Search

PR1, an HLA-A2-restricted peptide derived from both proteinase 3 and neutrophil elastase, is recognized on myeloid leukemia cells by cytotoxic T lymphocytes (CTLs) that preferentially kill leukemia and contribute to cytogenetic remission. To evaluate safety, immunogenicity and clinical activity of PR1 vaccination, a phase I/II trial was conducted. Sixty-six HLA-A2+ patients with acute myeloid leukemia (AML: 42), chronic myeloid leukemia (CML: 13) or myelodysplastic syndrome (MDS: 11) received three to six PR1 peptide vaccinations, administered subcutaneously every 3 weeks at dose levels of 0.25, 0.5 or 1.0 mg. Patients were randomized to the three dose levels after establishing the safety of the highest dose level. Primary end points were safety and immune response, assessed by doubling of PR1/HLA-A2 tetramer-specific CTL, and the secondary end point was clinical response. Immune responses were noted in 35 of 66 (53%) patients. Of the 53 evaluable patients with active disease, 12 (24%) had objective clinical responses (complete: 8; partial: 1 and hematological improvement: 3). PR1-specific immune response was seen in 9 of 25 clinical responders versus 3 of 28 clinical non-responders (P=0.03). In conclusion, PR1 peptide vaccine induces specific immunity that correlates with clinical responses, including molecular remission, in AML, CML and MDS patients.

Late-onset (LO) toxicities are a serious concern in many phase I trials. Since most dose-limiting toxicities occur soon after therapy begins, most dose-finding methods use a binary indicator of toxicity occurring within a short initial time period. If an agent causes LO toxicities, however, an undesirably large number of patients may be treated at toxic doses before any toxicities are observed. A method addressing this problem is the time-to-event continual reassessment method (TITE-CRM, Cheung and Chappell, 2000). We propose a Bayesian dose-finding method similar to the TITE-CRM in which doses are chosen using time-to-toxicity data. The new aspect of our method is a set of rules, based on predictive probabilities, that temporarily suspend accrual if the risk of toxicity at prospective doses for future patients is unacceptably high. If additional follow-up data reduce the predicted risk of toxicity to an acceptable level, then accrual is restarted, and this process may be repeated several times during the trial. A simulation study shows that the proposed method provides a greater degree of safety than the TITE-CRM, while still reliably choosing the preferred dose. This advantage increases with accrual rate, but the price of this additional safety is that the trial takes longer to complete on average. [ABSTRACT FROM AUTHOR]

There is a pressing need for allogeneic chimeric antigen receptor (CAR)-immune cell therapies that are safe, effective and affordable. We conducted a phase 1/2 trial of cord blood-derived natural killer (NK) cells expressing anti-CD19 chimeric antigen receptor and interleukin-15 (CAR19/IL-15) in 37 patients with CD19 + B cell malignancies. The primary objectives were safety and efficacy, defined as day 30 overall response (OR). Secondary objectives included day 100 response, progression-free survival, overall survival and CAR19/IL-15 NK cell persistence. No notable toxicities such as cytokine release syndrome, neurotoxicity or graft-versus-host disease were observed. The day 30 and day 100 OR rates were 48.6% for both. The 1-year overall survival and progression-free survival were 68% and 32%, respectively. Patients who achieved OR had higher levels and longer persistence of CAR-NK cells. Receiving CAR-NK cells from a cord blood unit (CBU) with nucleated red blood cells ≤ 8 × 10 7 and a collection-to-cryopreservation time ≤ 24 h was the most significant predictor for superior outcome. NK cells from these optimal CBUs were highly functional and enriched in effector-related genes. In contrast, NK cells from suboptimal CBUs had upregulation of inflammation, hypoxia and cellular stress programs. Finally, using multiple mouse models, we confirmed the superior antitumor activity of CAR/IL-15 NK cells from optimal CBUs in vivo. These findings uncover new features of CAR-NK cell biology and underscore the importance of donor selection for allogeneic cell therapies. ClinicalTrials.gov identifier: NCT03056339 ., (© 2024. The Author(s).)

Relapse is the major cause of failure of high-dose chemotherapy (HDC) with autologous stem cell transplantation (ASCT) for B cell non-Hodgkin lymphomas (B-NHL). Improvement strategies include use in combination with effective immunotherapies. We hypothesized that the combination of rituximab/HDC/ASCT with expanded cord blood (CB)-derived natural killer (NK) cells is safe and active in B-NHL. Patients with B-NHL age 15 to 70 years and appropriate ASCT candidates were eligible for the study. The CB units were selected without considering HLA match with the recipient. The CB NK cells were expanded from day -19 to day -5. Treatment included rituximab on days -13 and -7, BEAM (carmustine/etoposide/cytarabine/melphalan) on days -13 to -7, lenalidomide on days -7 to -2, CB NK infusion (10 8 /kg) on day -5, and ASCT (day 0). The primary endpoint was 30-day treatment-related mortality (TRM); secondary endpoints included relapse-free survival (RFS), overall survival (OS), and persistence of CB NK cells. We enrolled 20 patients. CB NK cells were expanded a median of 1552-fold with >98% purity and >96% viability. We saw no adverse events attributable to the CB NK cells and 0% 30-day TRM. At median follow-up of 47 months, the RFS and OS rates were 53% and 74%, respectively. CB NK cells were detectable in blood for 2 weeks, independent of HLA-mismatch status. CD16 expression in donor NK cells was correlated favorably with outcome, and homozygosity for the high-affinity CD16 variant (158 V/V) in CB, but not recipient, NK cells was correlated with better outcomes. Our data indicate that the combination of expanded and highly purified CB-derived NK cells with HDC/ASCT for B-NHL is safe. CD16 expression in donor NK cells, particularly if homozygous for the high-affinity CD16 variant, was correlated with better outcomes., (Copyright © 2023 The American Society for Transplantation and Cellular Therapy. Published by Elsevier Inc. All rights reserved.)

Conventional designs for choosing a dose for a new therapy may select doses that are unsafe or ineffective and fail to optimize progression-free survival time, overall survival time, or response/remission duration. We explain and illustrate limitations of conventional dose-finding designs and make four recommendations to address these problems. When feasible, a dose-finding design should account for long-term outcomes, include screening rules that drop unsafe or ineffective doses, enroll an adequate sample size, and randomize patients among doses. As illustrations, we review three designs that include one or more of these features. The first illustration is a trial that randomized patients among two cell therapy doses and standard of care in a setting where it was assumed on biological grounds that dose toxicity and dose-response curves did not necessarily increase with cell dose. The second design generalizes phase I-II by first identifying a set of candidate doses, rather than one dose, randomizing additional patients among the candidates, and selecting an optimal dose to maximize progression-free survival over a longer follow-up period. The third design combines a phase I-II trial and a group sequential randomized phase III trial by using survival time data available after the first stage of phase III to reoptimize the dose selected in phase I-II. By incorporating one or more of the recommended features, these designs improve the likelihood that a selected dose or schedule will be optimal, and thus will benefit future patients and obtain regulatory approval., (©2023 American Association for Cancer Research.)

