The advent of feeding based RNAi in Caenorhabditis elegans led to an era of gene discovery in aging research. Hundreds of gerogenes were discovered, and many are evolutionarily conserved, raising the exciting possibility that the underlying genetic basis for healthy aging in higher vertebrates could be quickly deciphered. Yet, the majority of putative gerogenes have still only been cursorily characterized, highlighting the need for high-throughput, quantitative assessments of changes in aging. A widely used surrogate measure of aging is lifespan. The traditional way to measure mortality in C. elegans tracks the deaths of individual animals over time within a relatively small population. This traditional method provides straightforward, direct measurements of median and maximum lifespan for the sampled population. However, this method is time consuming, often underpowered, and involves repeated handling of a set of animals over time, which in turn can introduce contamination or possibly damage increasingly fragile, aged animals. We have previously developed an alternative "Replica Set" methodology, which minimizes handling and increases throughput by at least an order of magnitude. The Replica Set method allows changes in lifespan to be measured for over one hundred feeding-based RNAi clones by one investigator in a single experiment- facilitating the generation of large quantitative phenotypic datasets, a prerequisite for development of biological models at a systems level. Here, we demonstrate through analysis of lifespan experiments simulated in silico that the Replica Set method is at least as precise and accurate as the traditional method in evaluating and estimating lifespan, and requires many fewer total animal observations across the course of an experiment. Furthermore, we show that the traditional approach to lifespan experiments is more vulnerable than the Replica Set method to experimental and measurement error. We find no compromise in statistical power for Replica Set experiments, even for moderate effect sizes, or when simulated experimental errors are introduced. We compare and contrast the statistical analysis of data generated by the two approaches, and highlight pitfalls common with the traditional methodology. Collectively, our analysis provides a standard of measure for each method across comparable parameters, which will be invaluable in both experimental design and evaluation of published data for lifespan studies.

Abstract Recent reporting of the 9 year follow-up for the TAILORx trial suggests that there may be no benefit with adjuvant chemotherapy for ER +, HER2 -, N(0) breast cancer patients with a Oncotype DX® (ODX) recurrence score (RS) <26. Since endocrine therapy for this group of patients who comply with treatment still results in distant recurrence (rMBC) in 3% and 5% of the ODX low and ODX intermediate risk groups at 9 years, respectively, we are motivated to help find early treatments for these patients by identifying their recurrence risk at diagnosis with improved risk stratification. Methods: Optical Prediction of Time Interval to Metastasis (OPTIM), a novel assay, prognostic for rMBC, is based on an intrinsic optical signature from collagen, derived from the average of point by point ratios of forward to backward (F/B) second harmonic generation (SHG) light scatter that is sensitive to form and structure of fibrillar collagen in the extracellular matrix of archival tissue microarray samples. (Burke et al. BMC Cancer 15 (2015): 929). The 125 patients in this cohort were part of a clinical trial, looking for genomic predictors of rMBC in untreated patients, so we were able to calculate a surrogate 21-gene RT-PCR assay (S-ODX) value based on gene expression data available through NCBI GEO database (Gyorffy et al. Breast Cancer Res Treat (2012) 132:1025). We analyzed these patient's rMBC outcomes using logistic regression and Kaplan-Meier (KM) analysis. Results: OPTIM alone stratified at 2.5X relative risk (RR) between quartiles Q1 and Q4, similar to S-ODX low vs high recurrence score (RS) groups (from TAILORx Trial) with 2.8X RR. Using quartiles of OPTIM vs S-ODX together we stratify patients to recurrence risk (rMBC/at Risk), with an improved risk stratification of 5X RR in the RS<26 low risk groups. OPTIM Quartiles vs RS Risk Groups in TAILORx TrialS-ODX →High (RS>25)Intermediate (RS 11-25)Low (RS <11)AllOPTIM↓↓↓↓Q17/97/12*5/1019/31*Q25/95/142/812/31Q38/12***1/81/11***10/31Q46/10**2/17*0/5**8/32*All26/40***15/518/34***49/125Recurrence at 10 years by KM analysis *p<0.05, **p<0.005, ***p<.0005 Combining S-ODX with OPTIM, low (L) or high (H) risk by assay, shows that they are independent and complementary. Notably 68%=85/125 are classified L by S-ODX (RS<26) and OPTIM effectively reclassifies H and L, and when combined with S-ODX H identifies 92%=45/49 of all rMBC at 10 years without treatment. Risk stratification improves to 6.8X RR comparing highest risk HH 66.7%=12/18 to lowest risk LL 9.8%=4/41. Distant Recurrence Identified by High Risk Group of Each AssayS-ODX AssayHHLLOPTIM AssayHLHLrMBC (total=49)1214194At Risk (total n=125)18224441rMBC at 10 yrs. S-ODX RS>25=H, RS<26=L; OPTIM Q1&Q2=H, Q3&Q4=L Conclusion: OPTIM as an independent prognostic optical bio-marker from collagen in intact tissue. Combination of OPTIM with the Oncotype DX® assay may produce a continuous risk estimator with higher dynamic range than either assay alone and will be the focus of future study, especially in a treated population, to determine if OPTIM might also predict response to treatment. Citation Format: Hill RL, Perry SW, Salzman P, Turner BM, Hicks DG, Brown EB. Optical Prediction of Time Interval to Metastasis (OPTIM): A rapid nondestructive optical assay applied to tissue microarray samples identifying high risk of distant recurrence in the lowest risk groups defined by the TAILORx trial [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P6-09-05.

