Biomedical subjects
Publications and source records attributed to Anastasia Ivanova.
In this paper, rules are given on how to construct escalation, A + B, and group designs for dose-finding trials. Operational characteristics of these designs are discussed and compared via simulations. Dose-finding designs for trials with ordinal toxicity outcome are considered.
BACKGROUND: The North Carolina-Louisiana Prostate Cancer Project (PCaP) is a multidisciplinary study of social, individual, and tumor-level causes of racial differences in prostate cancer aggressiveness. METHODS: A population-based sample of incident prostate cancer cases from North Carolina and Louisiana will include 1,000 African Americans and 1,000 Caucasian Americans. Study nurses administer structured questionnaires and collect blood, adipose tissue, urine, and toenail samples during an in-home visit. Clinical data are abstracted from medical records, diagnostic biopsies are reviewed and assayed, and tissue microarrays are constructed from prostatectomy samples. Prostate cancer aggressiveness is classified based on PSA, clinical stage, and Gleason grade. RESULTS: Preliminary data demonstrate between- and within-group differences in patient characteristics, screening, and treatment by race and state. Participation exceeds 70% in all groups. CONCLUSIONS: Preliminary data support the feasibility of this comprehensive study to help determine the focus of public health efforts to reduce racial disparities in prostate cancer mortality.
We consider the problem of dose-finding where subjects can be stratified into two populations with possibly different susceptibility to toxicity. The goal is to find the maximally tolerated dose for each population. We propose a non-parametric design for this problem. Toxicity is estimated using the bivariate isotonic regression estimator. The new design is compared with the two-sample continual reassessment method.
OBJECTIVE: To compare duration of mechanical ventilation for patients randomized to receive lorazepam by intermittent bolus administration vs. continuous infusions of propofol using protocols that include scheduled daily interruption of sedation. DESIGN: A randomized open-label trial enrolling patients from October 2001 to March 2004. SETTING: Medical intensive care units of two tertiary care medical centers. PATIENTS: Adult patients expected to require mechanical ventilation for >48 hrs and who required > or =10 mg of lorazepam or a continuous infusion of a sedative to achieve adequate sedation. INTERVENTIONS: Patients were randomized to receive lorazepam by intermittent bolus administration or propofol by continuous infusion to maintain a Ramsay score of 2-3. Sedation was interrupted on a daily basis for both groups. MEASUREMENTS AND MAIN RESULTS: The primary outcome was median ventilator days. Secondary outcomes included 28-day ventilator-free survival, intensive care unit and hospital length of stay, and hospital mortality. Median ventilator days were significantly lower in the daily interruption propofol group compared with the intermittent bolus lorazepam group (5.8 vs. 8.4, p = .04). The difference was largest for hospital survivors (4.4 vs. 9.0, p = .006). There was a trend toward greater ventilator-free survival for patients in the daily interruption propofol group (median 18.5 days for propofol vs. 10.2 for lorazepam, p = .06). Hospital mortality was not different. CONCLUSIONS: For medical patients requiring >48 hrs of mechanical ventilation, sedation with propofol results in significantly fewer ventilator days compared with intermittent lorazepam when sedatives are interrupted daily.
BACKGROUND: Adjuvant trastuzumab improves relapse-free survival in HER2-overexpressing breast cancer but is associated with cardiac toxicity. This phase II study was undertaken to determine the neoadjuvant clinical and pathologic response rate and the acute and chronic cardiac toxicity of trastuzumab given with weekly paclitaxel after AC (doxorubicin/cyclophosphamide). PATIENTS AND METHODS: Fifty-two women with newly diagnosed, stage II-IV, HER2-overexpressing breast cancer received AC for 4 cycles, followed by weekly TP (paclitaxel/trastuzumab) for 12 weeks, neoadjuvantly or adjuvantly, followed by 40 weeks of adjuvant trastuzumab. RESULTS: Congestive heart failure occurred in 4% of patients (95% confidence interval [CI], 0.5%-13.2%). Asymptomatic left ventricular ejection fraction (LVEF) decreases to < 50% occurred in 21% of patients (95% CI, 11.1%-34.7%); all but 1 recovered by 1.5 years. Median LVEF decreased progressively during therapy from 65% before therapy (95% CI, 63%-66%) to 62% after AC (95% CI, 59%-64%) and 58% after AC-TP (95% CI, 56%-64%; P < 0.01 for each decrease). The decrease in LVEF persisted 1.5 years after study entry at 57% (95% CI, 54%-60%), although all but 1 of the most severe decreases to < 50% recovered to normal. Clinical response rate among 37 patients treated neoadjuvantly was 86%, and the pathologic complete response rate was 19% (95% CI, 8%-35.2%). Because of withdrawals for toxicity, refractory disease, and patient preference, only 35% of patients completed the entire regimen. CONCLUSION: In this study, the AC-TP regimen resulted in a high clinical but moderate pathologic response rate, and although asymptomatic cardiac systolic dysfunction was common, most of the severe decreases recovered over time.
