Sample size calculation through the incorporation of heteroscedasticity and dependence for a penalized t-statistic in microarray experiments.

When identifying the differentially expressed genes (DEGs) in microarray data, we often observe heteroscedasticity between groups and dependence among genes. Incorporating these factors is necessary for sample size calculation in microarray experiments. A penalized t-statistic is widely used to improve the identifiability of DEGs. We develop a formula to calculate sample size with dependence adjustment for the penalized t-statistic. Sample size is determined on the basis of overall power under certain conditions to maintain a certain false discovery rate. The usefulness of the proposed method is demonstrated by numerical studies using both simulated data and real data.

Validation study of the in vitro skin irritation test with the LabCyte EPI-MODEL24.

A validation study on an in vitro skin irritation assay was performed with the reconstructed human epidermis (RhE) LabCyte EPI-MODEL24, developed by Japan Tissue Engineering Co. Ltd (Gamagori, Japan). The protocol that was followed in the current study was an optimised version of the EpiSkin protocol (LabCyte assay). According to the United Nations Globally Harmonised System (UN GHS) of classification for assessing the skin irritation potential of a chemical, 12 irritants and 13 non-irritants were validated by a minimum of six laboratories from the Japanese Society for Alternatives to Animal Experiments (JSAAE) skin irritation assay validation study management team (VMT). The 25 chemicals were listed in the European Centre for the Validation of Alternative Methods (ECVAM) performance standards. The reconstructed tissues were exposed to the chemicals for 15 minutes and incubated for 42 hours in fresh culture medium. Subsequently, the level of interleukin-1 alpha (IL-1 α) present in the conditioned medium was measured, and tissue viability was assessed by using the MTT assay. The results of the MTT assay obtained with the LabCyte EPI-MODEL24 (LabCyte MTT assay) demonstrated high within-laboratory and between-laboratory reproducibility, as well as high accuracy for use as a stand-alone assay to distinguish skin irritants from non-irritants. In addition, the IL-1α release measurements in the LabCyte assay were clearly unnecessary for the success of this model in the classification of chemicals for skin irritation potential.

Inter-laboratory validation of the modified murine local lymph node assay based on 5-bromo-2'-deoxyuridine incorporation.

The murine local lymph node assay (LLNA) is a well-established alternative to the guinea pig maximization test (GPMT) or Buehler test (BT) for the assessment of the skin sensitizing ability of a drug, cosmetic material, pesticide or industrial chemical. Instead of radioisotope using in this method, Takeyoshi M. et al. (2001) has developed a modified LLNA based on the 5-bromo-2'-deoxyuridine (BrdU) incorporation (LLNA:BrdU-ELISA). The LLNA:BrdU-ELISA is practically identical to the LLNA methodology excluding the use of BrdU, for which a single intraperitoneal injection of BrdU is made on day 4, and colorimetric detection of cell turnover. We conducted the validation study to evaluate the reliability and relevance of LLNA:BrdU-ELISA. The experiment involved 7 laboratories, wherein 10 chemicals were examined under blinded conditions. In this study, 3 chemicals were examined in all laboratories and the remaining 7 were examined in 3 laboratories. The data were expressed as the BrdU incorporation using an ELISA method for each group, and the stimulation index (SI) for each chemical-treated group was determined as the increase in the BrdU incorporation relative to the concurrent vehicle control group. An SI of 2 was set as the cut-off value for exhibiting skin sensitization activity. The results obtained in the experiments conducted for all 10 chemicals were sufficiently consistent with small variations in their SI values. The sensitivity, specificity, and accuracy of LLNA:BrdU-ELISA against those of GPMT/BT were 7/7 (100%), 3/3 (100%), and 10/10 (100%), respectively.

Statistical screening method for genetic factors influencing susceptibility to common diseases in a two-stage genome-wide association study.

A genome-wide association study (GWAS) is a standard strategy for detecting disease susceptibility genes, despite unsettled controversies on many aspects, including optimal study design and statistical analysis. As for study design, a two-stage design has been applied to maximize cost-effectiveness. However, there has been little consensus on appropriate statistical analysis for two-stage design. Thereby perplexing the researchers as to which statistical measures should be applied at the first stage, and how to determine the significance level of the differences at the second stage. Here, using simulation studies, we compared statistical operating characteristics of the screening in a two-stage GWAS by taking into consideration the proper balance of false-positive and false-negative error. As a result, the lower bound of confidence interval for odds ratios is recommended as the first stage measure, and then the second stage criteria should primarily depend on the purpose of the genome screen or its role in the overall gene-hunting scheme. Based on the simulation study, we suggest rules of thumb about which statistics to use in a given situation. An application of all operating characteristics of the screening method to an actual GWAS for gastric cancer illustrates the practical relevance of our discussion.