The Replica Set method is an approach to quantitatively measure lifespan or survival of Caenorhabditis elegans nematodes in a high-throughput manner, thus allowing a single investigator to screen more treatments or conditions over the same amount of time without loss of data quality. The method requires common equipment found in most laboratories working with C. elegans and is thus simple to adopt. The approach centers on assaying independent samples of a population at each observation point, rather than a single sample over time as with traditional longitudinal methods. Scoring entails adding liquid to the wells of a multi-well plate, which stimulates C. elegans to move and facilitates quantifying changes in healthspan. Other major benefits of the Replica Set method include reduced exposure of agar surfaces to airborne contaminants (e.g. mold or fungus), minimal handling of animals, and robustness to sporadic mis-scoring (such as calling an animal as dead when it is still alive). To appropriately analyze and visualize the data from a Replica Set style experiment, a custom software tool was also developed. Current capabilities of the software include plotting of survival curves for both Replica Set and traditional (Kaplan-Meier) experiments, as well as statistical analysis for Replica Set. The protocols provided here describe the traditional experimental approach and the Replica Set method, as well as an overview of the corresponding data analysis.

BACKGROUND: With the onset of next-generation sequencing technologies, we have made great progress in identifying recurrent mutational drivers of cancer. As cancer tissues are now frequently screened for specific sets of mutations, a large amount of samples has become available for analysis. Classification of patients with similar mutation profiles may help identifying subgroups of patients who might benefit from specific types of treatment. However, classification based on somatic mutations is challenging due to the sparseness and heterogeneity of the data. METHODS: Here we describe a new method to de-sparsify somatic mutation data using biological pathways. We applied this method to 23 cancer types from The Cancer Genome Atlas, including samples from 5805 primary tumours. RESULTS: We show that, for most cancer types, de-sparsified mutation data associate with phenotypic data. We identify poor prognostic subtypes in three cancer types, which are associated with mutations in signal transduction pathways for which targeted treatment options are available. We identify subtype-drug associations for 14 additional subtypes. Finally, we perform a pan-cancer subtyping analysis and identify nine pan-cancer subtypes, which associate with mutations in four overarching sets of biological pathways. CONCLUSIONS: This study is an important step toward understanding mutational patterns in cancer.

Background and aims Evidence continues to accumulate that athero-protective effects of high-density lipoprotein (HDL) depend to some degree on effective HDL functionality and that such functionality can become degraded in the setting of chronic inflammation. To investigate this issue, we have studied a group of post-myocardial infarction patients with high levels of C-reactive protein as an indicator of chronic inflammation and with concurrently high levels of HDL cholesterol. For these patients we have demonstrated high-risk for recurrent cardiac events as well as a strong association of risk with a polymorphism of the gene (SERPINB2) for plasminogen activator inhibitor-2 (PAI-2) presumptively reflective of an important role for fibrinolysis in risk. However, additional processes might be involved. The current work sought to characterize processes underlying how PAI-2 might be involved in the generation of risk. Methods Multivariate population data were leveraged using Bayesian network modeling, a graphical probabilistic approach for knowledge discovery, to generate networks reflective of influences on PAI-2 polymorphism-associated risk. Results Modeling results revealed three individual networks centering on the PAI-2 polymorphism with specific features providing information relating to how the polymorphism might associate with risk. These included racial dependency, platelet clot initiation and propagation, oxidative stress, inflammation effects on HDL metabolism and coagulation, and induction and termination of fibrinolysis. Conclusions Beyond direct association of a PAI-2 polymorphism with recurrent risk in post-myocardial infarction patients, results suggest that PAI-2 likely plays a key role leading to risk through multiple pathophysiologic processes. Such knowledge could potentially be valuable with individualization of patient care.

