We read with great interest the article by King et al, ‘‘Meta-analysis of 123I-MIBG cardiac scintigraphy for the diagnosis of Lewy body-related disorders.’’1 The study confirmed the usefulness of 123I-MIBG scintigraphy in discriminating Lewy body– from non-Lewy body– related disorders. We appreciated the review of the literature and the statistical analysis performed by the authors. Nevertheless, a critical point should be considered. The authors have identified an ‘‘ideal’’ heart-to-mediastinum ratio (H/M) threshold value (H/M ¼ 1.77) that may be useful in differentiating Lewy body– from non-Lewy body–related disorders. This threshold value has been obtained by a receiver operating curve analysis performed using H/M values of different studies.1 We believe that the indication of an ‘‘ideal’’ H/M threshold value in clinical practice may be incorrect. In fact, the H/M threshold value mainly depends on several technical factors (ie, gammacamera system, collimators, imaging time, radioactivity administered, shape and size of the regions of interest drawn on the heart and the mediastinum) that are different in various centers, as also previously demonstrated in patients with cardiac diseases submitted to 123I-MIBG scintigraphy.2–4 Furthermore, different physiological patient-related factors such as age, sex, or race may affect the H/M value. In fact, it has been demonstrated that H/M values may decrease with aging, and females usually show significantly higher H/ M values than males of the same age.5 Moreover, the H/M threshold value for a white population differs from that of a Japanese population.6,7 Therefore, caution should be exercised when applying an ‘‘ideal’’ H/M threshold value to individual centers.8 Every single nuclear medicine unit should use its own H/M threshold value, based on its own normal controls; this H/M threshold value is usually obtained by computing the 95th percentile of results in normal controls9 matched for age, sex, and race. Lastly, an alternative meta-analytical approach to that of King et al could be suggested. The computation of sensitivity and specificity of 123I-MIBG scintigraphy in each single study, based on the H/M ratio threshold value of each institution, could be performed; subsequently, a pooled sensitivity and specificity could be calculated, as recently performed by our group.
|Number of pages||2|
|Publication status||Published - 2011|