Bayesian Hierarchical Models for Subgroup Analysis.

In conventional subgroup analyses, subgroup treatment effects are estimated using data from each subgroup separately without considering data from other subgroups in the same study. The subgroup treatment effects estimated this way may be heterogenous with high variability due to small sample sizes in some subgroups and much different from the treatment effect in the overall population. A Bayesian hierarchical model (BHM) can be used to derive more precise, and less heterogenous estimates of subgroup treatment effects that are closer to the treatment effect in the overall population. BHM assumes exchangeability in treatment effect across subgroups after adjusting for effect modifiers and other relevant covariates. In this article, we will discuss the technical details for applying one-way and multi-way BHM using summary-level statistics, and patient-level data for subgroup analysis. Four case studies based on four new drug applications are used to illustrate the application of these models in subgroup analyses for continuous, dichotomous, time-to-event, and count endpoints.

Relational sanctification, communal coping, and depression among African American couples.

Despite the common use of religious buffers, African Americans are disproportionately affected by depressive symptoms. Communal coping may serve as one factor in helping religious African American couples alleviate the symptoms of depression. This study examines the association between relational sanctification and depressive symptoms as mediated by the communal coping of 467 African American married and cohabiting couples. Data from the sampled couples were analyzed using a common fate model, and analyses revealed higher scores on the measure of sanctification were associated with more communal coping; more communal coping was associated with fewer depressive symptoms among women and men, and communal coping acted as a mediator between relational sanctification and depressive symptoms in both partners. Findings from this study underscore the importance of considering how the religiosity and cooperative action of African American couples relate to depressive symptoms.

Perspectives on informative Bayesian methods in pediatrics.

Bayesian methods have been proposed as a natural fit for pediatric extrapolation, as they allow the incorporation of relevant external data to reduce the required sample size and hence trial burden for the pediatric patient population. In this paper we will discuss our experience and perspectives with these methods in pediatric trials. We will present some of the background and thinking underlying pediatric extrapolation and discuss the use of Bayesian methods within this context. We will present two recent case examples illustrating the value of a Bayesian approach in this setting and present perspectives on some of the issues that we have encountered in these and other cases.

Bayesian Methods in Human Drug and Biological Products Development in CDER and CBER.

The Center for Drug Evaluation and Research (CDER) and the Center for Biologics Evaluation and Research (CBER) of the U.S. Food and Drug Administration (FDA) have been leaders in protecting and promoting the U.S. public health by helping to ensure that safe and effective drugs and biological products are available in the United States for those who need them. The null hypothesis significance testing approach, along with other considerations, is typically used to demonstrate the effectiveness of a drug or biological product. The Bayesian framework presents an alternative approach to demonstrate the effectiveness of a treatment. This article discusses the Bayesian framework for drug and biological product development, highlights key settings in which Bayesian approaches may be appropriate, and provides recent examples of the use of Bayesian approaches within CDER and CBER.

Bayesian adaptive design for pediatric clinical trials incorporating a community of prior beliefs.

BACKGROUND: Pediatric population presents several barriers for clinical trial design and analysis, including ethical constraints on the sample size and slow accrual rate. Bayesian adaptive design methods could be considered to address these challenges in pediatric clinical trials. METHODS: We developed an innovative Bayesian adaptive design method and demonstrated the approach as a re-design of a published phase III pediatric trial. The innovative design used early success criteria based on skeptical prior and early futility criteria based on enthusiastic prior extrapolated from a historical adult trial, and the early and late stopping boundaries were calibrated to ensure a one-sided type I error of 2.5%. We also constructed several alternative designs which incorporated only one type of prior belief and the same stopping boundaries. To identify a preferred design, we compared operating characteristics including power, expected trial size and trial duration for all the candidate adaptive designs via simulation when performing an increasing number of equally spaced interim analyses. RESULTS: When performing an increasing number of equally spaced interim analyses, the innovative Bayesian adaptive trial design incorporating both skeptical and enthusiastic priors at both interim and final analyses outperforms alternative designs which only consider one type of prior belief, because it allows more reduction in sample size and trial duration while still offering good trial design properties including controlled type I error rate and sufficient power. CONCLUSIONS: Designing a Bayesian adaptive pediatric trial with both skeptical and enthusiastic priors can be an efficient and robust approach for early trial stopping, thus potentially saving time and money for trial conduction.

The Use of Drosophila to Understand Psychostimulant Responses.

