Permanent Inflammatory Effect of N-Butylcyanoacrylate on Vascular Wall Immunohistochemical Analysis

Summary

We report a case with atypical pattern and time course of inflammatory response after partial embolization of a cerebral arteriovenous malformation with N-butylcyanoacrylate (NBCA), examined by immunohistochemical analysis. This pathomorphological finding of permanent acute inflammation has not been previously reported in humans.

Introduction

Endovascular embolization has become a valuable tool in the treatment of cerebral arteriovenous malformations (AVMs) as a result of improvements in endovascular techniques. Embolization of AVM can be performed either as the definitive treatment of these vascular malformations, or more commonly, as a complementary modality in a combined therapeutic approach with radiosurgery and/or surgery. The histological findings of embolized AVMs have been the subject of numerous reports, and the typical time course of mural angionecrosis, inflammation, and, in some cases, repermeation of the cast intranidal structures have been described. AVMs may show further occlusion, or recanalization and expansion, or, in some cases, haemorrhage following embolization. It is still not quite clear what are the causative factors behind these different outcomes, especially what leads to delayed embolization-related haemorrhage.

Case Report

A 55-year-old man presented with signs of subarachnoid haemorrhage, confirmed on a CT scan. Neurologic exam revealed discrete right-sided hemiparesis. Intra-arterial digital subtraction angiography (IA-DSA) disclosed an AVM in the left posterior parietal region that was approximately 5 cm in diameter (figure 1), with feeders arising from the left pericallosal artery, left middie cerebral artery (MCA), left posterior cerebral artery (PCA) and right pericallosal artery. Venous drainage was through an enlarged cortical vein into the superior sagittal sinus. There was no evidence of deep venous drainage, and the AVM was assigned Spetzler-Martin grade 3

Figure 1.

Lateral view, arterial phase, DSA of the left common carotid artery injection shows a left parietal arteriovenous malformation (AVM) with dominant feeders arising from the pericallosal and middle cerebral arteries (MCA).

Two months later a partial embolization of the nidus with a standard mixture of NBCA and iodized oil (NBCA/oil ratio 1:2) was performed, through the most prominent feeder arising from the left pericallosal artery. The patient recovered uneventfully and remained asymptomatic. He was treated with dexamethasone for four days after the procedure.

Three months later the patient was admitted with left parietal parenchymal haematoma in the region of the AVM. IA-DSA demonstrated residual AVM (figure 2), and surgical extirpation of the lesion was performed. The resected partially embolized nidus was then examined by light microscopy and immunohistochemical methods. Because of partial embolisation, comparative histologic analysis was possible.

Figure 2.

Lateral view, arterial phase, DSA of the left internal carotid artery injection after haemorrhage from the residual AVM, three months after embolization. The size of the nidus is reduced in its superior aspect, in comparison with figure 1.

On a tissue specimen of 5 × 4 × 3 cm, light microscopic and immunohistochemical examination of both AVM parts (embolized and unembolized) were undertaken. All specimens were fixed in 10% buffered formalin and embedded in paraffin.

For the immunohistochemical study, 4 µm thick sections were cut from paraffin block, deparafinized in xylene for 10 minutes and rehydrated in graded alcohol (100%, 96%, 70%). The tissue sections were subjected to antigenretrieval in a microwave oven for 2 × 5 minutes in citrate buffer solution at pH 6.0 - Chem Mate Buffer for Antigen Retrieval, diluted 1:10. On termination of the antigen retrieval step, slides were left in a buffer for at least 20 minutes at room temperature, and stained immunohistochemically for CD 3 (Dako 1:100), CD 20 (Dako 1:100), CD 68 (Dako 1:100) and CD 34 (Dako 1:50) by labeled streptavidin biotin method (LSAB), on a Dako Mate automatic immunostainer, using a microwave streptavidin immunoperoxidase (MSIP) protocol. Histological examination of the previously embolized portion of the AVM revealed necrosis and acute inflammatory changes in the vascular walls, consisting predominantly of polymorphonuclear and eosinophilic cells (figure 3). Chronic inflammation, marked by the appearance of transmural lymphocytic infiltration and foreign body giant cells (FBGC) in pseudopapillar formations with CD 68 immunoreactivity, was also found (figure 4,5). Numerous blood vessels showed mural angionecrosis, and some demonstrated dense mural and perivascular lymphocytic infiltration, consisting predominantly of T cells with CD 3 immunoreactivity (figure 6). A number of embolized vessels without signs of repermeation were also seen. Endothelial proliferation or neoangiogenesis were not encountered. Blood vessels in the unembolized part of the AVM had intact endothelia with CD 34 immunoreactivity and showed no inflammatory changes (figure 7).

