Phase I/II Intra-patient Dose Escalation Study of Vorinostat in Children with Relapsed Solid Tumor, Lymphoma or Leukemia

Abstract Members of the histone deacetylase (HDAC) family exhibit great promise as potential drug targets in pediatric tumors including neuroblastoma, medulloblastoma, ependymoma and Ewing’s sarcoma. HDAC inhibitors of various structural classes have shown anti-tumoral effects in pre-clinical pediatric tumor models as single agents or in combination treatments. Suberoylanilidehydroxamic acid (SAHA=vorinostat) is the most clinical advanced compound of the class and was approved by the US FDA in October 2006 for the treatment of refractory cutaneous T-cell lymphoma. In this phase I/II trial, pediatric patients with relapsed solid tumors, lymphoma or leukemias are treated according to an individualized dose escalation concept ensuring each individual patient to receive his optimal dose with respect to toxicity and efficacy. The study is accompanied by an extensive pharmacokinetic, pharmacodynamic and biomarker program.


▼
Relapsed or progressive solid tumors, lymphomas and leukemias have a very poor prognosis in children despite great success of fi rst-line therapies in pediatric oncology through multi-modal, intense chemotherapy protocols [ 2 , 5 , 21 ] . The work of the investigator's laboratory has identifi ed several members of the histone deacetylase (HDAC) family as potential drug targets in neuroblastoma, medulloblastoma, and ependymoma [ 25 , 27 , 28 , 34 , 35 ] and HDAC inhibitors of various structural classes have shown anti-tumoral eff ects in preclinical pediatric tumor models as single agents or in combination treatments [ 3 , 7 -9 , 12 , 24 , 33 ] . HDAC inhibitors are able to induce apoptosis, cell In this phase I/II trial, pediatric patients with relapsed solid tumors, lymphoma or leukemias are treated according to an individualized dose escalation concept ensuring each individual patient to receive his optimal dose with respect to toxicity and effi cacy. The study is accompanied by an extensive pharmacokinetic, pharmacodynamic and biomarker program.  [ 18 ] . The maximum tolerated dose (MTD) was 400 mg qd for continuous daily dosing. This dose could be applied up to 2 years in individual patients. Observed dose limiting toxicities were anorexia, diarrhea, fatigue, nausea, and thrombocytopenia. The determined peak plasma levels following 400 mg qd dosing were in the range of 658 ± 439 ng/ml (corresponding to 2.5 ± 1.7 μM). These in vivo achievable concentrations show signifi cant antineoplastic eff ects against a variety of pediatric tumor cells. Furthermore, SAHA is crossing the blood brain barrier in mice [ 16 , 17 ] , thus making it a feasible compound also for the treatment of brain tumors. A Phase I of vorinostat and tretinoin in children with refractory or relapsed solid tumors or leukemias was recently published [ 11 ] .  [ 18 , 19 ] . Therefore, in contrast to previous trials, the current study involves an intra-individual dose escalation concept allowing each individual patient to receive his optimal dose with respect to toxicity and effi cacy.

Zusammenfassung
Design and duration of the study ▼ This is an open-label single-arm, intra-patient dose escalation, multi-center, Phase I/II clinical trial. 50 patients will be enrolled in the clinical trial. Recruitment and treatment of patients will be performed in 10 trial centers in Germany.
In the absence of treatment delays due to adverse events, treatment may continue for each patient for a maximum of 12 months after establishing the individual MTD.

Primary objective and primary endpoint
To determine a safe dose recommended (SDR) for the routine application of oral vorinostat (involving dose escalation) in children and adolescents (3-18 years) with relapsed/refractory solid tumor, lymphoma or leukemia. A SDR is defi ned as the highest dose with no ≥ grade 3 toxicity according to CTC criteria described below (Dose Limiting Toxicity) in no more than 1/50 patient in this study.