This article describes rationales and limitations for making inferences based on data from randomized controlled trials (RCTs). We argue that obtaining a representative random sample from a patient population is impossible for a clinical trial because patients are accrued sequentially over time and thus comprise a convenience sample, subject only to protocol entry criteria. Consequently, the trial's sample is unlikely to represent a definable patient population. We use causal diagrams to illustrate the difference between random allocation of interventions within a clinical trial sample and true simple or stratified random sampling, as executed in surveys. We argue that group-specific statistics, such as a median survival time estimate for a treatment arm in an RCT, have limited meaning as estimates of larger patient population parameters. In contrast, random allocation between interventions facilitates comparative causal inferences about between-treatment effects, such as hazard ratios or differences between probabilities of response. Comparative inferences also require the assumption of transportability from a clinical trial's convenience sample to a targeted patient population. We focus on the consequences and limitations of randomization procedures in order to clarify the distinctions between pairs of complementary concepts of fundamental importance to data science and RCT interpretation. These include internal and external validity, generalizability and transportability, uncertainty and variability, representativeness and inclusiveness, blocking and stratification, relevance and robustness, forward and reverse causal inference, intention to treat and per protocol analyses, and potential outcomes and counterfactuals.

We discuss how causal diagrams can be used by clinicians to make better individualized treatment decisions. Causal diagrams can distinguish between settings where clinical decisions can rely on a conventional additive regression model fit to data from a historical randomized clinical trial (RCT) to estimate treatment effects and settings where a different approach is needed. This may be because a new patient does not meet the RCT's entry criteria, or a treatment's effect is modified by biomarkers or other variables that act as mediators between treatment and outcome. In some settings, the problem can be addressed simply by including treatment-covariate interaction terms in the statistical regression model used to analyze the RCT dataset. However, if the RCT entry criteria exclude a new patient seen in the clinic, it may be necessary to combine the RCT data with external data from other RCTs, single-arm trials, or preclinical experiments evaluating biological treatment effects. For example, external data may show that treatment effects differ between histological subgroups not recorded in an RCT. A causal diagram may be used to decide whether external observational or experimental data should be obtained and combined with RCT data to compute statistical estimates for making individualized treatment decisions. We use adjuvant treatment of renal cell carcinoma as our motivating example to illustrate how to construct causal diagrams and apply them to guide clinical decisions.

Conditioning regimens play a major role in determining disease outcomes following allogeneic hematopoietic stem cell transplantation (allo-HSCT). The use of i.v. busulfan (Bu) as part of conditioning chemotherapy has been shown to be effective in controlling disease relapse; however, disease relapse remains a major cause of death following allo-HSCT. This study was conducted to determine the long-term outcomes of vorinostat with i.v. Bu plus dual nucleoside analogs clofarabine (Clo) and fludarabine (Flu) in the conditioning regimen for patients undergoing allo-HSCT. This was a rapid dose escalation phase I/II study designed to determine whether the addition of vorinostat would improve the efficacy of standard i.v. Bu/Flu/Clo conditioning regimen. This report presents the long-term disease outcomes of this combination in 68 patients with high-risk leukemia, including 31 (46%) with acute lymphoblastic leukemia (ALL) and 37 (54%) with acute myelogenous leukemia (AML) or myelodysplastic syndrome (MDS). Fifty-eight patients (85%) were in morphologic complete remission at time of transplantation, and 38 (56%) received a matched unrelated donor graft. Over the median follow-up of 37.6 months, 29 of the 68 patients died (43%), and the nonrelapse mortality (NRM) rate was 22% (n = 15). The median overall survival and median NRM were not reached. Nineteen patients (28%) experienced disease progression. The median progression-free survival was 36.8 months. Thirty-seven patients (57%) developed grade II-IV acute graft-versus-host disease (GVHD), and 20 patients (31%) developed chronic GVHD. Our results suggest a lack of benefit from adding a short course of vorinostat to i.v. Bu/Flu/Clo conditioning regimens for leukemia patients undergoing allo- HSCT., (Copyright © 2022 The American Society for Transplantation and Cellular Therapy. Published by Elsevier Inc. All rights reserved.)

Objective: This phase I trial (NCT01912625) evaluated the safety and pharmacokinetics of definitive concurrent chemoradiotherapy (cCRT) and the radiosensitizer trametinib (MEK1/2 inhibitor) for KRAS-mutated nonmetastatic non-small cell lung cancer (NSCLC)., Methods: Patients received cCRT (carboplatin/paclitaxel and 60 Gy/30 fractions radiotherapy); oral trametinib (7 days/week) commenced on day 1 and completed on the final day of radiotherapy. Dose-finding of trametinib was done using the time-to-event continual reassessment method (TiTE-CRM); dose levels were 0.5mg (level -1), 1mg (initial, level 1), 1.5mg (level 2), and 2mg (level 3). Progression-free (PFS) and overall survival (OS) times were also recorded., Results: Fifteen patients (stage III, variety of KRAS mutations) were treated, with 1/5/4/5 at dose levels -1/1/2/3, respectively. Five patients received dose reductions (n=2, levels 2 and 3; n=1, level 1). Twelve patients completed the full cCRT course. One patient (following 12d trametinib) was taken off protocol for an unrelated/unresolved grade 1 event and later experienced grade 5 sepsis/respiratory failure. There was one grade 4 retinal detachment; grade 3 events included skin rash (n=2) and ventricular dysfunction, pneumonitis, pain, fatigue, and diarrhea (n=1 each). The final dose selected by the TiTE-CRM of trametinib was 1.5 mg. Pharmacokinetic profiles were elucidated and extensively described. At median follow-up of 70 months, median PFS was 11 months and median OS was 38 months., Conclusions: The MTD for trametinib when combined with cCRT is 1.5 mg, with encouraging preliminary outcomes. This combination merits further study to combine with consolidation durvalumab in non-metastatic KRAS mutant NSCLC., (Copyright © 2022 The Authors. Published by Elsevier Ltd.. All rights reserved.)

This paper proposes randomized controlled clinical trial design to evaluate external cooling as a means to control fever and thereby reduce mortality in patients with septic shock. The trial will include concurrent external cooling and control arms while adaptively incorporating historical control arm data. Bayesian group sequential monitoring will be done using a posterior comparative test based on the 60-day survival distribution in each concurrent arm. Posterior inference will follow from a Bayesian discrete time survival model that facilitates adaptive incorporation of the historical control data through an innovative regression framework with a multivariate spike-and-slab prior distribution on the historical bias parameters. For each interim test, the amount of information borrowed from the historical control data will be determined adaptively in a manner that reflects the degree of agreement between historical and concurrent control arm data. Guidance is provided for selecting Bayesian posterior probability group-sequential monitoring boundaries. Simulation results elucidating how the proposed method borrows strength from the historical control data are reported. In the absence of historical control arm bias, the proposed design controls the type I error rate and provides substantially larger power than reasonable comparators, whereas in the presence bias of varying magnitude, type I error rate inflation is curbed.