Application of the improved BALB/c 3T3 cell transformation assay to the examination of the initiating and promoting activities of chemicals: the second interlaboratory collaborative study by the non-genotoxic carcinogen study group of Japan.

The Non-genotoxic Carcinogen Study Group in the Environmental Mutagen Society of Japan organised the second step of the inter-laboratory collaborative study on one-stage and two-stage cell transformation assays employing BALB/c 3T3 cells, with the objective of confirming whether the respective laboratories could independently produce results relevant to initiation or promotion. The method was modified to use a medium consisting of DMEM/F12 supplemented with 2% fetal bovine serum and a mixture of insulin, transferrin, ethanolamine and sodium selenite, at the stationary phase of cell growth. Seventeen laboratories collaborated in this study, and each chemical was tested by three to five laboratories. Comparison between the one-stage and two-stage assays revealed that the latter method would be beneficial in the screening of chemicals. In the test for initiating activity with the two-stage assay (post-treated with 0.1microg/ml 12-O-tetradecanoylphorbol-13-acetate), the relevant test laboratories all obtained positive results for benzo[a]pyrene and methylmethane sulphonate, and negative results for phenanthrene. Of those laboratories assigned phenacetin for the initiation phase, two returned positive results and two returned negative results, where the latter laboratories tested up to one dose lower than the maximum dose used by the former laboratories. In the exploration of promoting activity with the twostage assay (pretreated with 0.2microg/ml 3-methylcholanthrene), the relevant test laboratories obtained positive results for mezerein, sodium orthovanadate and TGF-beta1, and negative results for anthralin, phenacetin and phorbol. Two results returned for phorbol 12,13-didecanoate were positive, but one result was negative - again, the maximum dose to achieve the latter result was lower than that which produced the former results. These results suggest that this modified assay method is relevant, reproducible and transferable, provided that dosing issues, such as the determination of the maximum dose, are adequately considered. The application of this two-stage assay for screening the initiating and promoting potential of chemicals is recommended for consideration by other research groups and regulatory authorities.

Optimal two-stage designs allowing flexibility in number of subjects for phase II clinical trials.

Phase II clinical trials are conducted to test whether a drug has a minimum desired effect and to assess whether further development of the drug is warranted. They are often designed as one-arm trials with response rate as the primary endpoint, and a two-stage design is often used to ensure early termination of the trial for futility. To control the type I error rate and guarantee the specified power of the study, planned sample sizes for both stages must be rigidly followed, but a literature review suggests that actual sample size often differs from that planned. We propose to extend simple two-stage designs to allow more flexible sampling plans in both stages. Our designs are preferable to similar extensions proposed to control type I and II error probabilities. Additionally, our assumptions regarding distribution of the actual sample size at the end of stage 1 are more lenient. A list of optimal designs for typical error rates and the selected null and alternative response rates is presented.

A note on estimating treatment effect for time-to-event data in a literature-based meta-analysis.

OBJECTIVES: In a literature-based meta-analysis for time-to-event data, the hazard ratio in each trial is often estimated from the summary statistics described in the article. Several methods have been proposed: the direct method (Peto method); the indirect method using a p-value by the log-rank test and the number of total events; and the survival curve method using the Kaplan-Meier estimate. However, there has been no published report on a detailed investigation of these methods. We evaluated the performance of these methods by simulation. METHODS: In a set of simulation experiments, performance of five methods was evaluated by the bias of estimated log hazard ratio and coverage probability of the confidence interval. The methods evaluated were: 1) Cox regression analysis, 2) direct method, 3) indirect method, 4) survival curve method, and 5) modified survival curve method with an alternative weighting scheme. RESULTS: The direct method was confirmed to have a high degree of accuracy. Although the indirect method was also highly accurate, it tended to underestimate effect size when there was a strong effect. The survival curve method tended to underestimate effect size when event numbers were small and effect size was large. The modified survival curve method could alleviate this tendency toward underestimation of effect size found with the original survival curve method. CONCLUSIONS: When the Kaplan-Meier curve is used to estimate hazard ratios in trials with small sample size in the literature-based meta-analysis, we should check critically whether those trials' hazard ratios and overall hazard ratio are underestimated or not.

