Phase I Study of Decitabine in Combination with Vorinostat in Patients with Advanced Solid Tumors and Non-Hodgkin's Lymphomas

Hypermethylation of cytosines in CpG dinucleotides in the promoter regions of tumor-suppression genes and deacetylation of amino acid residues on the histone tails of nucleosomes, represent 2 epigenetic mechanisms of gene silencing that can contribute to tumor formation and progression (1, 2). Both events are considered reversible, and agents that inhibit the enzymes responsible for DNA methylation and histone deacetylation have been developed as anticancer agents (3).

Decitabine (5-aza-2′-deoxycytidine), a nucleoside analogue that is incorporated into DNA and acts as an hypomethylating agent by inhibiting DNA methyltransferase, and vorinostat (suberoylanilide hydroxamic acid), a small molecule that binds and directly inhibits histone deacetylase, are 2 agents with epigenetic effects that have shown clinical antitumor activity and are now approved for the treatment of myelodysplastic syndrome and cutaneous T-cell lymphoma, respectively (4–7). The validation of epigenetic treatments as anticancer strategies has supported an increasing number of trials evaluating epigenetic agents alone or in combination with other agents in both hematologic and solid malignancies (8, 9).

The combination of DNA methyltransferase inhibitor (DNMTi) with a histone deacetylase inhibitor (HDACi) represents an area that is gaining attention in the clinical development of epigenetic therapies. This concept is supported by preclinical evidence that DNA methylation and histone deacetylation are functionally linked, leading to transcriptional inactivation of genes critical for tumorigenesis (10, 11). Moreover, the in vitro combination of a DNMTi with an HDACi in hematologic and solid tumor cell lines have shown synergistic effects resulting in increased gene reexpression and superior antitumor activity (12–14).

The optimal schedule of the combination of a DNMTi with an HDACi has not been established yet. Although most of the preclinical studies performed have used a sequential administration of DNMTi followed by HDACi, it remains unclear whether different schedules of administration may have better clinical activity. In the phase I trial reported here, the combination decitabine and vorinostat was studied for the first time in patients with solid tumors and non-Hodgkin's lymphomas (NHL). Two different schedules of administration, sequential and concurrent, were evaluated. The principal objective of this study was to determine the safety and tolerability of the combination. Secondary objectives included the assessments of pharmacokinetics (PK) and preliminary antitumor efficacy.

Patients were eligible if they had a histologically or cytologically documented advanced solid malignancy or NHL, refractory to standard therapy or for which no standard therapy existed. Other key eligibility criteria included: Eastern Cooperative Oncology Group performance status 0 to 2; adequate hematologic, hepatic, and renal functions [white blood cell count ≥ 3 × 10⁹/L, absolute neutrophil count (ANC) ≥ 1.5 × 10⁹/L, platelets ≥ 100 × 10⁹/L, AST (aspartate aminotransferase)/ALT (alanine aminotransferase) ≤ 2.5 times upper limit of normal, bilirubin within normal limits, creatinine ≤ 150 mmol/L and creatinine clearance ≥ 60 mL/minute]; unlimited prior chemotherapy, radiotherapy or targeted agents with at least 3-week interval (6-week interval if prior nitrosoureas or mitomycin C) between study entry and any prior treatment; no valproic acid or other HDACi for at least 2 weeks before study entry; no prior decitabine; and no known brain metastases.

The institutional review board of both participating centers approved the study, which was conducted in accordance with federal and institutional guidelines.

This was a 2-center, open-label, phase I study in which intravenous decitabine administered on days 1 to 5 was combined with oral vorinostat either in a sequential (vorinostat starting on day 6) or a concurrent schedule (vorinostat starting on day 3), in 28-day cycles. The study began with dose escalation on the sequential schedule, and once the maximum tolerated dose (MTD) was established, accrual began on the concurrent schedule. On the sequential schedule, the starting dose of decitabine was 20 mg/m²/day on days 1 to 5, given as an intravenous infusion over 1 hour. Vorinostat was given at a starting dose of 100 mg twice a day on days 6 to 21. The starting dose of decitabine was chosen on the basis of published data showing good tolerability and higher response rates in patients with myelodysplastic syndrome treated with 20 mg/m²/day for 5 days (15). The starting dose of vorinostat was based on previous monotherapy studies showing that the maximum tolerated dose was 200 mg orally twice a day continuously in patients with solid tumors and 250 mg orally thrice a day for 14 days every 21 days in patients with hematologic malignancies (16, 17). No dose escalation or modification of the duration of treatment with decitabine was planned, whereas both dose escalation (up to 200 mg orally thrice a day) and evaluation of different durations of treatment (7, 16, or 21 days, starting on day 6) were initially planned for vorinostat. For the concurrent schedule, the starting doses of decitabine and vorinostat were based on the MTD established on the sequential schedule and dose escalation and evaluations of different durations of treatment (7 or 14 days, starting on day 3) were planned for vorinostat only.