Secondary objectives and secondary endpoints
▶ To determine the pharmacokinetics and the distribution of individual maximum tolerated doses (MTD), which is the maximum dose with no grade 3 or 4 toxicity according to CTC criteria. ▶ Antitumor eff ectiveness of vorinostat as measured by treatment response rate. Response will be evaluated in each patient 3 months after start of treatment with the individual MTD. ▶ Association of the histone deacetylase (HDAC)-inhibiting activity with the dose administered, toxicity, and treatment response. ▶ Feasibility and safety. ▶ Duration of response in responding patients. ▶ Additional biomarker studies for the prediction of vorinostat response (IL-6, IL-10, BMP4 induction, basal histone acetylation level) will be performed.

In/Exclusion criteria
The inclusion and exclusion criteria are listed in • ▶ Table 1 .

Course of the trial
During the phase I of the trial, each patient undergoes dose escalation every 2 weeks until he reaches his individual MTD ( • ▶ Fig. 1 ).
During this phase, PK and PD analysis will be performed. Patients Table 1 Eligibility criteria vorinostat in children.

Inclusion
-children and adolescents (3-18 years) with relapsed or therapy-refractory solid tumor, lymphoma or leukemia following standard fi rst-line or relapse protocols in pediatric oncology -diagnosis confi rmed by one of the Pathological, Radiological or Study Reference Centers recognized by the GPOH -no other simultaneous anti-neoplastic treatment or radiation during the study and 1 months before enrolment -suffi cient general condition (Lansky Score > 50 %) -life expectancy > 3 months -liver enzymes (ALT or AST) < 5x upper limit of normal reference value, bilirubin and creatinine < 3x upper limit of normal reference value -solid tumors: leukocytes > 2 000/μl, thrombocytes > 50 000/μl and adequate bone marrow function to permit evaluations of hematopoietic toxicity -no CTC grade 3 or 4 toxicity from previous treatments -normal ECG -written informed consent of the legal representatives and the patient if the patient is able to understand the study situation and to give consent (must be available before enrolment in the trial) -women with childbearing potential agree to use adequate contraception or to abstain from heterosexual activity throughout the study, starting with Visit 1 -sexually active male patient agrees to use an adequate method of contraception for the duration of the study -solid tumors: measurable disease activity according to RECIST criteria Exclusion -history of deep vein thrombosis or pulmonary embolism -pregnancy and lactation -patient with concomitant treatments and/or anti-neoplastic treatment such as chemotherapy, immune therapy, and diff erentiation therapy, other targeted therapy, radiation. The use of valproic acid as prior antiepileptic therapy is allowed with a 30-day washout period -prior exposure to Histone Deacetylase Inhibitors -known active HBV, HCV or HIV infection -patient with concomitant treatments such as amber [Hypericum perforatum], plant extracts, vitamins, and other anti-oxidative compounds -participation in other clinical trials or observation period of competing trials, respectively -patient is unable to swallow vorinostat suspension or capsules -patient on coumarin-derivative anticoagulants -any other medication which could accentuate known dose-dependent adverse eff ects of the study drug, for instance bone marrow depression or QT-prolongation will continue to phase II without interruption and receive vorinostat at their individual MTD dose for 3 months. At this point, response will be determined using MRI through central reference review. Patients showing at least stable disease will continue with the treatment for up to 9 months. Thereafter, a follow-up period of 3 months is scheduled. Figure 1 depicts the study fl ow chart.

Screening evaluations at baseline
At baseline, patients will receive a routine pediatric physical examination including a pregnancy test β-hCG if applicable, body weight and height. Blood will be taken for full blood count including differential, serum electrolytes (Na, K, Ca), glucose, AST, ALT, protein, bilirubin, lactate dehydrogenase (LDH) albumin, cholinesterase, INR and creatinine. In addition, tumor markers will be determined if appropriate (i. e., urine catecholamine-metabolites in case of neuroblastoma, β-HCG and AFP in germ cell tumors/hepatoblastoma) as well as pharmacodynamic biomarkers (IL-6, IL-10, BMP4).
Investigation of basal histone acetylation levels, expression of HDACs and HR23B in archived tumor samples will be investigated in order to determine their potential value as biomarkers for response prediction and future patient selection strategies. MRI-scan of tumor manifestations will be performed in patients with solid tumor or lymphoma in order to defi ne and quantify extension of disease at baseline prior to vorinostat treatment and a copy of scans will be submitted for reference radiology assessment. In patients with neuroblastoma, meta-iodbenzylguanidine (MIBG) scan will be additionally performed.
In case of leukemia, peripheral cytopenia or bone marrow infi ltration, baseline bone marrow aspiration will be done and unstained smears will be submitted to the central review pediatric hematology laboratory.

