Published ahead of print on January 6, 2010
doi: 10.3174/ajnr.A2012

American Journal of Neuroradiology 31:E35-E36, March 2010
© 2010 American Society of Neuroradiology

H. Meng^a, S.K. Natarajan^a, L. Gao^a, C. Ionita^a, J. Kolega^a and A.H. Siddiqui^a
aToshiba Stroke Research Center State University of New York at Buffalo Buffalo, New York

J. Mocco^b
bDepartment of Neurosurgery University of Florida Gainesville, Florida

We read the interesting report by Dai et al,¹ wherein the authors examined histology at the basilar terminus (BT) following right common carotid artery (RCCA) ligation as a part of a procedure to create an elastase aneurysm model in 30 consecutive New Zealand white rabbits at various time points. The authors concluded that in contrast to our previous report,² there was no bulgelike localized dilation associated with a missing internal elastic lamina (IEL) to suggest microaneurysm or nascent aneurysm formation at the BT, at any time point after RCCA ligation.

We find this report particularly interesting because the first aneurysmal changes at the rabbit BT that we noticed actually occurred in 2 rabbits that had undergone RCCA ligation.³ The rabbits underwent the same surgical procedure as that performed by Dai et al,¹ with the same objective of creating the elastase aneurysm model on the RCCA stump for an endovascular devicetesting.³ We examined whether there were any aneurysmal changes in the basilar bifurcation in these 2 rabbits through histologic staining of contiguous specimens with hematoxylin-eosin (HE), Van Gieson, and trichrome. In the first rabbit, sacrificed 10 weeks after RCCA ligation, we observed loss of IEL, endothelial cells, and smooth muscle cells, a thinned media, and an outwardconvex bulge at the BT (Fig 1). Fifteen consecutive sections of 15-µm thickness consistently presented this BT bulge with missing IEL and media thinning. We repeated the observation in the second rabbit, which had received incomplete RCCA ligation and was sacrificed 12 weeks later. We saw similar aneurysmalchanges, including IEL loss and medial thinning at the BT, albeit with a shallower and longer bulge (Fig 2). These vascular defects were clearly aneurysmal and could not be a staining or sectioning artifact, or misinterpreted branches.

View larger version (79K): [in this window] [in a new window] Fig 1. An aneurysm detected at the rabbit BT 10 weeks after RCCA ligation in an elastase aneurysm model. A, Basilar artery–posterior cerebral artery bifurcation shown at a low magnification. Arrow indicates blood flow direction. An aneurysm is observed between a and b, with IEL disruption (a total length of 300 µm) and media thinning occurring in the bulge (Van Gieson, original magnification x100). B–D, The aneurysm site is shown at a high magnification (original magnification x400). B, Van Gieson staining shows IEL loss. C, HE staining shows endothelial loss. D, Trichrome staining shows medial thinning.

View larger version (68K): [in this window] [in a new window] Fig 2. Aneurysmal degradation at the rabbit BT 12 weeks after incomplete RCCA ligation in an elastase aneurysm model. A and B, Basilar bifurcation shown at a low magnification (original magnification x100). Arrow indicates blood flow direction. A shallow bulge with a total disruption length of 600 µm is observed between a and b. A, Van Gieson staining. B, Trichrome staining. C and D, Aneurysmal degradation is shown at a high magnification (original magnification x400). C, Trichrome staining shows medial loss. D, Van Gieson staining shows IEL disruption.

These 2 cases prompted us to conduct a prospective study of nascent aneurysmal initiation at the BT induced by common carotid artery (CCA) ligation alone, which resulted in the article by Gao et al.² We performed unilateral or bilateral CCA ligation in otherwise unmanipulated New Zealand white rabbits and prospectively examined aneurysmal changes at the BT 12 weeks later. We observed nascent aneurysm formation and its dose dependence on basilar artery flow increase. Since then, we have shifted our model to perform bilateral CCA ligation only. We have seen prominent IEL loss at the BT in every one of our rabbits, as early as 2 days and 5 days post-CCA ligation. We are in the process of drafting these results.

Our experience with both the elastase model (reported here) and the nascent BT aneurysm model² obviously differs from that of Dai et al.¹ We are concerned that in their study, Van Gieson staining was done after washing off previous HE staining on the same specimen. The reliability of the de-staining and re-staining technique in detecting IEL loss has not been verified. Reports in prostate cancer specimens show that the reliability of de-staining HE and re-staining with cytokeratin could be as low as 58%.⁴ In our experience, more than 50 sections can be taken from 1 BT tissue specimen; thus, adjacent specimens can be used for different stains. Therefore, de-staining and re-staining of the same specimen do not seem necessary and could have confounded the interpretation of their results.

It is not clear whether Dai et al¹ created their animal models prospectively to examine the BT or performed a retrospectiveanalysis on stored specimens that had previous HE staining. In the latter scenario, there could be sectioning bias as wellas a need to de-stain and re-stain sections, rather than to use adjacent sections.

Finally, we respectfully disagree with the speculation by Dai et al¹ that our group could have misinterpreted branching vessels as aneurysmal bulges in Gao et al.² Branching vessels would have assumed completely different anatomy than what we have demonstrated, withno IEL loss or media thinning. In addition, multiple contiguous sections throughout the entire paraffin-embeddedbifurcation would have consistently revealed the branching vessels, instead of the bulges that spanned multiple sections and tapered off, as seen in our studies.