Geography is known to affect cost of care in surgical procedures. Understanding the relationship between geography and hospital costs is pertinent in the effort to reduce healthcare costs. We studied the geographic variation in cost for transsphenoidal pituitary surgery in hospitals across New York State.

Data is presented that was utilized as the basis for Bayesian network modeling of influence pathways focusing on the central role of a polymorphism of plasminogen activator inhibitor-2 (PAI-2) on recurrent cardiovascular disease risk in patients with high levels of HDL cholesterol and C-reactive protein (CRP) as a marker of inflammation, “Influences on Plasminogen Activator Inhibitor-2 Polymorphism-Associated Recurrent Cardiovascular Disease Risk in Patients with High HDL Cholesterol and Inflammation” (Corsetti et al., 2016; [1]). The data consist of occurrence of recurrent coronary events in 166 post myocardial infarction patients along with 1. clinical data on gender, race, age, and body mass index; 2. blood level data on 17 biomarkers; and 3. genotype data on 53 presumptive CVD-related single nucleotide polymorphisms. Additionally, a flow diagram of the Bayesian modeling procedure is presented along with Bayesian network subgraphs (root nodes to outcome events) utilized as the data from which PAI-2 associated influence pathways were derived (Corsetti et al., 2016; [1]).

BACKGROUND:Over-treatment of estrogen receptor positive (ER+), lymph node-negative (LNN) breast cancer patients with chemotherapy is a pressing clinical problem that can be addressed by improving techniques to predict tumor metastatic potential. Here we demonstrate that analysis of second harmonic generation (SHG) emission direction in primary tumor biopsies can provide prognostic information about the metastatic outcome of ER+, LNN breast cancer, as well as stage 1 colorectal adenocarcinoma.METHODS:SHG is an optical signal produced by fibrillar collagen. The ratio of the forward-to-backward emitted SHG signals (F/B) is sensitive to changes in structure of individual collagen fibers. F/B from excised primary tumor tissue was measured in a retrospective study of LNN breast cancer patients who had received no adjuvant systemic therapy and related to metastasis-free survival (MFS) and overall survival (OS) rates. In addition, F/B was studied for its association with the length of progression-free survival (PFS) in a subgroup of ER+ patients who received tamoxifen as first-line treatment for recurrent disease, and for its relation with OS in stage I colorectal and stage 1 lung adenocarcinoma patients.RESULTS:In 125 ER+, but not in 96 ER-negative (ER-), LNN breast cancer patients an increased F/B was significantly associated with a favorable MFS and OS (log rank trend for MFS: p = 0.004 and for OS: p = 0.03). On the other hand, an increased F/B was associated with shorter PFS in 60 ER+ recurrent breast cancer patients treated with tamoxifen (log rank trend p = 0.02). In stage I colorectal adenocarcinoma, an increased F/B was significantly related to poor OS (log rank trend p = 0.03), however this relationship was not statistically significant in stage I lung adenocarcinoma.CONCLUSION:Within ER+, LNN breast cancer specimens the F/B can stratify patients based upon their potential for tumor aggressiveness. This offers a "matrix-focused" method to predict metastatic outcome that is complementary to genomic "cell-focused" methods. In combination, this and other methods may contribute to improved metastatic prediction, and hence may help to reduce patient over-treatment.