The addictive properties of psychostimulants such as cocaine, amphetamine, methamphetamine, and methylphenidate are based on their ability to increase dopaminergic neurotransmission in the reward system. While cocaine and methamphetamine are predominately used recreationally, amphetamine and methylphenidate also work as effective therapeutics to treat symptoms of disorders including attention deficit and hyperactivity disorder (ADHD) and autism spectrum disorder (ASD). Although both the addictive properties of psychostimulant drugs and their therapeutic efficacy are influenced by genetic variation, very few genes that regulate these processes in humans have been identified. This is largely due to population heterogeneity which entails a requirement for large samples. Drosophila melanogaster exhibits similar psychostimulant responses to humans, a high degree of gene conservation, and allow performance of behavioral assays in a large population. Additionally, amphetamine and methylphenidate reduce impairments in fly models of ADHD-like behavior. Therefore, Drosophila represents an ideal translational model organism to tackle the genetic components underlying the effects of psychostimulants. Here, we break down the many assays that reliably quantify the effects of cocaine, amphetamine, methamphetamine, and methylphenidate in Drosophila. We also discuss how Drosophila is an efficient and cost-effective model organism for identifying novel candidate genes and molecular mechanisms involved in the behavioral responses to psychostimulant drugs.

Applying the Noninferiority Paradigm to Assess Exposure-Response Similarity and Dose Between Pediatric and Adult Patients.

The use of extrapolation of efficacy in pediatric drug development programs is possible when disease progression and treatment response are similar in adult and pediatric populations. Historically, the exposure-response (E-R) similarity was assessed by visual inspection of 2 E-R curves to support pediatric extrapolation. The aim of this study was to develop a quantitative framework to describe the E-R relationship and the difference in E-R between pediatric and adult patients based on accumulated experience in pediatric drug development programs. Using clinical data for 8 drugs with either a linear or nonlinear E-R relationship, we adapted the methodology used in noninferiority testing to assess the E-R similarity between adult and pediatric patients at the targeted drug exposure. We implemented bootstrap-based and Bayesian-based methodologies to estimate the probability of concluding noninferiority of the E-R relationship. This approach provides objective criteria that can be applied to an assessment of E-R noninferiority in 2 populations to support extrapolation of efficacy in drug development programs from adults to pediatric populations.

An Empirical Comparison of Statistical Methods for Missing Data in Randomized, Double-Blind, Placebo-Controlled, Phase 3 Clinical Trials for Chronic Pain and Lipid-Lowering Products.

BACKGROUND: Missing data are uncollected data but meaningful for the statistical analysis due to clinical relevancy of the data for properly specified estimands in clinical trials. Meanwhile the efforts to prevent or minimize missing data are commonly applied in clinical trials, in practice, missing data still occurs. Choosing a statistical method for imputation that deals with missing data targeting specified estimands provides the more reliable estimates of treatment effects. METHODS: We considered longitudinal clinical settings that have different degrees of missing data and treatment effects, and simulated different missing mechanisms using data from randomized, double-blind, placebo-controlled phase 3 confirmatory clinical trials of approved drugs. We compared four commonly used statistical methods to deal with missing data in clinical trials. RESULTS: We find that, when the data are missing not at random (MNAR) with higher missing rates, mixed model for repeated measurements (MMRM) method overestimates treatment difference. Pattern-mixture model estimates were seen to be more conservative in our studies than MMRM given MNAR assumptions, which are more realistic with missing data in clinical trials. CONCLUSIONS: We emphasize the importance of prevention of missing data and specifying the estimand based on trial objectives beforehand. The specified proper estimand and the proper statistical method might be key features to value the clinical trial results despite missing data.

Primary prevention of pediatric lead exposure requires new approaches to transfusion screening.

OBJECTIVE: To facilitate further assessment of transfusion-associated lead exposure by designing a procedure to test packed red blood cells (pRBCs) prepared for transfusion. STUDY DESIGN: The relationship between pRBCs and whole blood lead concentration was investigated in 27 samples using a modified clinical assay. Lead concentrations were measured in 100 pRBC units. RESULTS: Our sample preparation method demonstrated a correlation between whole blood lead and pRBC lead concentrations (R(2) = 0.82). In addition, all 100 pRBC units tested had detectable lead levels. The median pRBC lead concentration was 0.8 µg/dL, with an SD of 0.8 µg/dL and a range of 0.2-4.1 µg/dL. In addition, after only a few days of storage, approximately 25% of whole blood lead was found in the supernatant plasma. CONCLUSION: Transfusion of pRBCs is a source of lead exposure. Here we report the quantification of lead concentration in pRBCs. We found a >20-fold range of lead concentrations in the samples tested. Pretransfusion testing of pRBC units according to our proposed approach or donor screening of whole blood lead and selection of below-average units for transfusion to children would diminish an easily overlooked source of pediatric lead exposure.

