Link, Michael J.

Abstract Systematic reviews and meta-analyses in the neurosurgical literature have surged in popularity over the last decade. It is our concern that, without a renewed effort to critically interpret and appraise these studies as high or low quality, we run the risk of the quality and value of evidence-based medicine in neurosurgery being misinterpreted. Correspondingly, we have outlined 4 major domains to target in interpreting neurosurgical systematic reviews and meta-analyses based on the lessons learned by a collaboration of clinicians and academics summarized as 4 pearls. The domains of (1) heterogeneity, (2) modeling, (3) certainty, and (4) bias in neurosurgical systematic reviews and meta-analyses were identified as aspects in which the authors’ approaches have changed over time to improve robustness and transparency. Examples of how and why these pearls were adapted were provided in areas of cranial neuralgia, spine, pediatric, and neuro-oncology to demonstrate how neurosurgical readers and writers may improve their interpretation of these domains. The incorporation of these pearls into practice will empower neurosurgical academics to effectively interpret systematic reviews and meta-analyses, enhancing the quality of our evidence-based medicine literature while maintaining a critical focus on the needs of the individual patients in neurosurgery.

A 60-year-old man with a history of four prior operations for a left cerebellar/middle cerebellar peduncle hemangioblastoma presented with hearing loss, imbalance, and ataxia (de la Monte and Horowitz, 1989). Magnetic resonance imaging (MRI) demonstrated a 3-cm cystic mass with heterogeneous enhancement in the same location. We resected the mass via reopening of the retrosigmoid approach (Lee et al., 2014). Left cranial nerves IV, V, VII, VIII, IX, X, and XI were all well identified and preserved, and feeding arteries from the brainstem were meticulously coagulated and transected without violating the tumor-brainstem interface (Chen et al., 2013). Preoperative embolization greatly aided safe resection of the mass, whose pathology revealed recurrence of hemangioblastoma (Eskridge et al., 1996; Kim et al., 2006; Sakamoto et al., 2012). The video can be found here: https://youtu.be/3mZgY15xOZc .

Cavernous malformations (CMs) are low-pressure, focal, vascular lesions that may occur within the brainstem and require treatment, which can be a substantial challenge. Herein, we demonstrate the surgical resection of a hemorrhaged brainstem CM through a posterior petrosectomy approach. After dissection of the overlying vascular and meningeal structures, a safe entry zone into the brainstem is identified based on local anatomy and intraoperative neuronavigation. Small ultrasound probes can also be useful for obtaining real-time intraoperative feedback. The CM is internally debulked and resected in a piecemeal fashion through an opening smaller than the CM itself. As brainstem CMs are challenging lesions, knowledge of several surgical nuances and adoption of careful microsurgical techniques are requisite for success. The video can be found here: https://youtu.be/szB6YpzkuCo .

OBJECTIVE Tumor-associated macrophages (TAMs) have been implicated as pathologic actors in phenotypically aggressive vestibular schwannoma (VS), potentially mediated via programmed death-ligand 1 (PD-L1). The authors hypothesized that PD-L1 is a key regulator of the VS immune microenvironment. METHODS Forty-six consecutive, radiation-naïve, sporadic VSs that were subtotally resected at primary surgery were assessed via immunohistochemical analysis, including analysis of CD163 and PD-L1 expression. Pathologic data were correlated with clinical endpoints, including tumor control, facial nerve function, and complications. RESULTS Baseline parameters were equivalent between stable and progressive post–subtotal resection (STR) VS. CD163 percent positivity and M2 index were significantly increased among tumors that remained stable (34% vs 21%, p = 0.02; 1.13 vs 0.99, p = 0.0008), as well as patients with favorable House-Brackmann grade I or II facial nerve function (31% vs 13%, p = 0.04; 1.11 vs 0.97, p = 0.05). PD-L1 percent positivity was significantly associated with tumor progression (1% vs 11%, p = 0.01) and unfavorable House-Brackmann grade III–VI facial nerve function (1% vs 38%, p = 0.02). On multivariate analysis, PD-L1 was independently significant in all models (likelihood ratio 4.4, p = 0.04), while CD163 was dependent in all iterations. CONCLUSIONS In contrast to prior reports, in this study, the authors observed significantly increased levels of M1, CD163 + TAMs in association with VS that progressed after STR. Progressive tumors are characterized by increased PD-L1, potentially highlighting a mechanism of immune evasion that results in TAM deactivation, tumor growth, and further infiltration of anti-tumor immune cells. Targeting PD-1/PD-L1 may offer therapeutic promise, particularly in the setting of disease control after STR.