Bayesian decision procedures for binary and continuous bivariate dose-escalation studies

In this paper, Bayesian decision procedures are developed for dose-escalation studies based on binary measures of undesirable events and continuous measures of therapeutic benefit. The methods generalize earlier approaches where undesirable events and therapeutic benefit are both binary. A logistic regression model is used to model the binary responses, while a linear regression model is used to model the continuous responses. Prior distributions for the unknown model parameters are suggested. A gain function is discussed and an optional safety constraint is included. Copyright (C) 2006 John Wiley & Sons, Ltd.

Dose escalation and pharmacokinetic study of AEZS-108 (AN-152), an LHRH agonist linked to doxorubicin, in women with LHRH receptor-positive tumors

Receptors for luteinizing hormone-releasing hormone (LHRH) can be utilized for targeted chemotherapy of cytotoxic LHRH analogs. The compound AEZS-108 (previously AN-152) consists of [D-Lys?]LHRH linked to doxorubicin. The objectives of this first study in humans with AESZ-108 were to determine the maximum tolerated dose and to characterize the dose-limiting toxicity, pharmacokinetics, preliminary efficacy, and hormonal effects.

Favorable effect of priming with granulocyte colony-stimulating factor in remission induction of acute myeloid leukemia restricted to dose-escalation of cytarabine

The clinical value of chemotherapy sensitization of acute myeloid leukemia (AML) with G-CSF priming has remained controversial. Cytarabine is a key constituent of remission induction chemotherapy. The effect of G-CSF priming has not been investigated in relationship with variable dose levels of cytarabine. We randomized 917 AML patients to receive G-CSF (456 patients) or no G-CSF (461 patients) at the days of chemotherapy. In the initial part of the study, 406 patients were also randomized between 2 cytarabine regimens comparing conventional-dose (199 patients) versus escalated-dose (207 patients) cytarabine in cycles 1 and 2. We found that patients after induction chemotherapy plus G-CSF had similar overall survival (43% vs 40%, P = .88), event-free survival (37% vs 31%, P = .29), and relapse rates (34% vs 36%, P = .77) at 5 years as those not receiving G-CSF. However, patients treated with the escalated-dose cytarabine regimen benefited from G-CSF priming, with improved event-free survival (P = .01) and overall survival (P = .003), compared with patients without G-CSF undergoing escalated-dose cytarabine treatment. A significant survival advantage of sensitizing AML for chemotherapy with G-CSF was not apparent in the entire study group, but it was seen in patients treated with escalated-dose cytarabine during remission induction. The HOVON-42 study is registered under The Netherlands Trial Registry (www.trialregister.nl) as #NTR230.

Somatostatin-based radiotherapy with [90Y-DOTA]-TOC in neuroendocrine tumors: long-term outcome of a phase I dose escalation study

BACKGROUND We describe the long-term outcome after clinical introduction and dose escalation of somatostatin receptor targeted therapy with [90Y-DOTA]-TOC in patients with metastasized neuroendocrine tumors. METHODS In a clinical phase I dose escalation study we treated patients with increasing [90Y-DOTA]-TOC activities. Multivariable Cox regression and competing risk regression were used to compare efficacy and toxicities of the different dosage protocols. RESULTS Overall, 359 patients were recruited; 60 patients were enrolled for low dose (median: 2.4 GBq/cycle, range 0.9-7.8 GBq/cycle), 77 patients were enrolled for intermediate dose (median: 3.3 GBq/cycle, range: 2.0-7.4 GBq/cycle) and 222 patients were enrolled for high dose (median: 6.7 GBq/cycle, range: 3.7-8.1 GBq/cycle) [90Y-DOTA]-TOC treatment. The incidences of hematotoxicities grade 1-4 were 65.0%, 64.9% and 74.8%; the incidences of grade 4/5 kidney toxicities were 8.4%, 6.5% and 14.0%, and the median survival was 39 (range: 1-158) months, 34 (range: 1-118) months and 29 (range: 1-113) months. The high dose protocol was associated with an increased risk of kidney toxicity (Hazard Ratio: 3.12 (1.13-8.59) vs. intermediate dose, p = 0.03) and a shorter overall survival (Hazard Ratio: 2.50 (1.08-5.79) vs. low dose, p = 0.03). CONCLUSIONS Increasing [90Y-DOTA]-TOC activities may be associated with increasing hematological toxicities. The dose related hematotoxicity profile of [90Y-DOTA]-TOC could facilitate tailoring [90Y-DOTA]-TOC in patients with preexisting hematotoxicities. The results of the long-term outcome suggest that fractionated [90Y-DOTA]-TOC treatment might allow to reduce renal toxicity and to improve overall survival. (ClinicalTrials.gov number NCT00978211).