Natural history of adults with congenitally malformed aortic valves (unicuspid or bicuspid).

Appreciation of the frequency of the congenitally malformed aortic valve has come about during the last 50 years, a period during which aortic valve replacement became a predictably successful operation. Study of patients at necropsy with either a congenitally unicuspid (1 true commissure) or bicuspid (2 true commissures) valve in whom no aortic valve operation has been performed has not been conducted during these 50 years, to our knowledge. We studied 218 patients at necropsy with congenitally malformed aortic valves: 28 (13%) had a unicuspid valve and 190 (87%), a bicuspid valve. Their ages at death ranged from 21 to 89 years (mean, 55 yr), and 80% were men. Of the 218 adults, the aortic valve functioned normally during life in 54 (25%) and abnormally in 164 (75%): aortic stenosis in 142 (65%), pure aortic regurgitation without superimposed infective endocarditis (IE) in 2 (1%), and IE superimposed on a previously normally functioning aortic valve in 20 (9%). IE occurred in a total of 31 (14%) of the 218 patients: involving a previously normally functioning valve in 20 (65%) and a previously stenotic valve in 11 (35%). Of the 218 patients, at least 141 (65%) died as a consequence of aortic valve disease (124 patients) or ascending aortic tears with or without dissection (17 patients). An estimated 1% of the population, maybe higher in men, has a congenitally malformed aortic valve. Data from this study suggest that about 75% of them will develop a major complication. Conversely, and encouragingly, about 25% will go through life without a complication.

Acute aortic dissection with tear in ascending aorta not diagnosed until necropsy or operation (for another condition) and comparison to similar cases receiving proper operative therapy.

Acute aortic dissection (AD) with a tear in ascending aorta (AA) is recognized to masquerade occasionally as another condition, and therefore the proper diagnosis is never made or made too late. During a recent 3-year period (2009 to 2011) at Baylor University Medical Center at Dallas, 30 patients with acute AD with tear in AA were diagnosed: 23 (77%) were diagnosed clinically and underwent proper urgent operative therapy; the remaining 7 (23%) with AD with tear in AA were not diagnosed until operation (for another condition) or necropsy or retrospectively by review of antemortem computed tomographic scan after death. The number of patients with AD from tear in AA whose cardiac condition was not diagnosed clinically and who died in the hospital but did not have an autopsy is unclear. Because the autopsy rate in most United States hospitals, including Baylor University Medical Center, is now <5%, many cases of AD are probably not diagnosed and thus its frequency is being underestimated. In conclusion, ≥ 7 of 30 patients (23%) with acute AD with a tear in AA were not diagnosed until necropsy or operation (for another condition) or retrospectively by computed tomography during a 3-year period at a large tertiary medical center showing that this condition continues to be underdiagnosed, and when not diagnosed correctly is usually rapidly fatal. Because the autopsy rate today is so low, the percentage not diagnosed clinically is probably >23%.

Effect of coronary bypass and valve structure on outcome in isolated valve replacement for aortic stenosis.

Reports differ regarding the effect of concomitant coronary artery bypass grafting (CABG) in patients who undergo aortic valve replacement (AVR) for aortic stenosis (AS), and no reports have described the effect of aortic valve structure in patients who undergo AVR for AS. A total of 871 patients aged 24 to 94 years (mean 70) whose AVR for AS was their first cardiac operation, with or without first concomitant CABG, were included. Patients who underwent mitral valve procedures were excluded. In comparison with the 443 patients (51%) who did not undergo CABG, the 428 (49%) who underwent concomitant CABG were significantly older, were more often male, had lower transvalvular peak systolic pressure gradients and larger valve areas, had lower frequencies of congenitally malformed aortic valves, had lighter valves by weight, had higher frequencies of systemic hypertension, and had longer stays in the hospital after AVR. Early and late (to 10 years) mortality were similar by propensity-adjusted analysis in patients who did and did not undergo concomitant CABG. Congenitally unicuspid or bicuspid valves occurred in approximately 90% of those aged 21 to 50, in nearly 70% in those aged 51 to 70 years, and in just over 30% in those aged 71 to 95 years. Unadjusted and adjusted survival was significantly higher in patients with unicuspid or bicuspid valves compared to those with tricuspid valves. In conclusion, although concomitant CABG had no effect on the adjusted probability of survival, the type of aortic valve (unicuspid or bicuspid vs tricuspid) significantly affected the unadjusted and adjusted probability of survival.

