To the Editor,
We have read with great interest the recently-published article of Mercé et al1 “Implantable Loop Recorder: A New Tool in the Diagnosis of Cryptogenic Stroke”. Clearly, the active search for episodes of atrial fibrillation (AF) as a possible cause of stroke is a topic of current importance given that the underdiagnosis of cardioembolic sources could explain a significant percentage of episodes in the 30% of stroke cases considered cryptogenic. In the present letter, we would like raise two issues about the use of the implantable loop recorder to detect these episodes: Firstly, is it really as useful a tool as the article proposes? Secondly, should we use it in all patients?
In the aforementioned article, implantable loop recorders detected AF in 35.7% of the sample but we note that the mean age of the patients with AF episodes was significantly older (P=.01) than that of those with no detected episodes. Furthermore, we would point to the high percentage of men (71.4%) in the sample, and the high prevalence of cardiovascular risk factors (79% with high blood pressure, 36% with diabetes mellitus), which could, in turn, partly explain the high incidence of AF detected, even in a sample of only 14 patients. In a recent article on the use of implantable devices in a total of 51 patients, Cotter et al2 reported a 25.5% incidence of AF during follow-up—a lower rate than that reported by Mercé et al1 although we should point out that the mean age of the patients was lower (median, 52 years). In another recent study, Rojo-Martínez et al3 included a total of 101 patients with characteristics similar to those of the sample of Mercé et al. The incidence of hidden AF was 33.7%, demonstrating the high incidence of this entity in patients with what is termed cryptogenic stroke. Our own study used noninvasive methods (electrocardiogram performed in the center, Holter monitoring of admissions) in a cohort of 273 patients diagnosed with cryptogenic stroke in our center, with a mean age of 62 (13) years, and characteristics comparable to the samples of the 2 studies cited.4 The incidence of hidden AF was only 2.6% in a 3-year follow-up—significantly lower than in the other studies—with a mean age 67 (7) years slightly older than that of the total sample, which coincides with the data mentioned.
In view of the results, the implantable loop recorder clearly seems better at detecting hidden AF although, as we have mentioned, in a specific patient profile—older, with a greater prevalence of cardiovascular risk factors—detection of these episodes is more frequent. This finding could be explained by the etiology and pathogenesis of AF, which leads us question the cost-effectiveness of performing invasive studies such as these in younger patients with cryptogenic stroke and high suspicion of an embolic etiology. In this patient profile, we have studied the diagnostic effectiveness of imaging tests such as transesophageal echocardiography to detect patent foramen ovale5,6 but the value of the complementary use of the implantable loop recorder in detecting hidden AF remains unproven.
We believe the implantable loop recorder is an efficient diagnostic tool in cardioembolic profile cryptogenic stroke and that its cost-efficiency is greater in patients with an increased risk of AF, especially older and hypertensive patients. In young patients, its effectiveness should be determined in prospective studies with long series.