BACKGROUND:It is not understood why some pulmonary fibroses such as cryptogenic organizing pneumonia (COP) respond well to treatment, while others like usual interstitial pneumonia (UIP) do not. Increased understanding of the structure and function of the matrix in this area is critical to improving our understanding of the biology of these diseases and developing novel therapies. The objectives herein are to provide new insights into the underlying collagen- and matrix-related biological mechanisms driving COP versus UIP.METHODS:Two-photon second harmonic generation (SHG) and excitation fluorescence microscopies were used to interrogate and quantify differences between intrinsic fibrillar collagen and elastin matrix signals in healthy, COP, and UIP lung.RESULTS:Collagen microstructure was different in UIP versus healthy lung, but not in COP versus healthy, as indicated by the ratio of forward-to-backward propagating SHG signal (FSHG/BSHG). This collagen microstructure as assessed by FSHG/BSHG was also different in areas with preserved alveolar architecture adjacent to UIP fibroblastic foci or honeycomb areas versus healthy lung. Fibrosis was evidenced by increased col1 and col3 content in COP and UIP versus healthy, with highest col1:col3 ratio in UIP. Evidence of elastin breakdown (i.e. reduced mature elastin fiber content), and increased collagen:mature elastin ratios, were seen in COP and UIP versus healthy.CONCLUSIONS:Fibrillar collagen's subresolution structure (i.e. "microstructure") is altered in UIP versus COP and healthy lung, which may provide novel insights into the biological reasons why unlike COP, UIP is resistant to therapies, and demonstrates the ability of SHG microscopy to potentially distinguish treatable versus intractable pulmonary fibroses.

The objective of this work was to investigate whether fibrinolysis plays a role in establishing recurrent coronary event risk in a previously identified group of postinfarction patients. This group of patients was defined as having concurrently high levels of high-density lipoprotein cholesterol (HDL-C) and C-reactive protein (CRP) and was previously demonstrated to be at high-risk for recurrent coronary events. Potential risk associations of a genetic polymorphism of plasminogen activator inhibitor-2 (PAI-2) were probed as well as potential modulatory effects on such risk of a polymorphism of low-density lipoprotein receptor related protein (LRP-1), a scavenger receptor known to be involved in fibrinolysis in the context of cellular internalization of plasminogen activator/plansminogen activator inhibitor complexes. To this end, Cox multivariable modeling was performed as a function of genetic polymorphisms of PAI-2 (SERPINB, rs6095) and LRP-1 (LRP1, rs1800156) as well as a set of clinical parameters, blood biomarkers, and genetic polymorphisms previously demonstrated to be significantly and independently associated with risk in the study population including cholesteryl ester transfer protein (CETP, rs708272), p22phox (CYBA, rs4673), and thrombospondin-4 (THBS4, rs1866389). Risk association was demonstrated for the reference allele of the PAI-2 polymorphism (hazard ratio 0.41 per allele, 95% CI 0.20-0.84, p=0.014) along with continued significant risk associations for the p22phox and thrombospondin-4 polymorphisms. Additionally, further analysis revealed interaction of the LRP-1 and PAI-2 polymorphisms in generating differential risk that was illustrated using Kaplan-Meier survival analysis. We conclude from the study that fibrinolysis likely plays a role in establishing recurrent coronary risk in postinfarction patients with concurrently high levels of HDL-C and CRP as manifested by differential effects on risk by polymorphisms of several genes linked to key actions involved in the fibrinolytic process.

Purpose: To determine the plasma concentrations of acute responding cytokines/chemokines following 9-Gy ionizing radiation in C57BL/6 (radiation tolerant) and C3H/HeN (radiation sensitive) murine strains. Methods and materials: Mice (5/group) received 9-Gy total body irradiation (TB!), and the plasma from each mouse was collected at 6 h or 1, 2, 4, or 10 days after TBI. A multiplex bead array was used to assess the levels of 32 cytokines/chemokines in plasma to determine their common and strain-specific temporal responses.Results: The plasma levels of five cytokines/chemokines (AxI, FasL, ICAM-1, TARC, and TSLP) were beyond the detectable level. Five (VEGF, IL-2, IL-5, IL-17, and CD30) were unaffected by irradiation in either strain. Temporal patterns were similar in both murine strains for 10 of the cytokines tested, including G-CSF, IL-6, TCA-3, MCP-1, MIP-1 gamma, KC, CXCL 13, CXCL 16, MDC, and TIMP-1; the other 12 molecules (GM-CSF, IL-3, SCF, IL-1 beta, IL-4, IL-10, IL-12p70, MIP-1 alpha, Eotaxin, TNF-alpha, sTNF-R1, and CD40) showed strain-specific response patterns. While a number of cytokines had temporal response patterns following TBI, the strains exhibited quantitatively different results.Conclusions: The levels of 27 of the 32 plasma cytokines measured indicate the following: (1) different cytokine concentrations and temporal patterns in the two strains may partly explain different radiation sensitivities and sequelae following irradiation; (2) many of the cytokines/chemokines exhibit similar temporal responses in the two strains. These responses suggest the potential value of using a panel of cytokine/chemokine temporal patterns for radiation dosimetry. Although radiation doses will be difficult to quantitate due to the large variation in levels and temporal responses exhibited in the two murine strains, serial measurements of cytokines might help identify subjects exposed to radiation. (C) 2012 Elsevier Ltd. All rights reserved.

