Duke Translational Medicine Institute
Duke Translational Medicine Institute
Califf R.M.,Duke Translational Medicine Institute |
Zarin D.A.,U.S. National Institutes of Health |
Kramer J.M.,Duke Translational Medicine Institute |
Kramer J.M.,Duke Clinical Research Institute |
And 3 more authors.
JAMA - Journal of the American Medical Association | Year: 2012
Context: Recent reports highlight gaps between guidelines-based treatment recommendations and evidence from clinical trials that supports those recommendations. Strengthened reporting requirements for studies registered with ClinicalTrials.gov enable a comprehensive evaluation of the national trials portfolio. Objective: To examine fundamental characteristics of interventional clinical trials registered in the ClinicalTrials.gov database. Methods: A data set comprising 96 346 clinical studies from ClinicalTrials.gov was downloaded on September 27, 2010, and entered into a relational database to analyze aggregate data. Interventional trials were identified and analyses were focused on 3 clinical specialties - cardiovascular, mental health, and oncology - that together encompass the largest number of disability-adjusted life-years lost in the United States. Main Outcome Measures: Characteristics of registered clinical trials as reported data elements in the trial registry; how those characteristics have changed over time; differences in characteristics as a function of clinical specialty; and factors associated with use of randomization, blinding, and data monitoring committees (DMCs). Results: The number of registered interventional clinical trials increased from 28 881 (October 2004- September 2007) to 40 970 (October 2007-September 2010), and the number of missing data elements has generally declined. Most interventional trials registered between 2007 and 2010 were small, with 62% enrolling 100 or fewer participants. Many clinical trials were single-center (66%; 24 788/37 520) and funded by organizations other than industry or the National Institutes of Health (NIH) (47%; 17 592/37 520). Heterogeneity in the reported methods by clinical specialty; sponsor type; and the reported use of DMCs, randomization, and blinding was evident. For example, reported use of DMCs was less common in industry-sponsored vs NIH-sponsored trials (adjusted odds ratio [OR], 0.11; 95% CI, 0.09-0.14), earlier-phase vs phase 3 trials (adjusted OR, 0.83;95% CI, 0.76-0.91), and mental health trials vs those in the other 2 specialties. In similar comparisons, randomization and blinding were less frequently reported in earlier-phase, oncology, and device trials. Conclusion: Clinical trials registered in ClinicalTrials.gov are dominated by small trials and contain significant heterogeneity in methodological approaches, including reported use of randomization, blinding, and DMCs. ©2012 American Medical Association. All rights reserved.
Leonardi S.,Pratt Institute |
Thomas L.,Pratt Institute |
Neely M.L.,Pratt Institute |
Tricoci P.,Pratt Institute |
And 8 more authors.
Journal of the American College of Cardiology | Year: 2012
Objectives: This study compared prognoses of myocardial infarction related to percutaneous coronary intervention (PCI, procedural MI) using increasing creatine kinase-myocardial band (CK-MB) thresholds with spontaneous MI. Background: Procedural MI usually is defined by a CK-MB elevation of more than 3 times the upper limit of normal (ULN), but higher thresholds have been proposed. Methods: Patients from the EARLY-ACS (Early Glycoprotein IIb/IIIa Inhibition in Non-ST-Segment Elevation Acute Coronary Syndrome) study and the SYNERGY (Superior Yield of the New strategy of Enoxaparin, Revascularization and GlYcoprotein IIb/IIIa inhibitors) study treated with PCI were included. The primary end point was 1-year all-cause mortality from 24 h after PCI. To determine an enzymatic threshold for procedural MI with a prognosis similar to that of spontaneous MI, we redefined procedural MI using increasing CK-MB thresholds and compared corresponding hazard ratios with those of spontaneous MI (CK-MB more than twice the ULN). Hazard ratios for mortality for procedural and spontaneous MI were calculated using Cox proportional hazards regression and Global Registry of Acute Cardiac Events covariates for risk adjustment. Results: Nine thousand eighty-seven patients who underwent PCI (46.8%) were included; 773 procedural MI and 239 spontaneous MI occurred within 30 days. Adjusted hazard ratios for 1-year death were 1.39 (95% confidence interval [CI]: 1.01 to 1.89) for procedural MI and 5.37 (95% CI: 3.90 to 7.38) for spontaneous MI. The CK-MB threshold for procedural MI that achieved the same prognosis as spontaneous MI was 27.7 times the ULN (95% CI: 13.9 to 58.4), but this differed between the SYNERGY study (57.9 times the ULN, 95% CI: 17.9 to 63.6) and the EARLY-ACS study (20.4 times the ULN, 95% CI: 5.16 to 24.2). Of all procedural MI, 49 (6%) had CK-MB elevations of 27.7 or more times the ULN. Conclusions: The current enzymatic definition of procedural MI (CK-MB more than 3 times the ULN) used in clinical trials is less strongly associated with death than that of spontaneous MI. Procedural MI achieves similar prognosis for 1-year mortality when much higher CK-MB thresholds are applied. © 2012 American College of Cardiology Foundation.
