Sir

We read with great interest the article by Ma et al. [8] that was recently published in your journal. The authors focus the attention on contralateral hemodynamic changes after unilateral bypass surgery in patients with moyamoya disease (MMD); a feature that has been little analyzed but of particular interest given the complex cerebral hemodynamics resulting from MMD. As the authors' findings contradict with other work done on this topic, we would like to provide some comments on this study and highlight some recent literature.

Ma et al. reported their experience in the hemodynamic evaluation of contralateral (nonoperated and asymptomatic) hemispheres in 15 MMD patients who received one-sided direct bypass revascularization of the symptomatic hemisphere. Pre- and postoperative brain hemodynamic evaluations were performed by acetazolamide-enhanced Xenon-computed tomography. Cerebral revascularization consisted of double superficial temporal artery (STA) to middle cerebral artery bypass surgery (using both the frontal and the parietal branch of the STA). Postoperative digital subtraction angiography was used to confirm bypass patency, and postoperative hemodynamic studies were performed at 1, 3, and 6 months after surgery for evaluation of regional cerebral blood flow (rCBF) and cerebrovascular reserve (CVR). Among the 15 symptomatic (operated) hemispheres, all exhibited preoperatively impaired CVR, 14 hemispheres demonstrated decreased rCBF. Twelve patients presented with minor ischemic symptoms (transient ischemic attacks, TIAs) while 3 patients presented with a stroke. Each of the 15 nonoperated contralateral hemispheres showed good rCBF whereas 6 hemispheres exhibited impaired CVR. The preoperative rCBF and CVR in the operated hemispheres was much lower than the contralateral side (32.8 ± 5.5 ml 100 g−1 min−1 versus 50.4 ± 8.4 ml 100 g−1 min−1; P < 0.001, and 6.1 ± 13.2% versus 31.7 ± 9.0% respectively; P < 0.001). Revascularization was successful in all hemispheres, with no new signs of ischemia and significant improvement of rCBF and CVR at 1-month follow-up. However, 3 months postoperatively, CVR of the contralateral hemispheres decreased significantly (25.8 ± 8.1%; P = 0.003) and at 6-month follow-up, the rCBF in the contralateral hemispheres showed a downward trend (47.4 ± 8.0 ml 100 g−1 min−1; P = 0.05). At this time point, three patients had markedly decreased rCBF and impaired CVR in the non-treated hemisphere. Among them, two became symptomatic (TIAs) but already showed impaired CVR in the non-treated hemisphere preoperatively, while the other with normal CVR of the contralateral side preoperatively did not develop symptoms. Thus, the authors concluded that unilateral direct revascularization of the symptomatic hemisphere for MMD patients could lead to CVR impairment in the primarily asymptomatic contralateral hemisphere.

We believe that this conclusion may not be fully supported by the results. To begin with, the study design appears to be somewhat flawed since follow-up hemodynamic evaluation is reported inconsistently. For example, 1 month after surgery, only rCBF and CVR data from the operated hemispheres are shown. As for the contralateral hemodynamics—the data of interest for this study—only CVR data are reported at 3-month follow-up and only rCBF findings at 6-month follow-up which makes it hard to objectively identify differences in perfusion over time as well as between treated and non-treated hemispheres. Unfortunately, we have to disagree with the authors in their conclusion that progression of MMD cannot explain the postoperative contralateral hemodynamic worsening. The authors stated that progression of MMD in adults occurs slowly. For this they quoted a recent study reporting that MMD progression occurred in nearly 20% of patients during a mean follow-up period of 6 years [7]. Both in our practice and in the literature [4, 10–12], patients can experience disease progression more acutely (months), and generally MMD patients present with bilateral symptoms [2, 11, 12]. The fact that all patients in this case series exhibited unilateral symptoms might be related to sample size or the fact that patient selection was not randomized. Progression of the disease may very well have been responsible for their observations but cannot be confirmed since no angiography-based follow-up was done. Furthermore, both the hemodynamic trend of operated and nonoperated hemispheres should have been reported in order to completely evaluate hemodynamic and clinical progression of the disease. The authors also considered the possibility that after unilateral direct bypass revascularization, the increased blood flow through the anastomosis could have caused hemodynamic stress with subsequent impaired compensation from other collateral circulation (for example from the posterior circulation including the leptomeningeal collaterals from bilateral posterior cerebral arteries that is known to play a role in supplying the ischemic MMD parenchyma). They concluded that after successful bypass surgery, reduced compensation from the posterior circulation could have caused a decrease in collateral blood supply that may have contributed to the compromised hemodynamics of the contralateral side. To underline this, they reported a case that presented with decreased bilateral leptomeningeal collaterals from the posterior circulation 6 months after unilateral direct revascularization. Although the contralateral (nonoperated) hemisphere may provide collateral flow to the hemodynamically compromised hemisphere, it is likely that the preserved autoregulation in this hemisphere (nonoperated) will ensure enough perfusion despite the support to the compromised side. After restoration of blood flow in the compromised hemisphere by bypass surgery, this collateral support may indeed become obsolete but should not lead to a decrease in blood flow supply to the contralateral (nonoperated) hemisphere. In fact, more blood flow to the brain should actually have a beneficial effect on contralateral hemodynamics.

Recent literature shows such a contralateral hemodynamic benefit. For example, a review by Bacigaluppi et al. [1] on neurovascular imaging in diagnosis, preoperative assessment, and follow-up of MMD, frequently observed contralateral hemodynamic improvement after unilateral direct bypass revascularization procedures, as shown by improved CVR on quantitative MR imaging postoperatively. More so, another report from the same group, in collaboration with our institution, has shown that successful surgical revascularization can restore cortical thickness in patients with MMD [3]. These structural changes, albeit less consistent, were also observed in the non-treated contralateral hemispheres. Even perfusion studies in MMD patients, not necessarily focusing on contralateral hemodynamics, showed a positive effect in the non-treated hemisphere after unilateral bypass revascularization [5, 6, 9]. A recent study by Ideguchi et al. [5] reporting about the presence of ivy signs on MR imaging to gauge success of revascularization surgery in MMD patients demonstrated improved hemodynamics in the nonoperated contralateral hemisphere as measured with SPECT (Fig. 2 in the paper).

It is evident that after unilateral bypass surgery, postoperative cerebral hemodynamic changes may be expected in both treated and untreated hemisphere. Although Ma et al. did not observe a contralateral hemodynamic benefit from unilateral bypass revascularization, conflicting with recent the literature, we would like to underline the fact that these incongruities may be explained by the complexity of cerebral hemodynamic rearrangement occurring after bypass revascularization in MMD patients and our difficulty in understanding these phenomena.