Rootstock-regulated gene expression patterns associated with fire blight resistance in apple.

BACKGROUND: Desirable apple varieties are clonally propagated by grafting vegetative scions onto rootstocks. Rootstocks influence many phenotypic traits of the scion, including resistance to pathogens such as Erwinia amylovora, which causes fire blight, the most serious bacterial disease of apple. The purpose of the present study was to quantify rootstock-mediated differences in scion fire blight susceptibility and to identify transcripts in the scion whose expression levels correlated with this response. RESULTS: Rootstock influence on scion fire blight resistance was quantified by inoculating three-year old, orchard-grown apple trees, consisting of 'Gala' scions grafted to a range of rootstocks, with E. amylovora. Disease severity was measured by the extent of shoot necrosis over time. 'Gala' scions grafted to G.30 or MM.111 rootstocks showed the lowest rates of necrosis, while 'Gala' on M.27 and B.9 showed the highest rates of necrosis. 'Gala' scions on M.7, S.4 or M.9F56 had intermediate necrosis rates. Using an apple DNA microarray representing 55,230 unique transcripts, gene expression patterns were compared in healthy, un-inoculated, greenhouse-grown 'Gala' scions on the same series of rootstocks. We identified 690 transcripts whose steady-state expression levels correlated with the degree of fire blight susceptibility of the scion/rootstock combinations. Transcripts known to be differentially expressed during E. amylovora infection were disproportionately represented among these transcripts. A second-generation apple microarray representing 26,000 transcripts was developed and was used to test these correlations in an orchard-grown population of trees segregating for fire blight resistance. Of the 690 transcripts originally identified using the first-generation array, 39 had expression levels that correlated with fire blight resistance in the breeding population. CONCLUSIONS: Rootstocks had significant effects on the fire blight susceptibility of 'Gala' scions, and rootstock-regulated gene expression patterns could be correlated with differences in susceptibility. The results suggest a relationship between rootstock-regulated fire blight susceptibility and sorbitol dehydrogenase, phenylpropanoid metabolism, protein processing in the endoplasmic reticulum, and endocytosis, among others. This study illustrates the utility of our rootstock-regulated gene expression data sets for candidate trait-associated gene data mining.

Presence of a congenitally bicuspid aortic valve among patients having combined mitral and aortic valve replacement.

Although bicuspid aortic valve occurs in an estimated 1% of adults and mitral valve prolapse in an estimated 5% of adults, occurrence of the 2 in the same patient is infrequent. During examination of operatively excised aortic and mitral valves because of dysfunction (stenosis and/or regurgitation), we encountered 16 patients who had congenitally bicuspid aortic valves associated with various types of dysfunctioning mitral valves. Eleven of the 16 patients had aortic stenosis (AS): 5 of them also had mitral stenosis, of rheumatic origin in 4 and secondary to mitral annular calcium in 1; the other 6 with aortic stenosis had pure mitral regurgitation (MR) secondary to mitral valve prolapse in 3, to ischemia in 2, and to unclear origin in 1. Of the 5 patients with pure aortic regurgitation, each also had pure mitral regurgitation: in 1 secondary to mitral valve prolapse and in 4 secondary to infective endocarditis. In conclusion, various types of mitral dysfunction severe enough to warrant mitral valve replacement occur in patients with bicuspid aortic valves. A proper search for mitral valve dysfunction in patients with bicuspid aortic valves appears warranted.

Aortic medial elastic fiber loss in acute ascending aortic dissection.