News Article | September 7, 2016
In 1996, Kathy Giusti was diagnosed with multiple myeloma, a rare and often fatal cancer. Her first bone-marrow biopsy took place on a Friday night. Outside the room, a group of scientists waited with an ice chest to take her bone-marrow sample. She told her doctor that she was surprised to see them working late. “He said my tissue was precious,” she recalls. For precision medicine to live up to its potential, millions of people must share their genomic data, their health records, and their experiences. To researchers, all of it is precious. The richer the databases, the better patient care will become. This need gives ordinary people more power in medical research — not only to improve research quality by participating in greater numbers, but by speaking up and influencing what questions are asked in the first place. Despite lingering concerns over privacy (see page S70), it is patients and their loved ones who have been pushing for changes to the medical system that will enable personalized medicine, says Guisti, who founded the Multiple Myeloma Research Foundation with her sister in 1998. For individual patients, empowerment takes a lot of time, work, education and economic resources. If programmes such as the United States' planned million-volunteer Precision Medicine Initiative Cohort Program are to succeed (see page S69), they must build people's trust and bring in not only highly motivated, seriously ill people, but healthy volunteers too. Early patient-centred research projects are already showing that this can pay off for researchers, for drug companies and, more importantly, for patients. It can be difficult for even well-educated, financially secure people living close to major US medical centres to access the best medical care. In 1998, Marty Tenenbaum was diagnosed with melanoma that had metastasized to his liver, “which in those days had zero survivors”, he says. After a lot of searching, he was lucky enough not just to get onto a trial for an experimental therapy and surgery, but to be one of the few who responded. The trial failed, but Tenenbaum is still in remission. This experience, and the genome-sequencing boom, led Tenenbaum — a computer scientist and former professor at Stanford University — to found a personalized-medicine consultancy called CollabRX in 2008. The company used bioinformatics to suggest therapies to its wealthy customers at a cost of US$35,000–50,000. “My vision was to use information technology to close the loop between cancer research and clinical care,” he says. When CollabRX was acquired by Tegal in 2012, Tenenbaum wanted to start a non-profit organization that could serve more people. So he set up Cancer Commons, whose goal is to make the expertise of the best clinicians available to more cancer patients — especially those who cannot afford to travel to the best medical centres or pay for a personal consultation. Tenenbaum sees buried treasure in the scientific discussions that take place behind closed doors on tumour boards — the groups of doctors, geneticists and researchers who discuss individual cases and decide on the best course of therapy. There are millions of possible combinations of drugs, many more than could ever be tested in clinical trials. Instead, the best oncologists and tumour boards are in effect experimenting on their patients, says Tenenbaum, trying new drug cocktails and seeing what happens. Yet little is learned: the deliberations and failed hypotheses of the tumour boards are not included in individual patients' records, and none of the data are shared outside the hospital. “There's all this experimentation and no learning,” he says. “Every patient presents a vastly complicated data set that we're barely able to interpret,” says C. Anthony Blau, an oncologist at the University of Washington in Seattle who specializes in finding therapies for people with difficult-to-treat breast cancer. The ability to search a large pool of data on what has been tried with others would help oncologists to find the right treatment faster. Tenenbaum is promoting this data-sharing vision through an online portal called Ask Cancer Commons. Patients or their care-givers can upload whatever medical records or genetic tests they have and give a description of their case. A group drawn from more than 100 volunteer oncologists and geneticists, including Blau, then reviews the case and gives feedback, serving as a virtual tumour board. In the short term, Tenenbaum hopes that