Dose escalation in both schedules followed the standard 3 + 3 rule. The RP2D of this study was defined as the dose level at which 1 or less than 1 of 6 patients developed dose-limiting toxicity (DLT) and had the lowest frequency of treatment delays. Toxicity was graded using the National Cancer Institute Common Terminology Criteria for Adverse Events version 3.0. DLTs were defined as adverse events occurring during the first cycle of treatment and fulfilling one of the following criteria: ANC < 0.5 × 10⁹/L for 7 days or more, febrile neutropenia, platelets less than 25 × 10⁹/L or grade 3 thrombocytopenia associated with bleeding; any grade 3 or higher or intolerable grade 2 nonhematologic toxicity except alopecia, nausea, and vomiting responsive to antiemetics, diarrhea responsive to medications, and electrolyte abnormalities correctable with supportive therapy; and any toxicity resulting in treatment delay of more than 2 weeks. Response was assessed every 2 cycles using the Response Evaluation Criteria in Solid Tumors (18).

Pretreatment evaluation consisted of a complete medical history, physical examination, vital signs, electrocardiogram, complete blood count (CBC), serum chemistries, prothrombin time/INR, and activated partial thromboplastin time, serum or urine pregnancy test and baseline tumor measurements. On days 1, 8, and 15 of each cycle, evaluation consisted of a brief history and physical examination (day 1), vital signs (day 1), CBC, and serum chemistries.

Patients who experienced any DLT had treatment delayed by 1-week intervals until recovery and then may have continued on study with reduction of decitabine and vorinostat by one dose level. If no recovery occurred after a delay of 3 weeks of the next scheduled treatment cycle, patients were discontinued from protocol treatment. Blood counts must be recovered to begin a new treatment cycle.

Patients with an objective response or stable disease were allowed to remain on study until disease progression. Otherwise, study treatment continued until disease progression, unacceptable adverse event, patient's decision to withdraw consent, or changes in the patient's condition that rendered the continuation of study treatment unacceptable.

Blood samples for evaluation of decitabine were collected on days 1 and 5 of cycle 1, before dosing, 30 minutes after the infusion had started, at the end of infusion and at 5, 20, 35, 45, and 60 minutes from the end of infusion. Blood was collected into a sodium heparin Vacutainer tube and centrifuged at 1,500 × g for 15 minutes. The resulting plasma was transferred into polypropylene tubes and stored at −70°C until analyzed for decitabine concentration, using a validated high performance liquid chromatography with tandem mass spectrometry (LC-MS/MS) (19). On the sequential schedule, blood samples for vorinostat were collected on day 9 of cycle 1 before dosing and at 0.5, 1, 2, 2.5, 3, 4, 6. and 8 hours after dosing. On the concurrent schedule, blood samples were collected on day 4 of cycle 1 at the same time points. Samples were allowed to clot at 4°C for 20 to 30 minutes, and then centrifuged at 2,000 × g for 15 minutes at 4°C. The resulting serum was transferred to polypropylene cryotubes and stored at −70°C until analyzed for vorinostat concentrations with a validated LC-MS/MS assay (20). PK parameters were calculated by noncompartmental methods using WinNonlin (Version 5.2; Pharsight Corp.)

Exploratory analyses were performed for PK parameters and adjustments made for multiple comparisons. t Tests were used for independent group comparisons and paired t tests were used to compare PK parameters of decitabine on day 1 versus day 5 within a dose level.

Forty-four patients were enrolled into this study and 43 received treatment for a total of 136 cycles (median 2, range: 1–25; Table 1). One patient with neuroendocrine carcinoma of the pancreas did not receive treatment because baseline CT scans showed no significant growth of disease.

All patients had progressed from previous treatments for advanced disease. The median number of prior systemic treatment lines was 3. Three patients were treated previously with radiation therapy only and had not received systemic therapy.

A total of 43 patients received at least 1 cycle of study treatment. At the time of data cut off (March 2010), treatment was discontinued due to radiologically confirmed progression or due to symptomatic deterioration caused by underlying disease in 37 patients; 5 patients discontinued due to adverse events related to study treatment; 1 patient with stable disease after 25 cycles remains on study. There were no treatment-related deaths.