The cause of acute aortic dissection continues to be debated. One school of thought suggests that underlying aortic medial cystic necrosis is the common denominator. The purpose of the present study was to determine if there was loss and, if so, how much loss of medial elastic fibers in the ascending aorta in patients with acute aortic dissection with the entrance tear in the ascending aorta. We examined operatively excised ascending aortas in 69 patients having acute dissection with tears in the ascending aorta. Patients with previous aortotomy, healed dissection, and connective tissue disorders were excluded. The 69 patients' ages ranged from 31 to 88 years (mean 56); 49 were men and 20 were women. Loss of aortic medial elastic fibers was graded as 0 (no loss), 1+ (trace), 2+ (mild), 3+ (moderate), and 4+ (full thickness loss). Of these 69 patients, 56 (82%) had 0 or 1+ elastic fiber loss; 13 patients (18%), 2+ to 4+ loss including 4 with 2+, 6 with 3+, and 2 with 4+. Nearly all patients (97%) had a history of systemic hypertension and/or had received antihypertensive drug therapy. In conclusion, most patients (82% in this study) having acute aortic dissection with entrance tears in the ascending aorta have normal numbers or only trace loss of aortic medial elastic fibers. Thus, underlying abnormal ascending aortic structure uncommonly precedes acute dissection.

Effect of body mass index on survival in patients having aortic valve replacement for aortic stenosis with or without concomitant coronary artery bypass grafting.

The purpose of this report is to describe the effect of body mass index (BMI) on 30-day and late outcome in patients having aortic valve replacement (AVR) for aortic stenosis (AS) with or without concomitant coronary artery bypass grafting. From January 2002 through June 2010 (8.5 years), 1,040 operatively excised stenotic aortic valves were submitted to the cardiovascular laboratory at Baylor University Medical Center at Dallas. Of the 1,040 cases 175 were eliminated because they had a previous cardiac operation. The present study included 865 adults whose AVR for AS was their first cardiac operation. Propensity-adjusted analysis showed that 30-day and late mortality were strongly and significantly associated with BMI. Decreased risk of 30-day and long-term mortality was observed for patients with BMI in the low 30s compared to patients with BMI in the mid 20s or >40 kg/m(2). In conclusion, the findings in this study indicate a strong and significant adjusted association between BMI and 30-day and long-term mortality in patients having AVR for AS with or without concomitant coronary artery bypass grafting. Better survival was observed in patients with BMIs in the low 30s compared to patients with BMIs in the mid 20s and >40 kg/m(2).

Comparison of the structure of the aortic valve and ascending aorta in adults having aortic valve replacement for aortic stenosis versus for pure aortic regurgitation and resection of the ascending aorta for aneurysm.

BACKGROUND: There is debate concerning whether an aneurysmal ascending aorta should be replaced when associated with a dysfunctioning aortic valve that is to be replaced. To examine this issue, we divided the patients by type of aortic valve dysfunction-either aortic stenosis (AS) or pure aortic regurgitation (AR)-something not previously undertaken. METHODS AND RESULTS: Of 122 patients with ascending aortic aneurysm (unassociated with aortitis or acute dissection), the aortic valve was congenitally malformed (unicuspid or bicuspid) in 58 (98%) of the 59 AS patients, and in 38 (60%) of the 63 pure AR patients. Ascending aortic medial elastic fiber loss (EFL) (graded 0 to 4+) was zero or 1+ in 53 (90%) of the AS patients, in 20 (53%) of the 38 AR patients with bicuspid valves, and in all 12 AR patients with tricuspid valves unassociated with the Marfan syndrome. An unadjusted analysis showed that, among the 96 patients with congenitally malformed valves, the 38 AR patients had a significantly higher likelihood of 2+ to 4+ EFL than the 58 AS patients (crude odds ratio: 8.78; 95% confidence interval: 2.95, 28.13). CONCLUSIONS: These data strongly suggest that the type of aortic valve dysfunction-AS versus pure AR-is very helpful in predicting loss of aortic medial elastic fibers in patients with ascending aortic aneurysms and aortic valve disease.

Use of Gibberellic Acid for Management of Bunch Rot on Chardonnay and Vignoles Grape.

