Adaptive treatment allocation for comparative clinical studies with recurrent events data

In long-term clinical studies, recurrent event data are sometimes collected and used to contrast the efficacies of two different treatments. The event reoccurrence rates can be compared using the popular negative binomial model, which incorporates information related to patient heterogeneity into a data analysis. For treatment allocation, a balanced approach in which equal sample sizes are obtained for both treatments is predominately adopted. However, if one treatment is superior, then it may be desirable to allocate fewer subjects to the less-effective treatment. To accommodate this objective, a sequential response-adaptive treatment allocation procedure is derived based on the doubly adaptive biased coin design. Our proposed treatment allocation schemes have been shown to be capable of reducing the number of subjects receiving the inferior treatment while simultaneously retaining a test power level that is comparable to that of a balanced design. The redesign of a clinical study illustrates the advantages of using our procedure.     

Subset Selection Procedures for Randomized Designs

Starting from the principle that there exists a randomization procedure that assigns treatments to experimental units, four subset selection rules for the problem of selecting the best treatment from a set of different treatments are proposed. Two of these are extensions of already existing subset selection procedures, which were defined for unbalanced designs, and need a separate selection constant for each individual treatment. The other two rules proposed are new and need only one selection constant for all treatments. The various procedures are compared, and illustrated by application to a plant breeding variety trial.     

A Two-stage Design for Comparing Clinical Trials

A two-stage design is proposed to choose among several experimental treatments and a standard treatment in clinical trials. The first stage employs a selection procedure to select the best treatment, provided it is better than the standard. The second stage tests the hypothesis between the best treatment selected at the first stage (if any) and the standard treatment. All the treatments are assumed to follow normal distributions and the best treatment is the one with the largest population mean. The level and the power are defined and they are used to set up equations to solve unknown first stage sample size, second stage sample size, and procedure parameters. The optimal design is the one that gives the smallest average sample size. Numerical results are presented to illustrate the improvement of one design as compared to existing one stage design.     

A Selection and Testing Procedure for Comparing Several Experimental Treatments with a Control

We consider a selection and testing procedure for comparing k experimental treatments with a control treatment where the treatments are assumed to be normally distributed with unknown means and a common, unknown variance. Stein‐type sampling is used in the selection phase to screen for an experimental treatment that exhibits evidence of being better than the control treatment and each of the other experimental treatments, where better is defined in terms of the largest mean. In the testing phase, the best experimental treatment is compared to the control using a hypothesis test. If no experimental treatment indicates that it is an improvement over the control during the selection phase, our procedure allows for early termination. We provide definitions of level and power appropriate for our hybrid procedure and compute procedure parameters required to implement our procedure.     

Mechanisms of protein degradation in growing and non-growing L-cell cultures.

L-cells prelabelled with [14C]leucine and [3H]thymidine were placed in either fresh growth medium (minimal essential medium with 10% serum) or stepdown medium (minimal essential medium) for 3 days. The 14C/3H ratio remained constant in the growing cultures and decreased in the stationary-phase cultures, indicating no protein turnover in growing cultures and a degradative rate of 0.6%/h in the stationary-phase cultures. Media analysis, however, indicated that 14C-labelled proteins were being degraded at approx. 1.2%/h in growing cultures and 1.7%/h in stationary-phase cultures. Additional studies indicated that a subpopulation of L-cells in the monolayer, comprising approx. 20--30% of the total, were lost in the original processing procedure. Experiments in which recoveries approached 100% by fixation of the monolayer in situ indicated that a protein-degrading subpopulation accounted for all the observed proteolysis in the growing cultures. Proteolysis in these cultures was only partially inhibited with NH4Cl, indicating that only a small part of the protein degradation was occurring in an activated lysosomal-autophagic system. NaF produced a more effective inhibition of proteolysis, but we were not able to distinguish whether this effect was on an ATP-requiring basal-turnover mechanism or a direct effect on unregulated activity of proteinases in the cell hyaloplasm. However, NH4Cl inhibited the proteolysis induced when cells were placed in stepdown medium, suggesting that the induced proteolysis was occurring via the autophagic system. We conclude that L-cells exist in at least two states with respect to protein degradation: (a) a subpopulation that is actively replicating and does not degrade cellular proteins, and (b) a second subpopulation of cells, derived from the preceding one, which degraded most of their labelled proteins, are not capable of further replication, and are not sedimented in an iso-osmotic EDTA buffer solution. In addition, proliferating L-cells, when placed in stepdown medium, begin to degrade cell protein through a mechanism involving autophagolysosomes.     