A dose-finding study of lanreotide (a somatostatin analog) in patients with colorectal carcinoma

BACKGROUND. Laboratory data suggest that insulin-like growth factor-1 (IGE-1) may stimulate the growth of different human tumors. At least in acromegalic patients, somatostatin (SMS) analogs, such as lanreotide, suppress the serum levels of growth hormone (GH) and IGE-1. METHODS. To evaluate the tolerability and biologic activity of different doses of lanreotide in patients with advanced colorectal carcinoma, consecutive groups of 3 patients each were subcutaneous treated with lanreotide at doses of 1, 2, 3, 4, 5, or 6 mg three times a day for 2 months. In the event of Grade 3 side effects, 3 additional patients were treated with the same dose before the next dose escalation. Serum samples were obtained on Days 0, 15, 30, and 60 for serum GH, IGF-1, and lanreotide assessment. RESULTS. Twenty-four patients were enrolled and all were evaluable. Except for the 3 and 6 mg doses, for which the observation of a Grade 3 side effect required that an additional three patients be treated, it was sufficient to treat 3 patients at each dose. The overall incidence of side effects was as follows: changes in bowel habits, 83%; abdominal cramps, 79%; diarrhea, 17%; vomiting, 17%; nausea, 21%; steatorrhea, 78%; hyperglycemia, 35%; laboratory hypothyroidism, 39%; gallstones, 13%; and weight loss, 17%. No evidence of an increase in the incidence, intensity, or duration of side effects was observed with dose escalation. Serum IGF-1 levels were as follows: Day 13: 63%, 60%, and 67% of the baseline values for the low (12 mg), intermediate (3-4 mg), and high (5- 6 mg) dose groups, respectively; Day 30: 63%, 59%, and 51%, respectively; and Day 60: 73%, 69%, and 47%, respectively. Serum lanreotide levels declined during treatment in all of the dose groups (90 ng/mL on Day 15, and 35 ng/mL on Day 60 for the 5-6 mg group; 10 ng/mL on Day 15, and 1.5 ng/mL on Day 60 for the 1-2 mg group). No antitumor activity or tumor marker reduction was observed. CONCLUSIONS. No increase in toxicity was observed when subcutaneous lanreotide doses were escalated to 6 mg three times a day for 2 months. The highest doses seemed to maintain reduced serum IGF-1 levels; with the lowest doses, a 'rebound' in serum IGF-1 levels was observed during treatment. Nevertheless, intermittent subcutaneous injections do not ensure constant serum drug concentrations over time.

Rituximab, bendamustine, and lenalidomide in patients with aggressive B cell lymphoma not eligible for high-dose chemotherapy or anthracycline-based therapy: phase I results of the SAKK 38/08 trial

This phase I trial was designed to develop a new effective and well-tolerated regimen for patients with aggressive B cell lymphoma not eligible for front-line anthracycline-based chemotherapy or aggressive second-line treatment strategies. The combination of rituximab (375 mg/m(2) on day 1), bendamustine (70 mg/m(2) on days 1 and 2), and lenalidomide was tested with a dose escalation of lenalidomide at three dose levels (10, 15, or 20 mg/day) using a 3 + 3 design. Courses were repeated every 4 weeks. The recommended dose was defined as one level below the dose level identifying ≥2/6 patients with a dose-limiting toxicity (DLT) during the first cycle. Thirteen patients were eligible for analysis. Median age was 77 years. WHO performance status was 0 or 1 in 12 patients. The Charlson Comorbidity Index showed relevant comorbidities in all patients. Two DLTs occurred at the second dose level (15 mg/day) within the first cycle: one patient had prolonged grade 3 neutropenia, and one patient experienced grade 4 cardiac adverse event (myocardial infarction). Additional grade 3 and 4 toxicities were as follows: neutropenia (31 %), thrombocytopenia (23 %), cardiac toxicity (31 %), fatigue (15 %), and rash (15 %). The dose of lenalidomide of 10 mg/day was recommended for a subsequent phase II in combination with rituximab 375 mg/m(2) on day 1 and bendamustine 70 mg/m(2) on days 1 and 2.