Malignant cell transformation commonly results in the deregulation of thousands of cellular genes, an observation that suggests a complex biological process and an inherently challenging scenario for the development of effective cancer interventions. To better define the genes/pathways essential to regulating the malignant phenotype, we recently described a novel strategy based on the cooperative nature of carcinogenesis that focuses on genes synergistically deregulated in response to cooperating oncogenic mutations. These so-called 'cooperation response genes' (CRGs) are highly enriched for genes critical for the cancer phenotype, thereby suggesting their causal role in the malignant state. Here, we show that CRGs have an essential role in drug-mediated anticancer activity and that anticancer agents can be identified through their ability to antagonize the CRG expression profile. These findings provide proof-of-concept for the use of the CRG signature as a novel means of drug discovery with relevance to underlying anticancer drug mechanisms.

Pancreatic fistula continues to be a source of significant morbidity following distal pancreatic resections. The technique of pancreatic division varies widely among surgeons, and there is no evidence that identifies a single method as superior. In our practice, the technique of distal pancreatic resection has evolved from cut-and-sew to stapled technique with green and recently white cartridge. The aim of our study was to evaluate the rate of clinically significant fistulas [International Study Group on Pancreatic Fistula (ISGPF) grade B or C] following distal pancreatectomy and to identify variables associated with a low rate of fistula development.

Purpose: While secretagogue-induced diarrhea is rich in chloride (Cl-) and bicarbonate (HCO3-) anions, little is known about diarrhea or its anionic composition following irradiation. We performed studies to characterize the differences between cyclic adenosine monophosphate (cAMP)-stimulated anion secretions in irradiated and non-irradiated mice.Materials and methods: HCO3- secretion was examined in basal, cAMP-stimulated, and irradiated jejunal tissues from BALB/c (Bagg albino) mice. The abdomens of the mice were gamma-irradiated using a caesium-137 source.Results: Ussing-chamber experiments performed in an HCO3--containing, Cl -free solution on the bath side showed inhibition of HCO3- in irradiated mice. Non-irradiated mice exhibited bumetanide-sensitive and insensitive current, while irradiated mice displayed bumetanide-sensitive current. pH-stat experiments showed inhibition of basal and cAMP-stimulated HCO3- secretions in irradiated mice. Immunohistochemistry and Western blot analysis displayed a sodium-bicarbonate cotransporter expression in the villus and not the crypt of non-irradiated mice, while its expression and protein levels decreased in irradiated mice.Conclusions: Anion secretions in irradiated mice, being primarily Cl- and minimally HCO3-, differ from that of secretagogue-induced anion secretions. Understanding anion loss will help us correct electrolyte imbalances, while reduced HCO3- secretion in the upper-gastrointestinal tract might also have implications for irradiation-induced nausea and vomiting.