Harvest bunch rot of wine grape, caused primarily by Botrytis cinerea, is a perennial problem limiting the productivity of eastern vineyards, especially on cultivars with compact clusters. The aim of the present study was to evaluate the effectiveness of gibberellic acid (GA) sprays at reducing the compactness of Chardonnay and Vignoles clusters and minimizing bunch rot. Applications of GA reduced the number of berries per centimeter and the incidence and severity of bunch rots in Vignoles and, to a lesser extent, in Chardonnay over three consecutive years; however, the magnitude of GA effects often depended on the timing and rate of application. Bloom GA applications were more effective (P < 0.001) at reducing compactness and bunch rots than prebloom applications. Significantly, negative effects of GA applications on yield were negligible based on data from 4 years of trials on single vines and 2 years of data on 24-vine plots of Vignoles, provided the rates did not exceed 25 ppm. Regression analysis showed that berries per centimeter accounted for between 89 and 94% of variation in the incidence of Botrytis rot on Vignoles. On Chardonnay, compactness accounted for 53% of the variation in incidence, and the estimated compactness level at which no bunch rot would occur was 4.40 ± 1.05 (mean ± standard error) berries per centimeter. The relationship between cluster compactness and spray coverage of berries was also investigated in two separate experiments. Spray coverage of individual berries decreased linearly as cluster compactness increased within the range tested (3 to 18 berries per centimeter). Cluster compactness accounted for two-thirds of the variation in individual berry coverage, and coverage was reduced by 40 to 50% for clusters with about 18 berries per centimeter. These results strongly support the use of GA in integrated management of bunch rot on Vignoles and Chardonnay in eastern U.S. vineyards.

Virulence characteristics accounting for fire blight disease severity in apple trees and seedlings.

The gram-negative bacterium Erwinia amylovora is the causal agent of fire blight, the most destructive bacterial disease of rosaceous plants, including apple and pear. Here, we compared the virulence levels of six E. amylovora strains (Ea273, CFBP1367, Ea581a, E2002a, E4001a, and HKN06P1) on apple trees and seedlings. The strains produced a range of disease severity, with HKN06P1 producing the greatest disease severity in every assay. We then compared virulence characteristic expression among the six strains, including growth rates in immature apple fruit, amylovoran production, levansucrase activity, biofilm formation, carbohydrate utilization, hypersensitive cell death elicitation in tobacco leaves, and protein secretion profiles. Multiple regression analysis indicated that three of the virulence characteristics (amylovoran production, biofilm formation, and growth in immature apple fruit) accounted for >70% of the variation in disease severity on apple seedlings. Furthermore, in greenhouse-grown 'Gala' trees, >75% of the variation in disease severity was accounted for by five of the virulence characteristics: amylovoran production, biofilm formation, growth in immature apple fruit, hypersensitive cell death elicitation, and sorbitol utilization. This study demonstrates that virulence factor expression levels account for differences in disease severity caused by wild isolates of E. amylovora on apple trees.

Diversity and biogeography of sooty blotch and flyspeck fungi on apple in the eastern and midwestern United States.

Sooty blotch and flyspeck (SBFS) fungi on apple fruit were sampled from nine orchards in four midwestern U.S. states during 2000 and 30 orchards in 10 eastern U.S. states during 2005 in order to estimate taxonomic diversity and discern patterns of geographic distribution. Forty apple fruit per orchard were arbitrarily sampled and colonies of each mycelial phenotype were counted on each apple. Representative colonies were isolated, cultures were purified, and DNA was extracted. For representative isolates, the internal transcribed spacer (ITS) and large subunit (LSU) regions of ribosomal DNA were amplified and sequenced. In total, 60 SBFS putative species were identified based on ITS sequences and morphological characteristics; 30 of these were discovered in the 2005 survey. Modified Koch's postulates were fulfilled for all 60 species in an Iowa orchard; colonies resulting from inoculation of apple fruit were matched to the original isolates on the basis of mycelial type and ITS sequence. Parsimony analysis for LSU sequences from both surveys revealed that 58 putative SBFS species were members of the Dothideomycetes, 52 were members of the Capnodiales, and 36 were members of the Mycosphaerellaceae. The number of SBFS species per orchard varied from 2 to 15. Number of SBFS species and values of the Margalef and Shannon indexes were significantly (P < 0.05) lower in 21 orchards that had received conventional fungicide sprays during the fruit maturation period than in 14 unsprayed orchards. Several SBFS species, including Schizothyrium pomi, Peltaster fructicola, and Pseudocercosporella sp. RH1, were nearly ubiquitous, whereas other species, such as Stomiopeltis sp. RS5.2, Phialophora sessilis, and Geastrumia polystigmatis, were found only within restricted geographic regions. The results document that the SBFS complex is far more taxonomically diverse than previously recognized and provide strong evidence that SBFS species differ in geographic distribution. To achieve more efficient management of SBFS, it may be necessary to understand the environmental biology of key SBFS species in each geographic region.