Six dose levels were evaluated on the sequential and 3 on the concurrent schedule (Fig. 1). On the sequential schedule, 8 of 30 (27%) evaluable patients experienced at least 1 DLT. The nature of DLTs was as follows: hematologic in 5 patients, nonhematologic in 2, and both hematologic and nonhematologic in 1 patient. Among hematologic DLTs, grade 4 thrombocytopenia occurred in 4 patients, grade 4 neutropenia lasting 7 days or more occurred in 2 and febrile neutropenia occurred in 2 patients. There was a clear association between hematologic DLTs and higher doses of decitabine. In fact, among 14 patients treated at the 2 dose levels with decitabine given at 10 mg/m²/day, only 1 heavily pretreated patient with NHL who had a prior autotransplant developed a hematologic DLT, consisting of grade 4 thrombocytopenia. Regarding nonhematologic DLTs, 1 patient encountered grade 2 intolerable fatigue, anorexia, and dehydration, 1 had grade 3 fatigue, and 1 had grade 3 constipation (plus febrile neutropenia and grade 4 thrombocytopenia).

Among 12 patients treated on the concurrent schedule, four (33%) developed a DLT. Three patients had nonhematologic DLTs (1 grade 3 fatigue and 2 grade 3 fatigue and grade 3 dehydration), and 1 patient experienced a hematologic DLT (grade 3 febrile neutropenia).

Neutropenia and thrombocytopenia were the most frequent adverse events of at least possible attribution to study treatment (Table 2). Fatigue, nausea, diarrhea, and vomiting were the most frequent nonhematologic adverse events (Table 2). Seven patients experienced more than 1 episode of grade 3 fatigue whereas other nonhematologic adverse events were mostly grade 1 or 2.

Treatment delay due to related adverse events was calculated in patients who received 2 or more cycles of treatment: on the sequential schedule, a treatment delay for 1 occasion or more occurred in 4 of 4 (100%), 2 of 2 (100%), 2 of 2 (100%), 0 of 2 (0%), 1 of 5 (20%), and 3 of 5 (60%) patients, on dose levels 1, −1, 1a, 1b, −1b, and −2b, respectively. Among patients treated on the concurrent schedule, 1 or more treatment delay occurred in 2 of 6 (33%), 1 of 1 (100%), and 1 of 1 (100%) patients on dose levels −2b, −3b, and −3c respectively.

Dose reductions or omissions occurred as follows: on the sequential schedule 4 of 6 (66%), 1 of 3 (33%), 2 of 4 (50%), 1 of 4 (25%), 2 of 7 (29%), 1 of 7 (14%) patients on dose levels 1, −1, 1a, 1b, −1b, and −2b, respectively, and on the concurrent schedule 0 of 7 (0%), 2 of 3 (66%), and 2 of 2 (100%) patients on dose levels −2b, −3b, and −3c, respectively.

Three patients treated on the sequential schedule (10%) discontinued the study due to adverse events possibly related to the treatment. These consisted of fatigue and nausea in one, nausea and neutropenia in another, and nausea and vomiting in the third. On the concurrent schedule, 2 of 12 (17%) patients discontinued treatment due to treatment related adverse events. One event was fatigue, whereas the other patient developed skin changes that were biopsied and showed a neutrophilic dermatitis consistent with Sweet syndrome, a rare syndrome that has previously been described in patients with myelodysplastic syndrome/acute leukaemia treated with the combination of HDACi and DNMTi (21). After discontinuation of treatment and a course of oral prednisone, the skin changes resolved.

Based on the occurrence of DLTs reported previously, the MTD for the sequential combination was decitabine 10 mg/m²/day on days 1 to 5 with vorinostat 200 mg 3 times a day on days 6 to 12. The MTD for the concurrent schedule was decitabine 10 mg/m²/day on days 1 to 5 with vorinostat 200 mg 2 times a day on days 3 to 9. Among the sequential schedules, the highest dose with the least treatment delay was achieved in dose level −1b (decitabine 10 mg/m²/day on days 1–5 with vorinostat 200 mg twice daily, days 6–12). This is also the dose level with the highest percentage of patients with stable disease for 4 cycles or more. In fact, 4 of the 11 patients (36%) who remained on study for 4 or more cycles were treated at this dose level. Thus, this dose level represents the recommended phase II dose (RP2D) for the combination of the two drugs.

No objective tumor responses were observed. Among 38 patients evaluable for response, 11 (29%) with previously progressive cancers had stable disease for 4 or more cycles of treatment. Three of 11 had stable disease for 8 or more cycles, including 1 patient with colon cancer (8 cycles), 1 patient with thymoma (11 cycles), and 1 patient with mucinous adenocarcinoma of the appendix who remains on study after 25 cycles (Fig. 2).

