A previously healthy 46-year-old right-handed woman presented with left hemiparesis and hemisensory loss to an outside hospital. CT imaging demonstrated a right cervical internal carotid artery dissection with near-occlusion at the skull base, a left anterior cerebral artery perfusion deficit without intracranial occlusion, and an evolving right internal capsule infarct (ASPECTS score 9). Following dual antiplatelet therapy, she was transferred to our quaternary center for urgent endovascular treatment. Emergent carotid stenting resolved her neurologic deficits (NIHSS 3 to NIHSS 0 post-procedure).
Hours later, she acutely declined with left hemiplegia, forced right gaze deviation, dysarthria, and a fixed, dilated right pupil (NIHSS 26). Head CT revealed a large right frontal intraparenchymal hemorrhage with midline shift, prompting emergent decompressive craniotomy and evacuation. Hemostasis was achieved intraoperatively using a standard layer of microfibrillar collagen (Avitene) powder circumferentially within the cavity.
Postoperatively, her neurologic examination improved to 4+/5 (right/left; Medical Research Council [MRC] grades). She was NIHSS 13 on postoperative day one. However, over the subsequent two weeks, she developed worsening headaches despite remaining otherwise neurologically stable. Serial CT imaging demonstrated unexpected progressive right frontal edema and increasing mass effect despite hematoma evacuation. An MRI with and without gadolinium contrast revealed thick peripheral enhancement surrounding the surgical cavity with associated dural enhancement (Figure 1).
(A) Axial T2weighted with gadolinium contrast showing edema around the right frontal hematoma cavity. (B) Axial MPRAGE post-contrast showing ring enhancement. (C) Axial DWI showing no central restriction of the lesion.

Figure 1. Long description
An MRI scan of the brain displays three different views. The first image, an axial T2-weighted scan with gadolinium contrast, highlights edema around a right frontal hematoma cavity. The second image, an axial MPRAGE post-contrast scan, shows ring enhancement. The third image, an axial DWI scan, indicates no central restriction of the lesion.
Although the patient remained afebrile and lacked signs of meningismus, the combination of these imaging findings with recurrent leukocytosis (24.0 × 109/L) and tachycardia raised high suspicion for a deep surgical site infection or an infected hematoma. On MRI, there was no central diffusion restriction, a finding that typically argues against a mature abscess. However, diffusion restriction may be absent with early infection, partially treated infection, or postoperative changes, limiting its negative predictive value in this context.
Blood cultures were repeatedly negative, inflammatory markers such as CRP were minimally elevated (35.5), and the surgical incision remained clean without superficial signs of infection. Nonetheless, under guidance from infectious disease, empiric broad spectrum antimicrobial therapy was initiated as postoperative infected hematoma was felt to be the most likely diagnosis. The severity and progression of the vasogenic edema with worsening subfalcine herniation posed a risk of secondary ischemic injury, making prolonged observation unsafe.
In light of the diagnostic uncertainty and worsening radiographic findings, a decision was made to return to the operating room for decompression and definitive tissue diagnosis. Re-exploration revealed liquefied hematoma under moderate pressure without purulence. The walls of the prior hemorrhage cavity were lined with residual Avitene material.
Cultures and microbiological evaluation of subdural and intracapsular specimens showed no growth. Histopathologic examination demonstrated reactive and infarcted brain parenchyma with prominent macrophage infiltration, chronic inflammatory cells, neovascularization, and homogeneous eosinophilic material consistent with foreign material (Figure 2).
Histopathology of the resected tissue show necrotic and reactive brain parenchyma admixed with homogeneous eosinophilic material compatible with surgical hemostatic material (A) H&E at 4X. (B) H&E at 20X. Another area of the resected tissue shows foci of neovascularization and presence of chronic inflammatory cells (C: H&E at 20X) which stain strongly for CD68 (D, 20X) which is a marker of histocytes.

Figure 2. Long description
The images display brain tissue under a microscope. Image A shows necrotic and reactive brain parenchyma mixed with homogeneous eosinophilic material, likely surgical hemostatic material, at 4X magnification. Image B provides a closer view at 20X magnification. Image C highlights areas with neovascularization and chronic inflammatory cells at 20X magnification. Image D shows these cells staining strongly for CD68, a marker for histocytes, also at 20X magnification.
There was no neutrophilic inflammation, and no organisms were identified on special stains. These findings supported a diagnosis of foreign body granulomatous inflammation secondary to retained microfibrillar collagen hemostatic agent (Avitene), mimicking deep surgical site infection. The patient’s headaches and mass effects resolved within two days of Avitene foreign body removal. She showed substantial neurologic improvement during her stay, especially her headaches, and was transferred to another center for inpatient rehabilitation (NIHSS 0 at discharge).
