Tolerability to romidepsin in patients with relapsed/refractory T-cell lymphoma

Romidepsin—a structurally unique, potent, bicyclic class 1 selective histone deacetylase inhibitor [1–3]—is approved by the United States Food and Drug Administration for patients with cutaneous T-cell lymphoma (CTCL) who have received at least one prior systemic therapy and patients with peripheral T-cell lymphoma (PTCL) who have received at least one prior therapy [4].

CTCL is a primarily indolent, heterogeneous group of non-Hodgkin lymphoma (NHL) with a poor prognosis in advanced stage disease [5]. CTCL arises when CD4+ malignant T cells localize to the skin [6]; however, in later disease stages, patients may also have lymph node, blood, and/or visceral involvement [7]. Patients with CTCL often experience intolerable itching (pruritus), visual (cosmetic) skin changes, and frequent infections [8–10]. PTCL is an aggressive, uncommon form of NHL typically associated with a poor prognosis [11]. Disease arises from mature, post-thymic T cells or natural killer (NK) cells [12]. Clinical features vary widely in this heterogeneous group of diseases, with varying symptoms and organ involvement. However, hematologic abnormalities are common in patients with PTCL and may be due to disease involvement in the bone marrow or prior myelosuppressive chemotherapy [13]. Durable clinical responses in PTCL or advanced-stage CTCL are difficult to achieve [5, 12, 14].

A phase 1 trial conducted by the National Cancer Institute (NCI) demonstrated activity of romidepsin in T-cell lymphoma [15]. A phase 2 NCI trial was then initiated to evaluate the safety and efficacy of romidepsin in relapsed or refractory (R/R) CTCL or PTCL [16, 17]. Based on initial results from the NCI trial, separate pivotal registration trials were also conducted in each indication: GPI-04-0001 in R/R CTCL [18] and GPI-06-0002 in R/R PTCL [13].

In GPI-04-0001, single-agent romidepsin therapy resulted in durable responses in patients with R/R CTCL who had received at least one prior systemic therapy with an objective response rate (ORR) of 34% (33/96, including 6% [6/96] complete response [CR]) and median duration of response (DOR) of 15 months (range, < 1-20+; median follow-up not reported) [18]. Similar responses to romidepsin were observed in all stages of disease and across all disease compartments: skin, lymph nodes, and blood [18]. The most common romidepsin-related adverse events (AEs) in CTCL were gastrointestinal or asthenic conditions, primarily grade 1–2 [18].

In GPI-06-0002, patients with R/R PTCL achieved durable responses with romidepsin treatment, with an ORR of 25% (33/130, including 15% [19/130] confirmed/unconfirmed CR) [13] and median DOR of 28 months (range < 1-48+) with median follow-up of 22.3 months [19]. Romidepsin demonstrated comparable efficacy in the 3 most common PTCL subtypes: PTCL not otherwise specified, angioimmunoblastic T-cell lymphoma, and anaplastic lymphoma kinase–negative anaplastic large cell lymphoma. Patient baseline characteristics or prior treatments did not affect the response [13]. The most frequent romidepsin-related AEs were nausea and asthenia/fatigue, which were primarily grade 1–2 and did not result in drug discontinuation [13].

In both R/R CTCL and R/R PTCL, romidepsin was not correlated with clinically meaningful QTc prolongation [13, 18, 20], and a similar toxicity profile was observed [4, 13, 18]. The purpose of this analysis was to examine common AE data in detail from the pivotal trials of romidepsin in R/R CTCL and R/R PTCL and to evaluate the AE profile in these 2 groups of patients.

Single-agent romidepsin has demonstrated clinical activity in patients with R/R CTCL (34% ORR including 6% CR and median DOR of 15 months) [18] and PTCL (25% ORR including 15% confirmed/unconfirmed CR and median DOR of 28 months) [19]. The types of common AEs reported with romidepsin therapy were similar across the 2 indications, including hematologic toxicities, gastrointestinal events, asthenia/fatigue, or infections [13, 16–18]. Approved romidepsin dosing was generally well tolerated in patients with R/R CTCL or PTCL, with AEs resulting in < 20% of patients discontinuing therapy and < 15% of patients requiring dose reductions.