We argue that well-informed patient-specific decision-making may be carried out as three consecutive tasks: (1) estimating key parameters of a statistical model, (2) using prognostic information to convert these parameters into clinically interpretable values, and (3) specifying joint utility functions to quantify risk-benefit trade-offs between clinical outcomes. Using the management of metastatic clear cell renal cell carcinoma as our motivating example, we explain the role of prognostic covariates that characterize between-patient heterogeneity in clinical outcomes. We show that explicitly specifying the joint utility of clinical outcomes provides a coherent basis for patient-specific decision-making.

Background: COVID-19 Convalescent plasma (CCP) is safe and effective, particularly if given at an early stage of the disease. Our study aimed to identify an association between survival and specific antibodies found in CCP., Patients and Methods: Patients ≥18 years of age who were hospitalized with moderate to severe COVID-19 infection and received CCP at the MD Anderson Cancer Center between 4/30/2020 and 8/20/2020 were included in the study. We quantified the levels of anti-SARS-CoV-2 antibodies, as well as antibodies against antigens of other coronavirus strains, in the CCP units and compared antibody levels with patient outcomes. For each antibody, a Bayesian exponential survival time regression model including prognostic variables was fit, and the posterior probability of a beneficial effect (PBE) of higher antibody level on survival time was computed., Results: CCP was administered to 44 cancer patients. The median age was 60 years (range 37-84) and 19 (43%) were female. Twelve patients (27%) died of COVID-19-related complications. Higher levels of two non-SARS-CoV-2-specific antibodies, anti-HCoV-OC43 spike IgG and anti-HCoV-HKU1 spike IgG, had PBE = 1.00, and 4 SARS-CoV-2-specific antibodies had PBEs between 0.90 and 0.95. Other factors associated with better survival were shorter time to CCP administration, younger age, and female sex., Conclusions: Common cold coronavirus spike IgG antibodies anti-HCoV-OC43 and anti-HCoV-HKU1 may target a common domain for SARS-CoV-2 and other coronaviruses. They provide a promising therapeutic target for monoclonal antibody production., Competing Interests: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest., (Copyright © 2021 Greenbaum, Klein, Martinez, Song, Thall, Ramdial, Knape, Aung, Scroggins, Knopfelmacher, Mulanovich, Borjan, Adachi, Muthu, Leung, Medina, Champlin, Olson, Alousi, Rezvani and Shpall.)

This paper proposes a Bayesian adaptive basket trial design to optimize the dose-schedule regimes of an experimental agent within disease subtypes, called "baskets", for phase I-II clinical trials based on late-onset efficacy and toxicity. To characterize the association among the baskets and regimes, a Bayesian hierarchical model is assumed that includes a heterogeneity parameter, adaptively updated during the trial, that quantifies information shared across baskets. To account for late-onset outcomes when doing sequential decision making, unobserved outcomes are treated as missing values and imputed by exploiting early biomarker and low-grade toxicity information. Elicited joint utilities of efficacy and toxicity are used for decision making. Patients are randomized adaptively to regimes while accounting for baskets, with randomization probabilities proportional to the posterior probability of achieving maximum utility. Simulations are presented to assess the design's robustness and ability to identify optimal dose-schedule regimes within disease subtypes, and to compare it to a simplified design that treats the subtypes independently.

A Bayesian group sequential design is proposed that performs survival comparisons within patient subgroups in randomized trials where treatment-subgroup interactions may be present. A latent subgroup membership variable is assumed to allow the design to adaptively combine homogeneous subgroups, or split heterogeneous subgroups, to improve the procedure's within-subgroup power. If a baseline covariate related to survival is available, the design may incorporate this information to improve subgroup identification while basing the comparative test on the average hazard ratio. General guidelines are provided for calibrating prior hyperparameters and design parameters to control the overall Type I error rate and optimize performance. Simulations show that the design is robust under a wide variety of different scenarios. When two or more subgroups are truly homogeneous but differ from the other subgroups, the proposed method is substantially more powerful than tests that either ignore subgroups or conduct a separate test within each subgroup. Supplementary materials for this article are available online.

This study investigated the efficacy and safety of azacitidine maintenance in the posttransplant setting based on the encouraging phase 1/2 reports for azacitidine maintenance in patients with acute myeloid leukemia/myelodysplastic syndrome (AML/MDS). Between 2009 and 2017, a total of 187 patients aged 18 to 75 years were entered into a randomized controlled study of posttransplant azacitidine if they were in complete remission. Patients randomized to the treatment arm (n = 93) were scheduled to receive azacitidine, given as 32 mg/m2 per day subcutaneously for 5 days every 28 days for 12 cycles. The control arm (n = 94) had no intervention. Eighty-seven of the 93 patients started azacitidine maintenance. The median number of cycles received was 4; a total of 29 patients relapsed on study, and 23 patients withdrew from the study due to toxicity, patient's preference, or logistical reasons. Median relapse-free survival (RFS) was 2.07 years in the azacitidine group vs 1.28 years in the control group (P = .43). There was also no significant difference for overall survival, with a median of 2.52 years vs 2.56 years in the azacitidine and control groups (P = .85), respectively. Multivariate Cox regression analysis revealed no improvement in RFS or overall survival with the use of azacitidine as maintenance compared with the control group (hazard ratios of 0.73 [95% confidence interval, 0.49-1.1; P = .14] and 0.84 [95% confidence interval, 0.55-1.29; P = .43]) [corrected]. This randomized trial with azacitidine maintenance showed that a prospective trial in the posttransplant setting was feasible and safe but challenging. Although RFS was comparable between the 2 arms, we believe the strategy of maintenance therapy merits further study with a goal to reduce the risk of relapse in patients with AML/MDS. This trial was registered at www.clinicaltrials.gov as #NCT00887068., (© 2020 by The American Society of Hematology.)

This article introduces the adverse event (AE) burden score. The AE burden by treatment cycle is a weighted sum of all grades and AEs that the patient experienced in a cycle. The overall AE burden score is the total AE burden the patient experienced across all treatment cycles. AE data from two completed Alliance multi-center randomized double-blind placebo-controlled trials, with different AE profiles (NCCTG 97-24-51: 176 patients, and A091105: 83 patients), were utilized for illustration. Results of the AE burden score analyses corroborated the trials' primary results. In 97-24-51, the overall AE burden for patients on the treatment arm was 2.2 points higher than those on the placebo arm, with a higher AE burden for patients who went off treatment early due to AE. Similarly, in A091105, the overall AE burden was 1.6 points higher on the treatment arm. On the placebo arms, the AE burden in 97-24-51 remained constant over time; and increased in later cycles in A091105, likely attributable to the increase in disease morbidity. The AE burden score enables statistical comparisons analogous to other quantitative endpoints in clinical trials, and can readily accommodate different trial settings, diseases, and treatments, with diverse AE profiles.

