Journal Scan – This Month in Other Journals, July 2022

1. Huynh J, Donovan J, Phu NH, et al. Tuberculous meningitis: progress and remaining questions. Lancet Neurol 2022;21:450–64. Available from: http://dx.doi.org/10.1016/S1474-4422(21)00435-X

Tuberculosis affects 10 million people globally each year, of which an estimated 2–5% have tuberculous meningitis. The true incidence of tuberculous meningitis is unknown; however, tuberculous meningitis is the leading cause of bacterial brain infections in settings with a high tuberculosis burden, disproportionately affecting young children and individuals with HIV. Here, the authors review advances made in the past 7 years concerning the pathogenesis, diagnosis, and treatment of tuberculous meningitis, emphasizing areas of uncertainty and updating Reviews published in The Lancet Neurology in 2005 and 2013. This Review focuses mainly on adult tuberculous meningitis and briefly emphasizes novel research advances in pediatric tuberculous meningitis, including important clinical trials on its management.

Confirming a diagnosis of tuberculous meningitis is challenging because it requires detection of M tuberculosis in CSF. CSF Ziehl-Neelsen staining and microscopy is rapid, inexpensive, and can be performed in many laboratories with few resources. However, a study of 618 individuals with tuberculous meningitis in Vietnam, South Africa, and Indonesia reported that its sensitivity was generally poor (ie, approximately 30%) and was not improved by adaptations to enhance staining of intracellular bacteria. PCR-based tests, such as GeneXpert MTB/RIF and GeneXpert MTB/RIF Ultra (Cepheid, Sunnyvale, CA, USA), are rapid and offer identification of rifampicin resistance. Although these tests are useful when positive, the negative predictive values of GeneXpert MTB/RIF is insufficient to rule out tuberculous meningitis. Large-volume CSF sampling and meticulous processing steps are essential to optimize the performance of smear, culture, and nucleic acid amplification tests.

Brain MRI features that predict future outcome and treatment response are poorly defined. Tools based on artificial intelligence and machine learning are being developed to enable an unbiased and automated assessment of brain images and provide guidance to clinicians on treatment response and outcomes. Artificial intelligence and machine learning, particularly when applied to digital brain imaging, have been implemented to devise diagnostic, complication prediction, and outcome prediction systems for neurodegenerative disorders, demyelinating disorders of the brain, and acute ischemic strokes. Machine learning in tuberculous meningitis is likely to assist in diagnosis and individualized treatment decisions in the future.

1 figure, 3 tables, no imaging

2. Ehresman J, Catapano JS, Baranoski JF, et al. Treatment of spinal arteriovenous malformation and fistula. Neurosurg Clin N Am 2022;33:193–206. Available from: https://doi.org/10.1016/j.nec.2021.11.005

Multiple classification systems have been created to separate spinal arteriovenous malformations (AVMs) based on location, angioarchitecture, and size. The first classification system was created by Di Chiro in 1971 after Di Chiro and colleagues first reported 4 years earlier on the selective angiography of spinal AVMs. Lesions were classified into 3 categories: type I, single-vessel arteriovenous fistula (AVF); type II, glomus AVM; and type III, juvenile AVM. More than 15 years later, in 1987, Di Chiro’s group added a fourth type of AVM, those with direct feeders from the anterior spinal artery (ASA). In 2002, Spetzler and colleagues created a schema based on surgical experience that classified the vascular lesions by anatomic locations surrounding or within the spinal cord. This classification is the one primarily used in this review, and it includes (1) extradural AVFs, (2) intradural ventral AVFs, (3) intradural dorsal AVFs, (4) extradural-intradural AVMs, (5) intramedullary AVMs, and (6) conus medullaris AVMs.

Extradural AVFs involve a direct connection between a radicular artery branch and the epidural venous plexus. Venous engorgement or congestion can cause increased retrograde pressure in the medullary veins within the spinal cord, leading to compressive symptoms.

Intradural dorsal AVFs, also known as type I AVMs or spinal dural AVFs, are the most common type of spinal AV shunt. These fistulas are located at the dural nerve root sleeve and involve a radicular feeder artery with a direct connection to a medullary vein.

Intradural ventral AVFs, often referred to as perimedullary type IV AVFs, involve a direct connection between the ASA and the coronal (pial) venous plexus.

Extradural-intradural AVMs, otherwise known as type III, juvenile, or metameric AVMs, are often large and complex lesions spanning multiple tissue layers.

Intramedullary AVMs, or type II glomus AVMs, include a classic nidus within the spinal cord and may include multiple arterial branches of the ASA and the posterior spinal artery (PSA) that drain into the coronal venous plexus.

Conus medullaris AVMs are complex lesions with multiple direct feeder arteries from the ASA, PSA, and radicular arteries that drain into a complex venous network near the conus medullaris and cauda equina.