BAYESIAN PHASE I DOSE FINDING IN CANCER TRIALS

This dissertation explores phase I dose-finding designs in cancer trials from three perspectives: the alternative Bayesian dose-escalation rules, a design based on a time-to-dose-limiting toxicity (DLT) model, and a design based on a discrete-time multi-state (DTMS) model. We list alternative Bayesian dose-escalation rules and perform a simulation study for the intra-rule and inter-rule comparisons based on two statistical models to identify the most appropriate rule under certain scenarios. We provide evidence that all the Bayesian rules outperform the traditional ``3+3'' design in the allocation of patients and selection of the maximum tolerated dose. The design based on a time-to-DLT model uses patients' DLT information over multiple treatment cycles in estimating the probability of DLT at the end of treatment cycle 1. Dose-escalation decisions are made whenever a cycle-1 DLT occurs, or two months after the previous check point. Compared to the design based on a logistic regression model, the new design shows more safety benefits for trials in which more late-onset toxicities are expected. As a trade-off, the new design requires more patients on average. The design based on a discrete-time multi-state (DTMS) model has three important attributes: (1) Toxicities are categorized over a distribution of severity levels, (2) Early toxicity may inform dose escalation, and (3) No suspension is required between accrual cohorts. The proposed model accounts for the difference in the importance of the toxicity severity levels and for transitions between toxicity levels. We compare the operating characteristics of the proposed design with those from a similar design based on a fully-evaluated model that directly models the maximum observed toxicity level within the patients' entire assessment window. We describe settings in which, under comparable power, the proposed design shortens the trial. The proposed design offers more benefit compared to the alternative design as patient accrual becomes slower.

Improving the accuracy of radiation pneumonitis dose response models

The prognosis for lung cancer patients remains poor. Five year survival rates have been reported to be 15%. Studies have shown that dose escalation to the tumor can lead to better local control and subsequently better overall survival. However, dose to lung tumor is limited by normal tissue toxicity. The most prevalent thoracic toxicity is radiation pneumonitis. In order to determine a safe dose that can be delivered to the healthy lung, researchers have turned to mathematical models predicting the rate of radiation pneumonitis. However, these models rely on simple metrics based on the dose-volume histogram and are not yet accurate enough to be used for dose escalation trials. The purpose of this work was to improve the fit of predictive risk models for radiation pneumonitis and to show the dosimetric benefit of using the models to guide patient treatment planning. The study was divided into 3 specific aims. The first two specifics aims were focused on improving the fit of the predictive model. In Specific Aim 1 we incorporated information about the spatial location of the lung dose distribution into a predictive model. In Specific Aim 2 we incorporated ventilation-based functional information into a predictive pneumonitis model. In the third specific aim a proof of principle virtual simulation was performed where a model-determined limit was used to scale the prescription dose. The data showed that for our patient cohort, the fit of the model to the data was not improved by incorporating spatial information. Although we were not able to achieve a significant improvement in model fit using pre-treatment ventilation, we show some promising results indicating that ventilation imaging can provide useful information about lung function in lung cancer patients. The virtual simulation trial demonstrated that using a personalized lung dose limit derived from a predictive model will result in a different prescription than what was achieved with the clinically used plan; thus demonstrating the utility of a normal tissue toxicity model in personalizing the prescription dose.

Attitudes towards a maintenance (-agonist) treatment approach in high-dose benzodiazepine-dependent patients: a qualitative study

BACKGROUND High-dose benzodiazepine dependence constitutes a major clinical concern. Although withdrawal treatment is recommended, it is unsuccessful for a significant proportion of affected patients. More recently, a benzodiazepine maintenance approach has been suggested as an alternative for patients' failing discontinuation treatment. While there is some data supporting its effectiveness, patients' perceptions of such an intervention have not been investigated. METHODS An exploratory qualitative study was conducted among a sample of 41 high-dose benzodiazepine (BZD)-dependent patients, with long-term use defined as doses equivalent to more than 40 mg diazepam per day and/or otherwise problematic use, such as mixing substances, dose escalation, recreational use, or obtainment by illegal means. A qualitative content analysis approach was used to evaluate findings. RESULTS Participants generally favored a treatment discontinuation approach with abstinence from BZD as its ultimate aim, despite repeated failed attempts at withdrawal. A maintenance treatment approach with continued prescription of a slow-onset, long-acting agonist was viewed ambivalently, with responses ranging from positive and welcoming to rejection. Three overlapping themes of maintenance treatment were identified: "Only if I can try to discontinue…and please don't call it that," "More stability and less criminal activity…and that is why I would try it," and "No cure, no brain and no flash…and thus, just for everybody else!" CONCLUSIONS Some patients experienced slow-onset, long-acting BZDs as having stabilized their symptoms and viewed these BZDs as having helped avoid uncontrolled withdrawal and abstain from criminal activity. We therefore encourage clinicians to consider treatment alternatives if discontinuation strategies fail.