this will provide a lifeline for people who cannot access top-quality cancer care locally. But as more people take part, the database will grow. Patient data and the reports of the virtual tumour board are fed into a database that doctors can use to help future patients. Cancer Commons follows up to find out what the doctors did with their feedback, and how well it worked. Hospitals are also contributing to the growth of the database. In a pilot test in 2015, 50 cases from three tumour boards were summarized by volunteers and verified by doctors. Eventually, Cancer Commons will have enough of these hand-annotated data to construct algorithms that are capable of picking out important information from complex patient records and tumour-board deliberations without the need for volunteers at all. The project, known as the Insight Network, is currently fundraising and is expected to shift into full gear in the next few years. As such projects grow, the need for people to take charge of their own care will diminish. “Avid patients will lead the way,” says Blau. “The knowledge gained through them will be applicable to the population as a whole.” One of the largest online data-sharing health projects, PatientsLikeMe, began in 2004 with a focus on neurological disorders. Since then, it has expanded to include 2,500 diseases. PatientsLikeMe now has about 500,000 users, most of them seriously ill. Through the website, these people can track their symptoms, join discussions and complete research surveys. Jamie Heywood, a mechanical engineer and co-founder of the site PatientsLikeMe, calls it “a prospective epidemiology platform”. So far, PatientsLikeMe has published more than 75 studies, mostly in collaboration with academic or corporate researchers. It has partnerships with the US Food and Drug Administration and drug company AstraZeneca among others, and it is financed by sharing patients' data with drug companies and researchers. Heywood's younger brother, Stephen, was diagnosed with amyotrophic lateral sclerosis (ALS) in 1998 at the age of 29, spurring Jamie to start the ALS Therapy Development Institute. It took the unusual approach of publishing results in real time as it screened drugs in mice and conducted a stem-cell trial with three participants, including Stephen. People with a serious disease can feel very alone, says Heywood. They want to be treated like partners, not subjects, and this is what PatientsLikeMe tries to do. Sharing data and experiences that may help others, and knowing that they are going to help researchers and drug companies, can make people feel heard and empowered ( et al. Patient http://doi.org/bpqw; 2016). “My brother died eight years ago and he's still helping people,” says Heywood. To build trust and encourage people to share data, the site is designed to be as accessible as possible. Instead of medical terms, patients use phrases such as 'brain freeze' to describe how they feel; the vernacular is then automatically matched with the standard medical code. The company also gets to know its users and adapts how it interacts with them accordingly. Some need to make a decision quickly after a diagnosis, for instance, whereas others with degenerative diseases have more time. Unlike conventional top-down studies, in which data are collected on rigid timetables, PatientsLikeMe offers its users the flexibility to add data whenever they want, but this makes it less statistically rigorous than traditional research. “Real-world observational studies will never be able to match double-blind prospective studies in their ability to examine causality,” says Heywood, and understanding biases in data contributed by users is a big challenge for epidemiologists and others at the company. But when patients are in charge, he says, they supply data that researchers do not normally have access to. As an example of how patient-centred research can yield insights that benefit both patients and drug companies, Heywood points to a collaboration on insomnia between PatientsLikeMe, Northwestern University in Evanston, Illinois, and US drug company Merck & Co. After developing a new sleep drug, surovexant, Merck approached PatientsLikeMe and asked it to look at the sleep patterns of members. Based on an initial survey of 75,000 users, Heywood says that his organization came up with about 50 hypotheses. The team then narrowed both the set of questions and the study group. The resulting survey of just over 5,000 users in 2013 showed that only 13% had been