Potential efficacy of the oral histone deacetylase inhibitor vorinostat in a phase I trial in follicular and mantle cell lymphoma

Abstract

Vorinostat (suberoylanilide hydroxamic acid, SAHA, Zolinza) is a histone deacetylase inhibitor with clinical activity in cutaneous T‐cell lymphoma (CTCL). A phase I trial of oral vorinostat was conducted in Japanese patients with malignant lymphoma. Vorinostat 100 or 200 mg was administered twice daily for 14 consecutive days followed by a 1‐week rest interval. Of 10 patients enrolled, four had follicular lymphoma (FL), two mantle cell lymphoma (MCL), two diffuse large B‐cell lymphoma, and two CTCL (median age, 60 years; median number of prior regimens, 3). Vorinostat was well tolerated up to 200 mg with only one of six patients developing a dose‐limiting toxicity (DLT; Grade 3 anorexia/hypokalemia). Common Grade 3 events were reversible neutropenia (30%), thrombocytopenia, and hypermagnesemia (20% each). The median number of treatment cycles was five (range, 1–36); two patients were continuing treatment. The overall response rate was 40%, with two complete responses/unconfirmed (CRu) and one partial response among FL patients and one CRu among MCL patients. One FL patient maintained CRu for 18.0 months. The median time to achieve CRu among the three patients was 8 months. These data suggest that further investigations of vorinostat in non‐Hodgkin lymphoma, focusing on FL and MCL, are warranted. (Cancer Sci 2009)

In 2002, the worldwide incidence of non‐Hodgkin lymphoma (NHL) was estimated to be 300 571, corresponding to 171 820 deaths.^{( 1 )} Common types of NHL include diffuse large B‐cell lymphoma (DLBCL) and follicular lymphoma (FL). Because relapse is common in NHL, especially for indolent B‐cell lymphoma even after the advent of rituximab, novel effective treatments are clearly needed.

Vorinostat (Zolinza; Merck, Whitehouse Station, NJ, USA) is an inhibitor of class I and II histone deacetylase (HDAC) enzymes that regulate transcriptional and post‐transcriptional processes by removing acetyl groups from proteins.^{( 2 )} In preclinical studies, vorinostat exposure resulted in cell cycle arrest and antiproliferative and antitumor activities in lymphoma models.^{( 3} , ⁴ , ^{5 )} In October 2006, vorinostat was approved in the United States for the treatment of cutaneous manifestations in patients with cutaneous T‐cell lymphoma (CTCL) who have progressive, persistent, or recurrent disease on or following two systemic therapies.^{( 6} , ^{7 )} Clinical activity of vorinostat has also been observed in patients with transformed DLBCL,^{( 8} , ^{9 )} but it was limited against DLBCL.^{( 10 )}

Recently, the results of a phase I study of oral vorinostat in American patients with advanced leukemias and myelodysplastic syndromes were reported.^{( 11 )} The present study was conducted to evaluate vorinostat in Japanese patients with malignant lymphoma (ML). The primary objectives were to determine the maximum tolerated dose (MTD) and evaluate dose‐limiting toxicities (DLTs) and the overall safety profile of vorinostat in the first cycle. Secondary objectives were to assess the pharmacokinetics and the overall safety profiles in subsequent cycles and to determine the recommended phase II dose in patients with ML. Exploratory objectives were to evaluate response according to the International Workshop Response Criteria ^{( 12 )} and to evaluate the pharmacodynamic effects on normal cells via assessment of histone acetylation.

Materials and Methods

Patient eligibility.  Patients with ML who had received one to four prior regimens and had relapsed after complete response or partial response or were refractory to previous chemotherapies were eligible. Patients must have been ≥20 years, but ≤75 years in age, with Eastern Cooperative Oncology Group performance status of 0–1 and adequate hepatic (total bilirubin ≤1.5× the upper limit of normal [ULN]; aspartate aminotransferase and alanine aminotransferase ≤2.5 × ULN), hematologic (neutrophil count ≥1500/μL; platelet count ≥100 000/μL; hemoglobin ≥9.0 g/dL, prothrombin and activated partial thromboplastin time ≤1.5 × ULN), and renal function (creatinine ≤1.5 × ULN and creatinine clearance ≥50 mL/min). Exclusion criteria included: prior HDAC inhibitor treatment or allogeneic stem cell transplantation, prior therapy within 4 weeks of study, central nervous system involvement, uncontrolled concomitant illness, pregnancy, lactation, or concurrent malignancies. This open‐label, single‐arm, non‐randomized phase I trial (Protocol 030) was approved by the institutional review boards of the National Cancer Center Hospital, Tokyo, Japan, and Aichi Cancer Center Hospital, Nagoya, Japan. Prior to enrollment, all patients provided written informed consent in accordance with the Declaration of Helsinki.