Dose-painting intensity-modulated proton therapy for intermediate- and high-risk meningioma.

BACKGROUND Newly diagnosed WHO grade II-III or any WHO grade recurrent meningioma exhibit an aggressive behavior and thus are considered as high- or intermediate risk tumors. Given the unsatisfactory rates of disease control and survival after primary or adjuvant radiation therapy, optimization of treatment strategies is needed. We investigated the potential of dose-painting intensity-modulated proton beam-therapy (IMPT) for intermediate- and high-risk meningioma. MATERIAL AND METHODS Imaging data from five patients undergoing proton beam-therapy were used. The dose-painting target was defined using [68]Ga-[1,4,7,10-tetraazacyclododecane tetraacetic acid]- d-Phe(1),Tyr(3)-octreotate ([68]Ga-DOTATATE)-positron emission tomography (PET) in target delineation. IMPT and photon intensity-modulated radiation therapy (IMRT) treatment plans were generated for each patient using an in-house developed treatment planning system (TPS) supporting spot-scanning technology and a commercial TPS, respectively. Doses of 66 Gy (2.2 Gy/fraction) and 54 Gy (1.8 Gy/fraction) were prescribed to the PET-based planning target volume (PTVPET) and the union of PET- and anatomical imaging-based PTV, respectively, in 30 fractions, using simultaneous integrated boost. RESULTS Dose coverage of the PTVsPET was equally good or slightly better in IMPT plans: dose inhomogeneity was 10 ± 3% in the IMPT plans vs. 13 ± 1% in the IMRT plans (p = 0.33). The brain Dmean and brainstem D50 were small in the IMPT plans: 26.5 ± 1.5 Gy(RBE) and 0.002 ± 0.0 Gy(RBE), respectively, vs. 29.5 ± 1.5 Gy (p = 0.001) and 7.5 ± 11.1 Gy (p = 0.02) for the IMRT plans, respectively. The doses delivered to the optic structures were also decreased with IMPT. CONCLUSIONS Dose-painting IMPT is technically feasible using currently available planning tools and resulted in dose conformity of the dose-painted target comparable to IMRT with a significant reduction of radiation dose delivered to the brain, brainstem and optic apparatus. Dose escalation with IMPT may improve tumor control and decrease radiation-induced toxicity.

Radiobiological model comparison of 3D conformal radiotherapy and IMRT plans for the treatment of prostate cancer

The main aim of radiotherapy is to deliver a dose of radiation that is high enough to destroy the tumour cells while at the same time minimising the damage to normal healthy tissues. Clinically, this has been achieved by assigning a prescription dose to the tumour volume and a set of dose constraints on critical structures. Once an optimal treatment plan has been achieved the dosimetry is assessed using the physical parameters of dose and volume. There has been an interest in using radiobiological parameters to evaluate and predict the outcome of a treatment plan in terms of both a tumour control probability (TCP) and a normal tissue complication probability (NTCP). In this study, simple radiobiological models that are available in a commercial treatment planning system were used to compare three dimensional conformal radiotherapy treatments (3D-CRT) and intensity modulated radiotherapy (IMRT) treatments of the prostate. Initially both 3D-CRT and IMRT were planned for 2 Gy/fraction to a total dose of 60 Gy to the prostate. The sensitivity of the TCP and the NTCP to both conventional dose escalation and hypo-fractionation was investigated. The biological responses were calculated using the Källman S-model. The complication free tumour control probability (P+) is generated from the combined NTCP and TCP response values. It has been suggested that the alpha/beta ratio for prostate carcinoma cells may be lower than for most other tumour cell types. The effect of this on the modelled biological response for the different fractionation schedules was also investigated.