Purpose: Allogeneic hematopoietic stem-cell transplantation (HSCT) is a curative treatment for many hematologic cancers. Use of haploidentical (mismatched) donors increases HSCT availability but is limited by severe graft-versus-host disease (GvHD) and delayed immune reconstitution. Alloanergization of donor T cells is a simple approach to rebuild immunity while limiting GvHD after haploidentical HSCT, but the optimal T-cell dose and impact on immune reconstitution remain unknown. Patients and Methods: We performed a multicenter phase I trial of alloanergized donor lymphocyte infusion (aDLI) after CD34-selected myeloablative haploidentical HSCT. The primary aim was feasibility and safety with secondary aims of assessing the less frequently addressed issue of impact on immune reconstitution. Results: Nineteen patients with high-risk acute leukemia or myelodysplasia were enrolled. Engraftment occurred in 18 of 19 patients (95%). Pre-aDLI, 12 patients (63%) had bacteremia, nine of 17 at-risk patients (53%) reactivated CMV, and one developed acute GvHD. Sixteen patients received aDLI at dose levels 1 (10 3 T cells/kg, n = 4), 2 (10 4 , n = 8), and 3 (10 5 , n = 4). After aDLI, five patients developed clinically significant acute GvHD, and four of 14 at-risk patients (29%) reactivated CMV. T-cell recovery was significantly greater, and functional virus- and tumor-associated antigen-specific T cells were detectable earlier in patients receiving dose level 2 or 3 versus dose level 1/no aDLI. Alloanergization of donor cells expanded the CD4 + T-regulatory cell frequency within aDLI, which increased further in vivo without impeding expansion of virus- and tumor-associated antigen-specific T cells. Conclusions: These data demonstrate safety and a potential role for aDLI in contributing to immune reconstitution and expanding tolerogenic regulatory T cells in vivo after CD34-selected myeloablative haploidentical HSCT. Clin Cancer Res; 24(17); 4098-109. ©2018 AACR ., (©2018 American Association for Cancer Research.)

We conducted a prospective phase 2 trial of gemcitabine, busulfan and melphalan (Gem/Bu/Mel) with autologous stem cell transplantation (ASCT) in patients with primary refractory or poor-risk relapsed Hodgkin lymphoma (HL) (ie, extranodal relapse or within 1 year of frontline therapy). The trial was powered to detect an improvement in 2-year progression-free survival (PFS) from a historical 50% using a BEAM regimen (carmustine/etoposide/cytarabine/melphalan) to 65%. We compared the study population with all other concurrent patients who were eligible for the trial but instead received the BEAM regimen at our center. No patient received post-ASCT maintenance therapy. The Gem/Bu/Mel trial enrolled 80 patients with a median age of 31 years, 41% with primary refractory HL and 59% with relapsed HL (36% extranodal relapses), and 30% with positron emission tomography (PET)-positive lesions at ASCT. The concurrent BEAM (n = 45) and Gem/Bu/Mel cohorts were well balanced except for higher rates of bulky relapse and PET-positive tumors in the Gem/Bu/Mel cohort. There were no transplantation-related deaths in either cohort. At a median follow-up of 34.5 months (range, 26 to 72 months), Gem/Bu/Mel was associated with better 2-year PFS (65% versus 51%; P = .008) and overall survival (89% versus 73%; P = .0003). In conclusion, our data show that Gem/Bu/Mel is safe, in this nonrandomized comparison yielding improved outcomes compared with a concurrently treated and prognostically matched cohort of patients with primary refractory or poor-risk relapsed HL receiving BEAM., (Copyright © 2018 The American Society for Blood and Marrow Transplantation. Published by Elsevier Inc. All rights reserved.)

Background: Conventional phase I algorithms for finding a phase-2 recommended dose (P2RD) based on toxicity alone is problematic because the maximum tolerated dose (MTD) is not necessarily the optimal dose with the most desirable risk-benefit trade-off. Moreover, the increasingly common practice of treating an expansion cohort at a chosen MTD has undesirable consequences that may not be obvious., Patients and Methods: We review the phase I-II paradigm and the EffTox design, which utilizes both efficacy and toxicity to choose optimal doses for successive patient cohorts and find the optimal P2RD. We conduct a computer simulation study to compare the performance of the EffTox design with the traditional 3 + 3 design and the continuous reassessment method., Results: By accounting for the risk-benefit trade-off, the EffTox phase I-II design overcomes the limitations of conventional toxicity-based phase I designs. Numerical simulations show that the EffTox design has higher probabilities of identifying the optimal dose and treats more patients at the optimal dose., Conclusions: Phase I-II designs, such as the EffTox design, provide a coherent and efficient approach to finding the optimal P2RD by explicitly accounting for risk-benefit trade-offs underlying medical decisions.

Medical therapy often consists of multiple stages, with a treatment chosen by the physician at each stage based on the patient's history of treatments and clinical outcomes. These decisions can be formalized as a dynamic treatment regime. This paper describes a new approach for optimizing dynamic treatment regimes that bridges the gap between Bayesian inference and existing approaches, like Q-learning. The proposed approach fits a series of Bayesian regression models, one for each stage, in reverse sequential order. Each model uses as a response variable the remaining payoff assuming optimal actions are taken at subsequent stages, and as covariates the current history and relevant actions at that stage. The key difficulty is that the optimal decision rules at subsequent stages are unknown, and even if these decision rules were known the relevant response variables may be counterfactual. However, posterior distributions can be derived from the previously fitted regression models for the optimal decision rules and the counterfactual response variables under a particular set of rules. The proposed approach averages over these posterior distributions when fitting each regression model. An efficient sampling algorithm for estimation is presented, along with simulation studies that compare the proposed approach with Q-learning.

A variable selection procedure is developed for a semi-competing risks regression model with three hazard functions that uses spike-and-slab priors and stochastic search variable selection algorithms for posterior inference. A rule is devised for choosing the threshold on the marginal posterior probability of variable inclusion based on the Deviance Information Criterion (DIC) that is examined in a simulation study. The method is applied to data from esophageal cancer patients from the MD Anderson Cancer Center, Houston, TX, where the most important covariates are selected in each of the hazards of effusion, death before effusion, and death after effusion. The DIC procedure that is proposed leads to similar selected models regardless of the choices of some of the hyperparameters. The application results show that patients with intensity-modulated radiation therapy have significantly reduced risks of pericardial effusion, pleural effusion, and death before either effusion type.

We propose a Bayesian nonparametric utility-based group sequential design for a randomized clinical trial to compare a gel sealant to standard care for resolving air leaks after pulmonary resection. Clinically, resolving air leaks in the days soon after surgery is highly important, since longer resolution time produces undesirable complications that require extended hospitalization. The problem of comparing treatments is complicated by the fact that the resolution time distributions are skewed and multi-modal, so using means is misleading. We address these challenges by assuming Bayesian nonparametric probability models for the resolution time distributions and basing the comparative test on weighted means. The weights are elicited as clinical utilities of the resolution times. The proposed design uses posterior expected utilities as group sequential test criteria. The procedure's frequentist properties are studied by extensive simulations.