Study design, treatment, and evaluation.  Oral vorinostat 100 or 200 mg was administered twice daily (bd) for 14 consecutive days followed by a 7‐day rest in a 3‐week cycle. The study used a ‘3 + 3’ dose escalation design as described elsewhere.^{( 11 )} Dose‐limiting toxicities were: (i) Grade 4 neutropenia persisting for ≥5 days; (ii) Grade ≥3 infection with Grade 3 or 4 neutropenia; (iii) Grade ≥3 thrombocytopenia requiring blood transfusion; (iv) Grade 4 thrombocytopenia or anemia; (v) Grade ≥3 non‐hematologic toxicities other than anorexia, nausea/vomiting, and fatigue; or (vi) Grade ≥3 anorexia, nausea/vomiting, and fatigue refractory to treatment during the first cycle. The MTD was defined as the dose level at which all of the three patients initially enrolled or three or more of the six enrolled patients developed any DLT(s) in the first cycle. For patients on dose level 2 (200 mg), the dose was reduced to 100 mg if a DLT occurred during the first or subsequent cycles, or if a drug‐related adverse event (AE) resolved to Grade 1 or less within 1 week after cycle completion. Once a patient underwent a dose reduction, an increase to the previous dose level was not permitted even after resolution of the toxicity. Patients were to be discontinued for disease progression, unacceptable toxicity, withdrawal of consent, lost to follow‐up, or at the investigator’s discretion. Safety was assessed using National Cancer Institute Common Terminology Criteria for Adverse Events, version 3.0.^{( 12 )} Adverse events that were definitely, probably, or possibly related to vorinostat were categorized as drug‐related. Efficacy was assessed according to the International Workshop Response Criteria.^{( 13 )}

Pharmacokinetic and pharmacodynamic analyses.  The pharmacokinetics of vorinostat were investigated in Cycle 1. Patients received vorinostat once daily on Days 1 and 17, and bd on Days 3 to 16 using 100 mg on dose level 1 and 200 mg on dose level 2 as a schedule in Cycle 1. Intensive blood sampling was performed at pre‐dose and 0.25, 0.5, 1, 1.5, 2, 3, 4, 6, 8, 10, 12, and 24 h post‐dose on Days 1 and 17. Based on measured serum concentrations, area under the concentration‐time curves from time 0 to 12 h (AUC_0‐12hr), AUC from time zero to infinity (AUC_0‐∞), the maximum concentration (C _max), time to maximum concentration (T _max), and terminal half life (t _1/2) were determined on Days 1 and 17. The geometric mean ratio for AUC_0‐∞ or AUC_0‐12h on Day 1 and AUC_0‐12h on Day 17 were calculated. Trough values of vorinostat on Days 4 and 10 were also determined. Urine samples were collected at pre‐dose and during 0–6, 6–12, and 12–24 h post‐dose on Days 1 and 17. Based on measured urine concentrations, urinary recovery (% of dose) was calculated.

Blood samples (5 mL) were collected on Day 1 of Cycle 1 at pre‐dose and at 2, 4, 8, and 24 h. Peripheral blood mononuclear cells (PBMCs) were lysed and utilized to assess histone H3 acetylation by enzyme‐linked immunosorbent assay as previously described.^{( 11 )}

Results

Patient population.  The baseline patient characteristics of the 10 enrolled patients are shown in Table 1. The median number of prior regimens was three (range, 1–4). The median study duration was 3.1 months (range, 0.4–26.1+).

Table 1.

 Baseline patient characteristics (n = 10)

Characteristic	Number of patients
Median age, years (range)	60 (52–74)
Male	4
Female	6
ECOG performance status
0	8
1	2
Disease type
FL	4
MCL	2
DLBCL	2
CTCL	2
Stage at enrollment
II	4
III	3
IV	3
Number of prior regimens
1	1
2	2
3	4
4	3
Median	3

CTCL, cutaneous T‐cell lymphoma; DLBCL, diffuse large B‐cell lymphoma; ECOG, Eastern Cooperative Oncology Group; FL, follicular lymphoma; MCL, mantle cell lymphoma.