PK results (Tables 3 and 4) of decitabine were similar to those reported in previous studies (22). Decitabine is rapidly eliminated with a half-life of approximately 20 minutes. On the sequential schedule, increasing doses of decitabine did not produce dose proportional increases in maximum plasma concentration (C_max) and area under the plasma concentration–time curve (AUC). No significant differences were found when comparing PK parameters on day 1 and day 5 within the same dose level in either schedule. In addition, there were no statistically significant differences when comparing PK parameters for the dose of 10 mg/m²/day between the 2 schedules of administration.

A comparison of vorinostat PK parameters with increasing doses was possible only for the sequential schedule. AUC and C_max after a single vorinostat administration increased proportionaly when the dose of vorinostat was increased from 100 mg to 200 mg. Interestingly, C_max, AUC_inf ( = AUC to infinity), and AUC_t ( = AUC over dosing interval) were lower in the concurrent schedule than those in the sequential schedule (P = 0.09, 0.004, and 0.0005, respectively) at the 200 mg BID dose level. The difference in the AUC_t remains statistically significant after adjustment for multiple comparisons (Supplementary Figure A1).

Aberrant DNA methylation and histone deacetylation are involved in tumor formation and progression and have been evaluated as targets for the development of anticancer agents (1, 2). The possibility of optimally reexpressing methylated genes following treatment with the combination of a DNMTi with an HDACi has been confirmed in preclinical studies and formed the basis for clinical trials using combined epigenetic therapies (23).

Here we report for the first time the results of a phase I trial demonstrating the feasibility of delivering decitabine in combination with vorinostat in patients with advanced solid tumors or NHLs. Decitabine given for 5 days at a dose of 10 mg/m²/day as a 1-hour intravenous infusion can be combined with oral vorinostat either on a sequential (vorinostat 200 mg 3 times a day on days 6–12) or a concurrent schedule (vorinostat 200 mg 2 times a day on days 3–9). The toxicities observed were predictable and manageable at these stated MTD doses. In both schedules, the combination of decitabine and vorinostat appears to have a narrow therapeutic index and both drugs required dose reductions from their single-agent recommended doses used in previous studies of hematologic malignancies and solid tumors. However, the optimal single-agent doses of decitabine and vorinostat in patients with solid tumors remain unknown and there is no clear evidence that higher doses are associated with better outcome.

Among the 2 schedules evaluated in this study, the sequential schedule appears to be easier to deliver and more tolerable. In addition, PK analyses showed that for the same dose of vorinostat, AUC, and C_max were lower for the concurrent schedule, suggesting a possible unfavorable PK interaction between the 2 drugs. However, the number of patients enrolled in our study was small and the study was not designed to establish if this was due to increased metabolism or reduced absorption of vorinostat. Notably, escalation of vorinostat to 200 mg thrice a day on the concurrent schedule resulted in increased toxicities with 2 of 2 patients developing DLTs. Within the sequential schedule, dose level −1b (decitabine 10 mg/m²/day on days 1–5 and vorinostat 200 mg twice a day on days 6–12) was more favorable in dose delivery without delays, and therefore it represents the dose we recommend for further phase II evaluation. Dose level −1b on the sequential schedule also had the highest percentage of patients achieving stable disease for 4 cycles or more. Although this represents only a small number of patients, rendering it impossible to draw any definitive conclusion regarding superiority of any of the dose levels studied, we believe this dose level warrants further investigation in clinical trials.

Combination therapies employing DNMTi or HDACi with other agents are being pursued clinically (8, 9). A small number of clinical trials have evaluated different combinations of epigenetic agents in patients with hematologic and, more recently, solid malignancies (22, 24–27). There is optimism that combined epigenetic therapy can result in increased antitumor activity in comparison to the use of single-agent DNMTi or HDACi, but this needs validation in randomized studies. In our study we observed stable disease in previously progressing patients with different tumor types, but it is not possible to establish if this is due to the combination of the 2 agents and what the expected outcomes would be if each agent was used alone. Decitabine and vorinostat are active in hematologic malignancies and cutaneous T-cell lymphomas, respectively, but their role in the treatment of solid tumors remains undefined. Among the 2 agents, vorinostat has shown single-agent antitumor activity with reports of stable disease and a few cases of partial responses in patients with different types of solid tumors (16, 28).

This was a small phase I study and it is difficult to speculate based on its results in which tumor types this combination should be further explored. Recently reported studies of vorinostat in patients with solid tumors as single-agent or in combination with chemotherapy have shown preliminary evidence of activity in non–small-cell lung, breast, colorectal, mesothelioma, thyroid, and adenoid cystic carcinoma (16, 28–30). It is therefore reasonable to consider further clinical investigation of this combination in the context of the aforementioned malignancies.

No potential conflicts of interests were disclosed.

Supported by NCI Grants. U01CA132123, U01CA099168, and P30-CA47904

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.