54%, all CR) and skin (44%), with no CR in lymph nodes. Responses were observed in all clinical subtypes.
Given the study design, with 18 out of 24 patients (75%) on the investigator choice arm crossing over to the investigational arm, it was not possible to assess any OS benefit from mogamulizumab. Median PFS was poor in each arm (0.93 months for mogamulizumab vs. 0.88 months for chemotherapy), much worse than the Japanese pivotal study (PFS, 5.2 months; OS, 14.4 months).3 The authors concluded that the inclusion of primary refractory patients, stricter cORR criteria (8 weeks vs. 4 weeks), and a higher incidence of poor base- line prognostic factors may account for the inferior effica- cy of mogamulizumab in this trial compared to the Japanese studies. In addition, 40% of the patients on the mogamulizumab arm of this trial had received prior zidovudine/interferon-Alpha (IFNa) therapy, whereas no patient had received it in the Japanese studies, suggesting that mogamulizumab may be less effective after zidovu- dine/IFNa failure. Key differences in disease biology between western and Japanese ATL may also explain dif- ferences in response. For example, the presence of CCR4 gain-of-function mutations that have been associated with better outcomes following mogamulizumab therapy in some studies22 were not assessed.
Despite the somewhat disappointing results, this is an important study because it gives us the first prospective cohort of homogeneously-treated, non-Japanese ATL
haematologica | 2019; 104(5)