Figure 3.

Photomicrograph of the embolized portion of the resected AVM shows vessel wall with diffuse polymorphonuclear infiltration and damaged endothelium (H&E, original magnification × 400).

Figure 4.

Photomicrograph demonstrates chronic inflammation with transmural lymphocytic infiltration of foreign body giant cells (FBGC) and pseudopapillar formation (H&E, original magnification × 200).

Figure 5.

Photomicrograph reveals dense mural infiltration of the vessel wall with CD 68 reactivity in FBGC (arrows) (Immunohistochemistry CD 68, original magnification × 200).

Figure 6.

Photomicrograph of the previously embolized portion of the resected AVM demonstrates dense mural and perivascular T3 lymphocytic infiltration and angionecrosis (Immunohistochemistry CD 3, original magnification × 400).

Figure 7.

Photomicrograph of the unembolized portion of the previously resected AVM shows an intact endothelium and CD 34 reactivity (Immunohistochemistry CD 34, original magnification × 200).

Discussion

Surprisingly, histologic and immunohistochemical examinations indicate that the inflammatory effect of this embolization material was permanent. Necrosis, acute transmural inflammation, and strong granulomatous response characterizad by prominent proliferation of macrophages type FBGC, creating pseudopapillar formations, were detected three months after the embolization. These findings have been established in previous studies, but persistent acute inflammation has not been reported in humans ^1-3. The acute inflammation is expected to transform into a chronic lesion within one month after the embolization procedure ¹. The unembolized portion of the resected AVM in our report showed no evidence of inflammatory changes, which practically excludes possible other, embolisation-unrelated causes of the observed effect. The mechanism of this effect is probably based on immunologic tissue reaction ^2,4 with angionecrosis ^1-3,5 and dense infiltration of T3 Iymphocytes in a later phase, and on acute inflammation with polymorphonuclears, including numerous eosinophils, in the early phase. We did not find endothelial proliferation, nor neoangiogenesis as previously reported ^1,4,6,7. This is in agreement with the reported results showing that the perinidal capillary network, providing collateral blood supply to the nidus, starts to develop at three months after embolization¹. A mild histiocytic giant cell reaction has been reported at about the same time after embolization, which eventually evolved to end-state sclerotic arteritis⁸. Various substances and their combinations have been used for embolization in the reported studies, and this has to be taken into account, since the reaction intensity depends on the material employed⁹. Endovascular AVM embolization has been performed for about 20 years and during this period various embolic materials have been studied: chronic inflammatory response to a foreign body has always been encountered. Intensity of the inflammatory reaction may influence the complication rate and severity of embolization therapy ^3,10,11, such as angionecrosis or leakage of embolic material, with consequent inflammatory response in the surrounding tissue. Differences in findings may also be explained, in addition to the type of embolic material, to individual variations in tissue reaction, because of proven cell-mediated immunologic response. It may be beneficial to test the individual cell-mediated immunologic response to different embolizing agents, in order to avoid haemorrhage from embolized AVMs.

Conclusions

The presented case indicates that the inflammatory effect of NBCA and iodized oil embolic material may be permanent. The issue immunologic response, preceded by granulomatous inflammation of foreign-body type with FBGC antigen presenting cells, plays an important role. Individual differences in patient’s cell-mediated immune response may account for different outcomes with the same embolization material and technique.