To assess the effects of physician-patient racial concordance and continuity of care on hypertension outcomes, we described patterns of care for hypertension; we used cross-tabulations and repeated measures (generalized estimating equations) analyses with panel survey data from elderly persons interviewed and examined in 1987 and 1990. Continuity of care was associated with recognition of hypertension, receipt of medication, and lower incidence of undetected hypertension. Physician race had little effect, but continuity is important for successful management of hypertension in older persons.
A fully sequential procedure is proposed for comparing K > or =3 treatments with immediate binary responses. The procedure uses an adaptive urn design to randomize patients to the treatments and stopping rules are incorporated for eliminating less promising treatments. Simulation is used to assess the performance of the procedure for several adaptive urn designs, in terms of expected numbers of treatment failures and allocation proportions, and the effect on estimation at the end of the trial is also addressed. It is concluded that the drop-the-loser rule is more effective than equal allocation and all of the other designs considered. The practical benefits of the procedure are illustrated using the results of a three-treatment lung cancer study. It is then shown how the sequential elimination procedure may be used in dose-finding studies and its performance is compared with a recently proposed method. Several possible extensions to the work are briefly indicated.
A clinical trial is considered in which two treatments with binary responses are to be compared. A popular sequential stopping rule, the triangular test, is studied when various response-adaptive treatment allocation rules are applied, such as the recently proposed drop-the-loser rule, an urn randomization scheme. The paper extends previous work by Coad and Rosenberger, who combined the triangular test with the randomized play-the-winner rule. The purpose of the paper is to investigate to what extent the variability of an adaptive design affects the overall performance of the triangular test. The adaptive rules under consideration are described and some of their asymptotic properties are summarized. Simulation is then used to assess the performance of the triangular test when combined with the various adaptive rules. The main finding is that the drop-the-loser rule is the most promising of the adaptive rules considered in terms of a less variable allocation proportion and a smaller number of treatment failures. The use of this rule with the triangular test is beneficial compared with the triangular test with equal allocation, since it yields fewer treatment failures on average while providing comparable power with similar expected sample size. The results of an AIDS trial are used to illustrate the performance of the triangular test when combined with the drop-the-loser rule.
In this paper, we propose a model-based approach to detect and adjust for observable selection bias in a randomized clinical trial with two treatments and binary outcomes. The proposed method was evaluated using simulations of a randomized block design in which the investigator favoured the experimental treatment by attempting to enroll stronger patients (with greater probability of treatment success) if the probability of the next treatment being experimental was high, and enroll weak patients (with less probability of treatment success) if the probability of the next treatment being experimental was low. The method allows not only testing for the presence of observable selection bias, but also testing for a difference in treatment effects, adjusting for possible selection bias.
The objective of a Phase I trial with two agents is to find a set of maximum-tolerated dose combinations that yield a prespecified toxicity rate. In this article, we consider the case where several doses of one agent are fixed and the goal is to find the maximum-tolerated dose of the other agent to be used in combination with each of the doses of agent one. We propose a Bayesian design that uses a parsimonious working model for the dose-toxicity relationship. We show that the new design is more effective in identifying the maximum-tolerated combinations than one-dimensional designs applied at each dose level of one of the agents.
The goal of a phase II trial in oncology is to evaluate the efficacy of a new therapy. The dose investigated in a phase II trial is usually an estimate of a maximum-tolerated dose obtained in a preceding phase I trial. Because this estimate is imprecise, stopping rules for toxicity are used in many phase II trials. We give recommendations on how to construct stopping rules to monitor toxicity continuously. A table is provided from which Pocock stopping boundaries can be easily obtained for a range of toxicity rates and sample sizes. Estimation of the probability of toxicity and response is also discussed.
Proteasome inhibitors, a novel class of chemotherapeutic agents, enhance the antitumor efficacy of anthracyclines in vitro and in vivo. We therefore sought to determine the maximum tolerated dose (MTD) and dose-limiting toxicities of bortezomib and pegylated liposomal doxorubicin (PegLD). Bortezomib was given on days 1, 4, 8, and 11 from 0.90 to 1.50 mg/m2 and PegLD on day 4 at 30 mg/m2 to 42 patients with advanced hematologic malignancies. Grade 3 or 4 toxicities in at least 10% of patients included thrombocytopenia, lymphopenia, neutropenia, fatigue, pneumonia, peripheral neuropathy, febrile neutropenia, and diarrhea. The MTD based on cycle 1 was 1.50 and 30 mg/m2 of bortezomib and PegLD, respectively. However, due to frequent dose reductions and delays at this level, 1.30 and 30 mg/m2 are recommended for further study. Pharmacokinetic and pharmacodynamic studies did not find significant drug interactions between these agents. Antitumor activity was seen against multiple myeloma, with 8 of 22 evaluable patients having a complete response (CR) or near-CR, including several with anthracycline-refractory disease, and another 8 having partial responses (PRs). One patient with relapsed/refractory T-cell non-Hodgkin lymphoma (NHL) achieved a CR, whereas 2 patients each with acute myeloid leukemia and B-cell NHL had PRs. Bortezomib/PegLD was safely administered in this study with promising antitumor activity, supporting further testing of this regimen.
