TORONTO – Device positioning may help explain significant increases in pump thrombosis after left ventricular assist device implantation, according to a single-center study presented at the 2014 annual meeting of the American Association for Thoracic Surgery.
Dr. Jay Bhama, associate director of lung and heart transplantation at the University of Pittsburgh, found that more device-positioning issues coincided with more occurrences of pump thrombosis (PT). The adequacy of anticoagulation, major adverse events, and medical noncompliance were not found to be contributing factors, Dr. Bhama said.
His investigation joins recently published data indicating that left ventricular assist device (LVAD) thrombosis nearly quadrupled in less than 2 years in a multicenter study.
The purported mechanisms of PT in patients supported with the HeartMate II LVAD are thought to be multifactorial, but possibly related to design modifications, expansion of use to the destination therapy indication, nonuniform surgical implant technique, and nonuniform anticoagulation strategies across centers and over time.
"Starting in 2010, we started to notice a rapid and sudden increase in the rate of pump thrombosis, which has increased steadily over the last 3 years," said Dr. Bhama, who reported that PT occurred in 10 of 62 patients (16%) treated at the University of Pittsburgh Medical Center, with an overall event rate of 0.281 per patient-year.
In response to the increase, the group at the medical center investigated how potential contributing factors may have changed over time. They retrospectively assessed all primary LVAD implants in patients who survived hospitalization (62 of 74 total implants) between 2004 and 2012, grouping patients according to the era of implant: from June 2004 to December 2009 (era 1; n = 24) and from January 2010 (when FDA approval was given to expand use to destination therapy) to December 2012 (era 2; n = 38).
None of those who died during the index hospitalization experienced PT, Dr. Bhama noted.
PT was defined as either visualized thrombus within the pump at device exchange or significant hemolysis in the setting of heart failure symptoms or pump malfunction.
The actuarial freedom from PT at 24 months was significantly lower in era 2 than in era 1 (57% vs. 100%; P = .016).
Effective anticoagulation (percent of all international normalized ratio [INR] measurements greater than 1.8) was more reliably achieved in era 2 than in era 1 (50% vs. 34%; P less than .001).
To assess device positioning, the researchers looked at the angle of the inflow cannula, defining malposition as either less than the 5th or greater than the 95th percentile of the median of all the inflow cannula angles. Regarding the outflow cannula, they looked at patients who had bend-relief disconnects, either partial or complete, and those who had radiographic evidence of outflow graft malposition or kink.
Device positioning issues were significantly more prevalent during era 2 than during era 1 (29% vs. 4%). Most of this difference was driven by inflow cannula positioning problems, Dr. Bhama noted.
When the patients with concerns related to device positioning were excluded, the freedom from PT at 24 months no longer differed significantly between groups (P = .094).
The groups were demographically similar except for age, which was higher in the era 2 group (57 years, vs. 50 years for era 1; P = .037). More patients in the era 2 group received an LVAD for destination therapy, although this difference actually wasn’t significant (61% vs. 38%; P = .066).
The groups were also similar with regard to early major adverse events (right ventricular failure, bleeding, infection, and stroke) and medical noncompliance.
In the earlier multicenter study, Dr. Randall C. Starling, of the Cleveland Clinic, and his colleagues reported an abrupt increase in LVAD thrombosis: Between March 2011 and Jan. 1, 2013, the occurrence of PT at 3 months after implantation increased from 2.2% to 8.4% (N. Engl. J. Med. 2014;370:33-40).
"Dissecting the root cause of this problem is an extremely difficult task," said Dr. Nader Moazami, the invited discussant for Dr. Bhama’s presentation and the second author on the Starling paper. Dr. Moazami is surgical director of the Kaufman Center for Heart Failure at the Cleveland Clinic.
"While recent advances in LVAD technology with continuous-flow pumps have saved the lives of thousands of dying patients, issues related to adverse events and the associated morbidity and mortality are of immense importance, specifically as we consider the relevance of this technology to the more ambulatory heart failure patients," said Dr. Moazami, commenting on the study.
However, he suggested that the real cause of the recent increase in PT has not yet been discovered, and questioned the validity of assessing inflow cannula angulation based on a chest x-ray. "This to my knowledge has never been validated and was a concern in about half of the patients in the pump thrombosis group," Dr. Moazami said. Patients with "a demonstrable mechanical reason for pump thrombosis" were excluded from the Starling team’s analysis, he added.