Dose escalation and maintenance therapy
During vorinostat dose escalation and 3 months maintenance treatment at individual MTD, patients will receive a routine  Fig. 1 Flow chart of the trial limiting toxicity (DLT) is observed at a given dose d in 1/50 patients (this defi nes the safe dose for routine application) then the 95 % confi dence interval for the true rate r of DLT at this dose is [0.05; 10.65] %. If in the former statement d is not the starting dose chosen in this study, then, under any reasonable continuous model of dose-dependency, the upper limit of the 95 % confi dence interval for r will be even lower than 10.6 % because it will incorporate the additional information that, according the dose escalation design, no DLT can have been observed for any doses d' < d. Additionally, this is a feasible number of patients to be recruited by the participating trial centers within 2 years. Patients with discontinuation of treatment are not to be replaced.

Primary analysis
A continuous monitoring will be performed to determine whether or not the starting dose d1 used in this study should be lowered due to an unacceptable rate of toxicity (DLT) observed at this dose level. Let r be the true toxicity rate at d1. It is assumed that the maximum acceptable toxicity rate at d1 is 10 %. The following criterion for lowering the starting dose is adopted in this study: (1) Pr(r > 0.10) ≥ 95 % Where this probability is evaluated based on the posterior distribution of r. In order to calculate the posterior distribution of r, the binomialbeta model is used with a noninformative prior. Thus, if y cases of DLT have been observed among the fi rst m patients the posterior distribution of r is beta (1 + y, 1 + m − y). This allows the evaluation of criterion (1) and, in particular, the calculation of critical pediatric physical examination including vital signs, body weight, concomitant treatment and AEs once per week. Blood will be taken once per week for full blood counting including diff erential, serum electrolytes, glucose, AST, ALT, protein, bilirubin, LDH and creatinine. An ECG will be done at day 8 and 15 of vorinostat treatment and when the MTD is reached, to rule out QT-changes. Collection of blood and cerebrospinal fl uid for pharmacokinetic evaluation will be done after 1 week of vorinostat treatment and at the time when individual MTD is established. This material will be sent to the reference pharmacokinetic laboratory. 3 months after start of vorinostat treatment at MTD, response will be evaluated by MRI imaging using the RECIST criteria and bone marrow cytology if indicated. In patients with neuroblastoma, a MIBG scan will also be performed. Furthermore, full blood counting including diff erential, serum electrolytes, glucose, AST, ALT, protein, bilirubin, LDH, creatinine, and tumor marker (if indicated). Biomarker (IL-6, IL-10, BMP4) will be determined at month 1 and 3 of MTD.
Patients showing CR, PR, or SD on response evaluation will continue with vorinostat treatment for a maximum of 9 months or until disease progression. In these patients, tumor manifestations (MRI, tumor markers, bone marrow if indicated, MIBG) will be determined every 3 months until end of treatment (EOT). During this time, patients will receive routine pediatric oncological care including physical examination including body weight, full blood counting including diff erential, serum electrolytes, glucose, AST, ALT, protein, bilirubin, LDH and creatinine every 2 weeks.

Follow-up
After discontinuation of vorinostat treatment, patients will be followed for 3 months. During follow-up, all patients will receive routine supportive pediatric oncological care including physical examination including vital signs, body weight, Lansky performance status, concomitant treatment and AEs every 2 weeks. Blood will be taken every 2 weeks for full blood count including diff erential, serum electrolytes, glucose, AST, ALT, protein, bilirubin, LDH and creatinine. Since alternative salvage therapies are not available at the moment, patients will be off ered best supportive care after study termination.