Differential shaping of EEG theta rhythms

Heart rate, EEG, frontal EMG, and forearm EMG were recorded in 20 subjects for 3 baseline, 8 feedback, and 2 postbaseline sessions in order to compare two biofeedback methods of teaching subjects to increase theta EEG activity. Subjects were divided into high- and low-EMG groups. Five high-EMG subjects, and 5 low-EMG subjects then received 8 sessions of strictly theta feedback. The remaining 10 subjects, 5 from the high-EMG group, and 5 from the low-EMG group, received a “graduated” training which involved shaping the target response. This procedure consisted of 4 initial sessions of EMG feedback, followed by a second phase consisting of 4 sessions of theta feedback. Results showed a clear relationship between subjects' baseline frontal EMG levels and the effect of the training methods. Although subjects with high-EMG baseline increased their theta output only with the two-phase training, subjects with low-EMG baseline levels performed better when given theta feedback only. This result shows not only that amounts of theta can be reliably increased, but that training techniques should be adapted to the physiological characteristics of the individual—in this case, baseline levels of frontal EMG levels.     

Responses of protein synthesis and degradation in growth control of WI-38 cells

The overall rates of protein synthesis and degradation in perfusion-grown WI-38 cells were followed in the three days after a stepdown in the serum concentration of the culture medium, from 10% to 0.3%. Within three hours after the stepdown, the rate of protein synthesis had decreased and the rate of protein degradation had increased, the combined result being the cessation of protein accumulation. The degradation rate returned over the next three days to its original value, but a zero rate of accumulation was retained because the synthesis rate continued to decline. The rate of DNA synthesis remained constant for six hours after the stepdown. It then declined steadily until reaching a minimum about eight hours later. The results show that extracellular control of protein accumulation depends on adjustments in both protein synthesis and protein degradation, and that the adjustments take place rapidly. This behavior suggests that the cell cycle is arrested after a stepdown because post-mitotic cells are unable to accumulate additional protein. However, an alternative interpretation of the data is that at least part of the changed accumulation is the result, rather than the cause, of the cycle arrest, and that the arrest is caused by other, more specific, reactions than those of general protein metabolism.     

Multiple testing to establish superiority/equivalence of a new treatment compared with k standard treatments for unbalanced designs

In clinical studies, multiple superiority/equivalence testing procedures can be applied to classify a new treatment as superior, equivalent (same therapeutic effect), or inferior to each set of standard treatments. Previous stepwise approaches (Dunnett and Tamhane, 1997, Statistics in Medicine16, 2489-2506; Kwong, 2001, Journal of Statistical Planning and Inference 97, 359-366) are only appropriate for balanced designs. Unfortunately, the construction of similar tests for unbalanced designs is far more complex, with two major difficulties: (i) the ordering of test statistics for superiority may not be the same as the ordering of test statistics for equivalence; and (ii) the correlation structure of the test statistics is not equi-correlated but product-correlated. In this article, we seek to develop a two-stage testing procedure for unbalanced designs, which are very popular in clinical experiments. This procedure is a combination of step-up and single-step testing procedures, while the familywise error rate is proved to be controlled at a designated level. Furthermore, a simulation study is conducted to compare the average powers of the proposed procedure to those of the single-step procedure. In addition, a clinical example is provided to illustrate the application of the new procedure.     

Performance limitations from delay in human and mechanical motor control

We discuss natural limitations on motor performance caused by the time delay required for feedback signals to propagate within the human body or mechanical control systems. By considering a very simple delayed linear servomechanism model, we show there exists a best possible speed-accuracy trade-off similar to Fitts' law that cannot be exceeded when delay is present. This is strictly a delay effect and does not occur for the ideal case of instantaneous feedback. We then examine the performance of the vector integration to endpoint (VITE) circuit as a model of human movement and show that when this circuit is generalized to include delayed feedback the performance may not exceed that of the servomechanism with an equal delay. We suggest the existence of such a limitation may be a ubiquitous consequence of delay in motor control with the implication that the index of performance in Fitts' law cannot arbitrarily large.     

Ordered multiple comparisons with the best and their applications to dose-response studies

This article considers the problem of comparing several treatments (dose levels, interventions, etc.) with the best, where the best treatment is unknown and the treatments are ordered in some sense. Order relations among treatments often occur quite naturally in practice. They may be ordered according to increasing risks, such as tolerability or safety problems with increasing dose levels in a dose-response study, for example. We tackle the problem of constructing a lower confidence bound for the smallest index of all treatments being at most marginally less effective than the (best) treatment having the largest effect. Such a bound ensures at confidence level 1 -alpha that all treatments with lower indices are relevantly less effective than the best competitor. We derive a multiple testing strategy that results in sharp confidence bounds. The proposed lower confidence bound is compared with those derived from other testing strategies. We further derive closed-form expressions for power and sample size calculations. Finally, we investigate several real data sets to illustrate various applications of our methods.     