Technical modification of the Balb/c 3T3 cell transformation assay: the use of serum-reduced medium to optimise the practicability of the protocol.

The two-stage Balb/c 3T3 model of cell transformation can mimic the two-stage carcinogenicity bioassay, and has been recognised as a screening method for detecting potential tumour initiators and promoters. A technical modification to the original protocol (which involved the use of M10F medium, consisting of MEM plus 10% fetal bovine serum [FBS]) has been previously proposed, in order to increase its efficacy, namely: the introduction of enriched, serum-reduced medium (DF2F medium, comprising DMEM/F12 plus 2% FBS and other supplements). The aim of this study was to further modify the protocol, so as to attain higher practicability for the assay. The protocol was further optimised by: a) reducing the number of plates required, through the use of larger plates; b) reducing the cost of the assay by retaining the reduced serum concentration and by using 2microg/ml insulin, rather than the more-complex insulin-transferrin-ethanolamine-sodium selenite (ITES) supplement (i.e. DF2F2I medium); and c) extending the culture period from 24-25 days to 31-32 days, resulting in clearer foci (the number of medium changes did not increase, as less-frequent medium changes were performed during the extended culture period). Growth curve construction revealed that variations in the saturation densities of the parental Balb/c 3T3 cell line and its three transformed clones were highest when M10F medium was replaced with DF2F2I medium just before cells reached confluence. We applied this newly-optimised protocol to the assessment of: a) the tumour initiating activity of 3-methylcholanthrene (MCA), N-methyl-N'-nitro-N-nitrosoguanidine, mitomycin C, methylmethane sulphonate, CdCl(2) and phenacetin, combining a post-treatment of 100ng/ml 12-O-tetradecanoylphorbol-13-acetate at the promotion stage; and b) the tumour promoting activity of insulin, lithocholic acid, CdCl(2) and phenobarbital, with pre-treatment of 0.2microg/ml MCA at the initiation stage. In the present study, only phenobarbital was negative when tested by using the modified protocol.

Criteria revision and performance comparison of three methods of signal detection applied to the spontaneous reporting database of a pharmaceutical manufacturer.

BACKGROUND AND OBJECTIVE: Several statistical methods exist for detecting signals of potential adverse drug reactions in spontaneous reporting databases. However, these signal-detection methods were developed using regulatory databases, which contain a far larger number of adverse event reports than the databases maintained by individual pharmaceutical manufacturers. Furthermore, the composition and quality of the spontaneous reporting databases differ between regulatory agencies and pharmaceutical companies. Thus, the signal-detection criteria proposed for regulatory use are considered to be inappropriate for pharmaceutical industry use without modification. The objective of this study was to revise the criteria for signal detection to make them suitable for use by pharmaceutical manufacturers. METHODS: A model comprising 40 drugs and 1000 adverse events was constructed based on a spontaneous reporting database provided by a pharmaceutical company and used in a simulation to investigate appropriate criteria for signal detection. In total, 1000 pseudo datasets were generated with this model, and three statistical methods (proportional reporting ratio [PRR], Bayesian Confidence Propagation Neural Network [BCPNN] and multi-item gamma Poisson shrinker [MGPS]) for signal detection were applied to each dataset. The sensitivity and specificity of each method were evaluated using these pseudo datasets. The optimum critical value for signal detection (i.e. the value that achieved the highest sensitivity with 95% specificity) was identified for each method. The optimum values were also examined with the adverse events classified into two categories according to frequency. The three original detection methods and their revised versions were applied to a real pharmaceutical company database to detect 173 known adverse reactions of four drugs. RESULTS: The 1000 pseudo datasets consisted of an average of 81 862 reports and 11,407 drug-event pairs, including 1192 adverse drug reactions. The sensitivities of PRR, BCPNN and MGPS methods were 49%, 45% and 26%, respectively, whereas their specificities were 95%, 99.6% and 99.99%, respectively; these sensitivities were unacceptably low for pharmaceutical manufacturers, whereas the specificities were acceptable. The highest sensitivity for each method, obtained by changing critical values and maintaining specificity at 95%, was 44%, 62% and 62%, respectively. When adverse events were classified into two categories, sensitivities as high as 75% for regular events and 39% for rare events were achieved with the revised BCPNN method. The critical values of the information component minus two standard deviations (IC - 2SD) index of the revised BCPNN method were greater than -0.7 for regular events and greater than -0.6 for rare events. The revised BCPNN method yielded 51% sensitivity and 89% specificity for the real dataset. CONCLUSION: A lower critical value may be needed when signal-detection methodology is applied to the spontaneous reporting databases of pharmaceutical manufacturers. For example, it is recommended that pharmaceutical manufacturers use the BCPNN method with IC - 2SD criteria of greater than -0.7 for regular events and greater than -0.6 for rare events.