A design is presented for a randomized clinical trial comparing two second-line treatments, chemotherapy versus chemotherapy plus reirradiation, for treatment of recurrent non-small-cell lung cancer. The central research question is whether the potential efficacy benefit that adding reirradiation to chemotherapy may provide justifies its potential for increasing the risk of toxicity. The design uses two co-primary outcomes: time to disease progression or death, and time to severe toxicity. Because patients may be given an active third-line treatment at disease progression that confounds second-line treatment effects on toxicity and survival following disease progression, for the purpose of this comparative study follow-up ends at disease progression or death. In contrast, follow-up for disease progression or death continues after severe toxicity, so these are semi-competing risks. A conditionally conjugate Bayesian model that is robust to misspecification is formulated using piecewise exponential distributions. A numerical utility function is elicited from the physicians that characterizes desirabilities of the possible co-primary outcome realizations. A comparative test based on posterior mean utilities is proposed. A simulation study is presented to evaluate test performance for a variety of treatment differences, and a sensitivity assessment to the elicited utility function is performed. General guidelines are given for constructing a design in similar settings, and a computer program for simulation and trial conduct is provided.

Pretransplant conditioning regimens critically determine outcomes in the setting of allogeneic stem cell transplantation (allo-SCT). The use of nucleoside analogs such as fludarabine (Flu) in combination with i.v. busulfan (Bu) has been shown to be highly effective as a pretransplant conditioning regimen in acute myeloid leukemia (AML), chronic myeloid leukemia (CML), and myelodysplastic syndrome (MDS). Because leukemia relapse remains the leading cause of death after allo-SCT, we studied whether clofarabine (Clo), a nucleoside analog with potent antileukemia activity, can be used to complement Flu. In a preliminary report, we previously showed the safety and efficacy of Clo ± Flu with i.v. Bu in 51 patients with high-risk AML, CML, and MDS. The study has now been completed, and we present long-term follow-up data on the entire 70-patient population, which included 49 (70%), 8 (11%), and 13 (19%) patients with AML, MDS, and CML, respectively. Thirteen patients (19%) were in complete remission, and 41 patients (59%) received matched unrelated donor grafts. Engraftment was achieved in all patients. Sixty-three patients (90%) achieved complete remission. There were no deaths reported at day +30, and the 100-day nonrelapse mortality rate was 4% (n = 3). Thirty-one percent of patients (n = 22) developed grades II to IV acute graft-versus-host disease, and the median overall survival and progression-free survival times were 2.4 years and .9 years, respectively. Our results confirm the safety and overall and progression-free survival advantage of the arms with higher Clo doses and lower Flu doses, which was most prominent in the AML/MDS group., (Copyright © 2016 The American Society for Blood and Marrow Transplantation. Published by Elsevier Inc. All rights reserved.)

This paper is motivated by a phase I-II clinical trial of a targeted agent for advanced solid tumors. We study a stylized version of this trial with the goal to determine optimal actions in each of two cycles of therapy. A design is presented that generalizes the decision-theoretic two-cycle design of Lee and others (2015. Bayesian dose-finding in two treatment cycles based on the joint utility of efficacy and toxicity. Journal of the American Statistical Association, to appear) to accommodate ordinal outcomes. Backward induction is used to jointly optimize the actions taken for each patient in each of the two cycles, with the second action accounting for the patient's cycle 1 dose and outcomes. A simulation study shows that simpler designs obtained by dichotomizing the ordinal outcomes either perform very similarly to the proposed design, or have much worse performance in some scenarios. We also compare the proposed design with the simpler approaches of optimizing the doses in each cycle separately, or ignoring the distinction between cycles 1 and 2., (© The Author 2015. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.)

We analyze a dataset arising from a clinical trial involving multi-stage chemotherapy regimes for acute leukemia. The trial design was a 2 × 2 factorial for frontline therapies only. Motivated by the idea that subsequent salvage treatments affect survival time, we model therapy as a dynamic treatment regime (DTR), that is, an alternating sequence of adaptive treatments or other actions and transition times between disease states. These sequences may vary substantially between patients, depending on how the regime plays out. To evaluate the regimes, mean overall survival time is expressed as a weighted average of the means of all possible sums of successive transitions times. We assume a Bayesian nonparametric survival regression model for each transition time, with a dependent Dirichlet process prior and Gaussian process base measure (DDP-GP). Posterior simulation is implemented by Markov chain Monte Carlo (MCMC) sampling. We provide general guidelines for constructing a prior using empirical Bayes methods. The proposed approach is compared with inverse probability of treatment weighting, including a doubly robust augmented version of this approach, for both single-stage and multi-stage regimes with treatment assignment depending on baseline covariates. The simulations show that the proposed nonparametric Bayesian approach can substantially improve inference compared to existing methods. An R program for implementing the DDP-GP-based Bayesian nonparametric analysis is freely available at https://www.ma.utexas.edu/users/yxu/.

More active high-dose regimens are needed for refractory/poor-risk relapsed lymphomas. We previously developed a regimen of infusional gemcitabine/busulfan/melphalan, exploiting the synergistic interaction. Its encouraging activity in refractory lymphomas led us to further enhance its use as a platform for epigenetic modulation. We previously observed increased cytotoxicity in refractory lymphoma cell lines when the histone deacetylase inhibitor vorinostat was added to gemcitabine/busulfan/melphalan, which prompted us to clinically study this four-drug combination. Patients ages 12 to 65 with refractory diffuse large B cell lymphoma (DLCL), Hodgkin (HL), or T lymphoma were eligible. Vorinostat was given at 200 mg/day to 1000 mg/day (days -8 to -3). Gemcitabine was infused continuously at 10 mg/m(2)/minute over 4.5 hours (days -8 and -3). Busulfan dosing targeted 4000 μM-minute/day (days -8 to -5). Melphalan was infused at 60 mg/m(2)/day (days -3 and -2). Patients with CD20(+) tumors received rituximab (375 mg/m(2), days +1 and +8). We enrolled 78 patients: 52 DLCL, 20 HL, and 6 T lymphoma; median age 44 years (range, 15 to 65); median 3 prior chemotherapy lines (range, 2 to 7); and 48% of patients had positron emission tomography-positive tumors at high-dose chemotherapy (29% unresponsive). The vorinostat dose was safely escalated up to 1000 mg/day, with no treatment-related deaths. Toxicities included mucositis and dermatitis. Neutrophils and platelets engrafted promptly. At median follow-up of 25 (range, 16 to 41) months, event-free and overall survival were 61.5% and 73%, respectively (DLCL) and 45% and 80%, respectively (HL). In conclusion, vorinostat/gemcitabine/busulfan/melphalan is safe and highly active in refractory/poor-risk relapsed lymphomas, warranting further evaluation. This trial was registered at ClinicalTrials.gov (NCI-2011-02891)., (Copyright © 2015 American Society for Blood and Marrow Transplantation. Published by Elsevier Inc. All rights reserved.)