Testing marginal homogeneity against stochastic order in multivariate ordinal data

Summary .  Many assessment instruments used in the evaluation of toxicity, safety, pain, or disease progression consider multiple ordinal endpoints to fully capture the presence and severity of treatment effects. Contingency tables underlying these correlated responses are often sparse and imbalanced, rendering asymptotic results unreliable or model fitting prohibitively complex without overly simplistic assumptions on the marginal and joint distribution. Instead of a modeling approach, we look at stochastic order and marginal inhomogeneity as an expression or manifestation of a treatment effect under much weaker assumptions. Often, endpoints are grouped together into physiological domains or by the body function they describe. We derive tests based on these subgroups, which might supplement or replace the individual endpoint analysis because they are more powerful. The permutation or bootstrap distribution is used throughout to obtain global, subgroup, and individual significance levels as they naturally incorporate the correlation among endpoints. We provide a theorem that establishes a connection between marginal homogeneity and the stronger exchangeability assumption under the permutation approach. Multiplicity adjustments for the individual endpoints are obtained via stepdown procedures, while subgroup significance levels are adjusted via the full closed testing procedure. The proposed methodology is illustrated using a collection of 25 correlated ordinal endpoints, grouped into six domains, to evaluate toxicity of a chemical compound.     

Ocular and stabilization feedback: an evaluation of two EMG biofeedback control procedures

This study evaluated the adequacy of two novel EMG biofeedback control procedures. During a single training session, 36 subjects received either contingent EMG feedback from the frontal region (Veridical), contingent feedback for vertical eye movements (Ocular), or a feedback condition where the signal increased with deviations in any direction from baseline EMG levels (Stabilization). The results supported the use of Ocular but not Stabilization feedback as a control procedure in frontalis EMG biofeedback studies. Ocular feedback did not produce reductions in frontalis EMG but did lead to changes in subjective measures of nonspecific treatment effects that were at least comparable to those obtained with Veridical feedback. Stabilization subjects produced small but significant reductions in EMG, felt the most bored as a result of their feedback training, and were the most likely to rate themselves as having received false feedback. The implications of attribution theory and multiprocess relaxation theory for the evaluation of nonspecific treatment effects are discussed.     

Rapid purification and characterization of homoserine dehydrogenase from Saccharomyces cerevisiae

Homoserine dehydrogenase of Saccharomyces cerevisiae has been rapidly purified to homogeneity by heat and acid treatments, ammonium sulfate fractionation, and chromatography on Matrex Gel Red A and Q-Sepharose columns. The final preparation migrated as a single entity upon sodium dodecyl sulfate-polyacrylamide gel electrophoresis with a Mr of 40,000. The Mr of the native enzyme was 81,000 as determined by gel filtration, suggesting that the enzyme is composed of two identical subunits. This feature was also confirmed by cross-linking analysis using the bifunctional reagent dimethyl suberimidate. Feedback inhibition by L-methionine and L-threonine was observed using the purified enzyme. The enzyme was markedly stabilized against heat treatment at high salt concentrations. Additions of feedback inhibitors or high concentrations of salts failed to cause any dissociation or aggregation of the enzyme subunits unlike enzymes from other sources such as Rhodospirillum rubrum. The enzyme denatured in 3 M guanidine-HCl was refolded by simple dilution with a concomitant restoration of the activity. Cross-linking analysis of the renaturation process suggested that the formation of the dimer is required for activity expression. Amino acid sequence analysis of peptides obtained by digestion of the enzyme protein with Achromobacter lyticus protease I revealed that several amino acid residues are strictly conserved among homoserine dehydrogenases from S. cerevisiae, Escherichia coli, and Bacillus subtilis.     

