108 Chapter 6 inadequate occlusion ranged from 50 to 82% For detecting incomplete occlusion sensitivity was 96% in our study, ranging from 65 to 90% in the other studies.16-19,24 On the other hand specificity and PPV are lower in our study. The PPV for inadequate occlusion is 36.7%: this is significantly lower than in the other studies (range 67 – 90%), while the PPV for incomplete occlusion (74% in this study) was similar to that in the other studies, which values range from 67 to 86% One difference with these other studies is the standard of reference: In the study presented here, DSA is not assessed in a research setting but in a clinical setting. Class 3 in the R&R classification is used to describe an aneurysm in which a degree of recanalization has been demonstrated which would require retreatment. However, the decision to re-treat depends not only on the recanalization and its configuration in respect to the parent vessel, which determine its suitability for endovascular treatment, but also on secondary factors such as age and comorbidity of the patient. Therefore in a clinical setting the physician may be more reluctant to award a rating of R&R class 3 than when in a research setting, thus leading to the lower specificity and PPV for detecting inadequate occlusion status with CEMRA which was found in this study. The fact that Kappa is only moderate (0.52) underlines the uncertainty in deciding whether an aneurysm is adequately occluded or not. This uncertainty in deciding whether retreatment of coiled and recanalized aneurysms is necessary, is known and well described by Daugherty et al.25 The mean size of the initial aneurysms in this study is less than in the study by Schaafsma et al. (6.2 versus 7 mm). The decision whether a recanalization should be categorized as R&R class 2 or class 3 can depend on the size of the initial aneurysm: in a 15 mm aneurysm a 1 mm contrast collection at the neck would most likely be regarded as a small class 2 neck remnant, while a similar collection in a 3 mm aneurysm could easily be classified as a class 3 recanalization. In smaller aneurysms the distinction between class 2 and class 3 recanalization is therefore likely to be more difficult than in larger aneurysms (see figures 1 – 4). Another difference in our study compared with some others is the use of MRA at 1.5T instead of 3T. However the added value of 3T in this setting has not been established.16,19,26-28 In the second study of Pierot24 and in the studies of Kaufmann16 and Schaafsma19, 1.5T and 3T MRA studies were evaluated without significant differences being found. The high number of (apparent) false positive results in this study as well as in previous studies can probably be explained in part by the fact that a residual aneurysm lumen is sometimes masked by the surrounding coil mesh on DSA, a so called ‘helmet configuration’ which is impenetrable to X-rays. This is a recognized pitfall of DSA.29 In these cases the CEMRA results are in fact true positives, but are counted as false positives because DSA is considered the standard of reference. The dichotomization between ratings of aneurysms which are completely occluded (R&R class 1) on the one hand, and those which present a residual neck (R&R class 2) or a residual aneurysm (R&R class 3) on the other, leads to slightly less discordant results between observers, and better test results in our study than when the choice has to be made between adequately (R&R class 1 and 2) and inadequately (R&R class 3) coiled aneurysms. The clinical consequences of this differentiation between R&R class 1 and class 2, however, appears minimal, as additional treatment is not considered necessary in both these classes.
proefschrit Van Zwam inhoud met kaft en stellingen.indd
To see the actual publication please follow the link above