The R package catnet provides an inference framework for categorical Bayesian networks. Bayesian networks are graphical statistical models that represent causal dependencies between random variables. A Bayesian network has two components: a Directed Acyclic Graph (DAG) with nodes representing random variables and a probability structure specified by conditional distributions, one for each node in the graph. Any Bayesian network satisfies the so called local Markov property that requires each node in the network to be independent of its non-descendants given its parent nodes. This property implies a factorization of the joint distribution of the random variables that greatly facilitates the statistical inference. Two classes of Bayesian networks are among the most used in practice: linear Gaussian networks and categorical ones (also called discrete Bayesian networks). In a linear Gaussian network, the nodes represent continuous variables with Gaussian conditional distributions such that the expected value of each node is a linear combination of its parent nodes. Gaussian networks benefit from strong analytical properties and inference methodology. However, their usage is not justified when the linearity and normality assumptions on the observed variables are violated. In a categorical Bayesian network, each node takes values in a fixed set of categories and the conditional distributions are multinomial with no additional parametric constraints. Categorical Bayesian networks are capable of representing non-linear relationships between random variables and are suitable when the data is genuinely categorical or the marginal distributions follow some multi-modal distributions suitable for discretization. The problem of learning Bayesian networks has relatively long history with abundance of literature devoted to its subject. See for example [Heckeman et al.(1995)], [Cooper & Herskovitz(1992)], [Chickering(1996b)], [Friedman et al.(1999)], [Larranaga et al.(1996)], and some more recent articles, [Tsamardinos et al.(2003)], [Yaramakala & Margaritis(2005)], [Daly & Shen(2007)]. The main goal of the catnet package is to provide tools for inferring categorical Bayesian networks from data based on the maximum likelihood (ML) estimation. The inference employs a scoring-based network learning and does not include any assumptions on the marginal or conditional distributions of the nodes in form of priors. Two main techniques are implemented finding the best network fitting some data for a predefined node order and stochastic search of optimal networks without constraints on the order of the nodes. For a given node order, an efficient exhaustive search via Dynamic Programming (DP) is implemented and the exact MLE solution is given. The approach is similar to the one in [Friedman & Koller(2003)], the latter however employs full Bayesian inference. The stochastic search in the space of node orders is implemented by employing a Simulated Annealing (SA) algorithm. A distinct feature of catnet is its support of both incomplete data, with some of the records having missing values, and the so called perturbed data. In the latter case, some of the nodes are controlled by breaking the causal influence of their parents. Perturbed data originate naturally from gene expression studies where selected genes are knocked up/down. The package equips the user not only with structure learning but also with selection, estimation and prediction functions. For example, in catnet , one can perform asymptotically consistent model selection from incomplete or perturbed data, [Balov(2011)]. The motivating goal is to provide more

The aim of the study was to investigate the relationship between plasma levels of malondialdehyde (MDA), a routinely used marker of oxidative stress, and squamous cell carcinoma of the oral cavity and oropharynx (OSCC). The prospective cohort study comprised a total of 67 patients who underwent surgery for OSCC. MDA was assessed using high performance liquid chromatography. The MDA levels in the pooled T1-2 patients were lower than in the patients with T3-4 tumors. A negative correlation of MDA and tumor grade was shown. Seventeen patients who manifested recurrence during the 49.6 months follow-up had significantly increased MDA compared to those staying in complete remission. Kaplan-Meier analysis revealed that the median disease-free interval and overall survival in the group with MDA > median was 19.3 and 22.5 months respectively, in contrast to 31.5 and 31.6 months respectively, in patients with MDA < or = median. The prognostic value and low cost of MDA measurement could make it a versatile and useful prognostic tool for the identification of OSCC patients with a high risk of recurrence.

BACKGROUND: Although duty hours regulations (DHR) were introduced as a measure to improve patient safety and graduate medical education, new evidence suggests that the opposite might be happening. This study was designed to assess surgery resident perceptions of the impact that DHR have had on their education, the number of hours they believed would be ideal for their training, and to evaluate the effect of seniority on these opinions.STUDY DESIGN: An Internet-based survey was electronically distributed to all Resident and Associate members of the American College of Surgeons.RESULTS: Of 599 respondents, 247 (41%) believed that DHR were an important barrier to their education, and 266 (44%) believed that the ideal work week should have 80 to 100 hours. These two opinions were highly correlated, and were increasingly voiced with increased resident experience. Senior residents were more likely to view DHR as an important barrier to their education whether or not they were general surgery residents or were trained in small, medium, or large programs.CONCLUSIONS: A large subset of surgery residents, particularly senior residents, considered DHR an important barrier to their education and expressed a desire to work longer hours than restrictions allow. These Findings suggest that strict and uniform DHR do not allow for optimal training of residents at different levels who have disparate educational goals and needs. Introducing some flexibility into senior residents' limitations should be considered. (J Am Coll Sing 2009;209: 47-54. (C) 2009 by the American College of Surgeons)

We study the class of general step-down multiple testing procedures, which contains the usually considered procedures determined by a nondecreasing sequence of thresholds (we call them threshold step-down, or TSD, procedures) as a parametric subclass. We show that all procedures in this class satisfying the natural condition of monotonicity and controlling the family-wise error rate (FWER) at a prescribed level are dominated by one of them - the classical Holm procedure. This generalizes an earlier result pertaining to the subclass of TSD procedures (Lehmann and Romano, Testing Statistical Hypotheses, 3rd ed., 2005). We also derive a relation between the levels at which a monotone step-down procedure controls the FWER and the generalized FWER (the probability of k or more false rejections).