Reactive gait and postural adjustments following the first exposures to (un)expected stepdown

This study evaluated the reactive biomechanical strategies associated with both upper- and lower-body (lead and trail limbs) following the first exposures to (un)expected stepdown at comfortable (1.22 ± 0.08 m/s) and fast (1.71 ± 0.11 m/s) walking velocities. Eleven healthy adults completed 34 trails per walking velocity over an 8-m, custom-built track with two forceplates embedded in its center. For the expected stepdown, the track was lowered by 0-, −10- and −20-cm from the site of the second forceplate, whereas the unexpected stepdown was created by camouflaging the second forceplate (−10-cm). Two-way repeated-measurement ANOVAs detected no velocity-related effects of stepdown on kinematic and kinetic parameters during lead limb stance-phase, and on the trail limb stepping kinematics. However, analyses of significant interactions revealed greater peak flexion angles across the trunk and the trail limb joints (hip, knee and ankle) in unexpected versus expected stepdown conditions at a faster walking velocity. The −10-cm unexpected stepdown (main effect) had a greater influence on locomotor behavior compared to expected conditions due mainly to the absence of predictive adjustments, reflected by a significant decrease in peak knee flexion, contact time and vertical impulse during stance-phase. Walking faster (main effect) was associated with an increase in hip peak flexion and net anteroposterior impulse, and a decrease in contact time and vertical impulse during stepdown. The trail limb, in response, swung forward faster, generating a larger and faster recovery step. However, such reactive stepping following unexpected stepdown was yet a sparse compensation for an unstable body configuration, assessed by significantly smaller step width and anteroposterior margin-of-stability at foot-contact in the first-recovery-step compared with expected conditions. These findings depict the impact of the expectedness of stepdown onset on modulation of global dynamic postural control for a successful accommodation of (un)expected surface elevation changes in young, healthy adults.     

Two-Stage Procedures for Comparing Treatments with a Control: Elimination at the First Stage and Estimation at the Second Stage

We consider the problem of comparing a set of p₁ test treatments with a control treatment. This is to be accomplished in two stages as follows: In the first stage, N₁ observations are allocated among the p₁ treatments and the control, and the subset selection procedure of Gupta and Sobel (1958) is employed to eliminate “inferior” treatments. In the second stage, N₂ observations are allocated among the (randomly) selected subset of p₂(≤p₁) treatments and the control, and joint confidence interval estimates of the treatment versus control differences are calculated using Dunnett's (1955) procedure. Here both N₁ and N₂ are assumed to be fixed in advance, and the so-called square root rule is used to allocate observations among the treatments and the control in each stage. Dunnett's procedure is applied using two different types of estimates of the treatment versus control mean differences: The unpooled estimates are based on only the data obtained in the second stage, while the pooled estimates are based on the data obtained in both stages. The procedure based on unpooled estimates uses the critical point from a p₂-variate Student t-distribution, while that based on pooled estimates uses the critical point from a p₁-variate Student t-distribution. The two procedures and a composite of the two are compared via Monte Carlo simulation. It is shown that the expected value of p₂ determines which procedure yields shorter confidence intervals on the average. Extensions of the procedures to the case of unequal sample sizes are given. Applicability of the proposed two-stage procedures to a drug screening problem is discussed.     

An Integrated Formulation for Comparing Treatments with Binary Response with a Control

The problem of comparing k(≧2) bernoulli rates of success with a control is considered. An one-stage decision procedure is proposed for either (1) choosing the best among several experimental treatments and the control treatment when the best is significantly superior or (2) selecting a random size subset that contains the best experimental treatment if it is better than the control when the difference between the best and the remaining treatments is not significant. We integrate two traditional formulations, namely, the indifference (IZ) approach and the subset selection (SS) approach, by seperating the parameter space into two disjoint sets, the preference zone (PZ) and the indifference zone (IZ). In the PZ we insist on selecting the best experimental treatment for a correct selection (CS₁) but in the IZ we define any selected subset to be correct (CS₂) if it contains the best experimental treatment which is also better than the control. We propose a procedure R to guarantee lower bounds P₁* for P(CS₁≧PZ) and P₂* for P(CS₂≧IZ) simultaneously. A brief table on the common sample size and the procedure parameters is presented to illustrate the procedure R.     

High-throughput detection of multiple genetic polymorphisms influencing drug metabolism with mismatch primers in allele-specific polymerase chain reaction

Because of genetic polymorphisms of drug-metabolizing enzyme genes, the activities of the enzymes in humans vary widely and alter the metabolism of commonly used clinical agents. Severe adverse effects or resistance to therapy may result. We have developed a rapid and high-throughput genotyping method for detecting polymorphisms of the drug-metabolizing enzyme genes CYP2C9*3, CYP2C19*2, *3, CYP2D6*2, *4, *10, *14, *21, NAT2*5, *6, *7, and TPMT*3 using allele-specific polymerase chain reaction (PCR) with mismatch primers (ASPCR-MP) and CYP2D6*5, *36, and CYP2D6xN using stepdown PCR with detection by SYBR Green I. We analyzed genomic DNA from 139 Japanese volunteers. Identical genotyping results were obtained by using ASPCR-MP, stepdown PCR, and conventional PCR. We found that the methods clearly differentiate three specific profiles with no overlap in the signals. Moreover, both ASPCR-MP and stepdown PCR for genotyping took less than 3-4h. To our knowledge, this is the first report of successful simultaneous detection of multiple genetic polymorphisms with point mutations using ASPCR-MP or multiple genetic polymorphisms with large structural alterations using stepdown PCR. In conclusion, ASPCR-MP and stepdown PCR appear to be suitable for large clinical and epidemiological studies as methods that enable highly sensitive genotyping and yield a high-throughput.     