Despite the similar AE profiles, there were some notable differences across the indications. Patients with R/R PTCL experienced more frequent and more severe hematologic toxicities and more frequent grade ≥ 3 infections (all types pooled). Although it is recognized that the cause of the thrombocytopenia with romidepsin is not due to a direct myelosuppressive effect but rather to defective megakaryocytic budding [21], it seems likely that the hematologic AEs observed in patients with PTCL may be attributed to a reduced megakaryocyte pool due to prior myelosuppressive chemotherapy and/or bone marrow disease involvement [13]. In this study, patients with R/R PTCL who had received ≥ 3 prior systemic therapies had significantly higher treatment-related grade ≥ 3 thrombocytopenia compared with those who had < 3 prior systemic therapies. Prior monoclonal antibody exposure in patients with R/R PTCL (predominantly alemtuzumab or rituximab) was associated with a significantly higher incidence of treatment-related grade ≥ 3 infection and neutropenia. This analysis could not be performed for patients with R/R CTCL, as rates of treatment-related grade ≥ 3 infections were too low to assess differences, and only 4 patients had prior MAb exposure.

For this analysis, all infection types were pooled to accurately reflect the total infection risk. Because AE reporting separates each infection type, the incidence of each subset may fall below the cutoff of those reported as “common” AEs, and infection risk may be overlooked. Additionally, comparing infection risk across agents is difficult when they are not pooled. In the pivotal study of pralatrexate for the treatment of R/R PTCL, only the specific infections seen in ≥ 10% of patients (upper respiratory tract infection and sinusitis) were reported [22]. In patients with R/R CTCL treated with vorinostat, only upper respiratory infection was reported in ≥ 10% of patients [23]. No specific infection type was reported in ≥ 10% of patients with R/R PTCL or CTCL treated with romidepsin. However, when pooled, 55% of patients with R/R PTCL and 50% of patients with R/R CTCL (19% and 8% grade ≥ 3, respectively) experienced infection during romidepsin treatment. Notably, most of the infections were not considered related to romidepsin treatment and thus may have been due to the increased risk of infection related to disease state or prior therapies. As a result of compromised skin integrity, patients with CTCL experience frequent infections [9, 10, 24, 25].

Because of the risk of infection in patients with T-cell lymphoma, steps to minimize infectious complications should be taken. Romidepsin does not require a central IV catheter, which has been associated with an increased risk of infection in patients with CTCL [26]. Strict adherences to aseptic technique should be taken when establishing a peripheral IV line for drug administration. Additionally, for patients with poor immune function due to heavy pretreatment or prior monoclonal antibody therapy, prophylactic antibiotic or antiviral medication may be appropriate.

Electrocardiogram changes were described in studies with romidepsin [27–29]; however, there was no evidence of myocardial damage or significant changes in left ventricular ejection fraction in romidepsin clinical trials with routine cardiac monitoring [28, 29]. It was shown that clinically insignificant QTc effects observed with romidepsin treatment were likely related to antiemetic administration [20, 28, 30]. Changes in ECG parameters, including QTc intervals, are a class effect of 5-hydroxytryptamine 3 receptor agonist antiemetics [31, 32]. Granisetron may have less of an effect on the QT interval than other 5-hydroxytryptamine 3 receptor agonist antiemetics [32]. Despite the use of prophylactic antiemetics, > 50% of patients with R/R PTCL or CTCL treated with romidepsin experience drug-related nausea and/or vomiting, although the majority of events are grade 1–2 [4, 29].

The greatest incidence of all grade and grade ≥ 3 AEs and the majority of discontinuations due to AEs occurred during cycles 1–2. Early discontinuations primarily related to infection, thrombocytopenia, or ECG abnormalities confirm the need to closely monitor heavily pretreated patients, in particular those who have received prior monoclonal antibody therapy or those with comorbidities, with consideration of prophylactic antibiotic or antiviral medication, or a lower initial romidepsin dose of 10 or 12 mg/m² in such patients. Despite this, however, 28% and 36% of patients with R/R PTCL or CTCL, respectively, tolerated romidepsin for at least 6 cycles.

Single-agent romidepsin leads to durable response in a subset of patients with R/R PTCL or CTCL. This study demonstrated that patients with R/R PTCL or CTCL have similar AE profiles with romidepsin treatment, although patients with PTCL experienced more frequent and more severe hematologic toxicities and more frequent grade ≥ 3 infections. Extended dosing of romidepsin is feasible in responding patients without cumulative toxicities.