Long-term clinical effects of epalrestat, an aldose reductase inhibitor, on diabetic peripheral neuropathy: the 3-year, multicenter, comparative Aldose Reductase Inhibitor-Diabetes Complications Trial.

OBJECTIVE: We sought to evaluate the long-term efficacy and safety of epalrestat, an aldose reductase inhibitor, on diabetic peripheral neuropathy. RESEARCH DESIGN AND METHODS: Subjects with diabetic neuropathy, median motor nerve conduction velocity (MNCV) >or=40 m/s, and HbA(1c)

Mouse lymphoma thymidine kinase gene mutation assay: follow-up meeting of the International Workshop on Genotoxicity Testing--Aberdeen, Scotland, 2003--Assay acceptance criteria, positive controls, and data evaluation.

The Mouse Lymphoma Assay (MLA) Workgroup of the International Workshop on Genotoxicity Testing (IWGT), comprised of experts from Japan, Europe, and the United States, met on August 29, 2003, in Aberdeen, Scotland, United Kingdom. This meeting of the MLA Workgroup was devoted to reaching a consensus on the appropriate approach to data evaluation and on acceptance criteria for both the positive and negative/vehicle controls. The Workgroup reached consensus on the acceptance criteria for both the agar and microwell versions of the MLA. Recommendations include acceptable ranges for mutant frequency, cloning efficiency, and suspension growth of the negative/vehicle controls and on criteria to define an acceptable positive control response. The recommendation for the determination of a positive/negative test chemical response includes both the requirement that the response exceeds a defined value [the global evaluation factor (GEF)] and that there also be a positive dose-response (evaluated by an appropriate statistical method).

Minimization method for balancing continuous prognostic variables between treatment and control groups using Kullback-Leibler divergence.

This paper proposes a method for balancing prognostic variables between treatment and control groups in design of clinical trials. It assumes that some of prognostic variables are continuous and others are categorical and that they are independently distributed. The proposed method uses the Kullback-Leibler divergence (KLD) as the index of difference in distribution between two groups. It sequentially allocates each subject to a group using a biased coin method so as to reduce the estimate of KLD. That is, when first i subjects have been allocated to two groups and the (i+1)th subject is enrolled, the KLD is estimated if the (i+1)th subject was to be allocated to either of the groups, and the subject is then allocated with a certain probability, e.g. 0.80, so as to make the KLD small. Simulation studies based on the hypothetical prognostic variables and on the actual data of hyperlipidemia patients were carried out in order to compare the proposed method with the Pocock-Simon method, which transforms the continuous prognostic variables into categorical variables by dividing the whole scale into several categories. The p values of homogeneity test of means and variances were used to evaluate the achieved balance. The observed p values in the proposed method were better than those in the Pocock-Simon method. In addition to the balance, the precision of parameter estimates assuming analysis of covariance model was examined. The results showed the precision of estimators tended to be more stable in the proposed method than the Pocock-Simon method.

An inter-laboratory collaborative study by the Non-Genotoxic Carcinogen Study Group in Japan, on a cell transformation assay for tumour promoters using Bhas 42 cells.

The Bhas promotion assay is a cell culture transformation assay designed as a sensitive and economical method for detecting the tumour-promoting activities of chemicals. In order to validate the transferability and applicability of this assay, an inter-laboratory collaborative study was conducted with the participation of 14 laboratories. After confirmation that these laboratories could obtain positive results with two tumour promoters, 12-O-tetradecanoylphorbol-13-acetate (TPA) and lithocholic acid (LCA), 12 coded chemicals were assayed. Each chemical was tested in four laboratories. For eight chemicals, all four laboratories obtained consistent results, and for two of the other four chemicals, only one of the four laboratories showed inconsistent results. Thus, the rate of consistency was high. During the study, several issues were raised, each of which were analysed step-by-step, leading to revision of the protocol of the original assay. Among these issues were the importance of careful maintenance of mother cultures and the adoption of test concentrations for toxic chemicals. In addition, it is suggested that three different types of chemicals show positive promoting activity in the assay. Those designated as T-type induced extreme growth enhancement, and included TPA, mezerein, PDD and insulin. LCA and okadaic acid belonged to the L-type category, in which transformed foci were induced at concentrations showing growth-inhibition. In contrast, M-type chemicals, progesterone, catechol and sodium saccharin, induced foci at concentrations with little or slight growth inhibition. The fact that different types of chemicals similarly induce transformed foci in the Bhas promotion assay may provide clues for elucidating mechanisms of tumour promotion.