A comparison of two behavioral treatments in decreasing the orofacial movement of tardive dyskinesia

In a study with an elderly female subject, two behavioral treatments were evaluated in terms of their effectiveness in decreasing orofacial movement associated with tardive dyskinesia. Video feedback and discreet-discrete prompting, a self-control procedure using a portable audio signal generator, were compared by means of an alternating treatments experimental design. Video and instructional controls were included in the study. Results indicated that both procedures were effective in decreasing orofacial movement. In addition, during the concluding phase of the study, a prompting card was carried by the subject at all times as a reminder to control mouth movements on an ongoing basis. This concluding phase resulted in generalization of treatment effects to the nontreatment environment. Follow-up sessions indicated maintenance of treatment effects.     

Ocular and stabilization feedback: An evaluation of two EMG biofeedback control procedures

This study evaluated the adequacy of two novel EMG biofeedback control procedures. During a single training session, 36 subjects received either (1) contingent EMG feedback from the frontal region (Veridical), (2) contingent feedback for vertical eye movements (Ocular), or (3) a feedback condition where the signal increased with deviations in any direction from baseline EMG levels (Stabilization). The results supported the use of Ocular but not Stabilization feedback as a control procedure in frontalis EMG biofeedback studies. Ocular feedback did not produce reductions in frontalis EMG but did lead to changes in subjective measures of nonspecific treatment effects that were at least comparable to those obtained with Veridical feedback. Stabilization subjects produced small but significant reductions in EMG, felt the most bored as a result of their feedback training, and were the most likely to rate themselves as having received false feedback. The implications of attribution theory and multiprocess relaxation theory for the evaluation of nonspecific treatment effects are discussed.This research was supported in part by grants from the National Institutes of Health (AM31500) and the Robert Wood Johnson Foundation. Portions of this research were presented at the Sixth Annual Meeting of the Society of Behavioral Medicine, New Orleans, March 1985.     

Effect of one-step sucrose dilution of glycerol on in-vitro development of frozen-thawed mouse embyros

Several glycerol (GLY) dilution treatments were compared using frozen-thawed early blastocysts from Swiss Webster mice. These treatments consisted of 0.00 (0.00S n = 68), 0.25 (0.25S n = 67), 0.50 (0.50S n = 76), 0.75 (0.75S n = 66), 1.00 (1.00S n = 59), and 1.25 (1.25S n = 60) M of sucrose to remove GLY from embryos in one step. Then the one step procedure was compared with a three-step GLY dilution treatment (C n = 66). Embryos were exposed to 1.5 M of GLY in three-steps, frozen according to a standard freezing technique and stored at -196 degrees C. Embryos were thawed in a 37 degrees C water bath, pooled, and those graded good or excellent were randomly assigned to the experimental groups. The blastocysts were cultured in Whitten's medium microdrops under paraffin oil in a water saturated 5% CO(2) air atmosphere at 37 degrees C. The proportion (%) of embryos developing to expanded blastocysts was lowest (P < 0.05) in treatment 0.00S (63.1 +/- 4.0). The 0.25S (72.0 +/- 4.3) and 0.50S (75.0 +/- 3.1) 0.75S (79.0 +/- 4.4) treatments yielded a similar percentage of expanded blastocysts. The 1.00S treatment (87.0 +/- 4.0) was similar to 0.75S and 1.25S (98.3 +/- 4.0) treatments. The C treatment was superior (P < 0.05) to dilutions done with < 0.75 M sucrose, similar to 0.75S and 1.00S, and inferior (P < 0.05) to 1.25S. This later treatment yielded the highest percentage of expanded blastocysts. The percentage of embryos that hatched in treatments 0.00S, 0.25S, 0.50S, 0.75S and C was lower (P < 0.05) compared to 1.00S and 1.25S. The percentage of embryos and hatched blastocysts increased linearly (P < 0.01) from 0.0 to 1.25 M sucrose. Dilution of GLY with 1 or 1.25 M sucrose yielded better results compared with lower sucrose concentrations or the three-step GLY removal procedure.