diagnosed with insomnia, but that 73% of those who were undiagnosed also reported symptoms ( et al. Sleep Med. 16, 1332–1341; 2015). The data suggest that many people with serious illnesses have trouble sleeping — something that can exacerbate their condition, and doctors should be aware of the need to manage it. As the five-year study continues, participants will receive information about their individual sleep patterns. Another of the site's self-tracking tools has been particularly useful for Allison Silensky, who has been using PatientsLikeMe since 2008. Silensky, who has a form of bipolar disorder, is a member of the company's user advisory board. The site's mood tracker has helped Silensky to notice and remember trends in her mood that she might otherwise have neglected to mention to her doctor. If she feels great at the doctor's office, she says, “I don't realize that the three weeks prior were horrible because I'm living in the moment.” At first her doctors warned her against getting involved with something on the Internet, she says, but they soon came to see the benefit. After a few years, Silensky saw a trend that she had not noticed before: all her hospitalizations were in spring. Now her doctor adjusts her medication in January, and her therapist checks on her more frequently in spring. She has not been hospitalized since. People with serious and rare diseases, and their families, may be highly motivated to participate in data-sharing projects. But personalized medicine also needs healthy volunteers if researchers are to understand how diseases emerge. “This is a national movement, and we need everyone to participate,” says Bray Patrick-Lake, director of patient engagement at the Duke Translational Medicine Institute in Durham, North Carolina. Research and anecdote alike suggest that the possibility of helping others motivates people to share their data. Unpublished findings from a survey funded by the US National Institutes of Health suggest that proponents of precision medicine will have to convince people that they are contributing to the public good by sharing their data. Sandra Soo-Jin Lee, a biomedical ethicist, is looking at how diverse communities feel about projects that link electronic health records with biobanks for research. Her group at the Stanford Center for Biomedical Ethics in California is still analysing the results, which are based on surveys of 20 focus groups, including Hispanic, Asian and African American people, but already she has found “a tension” in people's attitudes. There is excitement that the data might lead to fresh targeted therapies or discoveries that are possible only with large pools of data. But some worry that the information will be used by the government for non-medical purposes, and others, says Lee, simply feel “a loss of control”. And many are nervous about who will profit. “There's concern about who's actually going to use the data,” she says — particularly that a third party will use the information to develop an expensive therapy. Donating data and tissue for the public good is one thing, but often the benefits are not distributed equally. The HeLa line of immortal cancer cells, for instance, derived without consent from the ovarian tumour of African American woman Henrietta Lacks, has long been a workhorse for cancer researchers. But the benefits of this research have not been distributed equally: cancer mortality rates for African Americans are still higher than those of any other ethnic group in the United States. “In a fundamentally unequal health system, it's harder to argue that everyone should share,” says Barbara Koenig, a medical anthropologist at the University of California, San Francisco. She studies the limits of informed consent in large public data-sharing projects. Without automatic enrolment for enterprises that serve the public good, most people will be motivated to opt into data sharing only when tragedy strikes in the form of a diagnosis for themselves or a loved one, says Koenig. “If people know they will benefit, they will share.” Heywood thinks that the people behind large government projects are misguided if they believe that “because they're trying to do the world good, the world will follow them”. Building trust with people takes time, he says. Continuing to carry out “onerous, top-down recruitment for clinical studies” is not working, says Sharon Terry, chief executive of the Genetic Alliance, a non-profit health advocacy organization. One of its projects, PEER, is