Intracardiac Shunts After TAVR: Rare but Potentially Lethal

Aseptic intracardiac shunts are not frequently seen after TAVR, estimated to occur in only about 0.5% of cases, according to a systematic review. But when conservative management is used, mortality is high, especially in patients who have symptoms of heart failure.

Symptomatic patients who undergo percutaneous closure fare better, however, lead author Paol Rojas, MD (Hospital Clínico Universitario, Valladolid, Spain), and colleagues report in a paper published online November 23, 2016, ahead of print in JACC: Cardiovascular Interventions.

They add that even though asymptomatic patients in the review seemed to do well with a conservative approach, management in this group “requires further evaluation,” particularly when shunts exhibit high pulmonary:systemic flow (Qp/Qs) values.

Importantly, the natural history of aseptic intracardiac shunts is unknown, Rojas et al say, calling for careful follow-up, possibly with transesophageal echo, in order to detect late complications. “Any data suggesting hemodynamic compromise (ie, right heart chamber dilatation, deterioration of biventricular function, increase in Qp/Qs, or symptom development) would strongly suggest the need for fistula closure,” they advise.

In an accompanying editorial, Gidon Perlman, MD, and John Webb, MD (St. Paul’s Hospital, Vancouver, Canada), say that rarer complications following TAVR will take on greater importance as more common problems like vascular injury, bleeding, stroke, and paravalvular leak become less frequent.

“With operator and institutional numbers rising, it can be anticipated that an infrequent complication with an occurrence rate of 0.5% will no longer seem so rare,” they write. “Many large-volume programs might encounter such a rare complication semiannually, and most operators might be expected to encounter this complication at some point during their careers.”

Limited Information

Rojas and colleagues searched for studies detailing the occurrence of aseptic intracardiac shunts published between December 2002 and April 2016, finding reports on just 31 patients. After exclusion of transapical fistulas because of the lack of shunting between cardiac chambers, 28 cases remained—an incidence of about 0.5%.

Median time from TAVR to diagnosis of the shunt was 21 days. The shunts appeared to arise during valve implantation or balloon inflation in all patients except one. In that case, it was attributed to unnoticed interventricular septal perforation with the introducer.

Affected patients often had severe calcification of the aortic valve (42.8%), associated coronary disease (39.0%), and a history of chest wall radiation therapy for cancer (14.3%). Most shunts (60.7%) were located in the interventricular septum. Concomitant TAVR-related complications were relatively common; the most frequent were atrioventricular block (21.4%), cardiac arrest (7.2%), and cardiac tamponade (7.2%).

Half of patients were asymptomatic, and 46.4% presented with heart failure. One additional patient presented with stroke.

Most patients (71.4%) were treated with medical therapy, but that approach was associated with poor clinical outcomes. Thirty-day mortality was 25% in conservatively treated patients, although outcomes were worse in those who were symptomatic or had higher Qp/Qs values according to Doppler echocardiography.

Cardiac surgery was used to correct the shunt in one patient, whereas seven patients underwent percutaneous closure with various devices. After a mean follow-up of 7.6 months, all patients who underwent surgery or an intervention had “satisfactory clinical outcomes,” according to the authors, who note that the long-term outcomes of patients surviving closure remains unknown.

Calcification a Recurring Theme

Three major mechanisms seem to be involved in the development of traumatic intracardiac shunts, Perlman and Webb write in their editorial. Most appear to be related to mechanical expansion of the aortic annulus, typically during aggressive balloon dilation, although direct trauma by components of the delivery system and interaction between the valve frame and the LV outflow tract also play a role, they say.

“A recurrent theme is that annular injury is most often associated with excessively large balloon-expandable valves implanted in the presence of extensive nonleaflet calcification,” they write. “Strategies to reduce the risk of annular injury must include recognition of patients at increased risk due to extensive or nodular annular or subannular calcification and accurate assessment of annular dimensions.”

They add that “annular tears may be less frequent with self-expandable valves, although a calcified annulus can compromise positioning and sealing with these devices.”

Balloon-expandable valves remain an option, the editorialists say, given that it’s now easier to avoid excessive oversizing with current-generation devices thanks to “their relatively effective external seals.” They suggest two strategies for patients at high risk: “underexpansion (balloon underfilling) of an oversized valve with subsequent nominal expansion if required” as well as “nominal expansion of an undersized valve with overexpansion (balloon overfilling) if required.”