Evaluation of treatment response
3 months after start of vorinostat treatment at the individual MTD, treatment response will be evaluated by comparing tumor manifestations with baseline measurements by central review according to the RECIST criteria [ 29 , 32 ] applying the updated Version 1.1 as described by Eisenhauer et al. 2009. In addition, brain tumor volumes will be determined using modifi ed Mac-Donald criteria (ellipsoid formula: axial × coronal × sagittal/2) as used in the German brain tumor network of the GPOH. In patients with neuroblastoma, treatment response will also be determined according to the INRC [ 4 ] incorporating qualitative results of MIBG-Imaging.

Sample size calculation
The sample size calculation in this study was based on the primary objective, namely, the accuracy requirements for the toxicity rate associated with the safe dose for the routine application. 50 pediatric patients ≥ 3 years will be included in the trial. If dose In single cases, pulmonary embolism, deep vein thrombosis, hyperglycemia, QTc prolongation, and electrolyte disturbances have been reported. Vorinostat was not evaluated in patients with hepatic impairment. As vorinostat is predominantly eliminated through metabolism, patients with hepatic impairment should be treated with caution. Vorinostat was not evaluated in patients with renal impairment. However, renal excretion does not seem to play a role in the elimination of vorinostat because < 1 % is recovered from the urine in adult patients and because the metabolites excreted by the kidneys are inactive [ 1 ] .

Confl ict of interest:
The authors have no confl ict of interest to disclose.
boundaries c (m) for the minimum observed numbers of DLT among the fi rst m patients such that inequality (1) is satisfi ed. With the specifi cation given above, the total probability Pr(L) that the starting dose must be lowered at least once in this study with n = 50 patients can be calculated by means of computer simulations. The probability P (L) is relevant because the lowering of the starting dose is a major event in the course of the trial (note that a repeated lowering of the starting dose even leads to the termination of the study). High values of P (L) would therefore indicate that the above criterion may be inappropriate.
The following values of P (L) were obtained (1 000 000 simulation runs per calculation); r = true toxicity rate at the planned starting dose of 180 mg/m².

Secondary analyses
Response will be summarized and estimated with exact (testbased) 2-sided 95 % confi dence bounds.
Caveat: The response rates determined in this study have only limited "transportability" given that, in order to apply to a diff erent patient cohort, not only would the patient have to be similar but also the application of Vorinostat (starting dose, excalation schedule) would need to be like the one used in this study.
All secondary variables will be analyzed using descriptive methods. Continuous variables will be summarized using standard summary statistics as appropriate and 95 % CI. Summary statistics for categorical variables will include frequency counts and percentages. If appropriate, graphical presentations of data will be created. Additional exploratory analyses will use multivariate models which will include covariates to explore the potential dependency of outcomes on these and investigate the potential relationship between outcome variables. Generally, results will be given by summaries of the data as indicated above together with the test statistics and their associated p-values. Appropriate confidence intervals of estimates of eff ect will be given to quantify the degree of uncertainty of these estimates.

Analysis of Adverse Events (AE):
Frequencies of patients experiencing at least one AE will be displayed. Detailed information collected for each AE will include: A description of the event, duration, whether the AE was serious, intensity, relationship to study drug, action taken, and clinical outcome. Severity of the AEs will be graded according to the CTCAEv4.0. Summary tables will present the number of patients observed with AEs, the corresponding percentages and 95 % CI.

Investigational medicinal product
Vorinostat is provided as 100 mg capsules by MSD SHARP & DOHME GMBH. For small children, or patients unable to swallow capsules, and to facilitate accurate dosing, an oral suspension is prepared by the Pharmacy Department, University Hospital Heidelberg. All study medication is centrally distributed through the Pharmacy Department, University Hospital Heidelberg.
Known side eff ects • ▶ Table 2 depicts the frequency of adverse events in adult patients with cutaneous T cell lymphoma receiving 400 mg/d vorinostat (adopted from ZOLINZA™ drug information).