Progressive postoperative cerebral edema with peripheral enhancement following hematoma evacuation raises immediate concern for deep surgical site infection or evolving abscess, particularly when accompanied by leukocytosis and mass effect. In this case, these findings prompted empiric antimicrobial therapy and ultimately repeat surgical exploration. Operative and histopathologic findings, however, demonstrated a sterile inflammatory process most consistent with a foreign body granulomatous reaction to microfibrillar collagen hemostatic material.
Granulomatous reactions to microfibrillar collagen have been reported across multiple organ systems, including intraabdominal, orbital, spinal, and central nervous system sites, establishing clear biologic plausibility for a foreign body-mediated inflammatory response. Reference Mcgregor, Macarthur and Carter1–Reference Roberto Giammalva, Brunasso and Costanzo4 In the neurosurgical literature, these reactions have most often been described as delayed mass lesions or tumor recurrence mimics, frequently in pediatric or oncologic contexts. Reference O’Shaughnessy, Schafernak, DiPatri, Goldman and Tomita5,Reference Ribalta, McCutcheon and Neto6 These typically present months to years after surgery and have been identified on surveillance imaging rather than during the acute postoperative period.
Reports of atypical imaging findings have been described as early as 1989 in a paper by Gondo et al., who noted persistent mass like postoperative CT abnormalities and surrounding edema after intracranial use of microfibrillar collagen that mimicked hemorrhage, tumor, or infection. Reference Gondo, Yamashita, Ishiwata, Hirata and Satoh7 Recent papers have described cases suspicious of granulomatous meningitis and necrotizing inflammatory reactions after craniotomy with microfibrillar collagen use. Renati et al. reported a case of suspected granulomatous meningitis presenting with meningeal enhancement and inflammatory CSF findings. Histopathology revealed chronic inflammation and granulomatous foreign body reaction surrounding amorphous eosinophilic material, with associated eosinophilic infiltrates suggestive of an allergic type of response. Reference Renati, Kaur, Kresak, Wicklund and Malaty8 Apel-Sarid et al. reported a series of three pediatric patients who developed a common inflammatory reaction at prior craniotomy sites following use of microfibrillar collagen. All patients presented several weeks after surgery with new or recurrent seizures, and postoperative imaging mimicked either abscess formation or tumor recurrence with one case also demonstrating a ring-enhancing lesion around the resection cavity. Histopathology demonstrated necrotizing granulomatous inflammation centered on microfibrillar collagen, with palisading macrophages and associated eosinophils. Reference Apel-Sarid, Cochrane, Steinbok, Byrne and Dunham9 Although the majority of previously reported cases were diagnosed after surgical resection, limiting the experience on non-operative management, other authors have advocated for conservative management if suspicion is high. Reference Apel-Sarid, Cochrane, Steinbok, Byrne and Dunham9
In this case, progressive edema and mass effect were identified within two weeks of hematoma evacuation. Although the lack of central diffusion restriction and minimal systemic inflammatory response argued against abscess, the degree of progressive vasogenic edema and midline shift necessitated surgical management. In addition, the inflammatory reaction occurred within a hemorrhage cavity, where postoperative imaging evolution is variable, further complicating interpretation.
This case highlights a critical diagnostic lesson that is rare but described in the literature. Hemostatic agents placed within parenchymal cavities can provoke robust macrophage driven inflammation with peripheral enhancement and progressive edema that closely mimics infection. This may lead to prolonged antimicrobial therapy, re-operation, or delayed decision-making in the setting of mass effect. Explicit consideration of a foreign body reaction related to operative materials should be part of postoperative diagnostic reasoning. Surgeons should aim to utilize the minimal amount of this agent necessary and irrigate out any excess to decrease the risk of reaction to hemostatic agent. If such a reaction is suspected after a thorough workup, a conservative approach has been advocated in the literature for clinically stable cases. Reference Apel-Sarid, Cochrane, Steinbok, Byrne and Dunham9
Author contributions
Conceptualization: K.O., A.R., G.R. Data curation: K.O., A.R., T.F., S.Y. Formal analysis: K.O. Investigation: K.O., A.R., T.F., S.F., G.R. Methodology: K.O., A.R. Project administration: K.O., A.R. Resources: S.Y., T.F. Supervision: T.F., A.R., G.R. Visualization: K.O., S.Y. Writing – original draft: K.O., T.F. Writing – review & editing: K.O., A.R., T.F., S.Y., G.R.
Funding statement
There is no funding to report.
Competing interests
All authors have no conflicts of interest or financial interests to disclose.