Detection of donor-specific anti-HLA antibodies (DSA) has been associated with graft rejection in all forms of transplantation. The mechanism by which DSA increase the risk of graft failure remains unclear. We hypothesized that complement-binding DSA are associated with engraftment failure in hematopoietic stem cell transplantation (HSCT) and analyzed 122 haploidentical transplant recipients tested prospectively for DSA. Retrospective analysis to detect C1q binding DSA (C1q+DSA) was performed on 22 allosensitized recipients. Twenty-two of 122 patients (18%) had DSA, 19 of which were women (86%). Seven patients with DSA (32%) rejected the graft. Median DSA level at transplant for patients who failed to engraft was 10,055 mean fluorescence intensity (MFI) versus 2065 MFI for those who engrafted (P = .007). Nine patients with DSA were C1q positive in the initial samples with median DSA levels of 15,279 MFI (range, 1554 to 28,615), compared with 7 C1q-negative patients with median DSA levels of 2471 MFI (range, 665 to 12,254) (P = .016). Of 9 patients who were C1q positive in the initial samples, 5 patients remained C1q positive at time of transplant (all with high DSA levels [median, 15,279; range, 6487 to 22,944]) and experienced engraftment failure, whereas 4 patients became C1q negative pretransplant and all engrafted the donor cells (P = .008). In conclusion, patients with high DSA levels (>5000 MFI) and complement-binding DSA antibodies (C1q positive) appear to be at much higher risk of primary graft failure. The presence of C1q+DSA should be assessed in allosensitized patients before HSCT. Reduction of C1q+DSA levels might prevent engraftment failure in HSCT., (Copyright © 2015 American Society for Blood and Marrow Transplantation. Published by Elsevier Inc. All rights reserved.)

We previously demonstrated vast expansion of hypoxic areas in the leukemic microenvironment and provided a rationale for using hypoxia-activated prodrugs. PR104 is a phosphate ester that is rapidly hydrolyzed in vivo to the corresponding alcohol PR-104A and further reduced to the amine and hydroxyl-amine nitrogen mustards that induce DNA cross-linking in hypoxic cells under low oxygen concentrations. In this phase I/II study, patients with relapsed/refractory acute myeloid leukemia (n=40) after 1 or 2 prior treatments or acute lymphoblastic leukemia (n=10) after any number of prior treatments received PR104; dose ranged from 1.1 to 4 g/m(2). The most common treatment-related grade 3/4 adverse events were myelosuppression (anemia 62%, neutropenia 50%, thrombocytopenia 46%), febrile neutropenia (40%), infection (24%), and enterocolitis (14%). Ten of 31 patients with acute myeloid leukemia (32%) and 2 of 10 patients with acute lymphoblastic leukemia (20%) who received 3 g/m(2) or 4 g/m(2) had a response (complete response, n=1; complete response without platelet recovery, n=5; morphological leukemia-free state, n=6). The extent of hypoxia was evaluated by the hypoxia tracer pimonidazole administered prior to a bone marrow biopsy and by immunohistochemical assessments of hypoxia-inducible factor alpha and carbonic anhydrase IX. A high fraction of leukemic cells expressed these markers, and PR104 administration resulted in measurable decrease of the proportions of hypoxic cells. These findings indicate that hypoxia is a prevalent feature of the leukemic microenvironment and that targeting hypoxia with hypoxia-activated prodrugs warrants further evaluation in acute leukemia. The trial is registered at clinicaltrials.gov identifier: 01037556., (Copyright© Ferrata Storti Foundation.)

A phase I/II clinical trial design is proposed for adaptively and dynamically optimizing each patient's dose in each of two cycles of therapy based on the joint binary efficacy and toxicity outcomes in each cycle. A dose-outcome model is assumed that includes a Bayesian hierarchical latent variable structure to induce association among the outcomes and also facilitate posterior computation. Doses are chosen in each cycle based on posteriors of a model-based objective function, similar to a reinforcement learning or Q-learning function, defined in terms of numerical utilities of the joint outcomes in each cycle. For each patient, the procedure outputs a sequence of two actions, one for each cycle, with each action being the decision to either treat the patient at a chosen dose or not to treat. The cycle 2 action depends on the individual patient's cycle 1 dose and outcomes. In addition, decisions are based on posterior inference using other patients' data, and therefore the proposed method is adaptive both within and between patients. A simulation study of the method is presented, including comparison to two-cycle extensions of the conventional 3+3 algorithm, continual reassessment method, and a Bayesian model-based design, and evaluation of robustness.

Introduction: To delineate an optimal clinical target volume (CTV) for gastroesophageal junction (GEJ) cancers by comparing locoregional versus extended locoregional radiation (RT) volumes., Materials: This retrospective analysis examined 222 patients (111 matched pairs treated with locoregional versus extended locoregional RT) with nonmetastatic GEJ carcinomas treated with concurrent chemoradiation ± surgery. The CTV for locoregional volumes was defined as gross tumor volume + elective coverage of first-echelon nodes and sometimes the celiac axis. The CTV for extended locoregional volumes was defined as gross tumor volume + elective coverage of celiac and splenic (± porta) nodes. Variables used for matching included sex, stage, performance status, histology, receipt of induction chemotherapy, type of concurrent chemotherapy, RT modality, receipt of surgery, type of surgery, and Siewert classification. Regression models stratified by matched pairs were fit to estimate effect of RT volume on clinical endpoints., Results: Adjusting p values for multiple testing, patients treated with extended locoregional versus locoregional RT had increased odds of grade 2+ acute chemoradiation-associated gastrointestinal toxicity (odds ratio: 2.92, adj. p = 0.0447). However, differing RT volumes were not significantly associated with postoperative complication rates, pathologic T-stage, frequency of positive perigastric/regional nodes on surgical specimen, distant-metastases progression-free survival, locoregional progression-free survival, or overall survival (adj. p > 0.05). Of the patients who did (n = 124) and did not (n = 72) receive elective celiac RT, two (1.6%) and two patients (2.8%), respectively, relapsed in the celiac nodes. No patients failed in the splenic or porta nodes., Conclusions: Most GEJ cancers can be safely treated without elective inclusion of splenic/porta nodes.

The Intubation-Surfactant-Extubation (INSURE) procedure is used worldwide to treat pre-term newborn infants suffering from respiratory distress syndrome, which is caused by an insufficient amount of the chemical surfactant in the lungs. With INSURE, the infant is intubated, surfactant is administered via the tube to the trachea, and at completion the infant is extubated. This improves the infant's ability to breathe and thus decreases the risk of long term neurological or motor disabilities. To perform the intubation safely, the newborn infant first must be sedated. Despite extensive experience with INSURE, there is no consensus on what sedative dose is best. This paper describes a Bayesian sequentially adaptive design for a multi-institution clinical trial to optimize the sedative dose given to pre-term infants undergoing the INSURE procedure. The design is based on three clinical outcomes, two efficacy and one adverse, using elicited numerical utilities of the eight possible elementary outcomes. A flexible Bayesian parametric trivariate dose-outcome model is assumed, with the prior derived from elicited mean outcome probabilities. Doses are chosen adaptively for successive cohorts of infants using posterior mean utilities, subject to safety and efficacy constraints. A computer simulation study of the design is presented.