an online resource that not only allows patients to make their information available to researchers, but also gives them control over how much of their health information is shared, and with whom, on a case-by-case basis. “Our experiment is to use the tools of social media to engage people,” says Terry. She thinks that the health-care industry does not do outreach as well as community organizations. So the Genetic Alliance is trying to learn from people in education and social services. Today, precision-medicine portals, whether patient-driven or not, are fragmented, and that can make it hard to reach people, admits Giusti. But once people believe they are “on the path to a cure”, they want to participate, she says. The progress made by the Multiple Myeloma Research Foundation towards treating that particular cancer is due in part to the patients who willingly donated their tissue to a biobank in the hope of accelerating research. That project, called CoMMpass, was launched in 2011, and since then has validated several drug targets and treatment strategies. It has shown, for instance, that people treated with a combination of three drugs live longer without the disease progressing than those who are given only two. But these projects are expensive: CoMMpass cost more than $40 million. “We struggle to see how this can be sustained,” says Terry. Large national, and ideally international, projects are the only way to make precision medicine work, says Kathryn North, leader of the Australian Genomics Health Alliance — and most people in the field accept that, she adds. Projects such as the US Precision Medicine Initiative are a great start, but they are only a start. There are economic, political and bureaucratic barriers to overcome, and probably only patients can make it happen. “The biggest advocates for this are the patient groups, because they can see how it transforms health care,” says North. The question facing patient advocates who want to see personalized medicine, says Terry, is this: “Can we impact the culture of large academic institutions, behemoth drug companies, and staid federal agencies?” Even if they want to change, there must be incentives driving them to do so. “I keep banking on public pressure, interest and rallying,” she says. If precision medicine one day comes to benefit broad swathes of the population, it may well be thanks to a few patients who took it on themselves to push for that kind of future.
Hirsch B.R.,Duke Cancer Institute |
Hirsch B.R.,Duke Clinical Research Institute |
Califf R.M.,Duke Translational Medicine Institute |
Cheng S.K.,Oregon Health And Science University |
And 7 more authors.
JAMA Internal Medicine | Year: 2013
Importance: Clinical trials are essential to cancer care, and data about the current state of research in oncology are needed to develop benchmarks and set the stage for improvement. Objective: To perform a comprehensive analysis of the national oncology clinical research portfolio. Design: All interventional clinical studies registered on ClinicalTrials.gov between October 2007 and September 2010 were identified using Medical Subject Heading terms and submitted conditions. They were reviewed to validate classification, subcategorized by cancer type, and stratified by design characteristics to facilitate comparison across cancer types and with other specialties. Results: Of 40 970 interventional studies registered between October 2007 and September 2010, a total of 8942 (21.8%) focused on oncology. Compared with other specialties, oncology trials were more likely to be single arm (62.3% vs 23.8%; P < .001), open label (87.8% vs 47.3%; P < .001), and nonrandomized (63.9% vs 22.7%; P < .001). There was moderate but significant correlation between number of trials conducted by cancer type and associated incidence and mortality (Spearman rank correlation coefficient, 0.56 [P =.04] and 0.77 [P =.001], respectively). More than one-third of all oncology trials were conducted solely outside North America. Conclusions and Relevance : There are significant variations between clinical trials in oncology and other diseases, as well as among trials within oncology. The differences must be better understood to improve both the impact of cancer research on clinical practice and the use of constrained resources. ©2013 American Medical Association. All rights reserved.
Mehta R.H.,Duke University |
O'Shea J.C.,Bon Secours Hospital |
Stebbins A.L.,Duke University |
Granger C.B.,Duke University |
And 6 more authors.
