Below is information about an article being published in the June 11 issue of Annals of Internal Medicine. The information is not intended to substitute for the full article as a source of information. Annals of Internal Medicine attribution is required for all coverage.
Two independent reanalyses of trial data, based on an unprecedented data release by Medtronic, Inc., indicate that the current evidence shows little benefit and potential harms are associated with the use of recombinant human bone morphogenetic protein-2 (rhBMP-2), an orthobiologic agent used to promote bone growth, in spinal surgery. As part of the Yale University Open Data Access (YODA) project, two teams of researchers at Oregon Health & Science University (OHSU) and the University of York were selected via an open, competitive process to perform independent analyses of all rhBMP-2 trial data. Medtronic, the medical device company that manufactures rhBMP-2, provided the clinical trial data and the funds for the studies to YODA. The rhBMP-2 trial data will now be available for general release to others who wish to conduct studies.
The research team at OHSU conducted a systematic review of individual patient data (IPD) from all Medtronic-sponsored studies, related internal documents, documents from the FDA, and other published research to assess the effectiveness and harms of rhBMP-2 in spinal fusion as compared to iliac crest bone graft (ICBG) or other bone grafts. The researchers also looked for reporting bias in the industry-sponsored journal publications. They found that in spinal fusion, rhBMP-2 had no efficacy advantage over ICBG, and was associated with an increased risk of cancer at 24 months though the overall cancer risk was low and many different types of cancer were included in the analysis. They cited lack of blinding of surgeons, patients, and outcome assessors as the main sources of bias for Medtronic studies. According to the OHSU reviewers, early journal publications underreported adverse events and emphasized results favorable to rhBMP-2.
Researchers at the Centre for Reviews and Dissemination (CRD), York reviewed IPD from all Medtronic-sponsored randomized, controlled trials (RCTs) and one other RCT, and extracted adverse event data from 35 published observational studies. The reviewers sought to determine the effectiveness and safety of rhBMP-2 for pain and spinal fusion compared to ICBG. They found that rhBMP-2 was associated with a small increase in fusion but greater immediate postoperative pain compared to ICBG. At two years, rh-BMP-2 offered no clinically important pain reduction and rhBMP-2 was associated with a possible increased risk for cancer. While rhBMP-2 recipients had nearly double the number of new cancers compared with ICBG recipients, the overall absolute risk for cancer was low in both groups. The investigators could not rule out a bias in pain assessment because participants were not blinded to the treatment received or their fusion status. The CRD team will publish results of their investigation of reporting bias separately.
‘Reviews highlight importance of evidence synthesis, data sharing, peer review, and reproducible research’
The OHSU and York systematic reviews of rhBMP-2 data were conducted separately and were reviewed independently by Annals of Internal Medicine. Annals’ review involved two internal teams of physician and statistical editors, and distinct sets of external reviewers. Neither team had access to the other manuscript until both were accepted for publication. Christine Laine, MD, MPH, Editor-in-Chief of Annals of Internal Medicine, writes that the fact that two independent groups armed with the same patient-level data arrived at essentially the same conclusions should greatly temper enthusiasm for rhBMP-2. She writes that beyond replicating results, this exercise demonstrates additional value in having different scientists tackle the same data. According to Dr. Laine, this first YODA project “is a novel exercise that illustrates the value of evidence synthesis, data sharing, peer review and editing, and reproducible research in helping us get closer to the truth.”
‘The role of rhBMP-2 in spinal surgery is still being defined’
According to Daniel Resnick, MD, MS, and Kevin J. Bozic, MD, orthopedic surgeons and the authors of an accompanying editorial, the findings of the OHSU and York reviews are important for guiding clinical decision making. In some procedures, where a second incision would be required for graft harvest, rhBMP-2 may be worth the very small risk for cancer. In other procedures, where the locally harvested graft and the ICBG are available through the same incision, rhBMP-2 may not be worth the small risk. Only in anterior cervical surgery do the authors suggest that rhBMP-2 not be used (unless there is a compelling reason) because of its higher complication rates. The authors caution that patients in the rhBMP-2 studies received autologous ICBG, the gold standard bone graft. Other graft materials could have lower fusion rates. They write that patients should be counseled on the relative benefits and harms of each option so they can actively participate in decision-making. Since rhBMP-2 is a new technology, the authors believe that its role in spinal surgery is still being defined. The two reviews, along with other data sources, will help physicians make the highest-value treatment decisions for their patients.
After a two-years-long process, the Yale University Open Data Access (YODA) project completes its first initiative with the publication of two independent reviews of Medtronic data on recombinant human bone morphogenetic protein-2 (rhBMP-2) in Annals of Internal Medicine. YODA seeks to address the problem of unpublished and selectively published clinical evidence. Medtronic is the first drug or medical device company to contract with YODA to allow access to all of its clinical trial data for independent reanalysis. According to an editorial authored by YODA project leader Harlan Krumholz, MD, SM and members of the YODA team, the Yale group invited Medtronic to be the first company to participate in the project to address questions that had arisen about the objectivity of Medtronic’s rhBMP-2 clinical trial results. Medtronic agreed to complete data transparency and the YODA team contracted with two independent research groups to reanalyze the data. Following peer-reviewed publication, Medtronic’s data will remain available to others who wish to address additional scientific questions. To address Medtronic’s and future participants’ concerns about data misuse, false positive inferences from data mining, legal ramifications, privacy issues, and commercial advantage for competitors, the YODA team developed a data release policy that will be followed for all projects. The YODA team writes that they hope that this project marks a major step toward a cultural shift in thinking about data sharing. According to them, publication of the rhBMP-2 reviews in Annals of Internal Medicine is “an historic moment in the emerging era of open science.”
Richard E. Kuntz, MD, MSc, Chief Scientific, Clinical & Regulatory Officer at Medtronic, Inc., writes that the company agreed to pioneer open access of industry-sponsored research through the YODA Project because it has a commitment to explore objective evaluation processes of Medtronic products. While open analyses may not reinforce the company’s own evaluations and conclusions, “all stakeholders should benefit from a transparent process that produces reliable information about the benefits and harms of the products being studied.” Dr. Kuntz writes that industry may move away from the paradigm in which a single research entity exclusively possesses and analyzes the data, however that traditional model does offer some consistency in analytic method and interpretation. Ultimately, open data access should lead to open discussion and exchange of information that can lead to collective agreement on the benefits and harms of medical products.
Older patients taking statins should not be prescribed the antibiotics clarithromycin or erythromycin due to increased risk for statin toxicity. Each year millions of patients are prescribed statins to treat dyslipidemia and prevent cardiovascular disease. While generally considered safe, the U.S. Food and Drug Administration (FDA) has cautioned prescribers and patients about a potential drug-drug interaction between commonly prescribed statins and medications used to treat HIV and hepatitis. The FDA also warned that antibiotics that inhibit cytochrome P450 isoenzyme 3A4, or CYP3A4 (a liver enzyme), may increase blood concentrations of statins that are metabolized by CYP3A4. The antibiotics clarithromycin and erythromycin are CYP3A4 inhibitors, but azithromycin is not. Researchers measured the frequency of statin toxicity for continuous statin users older than 65 years who were coprescribed clarithromycin (n=72,591) or erythromycin (n=3,267) compared with those who were prescribed azithromycin (n=68,478). They found that patients on statins who were also prescribed clarithromycin or erythromycin had a higher risk for hospitalization with rhabdomylosis, hospitalization with acute kidney injury, and all-cause mortality. The authors suggest that physicians avoid coprescribing these medications whenever possible.