Donor chimerism after allogeneic stem cell transplantation (allo-SCT) is commonly used to predict overall survival (OS) and disease-free survival (DFS). Because chimerism is observed at 1 or more times after allo-SCT and not at baseline, if chimerism is in fact associated with OS or DFS, then the occurrence of either disease progression or death informatively censors (terminates) the observed chimerism process. This violates the assumptions underlying standard statistical regression methods for survival analysis, which may lead to biased conclusions. To assess the association between the longitudinal post-allo-SCT donor chimerism process and OS or DFS, we analyzed data from 195 patients with acute myelogenous leukemia (n = 157) or myelodysplastic syndrome (n = 38) who achieved complete remission after allo-SCT following a reduced-toxicity conditioning regimen of fludarabine/intravenous busulfan. Median follow-up was 31 months (range, 1.1 to 105 months). Fitted joint longitudinal-survival time models showed that a binary indicator of complete (100%) donor chimerism and increasing percent of donor T cells were significantly associated with longer OS, whereas decreasing percent of donor T cells was highly significantly associated with shorter OS. Our analyses illustrate the usefulness of modeling repeated post-allo-SCT chimerism measurements as individual longitudinal processes jointly with OS and DFS to estimate their relationships., (Copyright © 2014 American Society for Blood and Marrow Transplantation. Published by Elsevier Inc. All rights reserved.)

A practical impediment in adaptive clinical trials is that outcomes must be observed soon enough to apply decision rules to choose treatments for new patients. For example, if outcomes take up to six weeks to evaluate and the accrual rate is one patient per week, on average three new patients will be accrued while waiting to evaluate the outcomes of the previous three patients. The question is how to treat the new patients. This logistical problem persists throughout the trial. Various ad hoc practical solutions are used, none entirely satisfactory. We focus on this problem in phase I-II clinical trials that use binary toxicity and efficacy, defined in terms of event times, to choose doses adaptively for successive cohorts. We propose a general approach to this problem that treats late-onset outcomes as missing data, uses data augmentation to impute missing outcomes from posterior predictive distributions computed from partial follow-up times and complete outcome data, and applies the design's decision rules using the completed data. We illustrate the method with two cancer trials conducted using a phase I-II design based on efficacy-toxicity trade-offs, including a computer stimulation study.

Background: The contribution of induction chemotherapy (IC) before preoperative chemoradiation for esophageal cancer (EC) is not known. We hypothesized that IC would increase the rate of pathologic complete response (pathCR)., Methods: Trimodality-eligibile patients were randomized to receive no IC (Arm A) or IC (oxaliplatin/FU; Arm B) before oxaliplatin/FU/radiation. Surgery was attempted ∼5-6 weeks after chemoradiation. The pathCR rate, post-surgery 30-day mortality, overall survival (OS), and toxic effects were assessed. Bayesian methods and Fisher's exact test were used., Results: One hundred twenty-six patients were randomized dynamically to balance the two arms for histology, baseline stage, gender, race, and age. Fifty-five patients in Arm A and 54 in Arm B underwent surgery. The median actuarial OS for all patients (54 deaths) was 45.62 months [95% confidence interval (CI), 27.63-NA], with median OS 45.62 months (95% CI 25.56-NA) in Arm A and 43.68 months (95% CI 27.63-NA) in Arm B (P = 0.69). The pathCR rate in Arm A was 13% (7 of 55) and 26% (14 of 54) in Arm B (two-sided Fisher's exact test, P = 0.094). Safety was similar in both arms., Conclusions: These data suggest that IC produces non-significant increase in the pathCR rate and does not prolong OS. Further development of IC before chemoradiation may not be beneficial. Clinical trial no.: NCT 00525915 (www.clinicaltrials.gov).

Allogeneic hematopoietic stem cell transplantation (allo-HCT) is a potentially curative treatment for multiple myeloma (MM); however, because of high treatment-related mortality (TRM), its role is not well defined. Patients with newly diagnosed, relapsed, or primary refractory myeloma were enrolled in a randomized phase II trial of 2 reduced-intensity conditioning regimens: fludarabine 120 mg/m(2) + melphalan 100 mg/m(2) (FM100) versus fludarabine 120 mg/m(2) + melphalan 140 mg/m(2) (FM140) before allo-HCT from related or unrelated donors. Fifty patients underwent allo-HCT using FM100 (n = 23) or FM140 (n = 27) conditioning between April 2002 and 2011. There were no significant differences between FM100 and FM140 in time to neutrophil engraftment (P = .21), acute grade II to IV graft-versus-host disease (GVHD) (P = 1.0), chronic GVHD (P = .24), response rate (P = 1.0), TRM (13% versus 15%, P = 1.0), median progression-free survival (PFS), 11.7 versus 8.4 months, P = .12, and median overall survival (OS), 35.1 versus 19.7 months, P = .38. Cumulative incidence of disease progression in FM100 and FM140 was 43% and 70%, respectively (P = .08). Recurrent disease was the most common cause of death for both FM100 (26%) and FM140 (44%), P = .24. On multivariate analysis, disease status at allo-HCT, complete response or very good partial response (VGPR) was significantly associated with longer PFS (15.6 versus 9.6 months in patients with

Purpose: Clinical features characteristic of small-cell prostate carcinoma (SCPC), "anaplastic," often emerge during the progression of prostate cancer. We sought to determine the efficacy of platinum-based chemotherapy in patients meeting at least one of seven prospectively defined "anaplastic" clinical criteria, including exclusive visceral or predominantly lytic bone metastases, bulky tumor masses, low prostate-specific antigen levels relative to tumor burden, or short response to androgen deprivation therapy., Experimental Design: A 120-patient phase II trial of first-line carboplatin and docetaxel (CD) and second-line etoposide and cisplatin (EP) was designed to provide reliable clinical response estimates under a Bayesian probability model with early stopping rules in place for futility and toxicity., Results: Seventy-four of 113 (65.4%) and 24 of 71 (33.8%) were progression free after four cycles of CD and EP, respectively. Median overall survival (OS) was 16 months [95% confidence interval (CI), 13.6-19.0 months]. Of the seven "anaplastic" criteria, bulky tumor mass was significantly associated with poor outcome. Lactic acid dehydrogenase strongly predicted for OS and rapid progression. Serum carcinoembryonic antigen (CEA) concentration strongly predicted OS but not rapid progression. Neuroendocrine markers did not predict outcome or response to therapy., Conclusion: Our findings support the hypothesis that patients with "anaplastic" prostate cancer are a recognizable subset characterized by a high response rate of short duration to platinum-containing chemotherapies, similar to SCPC. Our results suggest that CEA is useful for selecting therapy in men with castration-resistant prostate cancer and consolidative therapies to bulky high-grade tumor masses should be considered in this patient population., (©2013 AACR.)