Safety and toxicity.  Nine patients completed the first treatment cycle and were evaluated for DLTs; one patient was excluded due to disease progression during the first cycle. None of the three evaluable patients who received vorinostat 100 mg bd developed a DLT. One of six patients treated at the 200 mg bd dose level developed a DLT (Grade 3 anorexia and hypokalemia). In an ex‐Japanese trial, vorinostat administered at a dose of 300 mg bd for 14 consecutive days every 3 weeks exceeded the MTD, and 200 mg was defined as the MTD on this schedule.^{( 11 )} Therefore, the dosing regimen of this trial was changed so that the maximum dose level was 200 mg bd Although the MTD was not reached, vorinostat 200 mg bd, which was the maximum dose in this study, for 14 consecutive days every 3 weeks is the recommended phase II dose and schedule of vorinostat in Japanese patients with NHL.

The most common drug‐related AEs for the first and subsequent cycles are shown in Table 2. All 10 patients received at least one dose of vorinostat and were evaluable for safety. Hematologic drug‐related adverse events included neutropenia, thrombocytopenia, anemia (70% each), and leukopenia (60%). Non‐hematologic drug‐related adverse events included anorexia, hyperglycemia, proteinuria (70% each); and fatigue, hypocalcemia, nausea, and diarrhea (60% each). Most of these events were Grade ≤2, and all were reversible and rapidly resolved during the rest period. Grade 3 drug‐related adverse events were neutropenia (30%), thrombocytopenia, hypermagnesemia (20% each), anorexia, diarrhea, hypokalemia, and hyponatremia (10% each).

Table 2.

 Most common drug‐related adverse events in all cycles (n = 10)

Events	Grade 1 or 2	Grade 3
	n	n
Hematologic
Neutropenia	4	3
Thrombocytopenia	5	2
Anemia	7	0
Leukopenia	6	0
Non‐hematologic
Anorexia	6	1
Hyperglycemia	7	0
Proteinuria	7	0
Fatigue	6	0
Hypocalcemia	6	0
Nausea	6	0
Diarrhea	5	1
Increased alanine aminotransferase	5	0
Dysgeusia	5	0
Weight loss	5	0
Hypermagnesemia	3	2
Hypokalemia	4	1
Hyponatremia	3	1
Vomiting	4	0
Headache	4	0
Nail disorder	4	0
Rash	4	0
Decreased albumin	4	0
Increased creatinine	4	0
Increased lactate dehydrogenase	4	0
Decreased total protein	4	0

Drug‐related adverse events for all grades reported in 40% or above of patients are listed.

Four patients had dose modifications in the second cycle or later due to AEs (anorexia, Cycle 1; leukopenia and neutropenia, Cycle 1; neutropenia, Cycle 3; hypermagnesemia, Cycle 2). Six patients discontinued vorinostat in Cycles 1–5 due to disease progression, and two patients were continuing on vorinostat treatment as of March 2008. There were no discontinuations due to AEs and no sudden deaths. However, two patients (patients #6 and #7 in Table 3) who responded to vorinostat and therefore received it for several months developed Grade 1 nail brittleness (Fig. 1).

Table 3.

 Clinical outcomes of patients

Patient number	Disease type	Dose (mg)	Number of prior regimens	Best overall response†	Duration of best response‡ (month)	On‐study duration (month)
1	DLBCL	100 bd (14/21 days)	3	PD	–	0.8
2	CTCL		3	PD	–	0.4
3	MCL		4	SD	–	2.7
4	CTCL		3	SD	–	3.6
5	FL	200 bd (14/21 days)	3	SD	–	2.4
6	FL		4	CRu	18.0+	26.1+
7	FL		1	PR	6.2	25.9+
8	MCL		2	CRu	1.7	23.1
9	FL		2	CRu	1.3	12.7
10	DLBCL		4	PD	–	1.4

†Responses were determined according to the International Workshop Response Criteria^{( 13 )}. ‡For patients who achieved CRu, only the duration of CRu is shown (duration of PR is not shown). −, not applicable; +, ongoing; bd, twice daily; CRu, complete response/unconfirmed; CTCL, cutaneous T‐cell lymphoma; DLBCL, diffuse large B‐cell lymphoma; FL, follicular lymphoma; MCL, mantle cell lymphoma; PD, progressive disease; PR, Partial response; SD, stable disease.

Figure 1.

 Chips and streaks of the nails observed several months after oral intake of vorinostat in the two patients (patients #6 and #7 in Table 3) who responded to vorinostat.

