Becker's Spine Review

22 Executive Briefing: Sponsored by The Evidence-Based Medicine Behind Vertebral Augmentation: Changing the Discussion Post-2009 I n 2009, The New England Journal of Med- icine published two studies, the INVEST and Australian trials, that cast doubt on the benefits of vertebral augmentation. 1,2 The analyses found no significant differences in pain relief between vertebroplasty and a sham intervention. These studies fueled confusion among providers, leaving many in the industry unsure of the potential benefits associated with vertebral augmentation. Patients with a VCF have a five-fold increased risk of suffering a subsequent vertebral frac- ture compared with their pre-morbid condi- tion or age matched controls. 3, 4 Each addition- al VCF increases a patient's mortality risk. 5 Several recent large clinical studies followed for at least 12 months after vertebral com- pression fracture (VCF) have concluded that mortality rates following VCFs are significantly higher for patients treated conservatively ver- sus VP or BKP, while other studies have con- cluded no difference. For more information, visit www.medtronic.com/bkpmortality. Since 2009, six retrospective claims studies have been published, where researchers fo- cused on mortality risk of balloon kyphop- lasty and vertebroplasty compared to that of non-surgical management. 1. The 2017 Ong et al. analysis investigated if VCF patients were at a higher risk of mor- tality in the years following the publication of the 2009 trials. The study included more than 2 million patients, broken down as follows: 261,756 BKP patients, 117,232 VP patients and 1,698,956 NSM patients. 6 Overall, the pro- pensity-adjusted 10-year mortality risk for the VCF population was 85.1 percent; procedure comparisons at 10-year follow up showed: • 24 percent higher mortality risk for NSM ver- sus BKP, p<0.001 • 8 percent higher mortality risk for NSM ver- sus VP, p<0.001 • 13 percent lower mortality risk for BKP versus VP, p<0.001 2. The Edidin et al. 2015 study examined a total of 1,038,956 VCF patients with up to four years follow up. Of the patients, 141,343 patients underwent BKP and 75,364 under- went VP. The non-operated patients had a 55 percent higher propensity-adjusted mortal- ity risk (p<0.001) than the BKP patients and a 25 percent higher mortality risk (p<0.001) than the VP patients. Researchers found the non-operated patients experienced signifi- cantly higher adjusted risks of pneumonia, myocardial infarction/cardiac complications, deep vein thrombosis and urinary tract infec- tion than the BKP patients experienced. Also, the non-operated group had lower adjusted risks of subsequent augmentation/fusion, sub- sequent augmentation and pulmonary/respi- ratory complications. 7 3. The 2014 Lange et al. study analyzed 3,607 patients with osteoporotic VCFs, of which 598 underwent BKP or VP with a five-year follow up. Using propensity score matching, researchers found patients in the operated group were 43 percent less likely to die compared to the non-operated cohort (p<0.001). Further, those patients receiving BKP had a 66.7 percent 60-month adjusted survival rate compared to the 58.7 percent survival rate for VP (p=0.68). 8 4. In 2013, McCullough et al. studied one-year mortality risk among 10,541 augmented pa- tients and 115,851 NSM patients. The study, utilizing a 20 percent sample of Medicare data, found mortality was lower in the aug- mented group (5.2 percent) than in the control group (6.7 percent) using traditional covariate adjustments (p<0.001). After propensity score matching, accounting for selection bias, how- ever, the researchers found no significant dif- ference between the augmented (5.6 percent) and control patients (5.2 percent) for one-year morality (p=0.18). 9 5. The 2013 Chen et al. study compared VP, BKP and non-operated management for VCF patients with total follow up of 129,783 per- son-years. The study involved 68,752 VCF pa- tients broken into the following cohorts: 55.6 percent nonoperative, 11.2 percent VP and 33.2 percent BKP. The study demonstrated kyphoplasty yielded longer three-year patient survival at 59.9 percent compared with VP at 49.7 percent and non-operated treatment at 42.3 percent (p<0.001). 10 6. The 2011 Edidin A. et al. study analyzed Medi- care data from 2005 to 2008 to assess the mor- tality risk for VCF patients receiving non-operat- ed management, BKP or VP with up to four years follow up. Of the 858,978 total patients newly diagnosed with VCFs, 13.9 percent (119,253 patients) received BKP and 7.4 percent (63,693 patients) received VP. Researchers found adjust- ed survival rates of 60.8 percent in the operated cohort to be higher compared to 50 percent in the non-operated cohort (p<0.001). In compar- ing the operated subgroups, survival of 57.3 percent for VP patients, was lower than 62.8 per- cent for BKP patients (p<0.001). 11 Despite much clinical evidence supporting vertebral augmentation, 12,13 some providers strayed away from the procedures. For exam- ple, vertebral augmentation volume fell from 24 percent in 2009 to 14 percent in 2014 (p<0.001), according to the 2017 analysis. 6 Payers also harnessed the two 'sham' trials to deny payment for vertebral augmentation. In the U.S., a Medicare contractor published neg- ative coverage, later revising that policy with restrictive requirements. "The 2009 studies caused significant confu- sion as to the effectiveness of the procedure and what to do with the patient," says Douglas Beall, MD, of Edmond-based Oklahoma Spine Hospital. Josh Hirsch, MD, Boston-based Massachusetts General Hospital's Neuro Interventional Ra- diology director and the American Society of Spine Radiology's immediate past president, noted, "Trials should inform our decision-mak- ing, not make the decision for us. That is the key to patient-centric care." Did the 2009 studies shift VCF treatment patterns? The most recent of the retrospective claims studies on VCFs and mortality risk, the 2017 Ong KL, et al. analysis is titled, "Were VCF Patients at Higher Risk of Mortality Following the 2009 Publication of the Vertebroplasty 'Sham' Trials?" Medtronic funded the analysis and offered minimal input into the study's de- sign, but did not engage in the data collection, management, analysis or interpretation. Kevin Ong, PhD, PE, principal engineer of Ex- ponent and adjunct member of Drexel Uni- versity's School of Biomedical Engineering in Philadelphia, served as principal researcher for the Exponent-led study. Drs. Beall, Hirsch and others also served as researchers. "[In addition to other factors,] we wanted to ex- amine whether the publication of the 2009 'sham' control studies resulted in lower use of vertebral augmentation. We found that, in the five-year pe- riod following 2009, patients had greater mortal- ity risk for VCF patients," says Dr. Ong. Investigators used Medicare data to examine U.S. vertebral augmentation rates between 2010 and 2014 compared to the previous five-year period. The study examined the dif- ference in mortality and morbidity of VCF pa- tients pre- and post-2009 and compared risks between vertebral augmentation and NSM. In addition to the mortality risk findings listed above, the analysis revealed the following key data points: 1. Vertebral augmentation patients accounted for 20 percent of VCF patients in 2005, peak- ing at 24 percent between 2007 and 2008 be- fore dropping to 14 percent in 2014.

• Cover