PHOENIX (FRONTLINE MEDICAL NEWS) – Cardiothoracic surgeons who implant left ventricular assist devices in patients with failing hearts remain at a loss to fully explain why they started seeing a sharp increase in thrombus clogging in these devices in 2012, but nevertheless they are gaining a better sense of how to minimize the risk.

Three key principles for minimizing thrombosis risk are selecting the right patients to receive left ventricular assist devices (LVAD), applying optimal management strategies once patients receive a LVAD, and maintaining adequate flow of blood through the pump, Dr. Francis D. Pagani said in a talk at a session devoted to pump thrombosis at the annual meeting of the Society of Thoracic Surgeons.

Other critical aspects include optimal implantation technique, quick work-up of patients to rule out reversible LVAD inflow or outflow problems once pump thrombosis is suspected, and ceasing medical therapy of the thrombosis if it proves ineffective and instead progress to surgical pump exchange, pump explantation, or heart transplant when necessary, said Dr. Ahmet Kilic , a cardiothoracic surgeon at the Ohio State University, Columbus.

Another key issue is that, now that the pump thrombosis incidence is averaging about 10% of LVAD recipients, with an incidence rate during 2-year follow-up as high as 24% reported from one series, surgeons and physicians who care for LVAD patients must have a high index of suspicion and routinely screen LVAD recipients for early signs of pump thrombosis. The best way to catch pump thrombosis early seems to be by regularly measuring patients’ serum level of lactate dehydrogenase (LDH), said Dr. Robert L. Kormos , professor of surgery and director of the artificial heart program at the University of Pittsburgh.

“We measure LDH on most clinic visits, whether or not the patient has an indication of pump thrombosis. We need to screen [LDH levels] much more routinely than we used to,” he said during the session. “Elevated LDH is probably the first and most reliable early sign, but you need to also assess LDH isoenzymes because we’ve had patients with an elevation but no sign of pump thrombosis, and their isoenzymes showed that the increased LDH was coming from their liver,” Dr. Kormos said in an interview.

Although serial measurements and isoenzyme analysis can establish a sharp rise in heart-specific LDH in an individual patient, a report at the meeting documented that in a series of 53 patients with pump thrombosis treated at either of two U.S. centers, an LDH level of at least 1,155 IU/L flagged pump thrombosis with a fairly high sensitivity and specificity. This LDH level is roughly five times the upper limit of normal, noted Dr. Pagani, professor of surgery and surgical director of adult heart transplantation at the University of Michigan, Ann Arbor, and a senior author on this report.

But prior to this report Dr. Kormos said that he regarded a LDH level of 600-800 IU/L as enough of an elevation above normal to prompt concern and investigation. And he criticized some LVAD programs that allow LDH levels to rise much higher.

“I know of clinicians who see a LDH of 1,500-2,000 IU/L but the patient seems okay and they wonder if they should change out the pump. For me, it’s a no brainer. Others try to list a patient like this for a heart transplant so they can avoid doing a pump exchange. I think that’s dangerous; it risks liver failure or renal failure. I would not sit on any LVAD that is starting to produce signs of hemolysis syndrome, but some places do this,” Dr. Kormos said in an interview.

“Pump thrombosis probably did not get addressed in as timely a fashion as it should have been” when it was first seen on the rise in 2012, noted Dr. James K. Kirklin , professor of surgery and director of cardiothoracic surgery at the University of Alabama, Birmingham. “It is now being addressed, and we realize that this is not just a pump problem but also involves patient factors and management factors that we need to learn more about. We are quite ignorant of the patient factors and understanding their contributions to bleeding and thrombosis,” said Dr. Kirklin. He also acknowledged that whatever role the current generation of LVAD pumps play in causing thrombosis will not quickly resolve.

“I’m looking forward to a new generation of pumps, but the pumps we have today will probably remain for another 3-5 years.”

The issue of LVAD pump thrombosis first came into clear focus with publication at the start of 2014 of a report that tracked its incidence from 2004 to mid-2013 at three U.S. centers that had placed a total of 895 LVADs in 837 patients. The annual rate of new episodes of pump thrombosis jumped from about 1%-2% of LVAD recipients throughout the first part of the study period through the end of 2011, to an annual rate of about 10% by mid 2013 ( N Engl J Med. 2014 Jan 2;370[1]:33-40 ).

“The inflection occurred in about 2012,” noted Dr. Nicholas G. Smedira , a cardiothoracic surgeon at the Cleveland Clinic. “No one has figured out why” the incidence suddenly spiked starting in 2012 and intensified in 2013, he said. This epidemic of pump thrombosis has produced “devastating complications” that have led to multiple readmissions and reduced cost-effectiveness of LVADs and has affected how the heart transplant community allocates hearts, Dr. Smedira said during his talk at the session. He noted that once the surge in pump thrombosis started, the timing of the appearance of significant thrombus shifted earlier, often occurring within 2-3 months after LVAD placement. There now is “increasing device-related pessimism” and increasing demoralization among clinicians because of this recurring complication, he said.

More recent data show the trend toward increasingly higher rates of pump thrombosis continuing through the end of 2013, with the situation during 2014 a bit less clear. Late last year, data from 9,808 U.S. patients who received an LVAD and entered the Interagency Registry for Mechanically Assisted Circulatory Support ( INTERMACS ) showed that the incidence of pump thrombosis during the first 6 months following an implant rose from 1% in 2008 to 2% in 2009 and in 2010, 4% in 2011, 7% in 2012, 8% in 2013, and then eased back to 5% in the first half of 2014 ( J Heart Lung Transplant. 2015 Dec;34[12]:1515-26 ). The annual rate rose from 2% in 2008 to a peak of 11% in 2013, with 12-month data from 2014 not yet available at the time of this report.

“The modest reduction of observed pump thrombosis at 6 months during 2014 has occurred in a milieu of heightened intensity of anti-coagulation management, greater surgical awareness of optimal pump implantation and positioning and pump speed management. Thus, one may speculate that current thrombosis risk-mitigation strategies have contributed to reducing but not eliminating the increased thrombosis risk observed since 2011,” concluded the authors of the report.

Surgeons and cardiologists must now have a high index of suspicion for pump thrombosis in LVAD recipients, and be especially on the lookout for four key flags of a problem, said Dr. Kormos. The first is a rising LDH level, but additional flags include an isolated power elevation that doesn’t correlate with anything else, evidence of hemolysis, and new-onset heart failure symptoms. These can occur individually or in some combination. He recommended following a diagnostic algorithm first presented in 2013 that remains very valid today ( J Heart Lung Transplant. 2013 July;32[7]:667-70 ).

Dr. Kormos also highlighted that the presentation of pump thrombosis can differ between the two LVADs most commonly used in U.S. practice, the HeartMate II and the HeartWare devices. A LDH elevation is primarily an indicator for HeartMate II, while both that model and the HeartWare device show sustained, isolated power elevations when thrombosis occurs.

Dr. Pagani, Dr. Kirklin, and Dr. Smedira had no disclosures. Dr. Kormos has received travel support from HeartWare. Dr. Kilic has been a consultant to Thoratec and a speaker on behalf of Baxter International.

On Twitter @mitchelzoler