Efficacy.  The clinical outcomes of the patients are shown in Table 3. The overall response rate was 40%. Of the four patients with FL assigned to 200 mg bd, two achieved complete responses/unconfirmed (CRu), and one had a partial response (PR). One of the two patients with mantle cell lymphoma (MCL) received 200 mg bd and achieved a CRu. In addition, three patients (MCL, CTCL, and FL) had stable disease. The median time to achieve CRu among the three patients was 8 months.

Pharmacokinetics and pharmacodynamics.  Maximum plasma concentrations were observed in the range of 1.50–3.44 h after oral intake in both single and multiple doses of vorinostat at both dose levels. Vorinostat was then rapidly eliminated with an apparent t_1/2 of 1.25 to 1.40 h (Table 4, Fig. 2). The accumulation ratios of vorinostat following 14 days of multiple dosing at dose levels 1 and 2 were 1.35 and 1.34, respectively (Table 4). Two inactive metabolites (the O‐glucuronide of vorinostat and 4‐anilino‐4‐oxobutanoic acid) were measured in serum. AUCs of the O‐glucuronide and 4‐anilino‐4‐oxobutanoic acid metabolites following administration of vorinostat in the fed state were 2‐ to 4‐fold and 8‐ to 10‐fold higher, respectively, than that of vorinostat (data not shown). The O‐glucuronide and 4‐anilino‐4‐oxobutanoic acid metabolites were the main compounds recovered in the urine following administration of vorinostat. Only a small amount of vorinostat (less than 1% of the dose) was observed in the urine (Fig. 3).

Table 4.

 Summary of pharmacokinetic parameters of vorinostat†

Parameter	n	Day 1	n‡	Day 17
		SD, fed		MD, fed
		Geometric mean ± SD		Geometric mean ± SD
Dose level 1 (100 mg bd)
AUC0–∞ (μM h)	4	1.03 ± 0.45	3	1.42 ± 0.61
AUC0–12h (μM h)	4	1.01 ± 0.44	3	1.41 ± 0.59
C max (μM)	4	0.28 ± 0.10	3	0.34 ± 0.12
T max (h), median	4	3.44	3	1.50
t 1/2 (h), harmonic mean	4	1.27	3	1.34
Accumulation ratio§	–	–	3	1.35
Dose level 2 (200 mg bd)
AUC0–∞ (μM h)	6	2.67 ± 0.88	5	3.77 ± 1.29
AUC0–12h (μM h)	6	2.65 ± 0.87	5	3.30 ± 0.27
C max (μM)	6	0.66 ± 0.23	5	0.83 ± 0.28
T max (h), median	6	2.98	5	3.05
t 1/2 (h), harmonic mean	6	1.25	5	1.40
Accumulation ratio§	–	–	5	1.34

†Patients received vorinostat once daily on Days 1 and 17, and twice daily on Days 3 to 16 in the 1st cycle for pharmakokinetic data analysis. ‡One patient received 100 mg MD bd was excluded due to incompletion of the 1st cycle, and one patient received 200 mg MD bd was excluded due to occurrence of dose‐limiting toxicity. §AUC_{0–12h, Day 17}/AUC_{0–12h, Day 1}. AUC_{0 – ∞} , area under the concentration–time curve from time zero to infinity; AUC_0–12h, AUC from time zero to 12 h; bd, twice daily; C _max, maximum concentration; T _max, time to maximum concentration; MD, multiple doses; SD, single dose; t _1/2, terminal half life.

Figure 2.

 Mean serum concentration‐time profiles following administration of single/multiple oral doses of vorinostat. In each dose level (Lv), vorinostat was administered with single dose on Day 1, and then with multiple doses of twice daily (bd) on Days 3–16 and once daily (qd) Day 17. All of the doses were performed in the fed state. In Lv 1, pharmacokinetic data of vorinostat were obtained from four patients on Days 1, 4 and 10, and from three patients on Day 17 (one subject discontinued on Day 11 because of impossibility of dose‐limiting toxicity [DLT] assessment). In Lv 2, pharmacokinetic data of vorinostat were obtained from six patients on Days 1, 4, and 10, and from five patients on Day 17 (one subject discontinued on Day 11 because of DLT).

Figure 3.

