Incidence of and risk factors for hallucinations and delusions in patients with probable AD

Reply from the Authors: We thank Dr. Yekhlef et al. for their interest in our work. However, we are surprised by their finding that 47% of 19 patients with PSP and 29% of 17 patients with CBD showed a “hot-cross-bun” sign (HCB). This contradicts not only our findings in 35 patients with PSP and five with CBD, but also those of previous studies from two other centers.6-8 Pathologic findings have correlated this MRI abnormality with degeneration of transverse pontine fibers in MSA2,9 and in other conditions involving selective degeneration of these fibers,10 which is not a feature of PSP. An illustrative example was published by Savoiardo et al.9 In our study, the HCB was seen in only one of 35 patients with PSP by one of the two blinded neuroradiologists, and in none of five patients with CBD, but was seen in 27 of 54 patients with MSA. We only rated this abnormality as present if it was cruciform and unequivocal on T2-weighted and proton density images. Thus the finding of a hyperintense raphe, which is a common normal finding, and equivocal findings on proton density scans not confirmed on T2-weighted images were not rated. We also only included patients with an unequivocal clinical diagnosis, and had additional pathologic confirmation in nine of them (PSP 11%, MSA 4%, CBD 80%). It initially seems difficult to reconcile the French group’s findings with ours. However, one difference that might play a role is that their patients were imaged on 11 different scanners, whereas we only studied scans performed in our own hospital on either a single 0.5-T or a single 1.5-T machine. Additionally, it would appear that only one of the French cases (who had MSA) was pathologically confirmed. None of the MRI findings in these conditions are pathognomonic. The HCB, for instance, can be seen in patients with spinocerebellar ataxia type 3 and other conditions selectively involving the pontocerebellar system.11 As we emphasized, this sign should always be interpreted in the context of the clinical setting and other MRI findings. Yekhlef et al. make an important point by emphasizing the high correlation between the finding of a HCB and pontine atrophy; in our patients there was also an even stronger correlation with atrophy of the middle cerebellar peduncles in MSA patients. On the other hand, the only patient of ours with a clinical diagnosis of PSP in whom a HCB was seen by one of the two blinded neuroradiologists also had marked midbrain atrophy and midbrain hyperintensity. The combination of these abnormalities is likely to be more helpful in a clinical setting than if used as an isolated finding. However, in our patients we continue to find the HCB a particularly useful sign with high specificity for MSA.