Intrapulmonary shunts and its implications for assessment in stroke and divers with type II decompression illness.

The recent article “Intrapulmonary Shunt Is a Potentially Unrecognized Cause of Ischemic Stroke and Transient Ischemic Attack” (TIA) by Abushora et al. raises several questions in our daily clinical evaluation of patients with stroke, and also divers with unexplained Type II decompression sickness (DCS). A cohort of patients (stroke group) with either nonhemorrhagic stroke (CVA) or TIA (n = 321) was compared with ageand gendermatched control patients (n = 321). A subgroup of patients with CVA or TIA had what was defined as cryptogenic stroke (n = 71). These patients underwent transesophageal echocardiography, and were evaluated for visualization of saline contrast bubbles within the left atrium (LA) via either the pulmonary veins or through a patent foramen ovale (PFO). An intrapulmonary shunt was defined as direct visualization of bubbles within one or more pulmonic veins entering the LA. A PFO was defined as direct visualization of contrast crossing the foramen ovale, or visualization of contrast with the LA within three cardiac cycles of its appearance within the right atrium. Studies were performed with normal respirations, with no provocative maneuvers for detection of PFO. The authors mention that the study was performed without provocative maneuvers, to allow for a definitive differentiation between shunting via a pulmonary vein or PFO. For both detected intrapulmonary shunts and also PFOs, severity of shunting was graded based on the amount of contrast in a pulmonary vein or that entering via a PFO. Mild, intermediate, or large shunts were defined as 1–4, 5–20, or >20 visualized within any single image frame. Intrapulmonary shunts occurred more often in the stroke group than controls (72 of 321 or 22% vs. 32 of 321 or 10%). The severity of shunting was similar in both groups, mostly mild or intermediate. The number of pulmonary veins demonstrating shunting appeared to occur in 1, 2, or all 4 veins for both groups. The stroke group with detected intrapulmonary shunting did have a tendency toward contrast appearing in all 4 veins (38% incidence) more often than in controls (28%). Patent foramen ovale detection was higher (73 of 321 or 23%) for the stroke group versus controls (54 of 321 or 17%), but did not achieve statistical significance. The stroke group tended toward larger shunts compared to controls. Patent foramen ovale and/or an intrapulmonary shunt were noted more often in the stroke group (136 of 321 or 42%) than in controls (54 of 321 or 17%). The incidence of both an intrapulmonary shunt and a PFO in the same patient was not mentioned. From the stroke group a subgroup of patients was identified as having either a cryptogenic CVA (n = 50) or cryptogenic TIA (n = 21). This cryptogenic stroke group (n = 71) was then compared with an ageand gender-matched control group (n = 71). In this cohort, intrapulmonary shunting occurred more often (25 of 71 or 35%) than in controls (5 of 71 or 7%). Shunting in the cryptogenic group occurred in 1 or 2 veins 2/3 of the time and 1/4 of the time in all 4 veins, whereas in the control group, shunting occurred exclusively in 1 or 2 veins. Severity of shunting Address for correspondence and reprint requests: Edmund Kenneth Kerut, M.D., Heart Clinic of Louisiana, Marrero, Louisiana 70072. Fax: 6043496621; E-mail: [email protected]