 Mean urinary excretion following administration of single/ multiple oral doses of vorinostat. In each dose level (Lv), vorinostat was administered with a single dose on Day 1, and then with multiple doses twice daily (bd) on Days 3–16 and once daily (qd) on Day 17. All of the doses were performed in the fed state. β‐Oxidative form and glucuronide indicate 4‐anilino‐4‐oxobutanoic acid and O‐glucuronide of vorinostat, respectively.

PBMCs from all 10 patients enrolled were assessed for pharmacodynamics. Histone H3 acetylation was induced 1.8‐fold relative to baseline by 8 h after treatment and returned to baseline by 24 h in PBMCs of patients receiving the 200 mg, but not the 100 mg, vorinostat dose level (Fig. 4).

Figure 4.

 Average histone H3 acetylation on Day 1 by dose group. Peripheral blood mononuclear cells from all 10 patients were assessed.

Discussion

In an ex‐Japanese trial, vorinostat administered at a dose of 300 mg bd for 14 consecutive days every 3 weeks exceeded the MTD, and 200 mg was defined as the MTD for that schedule.^{( 11 )} Therefore, the dosing regimen of the present trial was changed so that the maximum dose level was 200 mg bd for 14 consecutive days every 3 weeks. Although two DLTs occurred in one of 10 patients (Grade 3 anorexia and Grade 3 hypokalemia), the MTD was not reached in this trial. Therefore, the maximum acceptable dose was judged as 200 mg bd for 14 consecutive days every 3 weeks.

The safety profile of vorinostat in Japanese patients was similar to that in non‐Japanese patients.^{( 6} , ⁷ , ⁸ , ⁹ , ¹⁰ , ¹¹ , ¹⁴ , ^{15 )} Also, the pharmacokinetic results were consistent with those in non‐Japanese patients.^{( 8} , ^{14 )} Vorinostat was rapidly eliminated from the systemic circulation with slight accumulation following multiple doses. The hematologic AEs with vorinostat were reversible, as the patients quickly recovered from these events during the rest period. Therefore, it may be appropriate to study this agent in combination with traditional chemotherapy drugs, which often cause hematologic toxicities in patients with hematological malignancies. Another HDAC inhibitor, depsipeptide, has been associated with cardiotoxicity, such as ventricular tachycardia and QT interval corrected for heart rate (QTc) interval prolongation.^{( 16} , ^{17 )} However, no patients in this study had a clinically significant cardiac arrhythmia or QTc interval prolongation. An interesting finding worth mentioning is that patients who responded to treatment and received vorinostat for a longer duration developed Grade 1 nail brittleness. To our knowledge, this finding has not been previously reported in trials with vorinostat.

The overall response rate was 40%, including three CRus and one PR. One of the CRus occurred in a heavily pretreated patient who had received four prior regimens. Two CRus and one PR occurred in FL patients, and one CRu occurred in an MCL patient. It should be noted that activity has been achieved at doses up to 200 mg on this schedule, as was reported previously in other solid and hematologic malignancies.^{( 11} , ^{18 )} The median time to achieve CRu among the three patients was 8 months. Therefore, it may be important that the establishment of the optimal dose and manner without either severe toxicities or disease progression enables patients to take this kind of agent with mild activity for a longer duration and to achieve favorable outcomes. Previous pre‐clinical studies showed that vorinostat suppressed translation of cyclin D1 in MCL cells.^{( 19 )} In the present study, histone acetylation was induced 1.8‐fold in PBMCs from patients on the 200 mg bd dose level. Although the MTD was not reached, favorable responses were achieved with vorinostat as monotherapy in patients with FL and MCL. The vorinostat dose was not increased above 200 mg bd for 14 consecutive days every 3 weeks because of known toxicity data from ex‐Japanese studies.

In conclusion, oral vorinostat was well tolerated up to 200 mg bd for 14 consecutive days every 3 weeks in Japanese patients with NHL. Although evaluation of response was an exploratory objective in this phase I trial, the findings of this study highlight the potential single‐agent activity of vorinostat in FL or MCL. This study provides support for further investigation of vorinostat as monotherapy or in combination with other active agents in NHL. Thus, these data suggest that phase II studies, focusing on FL and MCL, are warranted and a phase II study for indolent lymphoma including FL and MCL has been started in Asia.

Disclosure Statement

Justin L. Ricker was an employee of Merck & Co., Inc., at the time this manuscript was written. Tetsuya Otsuki is an employee of Banyu Pharmaceutical Co., Ltd., and holds stock and stock options of Merck & Co., Inc. The remaining authors report no potential conflicts of interest.