This paper investigates the design and analysis of dose-finding trials with two agents. The set of doses for each agent is fixed in advance. The goal of the trial is to find the set of dose combinations with probability of toxicity closest to a pre-specified value. For each of the two agents we assume that the probability of toxicity of an agent is non-decreasing with dose when the dose of the other agent is fixed. Using this assumption we construct a new non-parametric design that operates on a two-dimensional grid of all possible dose combinations. A bivariate isotonic regression estimator of the maximum tolerated combinations is described. We conclude that the new design and the bivariate isotonic estimator are superior to the procedure where several independent dose-finding trials are run.
PURPOSE: A biomarker of oxidant status applicable to epidemiological research is essential to studying the relationship between free radicals and chronic disease risk. Gas chromatography with mass-spectrometry detection (GC/MS) is the gold standard for measurement of urinary F2-isoprostanes (F2-isoPs), a non-invasive marker of oxidant status. However, this method is laborious and costly, which prohibits its use in large epidemiological studies. METHODS: We compared GC/MS assay with an inexpensive quick enzyme-linked immunoassay (ELISA) in measurements of 2,3-dinor-5,6-dihydro-15-F2t-isoprostane (F2-isoPM), an abundant beta-oxidation metabolite of 8-iso-prostaglandin-F2alpha. We measured F2-isoPM in urine of 52 participants of the Insulin Resistance Atherosclerosis Study by both methods. RESULTS: The ELISA measurements showed approximately 30-fold greater mean and median (22.10, SD 12.92, and 18.49 ng/mg creatinine) than the GC/MS measurements (0.703, SD 0.468, and 0.597 ng/mg creatinine). We found low linear correlation (Pearson correlation coefficient 0.51; 95% CI, 0.28-0.70) and weak agreement in ranking subjects by tertiles (weighted Kappa statistic 0.34) between a GC/MS and ELISA. CONCLUSIONS: We conclude that the current ELISA method is not a valid substitute for the GS/MS assay.
BACKGROUND: A prospective, randomized study was performed to compare the effects of early versus late enteral feeding on postburn metabolism. METHODS: Burn patients were randomized to receive enteral feedings either within 24 hours (early) or 7 days (late) of injury. Basal energy expenditure (BEE) was calculated from Harris-Benedict equations and resting energy expenditure (REE) was obtained from indirect calorimetry. The average daily energy expenditure (DEE) was expressed as REE/BEE. RESULTS: Average age, burn size, infections, and length of stay were similar between groups. Mortality between groups was similar (early, 28%; late, 38%) and not significantly influenced by inhalation injury. When controlled for percentage of total body surface area burn, inhalation injury, and age, the early group had an increased rather than decreased DEE, with a mean DEE calorie 0.17 more than the late group (p = 0.07). CONCLUSION: Early enteral feeding does not decrease the average energy expenditure associated with burn injury.
The interpretation of between-group comparisons is facilitated by the creation of treatment groups that are similar to each other in baseline composition. To prevent treatment effects from being confounded with time effects, most trials use restricted randomization to force balance. An unintended consequence of these restrictions is that they create patterns that allow for the prediction of future treatment allocations, and hence selection bias, especially in unmasked trials. In fact, the more restrictive the allocation procedure, the greater the potential for selection bias. It was decided, in the context of a recent clinical trial comparing two dosing schedules of paclitaxel and carboplatin for advanced stage IIIB/IV non-small-cell lung cancer, that the randomized block procedure could not simultaneously protect sufficiently against both selection and chronological bias. In this paper we detail our development of the maximal procedure. The maximal procedure takes as input the extent of chronological bias allowed by the randomized block procedure, then matches it, but does so with fewer restrictions. This feature makes the maximal procedure more resistant to selection bias than the randomized block procedure is.
We consider several designs from the family of up-and-down rules for the sequential allocation of dose levels to subjects in a dose-response study. We show that an up-and-down design can be improved by using more information than the most recent response. For example, the k-in-a-row rule uses up to the k most recent responses. We introduce a new design, the Narayana rule, which uses a local estimate of the probability of toxicity calculated from all previous responses. For the Narayana rule, as the sample size gets large, the probability of assignment goes to zero for dose levels not among the two (or three) closest to the target. Different estimators of the target dose are compared. We find that the isotonic regression estimator is superior to other estimators for small to moderate sample sizes.
For some drugs, toxicity events lead to early termination of treatment before a therapeutic response is observed. That is, there are three possible outcomes: toxicity (therapeutic response unknown), therapeutic response without toxicity, and no response with no toxicity. The optimal dose is the dose that maximizes the probability of the joint event, response, and no toxicity. The optimal safe dose is the dose, from among the doses with toxicity rate less than the maximum tolerable level, that maximizes the probability of response and no toxicity. We present a new sequential design to maximize the number of subjects assigned in the neighborhood of the optimal safe dose in a dose-finding trial with two outcomes.