Journal of the American College of Cardiology | Year: 2011
Objectives The purpose of this study was to examine the association between lower socioeconomic status (SES), as ascertained by years of education, and outcomes in patients with acute ST-segment elevation myocardial infarction (STEMI). Background Previous studies have shown an inverse relationship between SES and coronary heart disease and mortality. Whether a similar association between SES and mortality exists in STEMI patients is unknown. Methods We evaluated 11,326 patients with STEMI in the GUSTO-III (Global Use of Strategies to Open Occluded Coronary Arteries) trial study from countries that enrolled >500 patients. We evaluated clinical outcomes (adjusted using multivariate regression analysis) according to the number of years of education completed. Results One-year mortality was inversely related to years of education and was 5-fold higher in patients with <8 years compared with those with >16 years of education (17.5% vs. 3.5%, p < 0.0001). The strength of the relationship between education and mortality varied among different countries. Nonetheless, years of education remained an independent correlate of mortality at day 7 (hazard ratio per year of increase in education: 0.86; 95% confidence interval: 0.83 to 0.88) and also between day 8 and 1 year (hazard ratio per year of increase in education: 0.96; 95% confidence interval: 0.94 to 0.98), even after adjustment for baseline characteristics and country of enrollment. Conclusions When the number of years of education was used as a measure of SES, there was an inverse relationship such that significantly higher short-term and 1-year mortality existed beyond that accounted for by baseline clinical variables and country of enrollment. Future studies should account for and investigate the mechanisms underlying this link between SES and cardiovascular disease outcomes. © 2011 American College of Cardiology Foundation.
Syed A.,Stanford University |
Garcia M.A.,Stanford University |
Lyu S.-C.,Stanford University |
Bucayu R.,Stanford University |
And 8 more authors.
Journal of Allergy and Clinical Immunology | Year: 2014
Background The mechanisms contributing to clinical immune tolerance remain incompletely understood. This study provides evidence for specific immune mechanisms that are associated with a model of operationally defined clinical tolerance. Objective Our overall objective was to study laboratory changes associated with clinical immune tolerance in antigen-induced T cells, basophils, and antibodies in subjects undergoing oral immunotherapy (OIT) for peanut allergy. Methods In a phase 1 single-site study, we studied participants (n = 23) undergoing peanut OIT and compared them with age-matched allergic control subjects (n = 20) undergoing standard of care (abstaining from peanut) for 24 months. Participants were operationally defined as clinically immune tolerant (IT) if they had no detectable allergic reactions to a peanut oral food challenge after 3 months of therapy withdrawal (IT, n = 7), whereas those who had an allergic reaction were categorized as nontolerant (NT; n = 13). Results Antibody and basophil activation measurements did not statistically differentiate between NT versus IT participants. However, T-cell function and demethylation of forkhead box protein 3 (FOXP3) CpG sites in antigen-induced regulatory T cells were significantly different between IT versus NT participants. When IT participants were withdrawn from peanut therapy for an additional 3 months (total of 6 months), only 3 participants remained classified as IT participants, and 4 participants regained sensitivity along with increased methylation of FOXP3 CpG sites in antigen-induced regulatory T cells. Conclusion In summary, modifications at the DNA level of antigen-induced T-cell subsets might be predictive of a state of operationally defined clinical immune tolerance during peanut OIT. © 2013 American Academy of Allergy, Asthma & Immunology.
Fiuzat M.,Duke University |
Califf R.M.,Duke University |
Califf R.M.,Duke Translational Medicine Institute
Heart Failure Clinics | Year: 2011
Clinical trials are often conducted globally. Differences in standard of care, patient populations including genetic and phenotypic differences, disease etiologies, rates of comorbidities, ascertainment of endpoints, and differences in concomitant therapies and medical culture may influence subsequent outcomes. There has been little consensus on how clinical trial results should be evaluated. This article reviews the differences in cardiovascular trial results by geographic region, offers potential explanations for these differences, and suggests methods for standardization of trial results. © 2011 Elsevier Inc.
Vickery B.P.,University of North Carolina at Chapel Hill |
Scurlock A.M.,University of Arkansas for Medical Sciences |
Kulis M.,University of North Carolina at Chapel Hill |
Steele P.H.,University of North Carolina at Chapel Hill |
And 15 more authors.