References

1. Dai D, Ding YH, Kadirvel R, et al. Experience with microaneurysm formation at the basilar terminus in the rabbit elastase aneurysm model. AJNR Am J Neuroradiol 2010;31:300–03. Epub 2009 Oct 1[Abstract/Free Full Text]
2. Gao L, Hoi Y, Swartz DD, et al. Nascent aneurysm formation at the basilar terminus induced by hemodynamics.Stroke 2008;39:2085–90[Abstract/Free Full Text]
3. Ionita CN, Paciorek AM, Dohatcu A, et al. The asymmetric vascular stent: efficacy in a rabbit aneurysm model.Stroke 2009;40:959–65[Abstract/Free Full Text]
4. Dardik M, Epstein JI. Efficacy of restaining prostate needle biopsies with high-molecular weight cytokeratin. Hum Pathol 2000;31:1155–61[CrossRef][Medline]

Reply

Published ahead of print on January 6, 2010
doi: 10.3174/ajnr.A2019

American Journal of Neuroradiology 31:E37, March 2010
© 2010 American Society of Neuroradiology

D. Dai^a and D.F. Kallmes^a
aNeuroradiology Research Laboratory Department of Radiology Mayo Clinic Rochester, Minnesota

We greatly appreciate the correspondence from Meng et al regarding our recent article about the basilar bifurcation in rabbits following carotid ligation.¹We remain absolutely confident that the focal excrescences along the basilar apex in our model represent branch arteries and not aneurysms. Dr. Meng and colleagues raised concern that de- and re-staining our histologic sectionsmight have compromised the ability to identify the internal elastic lamellae. However, the article they cite regarding such loss of accuracy was for immunohistochemical techniques,² not the histochemical staining, Verhoeff Elastic-Van Gieson (VVG), used in our study. We have re-reviewed all of our slides in detail. There are no aneurysms.

We are grateful to Dr. Meng and colleagues for providing new histologic images in their letter, different from those published previously.³ We note with interest that in Fig 1A in their letter, the artery denoted as the basilar artery is much much smaller than the P1 segment (the first segment of the posterior cerebral artery). From our experience, we have found that it is very easy to lose track during histologic processing of which branch is the basilar and which are the P1s; as such, simple rotation of the slide by 90° would place the bifurcation of the superior cerebellar artery/P1 at the “basilar tip.” Thus, the “aneurysm”noted in their Fig 1 looks similar to the origins of superior cerebellar arteries in a subject from our lab (Fig 1).

View larger version (26K): [in this window] [in a new window] Fig 1. A subject at 8 weeks after right common carotid artery ligation in the elastase aneurysm model. Photomicrograph shows the basilar artery, superior cerebellar artery bifurcation, and the posterior cerebral artery at a low magnification. Arrow indicates the blood flow direction (VVG, original magnification x40).

Given the disparate results between our study and their previous work, we speculated in our article that the apparent aneurysms noted by them might in fact have been branch arteries. Certainly there may be other potential explanations for different results between our study and theirs. In any event, because they identify aneurysms in 100% of cases, it should be relatively easy for them to confirm, by using complementary methods in addition to serial histologic sections, that they have induced aneurysms. These complementary and potentially confirmatory techniques could include angiographic or micro-CT imaging,^4,5 or casting techniques such Microfil perfusion (MTS Medication Technologies, St. Petersburg, Florida).^6–9 In addition, we would assume that the microaneurysms induced in their model might, with time, grow to become large and thus intuitively obvious to all observers, including us.

Perhaps the most efficient way to move forward in clarifying the apparent disparity in conclusions between our 2 researchgroups would be to provide the histologic data to an independent vascular pathologist. We would be delighted to supply all of our serial histologic slides to such an expert in hopes of advancing this important field.

References

1. Dai D, Ding YH, Kadirvel R, et al. Experience with microaneurysm formation at the basilar terminus in the rabbit elastase aneurysm model. AJNR Am J Neuroradiol 2010;31:300–03. Epub 2009 Oct 1[Abstract/Free Full Text]
2. Dardik M, Epstein J. Efficacy of restaining prostate needle biopsies with high-molecular weight cytokeratin. Hum Pathol 2000;31:1155–61[CrossRef][Medline]
3. Gao L, Hoi Y, Swartz DD, et al. Nascent aneurysm formation at the basilar terminus induced by hemodynamics.Stroke 2008;39:2085–90[Abstract/Free Full Text]
4. Dorr A, Sled JG, Kabani N. Three-dimensional cerebral vasculature of the CBA mouse brain: a magnetic resonance imaging and micro computed tomography study. Neuroimage 2007;35:1409–23. Epub 2007 Jan 23[CrossRef][Medline]
5. Schambach SJ, Bag S, Groden C, et al. Vascular imaging in small rodents using micro-CT. Methods 2010;50:26–35. Epub 2009 Sep 20[CrossRef][Medline]
6. Krucker T, Schuler A, Meyer EP, et al. Magnetic resonance angiography and vascular corrosion casting as tools in biomedical research: application to transgenic mice modeling Alzheimer’s disease. Neurol Res 2004;26:507–16[CrossRef][Medline]
7. Heinzer S, Krucker T, Stampanoni M, et al. Hierarchical microimaging for multiscale analysis of large vascular networks. Neuroimage 2006;32:626–36[CrossRef][Medline]
8. Krucker T, Lang A, Meyer EP. New polyurethane-based material for vascular corrosion casting with improved physical and imaging characteristics. Microsc Res Tech 2006;69:138–47[CrossRef][Medline]
9. Cruise GM, Rivera EA, Jones RM, et al. A comparison of experimental aneurysm occlusion determination by angiography, scanning electron microscopy, MICROFIL perfusion, and histology. J Biomed Mater Res B Appl Biomater 2009;91:669–78[Medline]