Background: Visceral pleural invasion (VPI) is used as an indicator of adverse prognosis in non-small cell lung cancer (NSCLC). The purpose of this retrospective study was to evaluate the impact of VPI on disease-free survival (DFS) and overall survival (OS) in patients with node-negative NSCLC., Methods: Between 1998 and 2009, 1,166 patients with pathologic N0M0 NSCLC underwent surgical resection by lobectomy. Two hundred fourteen patients with VPI were compared with 952 patients without VPI., Results: Median follow-up was 59 months. In multivariate analysis, VPI, larger tumor size, older age, female sex, and poor performance status were significantly associated with decreased OS. In contrast, larger tumor size, female sex, and poor performance, but notably not VPI, were associated with decreased DFS. After examining interactive effects of VPI and T stage subgroups, we found that VPI did not significantly affect either OS or DFS in the subgroups of patients with smaller tumor sizes-stage T1a, stage T1b, or stage T2a. In contrast, a deleterious effect of VPI on DFS was seen for tumors larger than 5 cm-stages T2b and T3-with the VPI-stage T3 interaction effect being statistically significant for DFS but not for OS., Conclusions: The effect of VPI on survival in NSCLC varies greatly with tumor size, with VPI not strongly associated with OS or DFS in tumors smaller than 5 cm, but showing large negative effects on DFS for stage T2b and stage T3 tumors. Using VPI to upstage T1 tumors to a higher T stage is not warranted because it would misrepresent these VPI-T stage subgroup effects., (Copyright © 2013 The Society of Thoracic Surgeons. Published by Elsevier Inc. All rights reserved.)

Prior effective sample size (ESS) of a Bayesian parametric model was defined by Morita, et al. (2008, Biometrics, 64 , 595-602). Starting with an ε -information prior defined to have the same means and correlations as the prior but to be vague in a suitable sense, the ESS is the required sample size to obtain a hypothetical posterior very close to the prior. In this paper, we present two alternative definitions for the prior ESS that are suitable for a conditionally independent hierarchical model. The two definitions focus on either the first level prior or second level prior. The proposed methods are applied to important examples to verify that each of the two types of prior ESS matches the intuitively obvious answer where it exists. We illustrate the method with applications to several motivating examples, including a single-arm clinical trial to evaluate treatment response probabilities across different disease subtypes, a dose-finding trial based on toxicity in this setting, and a multicenter randomized trial of treatments for affective disorders.

We present new statistical analyses of data arising from a clinical trial designed to compare two-stage dynamic treatment regimes (DTRs) for advanced prostate cancer. The trial protocol mandated that patients were to be initially randomized among four chemotherapies, and that those who responded poorly were to be rerandomized to one of the remaining candidate therapies. The primary aim was to compare the DTRs' overall success rates, with success defined by the occurrence of successful responses in each of two consecutive courses of the patient's therapy. Of the one hundred and fifty study participants, forty seven did not complete their therapy per the algorithm. However, thirty five of them did so for reasons that precluded further chemotherapy; i.e. toxicity and/or progressive disease. Consequently, rather than comparing the overall success rates of the DTRs in the unrealistic event that these patients had remained on their assigned chemotherapies, we conducted an analysis that compared viable switch rules defined by the per-protocol rules but with the additional provision that patients who developed toxicity or progressive disease switch to a non-prespecified therapeutic or palliative strategy. This modification involved consideration of bivariate per-course outcomes encoding both efficacy and toxicity. We used numerical scores elicited from the trial's Principal Investigator to quantify the clinical desirability of each bivariate per-course outcome, and defined one endpoint as their average over all courses of treatment. Two other simpler sets of scores as well as log survival time also were used as endpoints. Estimation of each DTR-specific mean score was conducted using inverse probability weighted methods that assumed that missingness in the twelve remaining drop-outs was informative but explainable in that it only depended on past recorded data. We conducted additional worst-best case analyses to evaluate sensitivity of our findings to extreme departures from the explainable drop-out assumption.

In oncology, progression-free survival time, which is defined as the minimum of the times to disease progression or death, often is used to characterize treatment and covariate effects. We are motivated by the desire to estimate the progression time distribution on the basis of data from 780 paediatric patients with choroid plexus tumours, which are a rare brain cancer where disease progression always precedes death. In retrospective data on 674 patients, the times to death or censoring were recorded but progression times were missing. In a prospective study of 106 patients, both times were recorded but there were only 20 non-censored progression times and 10 non-censored survival times. Consequently, estimating the progression time distribution is complicated by the problems that, for most of the patients, either the survival time is known but the progression time is not known, or the survival time is right censored and it is not known whether the patient's disease progressed before censoring. For data with these missingness structures, we formulate a family of Bayesian parametric likelihoods and present methods for estimating the progression time distribution. The underlying idea is that estimating the association between the time to progression and subsequent survival time from patients having complete data provides a basis for utilizing covariates and partial event time data of other patients to infer their missing progression times. We illustrate the methodology by analysing the brain tumour data, and we also present a simulation study.

We consider treatment regimes in which an agent is administered continuously at a specified concentration until either a response is achieved or a predetermined maximum infusion time is reached. Response is an event defined to characterize therapeutic efficacy. A portion of the maximum planned total amount administered is given as an initial bolus. For such regimes, the amount of the agent received by the patient depends on the time to response. An additional complication when response is evaluated periodically rather than continuously is that the response time is interval censored. We address the problem of designing a clinical trial in which such response time data and a binary indicator of toxicity are used together to jointly optimize the concentration and the size of the bolus. We propose a sequentially adaptive Bayesian design that chooses the optimal treatment for successive patients by maximizing the posterior mean utility of the joint efficacy-toxicity outcome. The methodology is illustrated by a trial in which tissue plasminogen activator is infused intraarterially as rapid treatment for acute ischemic stroke., (© 2011, The International Biometric Society.)

Anti-HLA donor-specific Abs (DSAs) have been reported to be associated with graft failure in mismatched hematopoietic stem cell transplantation; however, their role in the development of graft failure in matched unrelated donor (MUD) transplantation remains unclear. We hypothesize that DSAs against a mismatched HLA-DPB1 locus is associated with graft failure in this setting. The presence of anti-HLA Abs before transplantation was determined prospectively in 592 MUD transplantation recipients using mixed-screen beads in a solid-phase fluorescent assay. DSA identification was performed using single-Ag beads containing the corresponding donor's HLA-mismatched Ags. Anti-HLA Abs were detected in 116 patients (19.6%), including 20 patients (3.4%) with anti-DPB1 Abs. Overall, graft failure occurred in 19 of 592 patients (3.2%), including 16 of 584 (2.7%) patients without anti-HLA Abs compared with 3 of 8 (37.5%) patients with DSA (P = .0014). In multivariate analysis, DSAs were the only factor highly associated with graft failure (P = .0001; odds ratio = 21.3). Anti-HLA allosensitization was higher overall in women than in men (30.8% vs 12.1%; P < .0001) and higher in women with 1 (P = .008) and 2 or more pregnancies (P = .0003) than in men. We conclude that the presence of anti-DPB1 DSAs is associated with graft failure in MUD hematopoietic stem cell transplantation.