Journal of Allergy and Clinical Immunology | Year: 2014
Background Although peanut oral immunotherapy (OIT) has been conclusively shown to cause desensitization, it is currently unknown whether clinical protection persists after stopping therapy. Objective Our primary objective was to determine whether peanut OIT can induce sustained unresponsiveness after withdrawal of OIT. Methods We conducted a pilot clinical trial of peanut OIT at 2 US centers. Subjects age 1 to 16 years were recruited and treated for up to 5 years with peanut OIT. The protocol was modified over time to permit dose increases to a maximum of 4000 mg/d peanut protein. Blood was collected at multiple time points. Clinical end points were measured with 5000-mg double-blinded, placebo-controlled food challenges once specific criteria were met. Results Of the 39 subjects originally enrolled, 24 completed the protocol and had evaluable outcomes. Twelve (50%) of 24 successfully passed a challenge 1 month after stopping OIT and achieved sustained unresponsiveness. Peanut was added to the diet. At baseline and the time of challenge, such subjects had smaller skin test results, as well as lower IgE levels specific for peanut, Ara h 1, and Ara h 2 and lower ratios of peanut-specific IgE/total IgE compared with subjects not passing. There were no differences in peanut IgG4 levels or functional activity at the end of the study. Conclusions This is the first demonstration of sustained unresponsiveness after peanut OIT, occurring in half of subjects treated for up to 5 years. OIT favorably modified the peanut-specific immune response in all subjects completing the protocol. Smaller skin test results and lower allergen-specific IgE levels were predictive of successful outcome. © 2013 American Academy of Allergy, Asthma & Immunology.
Haley N.R.,Duke Translational Medicine Institute
Regenerative Medicine | Year: 2011
The Phacilitate Cell and Gene Therapy Forum was a showcase of the steps required for translational medicine. The major focus was on requirements from early phase discovery through to the regulatory and business processes to bring treatments to patients. Talks on discovery and new methodologies and technologies were presented in light of adaptability to treatment regimens. Venture capitalists and bank representatives, external evaluators of these new treatments, gave valuable financial process insight. US FDA officials offered explanation and advice on successful approaches to the regulatory challenges to this new and emerging field. Completing the picture were talks from contract suppliers and manufacturers of necessary materials and equipment. © 2011 Future Medicine Ltd.
Hess C.N.,Duke Clinical Research Institute |
Lopes R.D.,Duke Clinical Research Institute |
Gibson C.M.,Harvard University |
Hager R.,Duke Clinical Research Institute |
And 7 more authors.
Circulation | Year: 2014
Background: Coronary artery bypass grafting success is limited by vein graft failure (VGF). Understanding the factors associated with VGF may improve patient outcomes. Methods and Results: We examined 1828 participants in the Project of Ex Vivo Vein Graft Engineering via Transfection IV (PREVENT IV) trial undergoing protocol-mandated follow-up angiography 12 to 18 months post-coronary artery bypass grafting or earlier clinically driven angiography. Outcomes included patient- and graft-level angiographic VGF (≥75% stenosis or occlusion). Variables were selected by using Fast False Selection Rate methodology. We examined relationships between variables and VGF in patient- and graft-level models by using logistic regression without and with generalized estimating equations. At 12 to 18 months post-coronary artery bypass grafting, 782 of 1828 (42.8%) patients had VGF, and 1096 of 4343 (25.2%) vein grafts had failed. Demographic and clinical characteristics were similar between patients with and without VGF, although VGF patients had longer surgical times, worse target artery quality, longer graft length, and they more frequently underwent endoscopic vein harvesting. After multivariable adjustment, longer surgical duration (odds ratio per 10-minute increase, 1.05; 95% confidence interval, 1.03-1.07), endoscopic vein harvesting (odds ratio, 1.41; 95% confidence interval, 1.16-1.71), poor target artery quality (odds ratio, 1.43; 95% confidence interval, 1.11-1.84), and postoperative use of clopidogrel or ticlopidine (odds ratio, 1.35; 95% confidence interval, 1.07-1.69) were associated with patient-level VGF. The predicted likelihood of VGF in the graft-level model ranged from 12.1% to 63.6%. Conclusions: VGF is common and associated with patient and surgical factors. These findings may help identify patients with risk factors for VGF and inform the development of interventions to reduce VGF.