Isofuranonaphthoquinone derivatives from cultures of the lichen Arthonia cinnabarina (DC.) Wallr.

Two isofuranonaphthoquinone derivatives, named arthoniafurones A (1-acetyl-8-hydroxynaphtho[2,3-c]furan-4,9-dione) and B [1-acetyl-4,8-dihydroxynaphtho[2,3-c]furan-9(4H)-one], were isolated from a spore-derived culture of the mycobiont of the lichen Arthonia cinnabarina, that is new to Japan. Bostrycoidin and 8-O-methylbostrycoidin were also identified in the A. cinnabarina culture.

A new statistical method for evaluation of L5178Ytk(+/-) mammalian cell mutation data using microwell method.

A statistical method to evaluate data from the mouse lymphoma L5178Y/tk assay (MLA) using microwell method is proposed. This proposed method is designed for data obtained from a single culture protocol instead of the duplicate culture recommended by United Kingdom Environmental Mutagen Society (UKEMS). The proposed method consists of the following three steps: (1) to apply Dunnett type test for identifying clear negative; (2) to apply a Simpson-Margolin procedure for detecting downturn data; and (3) to apply a trend test to evaluate the dose-dependent increase in mutant frequency (MF). The performance of the proposed method was evaluated through a Monte Carlo study and a case study. False positive rates realized in the Monte Carlo study were comparable with the UKEMS method modified for a single culture protocol with the heterogeneity factors being kept at 1.0. False negative rates were less than those of the modified UKEMS method for dose response patterns with a sharp uprise in higher dose groups, whereas, they were comparable for other patterns. The results of evaluating the data from an International Collaborative Study by the proposed method seem comparable with the UKEMS method. The proposed method enables us to evaluate data from the microwell MLA with a single culture protocol.

Mouse lymphoma thymidine kinase gene mutation assay: International Workshop on Genotoxicity Tests Workgroup report--Plymouth, UK 2002.

The Mouse Lymphoma Assay (MLA) Workgroup of the International Workshop on Genotoxicity Tests (IWGT) met on June 28th and 29th, 2002, in Plymouth, England. This meeting of the MLA group was devoted to discussing the criteria for assay acceptance and appropriate approaches to data evaluation. Prior to the meeting, the group conducted an extensive analysis of data from both the microwell and soft agar versions of the assay. For the establishment of criteria for assay acceptance, 10 laboratories (6 using the microwell method and 4 using soft agar) provided data on their background mutant frequencies, plating efficiencies of the negative/vehicle control, cell suspension growth, and positive control mutant frequencies. Using the distribution curves generated from this data, the Workgroup reached consensus on the range of values that should be used to determine whether an individual experiment is acceptable. In order to establish appropriate approaches for data evaluation, the group used a number of statistical methods to evaluate approximately 400 experimental data sets from 10 laboratories entered into a database created for the earlier MLA Workshop held in New Orleans [Environ. Mol. Mutagen. 40 (2002) 292]. While the Workgroup could not, during this meeting, make a final recommendation for the evaluation of data, a general strategy was developed and the Workgroup members agreed to evaluate this new proposed approach using their own laboratory data. This evaluation should lead to a consensus global approach for data evaluation in the near future.

Optimal patch application time in the evaluation of skin irritation.

We investigated the optimum application for evaluating skin irritation response by using samples of irritants commonly used as additives in cosmetics and other common household products. We studied 47 volunteers (16 men and 31 women). We selected three types of surfactant, one moisturizer, one anti-infective agent and one oil solution. Using Finn chambers on Scanpor tape, we performed the patch test. A total of 0.015 mL of each sample was applied to the Finn chamber. For liquids, circular filter paper was soaked in 0.015 mL of the sample. Samples were placed on the upper back of participants, and closed for 4, 24 or 48 h. A patch application time of 24 h is sufficient to detect primary skin irritation from irritants in cosmetics and other common household products. In addition, we found that skin irritation reactions were strongest at 24 h after patch removal and that the reaction tended to be weaker at 48 h after patch removal. Patch testing to evaluate irritants should be performed by means of a 24-h patch test with a follow-up reading at 24 h after patch removal. An application time of 24 h places less of a burden on patients than a 48-h patch test.

A new test statistic based on shrunken sample variance for identifying differentially expressed genes in small microarray experiments.

Choosing an appropriate statistic and precisely evaluating the false discovery rate (FDR) are both essential for devising an effective method for identifying differentially expressed genes in microarray data. The t-type score proposed by Pan et al. (2003) succeeded in suppressing false positives by controlling the underestimation of variance but left the overestimation uncontrolled. For controlling the overestimation, we devised a new test statistic (variance stabilized t-type score) by placing shrunken sample variances of the James-Stein type in the denominator of the t-type score. Since the relative superiority of the mean and median FDRs was unclear in the widely adopted Significance Analysis of Microarrays (SAM), we conducted simulation studies to examine the performance of the variance stabilized t-type score and the characteristics of the two FDRs. The variance stabilized t-type score was generally better than or at least as good as the t-type score, irrespective of the sample size and proportion of differentially expressed genes. In terms of accuracy, the median FDR was superior to the mean FDR when the proportion of differentially expressed genes was large. The variance stabilized t-type score with the median FDR was applied to actual colorectal cancer data and yielded a reasonable result.

Estimating the false discovery rate using mixed normal distribution for identifying differentially expressed genes in microarray data analysis.

The recent development of DNA microarray technology allows us to measure simultaneously the expression levels of thousands of genes and to identify truly correlated genes with anticancer drug response (differentially expressed genes) from many candidate genes. Significance Analysis of Microarray (SAM) is often used to estimate the false discovery rate (FDR), which is an index for optimizing the identifiability of differentially expressed genes, while the accuracy of the estimated FDR by SAM is not necessarily confirmed. We propose a new method for estimating the FDR assuming a mixed normal distribution on the test statistic and examine the performance of the proposed method and SAM using simulated data. The simulation results indicate that the accuracy of the estimated FDR by the proposed method and SAM, varied depending on the experimental conditions. We applied both methods to actual data comprised of expression levels of 12,625 genes of 10 responders and 14 non-responders to docetaxel for breast cancer. The proposed method identified 280 differentially expressed genes correlated with docetaxel response using a cut-off value for achieving FDR <0.01 to prevent false-positive genes, although 92 genes were previously thought to be correlated with docetaxel response ones.

Interlaboratory validation of the modified murine local lymph node assay based on adenosine triphosphate measurement.

INTRODUCTION: The murine local lymph node assay (LLNA) is a well-established alternative to the guinea pig maximization test (GPMT) or Buehler test (BT) for the assessment of the skin sensitizing ability of drugs and chemicals. Daicel Chemical Industries Ltd. has developed a modified LLNA based on the adenosine triphosphate (ATP) content (LLNA-DA). We conducted 2 interlaboratory validation studies to evaluate the reliability and relevance of LLNA-DA. METHODS: The experiment involved 17 laboratories, wherein 14 chemicals were examined under blinded conditions. In the first study, 3 chemicals were examined in 10 laboratories and the remaining 9 were examined in 3 laboratories. In the second study, 1 chemical was examined in 7 laboratories and the remaining 4 chemicals were examined in 4 laboratories. The data were expressed as the ATP content for each chemical-treated group, and the stimulation index (SI) for each chemical-treated group was determined as the increase in the ATP content relative to the concurrent vehicle control group. An SI of 3 was set as the cut-off value for exhibiting skin sensitization activity. RESULTS: The results of the first study obtained in the experiments conducted for the 3 chemicals that were examined in all the 10 laboratories and for 5 of the remaining 9 chemicals were sufficiently consistent with small variations in their SI values. The sensitivity, specificity, and accuracy of LLNA-DA against those of GPMT/BT were 7/8 (87.5%), 3/3 (100%), and 10/11 (90.9%), respectively. In the second study, all the 5 chemicals studied demonstrated acceptably small interlaboratory variations. DISCUSSION: In the first study, a large variation was observed for 2 chemicals; in the second study, this variation was small. It was attributed to the application of dimethylsulfoxide as the solvent for the metallic salts. In conclusion, these 2 studies provide good evidence for the reliability of the LLNA-DA.