6 figures (illustrations) and 6 tables listing relevant publications for each vascular malformation

3. Dono A, Alfaro-Munoz K, Yan Y, et al. Molecular, histological, and clinical characteristics of oligodendrogliomas: a multi-institutional retrospective study. Neurosurgery 2022;90:515–22

Oligodendrogliomas comprise 5.3% of gliomas. Based on the 2016 World Health Organization (WHO) fourth revised edition, the diagnosis of oligodendroglioma requires IDH1/IDH2 mutation and 1p/19q-co- deletion.  Recent studies have shown that genetic alterations can correlate with outcomes in molecularly defined gliomas (eg, CDKN2A/B loss in isocitrate dehydrogenase [IDH]-mutant astrocytomas and EGFR amplification or TERT promoter mutation in IDH wild-type astrocytomas). However, studies analyzing mutations that correlate with survival in molecular oligodendrogliomas (mODG) are limited. One hundred seven mODGs (2008-2019) diagnosed at 2 institutions were included. A retrospective review of clinical characteristics, molecular alterations, treatments, and outcomes was performed.

The authors results show a benefit of TMZ vs observation in progression-free survival (PFS). No difference in PFS or overall survival (OS) was observed between radiation and radiation/TMZ. PIK3CA mutations were detected in 15 (14%) mODG, and shorter OS was observed in PIK3CA-mutant compared with PIK3CA wild-type mODGs (10.7 years vs 15.1 years, P = .009).

In this study, there was no benefit of RT, TMZ, or the combination of both in OS. Despite the small sample size, the results are concordant with a recent retrospective study which showed no benefit from TMZ, RT, or their combination in WHO grade 2 mODG.20 Importantly, recent studies have demonstrated that TMZ is an independent risk factor for malignant transformation and mismatch repair pathway defects, which in turn could lead to worse survival.

2 tables, 1 figure, no imaging

4. Delgardo M, Higgins D, McCormick KL, et al. Clinical characteristics, outcomes, and pathology analysis in patients with dorsal arachnoid web. Neurosurgery 2022;90:581–87. Available from: https://journals.lww.com/10.1227/neu.0000000000001884

Seventeen cases of DAW between 2015 and 2019 at a tertiary medical center were retrospectively identified through a case log search. Patient characteristics, preoperative imaging, operative notes, and pathology reports were collected.

The mean age of the cohort was 50.5 years (IQR = 16) and presented primarily with back pain (64.7%). On imaging, all patients were found to have the “scalpel sign,” and nearly half had a syrinx present (41.2%). All DAWs were located in the thoracic spine, with the most common location being the midthoracic (70.6%). The mean follow-up length for all patients was 4.3 months. There were no preoperative symptoms significantly associated with postoperative symptom resolution; however, a trend was noted with the presence of a preoperative syrinx. Pathology samples consistently demonstrated fibroconnective or collagenous tissue with no evidence of inflammation or neoplasm.

3 figures, 4 tables with MRI

See also from the March podcast: Voglis S, Romagna A, Germans MR, Carreno I, Stienen MN, Henzi A, et al. Spinal arachnoid web—a distinct entity of focal arachnopathy with favorable long-term outcome after surgical resection: analysis of a multicenter patient population. Spine J [Internet]. 2022 Jan;22(1):126–35.

5. Cohen-Cohen S, Helal A, Yin Z, et al. Predicting pituitary adenoma consistency with preoperative magnetic resonance elastography. J Neurosurg 2021 Oct;1–8. Available from: https://thejns.org/view/journals/j-neurosurg/aop/article-10.3171-2021.6.JNS204425/article-10.3171-2021.6.JNS204425.xml

This study aimed to evaluate the efficacy of MR elastography (MRE) in predicting tumor consistency and its potential use in a series of patients with pituitary adenomas. Thirty-eight patients with pituitary adenomas (≥ 2.5 cm) were prospectively evaluated with MRI and MRE before surgery. Absolute MRE stiffness values and relative MRE stiffness ratios, as well as the relative ratio of T1 signal, T2 signal, and diffusion-weighted imaging apparent diffusion coefficient (ADC) values were determined prospectively by calculating the ratio of those values in the tumor to adjacent left temporal white matter. Tumors were classified into three groups according to surgical consistency (soft, intermediate, and firm).

The mean maximum tumor diameter was 3.7 cm, and the mean preoperative tumor volume was 16.4 cm3. Cavernous sinus invasion was present in 20 patients (52.63%). A gross-total resection was possible in 9 (23.68%) patients. The entire cohort’s mean absolute tumor stiffness value was 1.8  kilopascals (kPa) (range 1.1–3.7 kPa), whereas the mean tumor stiffness ratio was 0.66 (range 0.37–1.6). Intraoperative tumor consistency was significantly correlated with absolute and relative tumor stiffness. Tumor consistency alone was not a significant factor for predicting gross-total resection. Patients with intermediate and firm tumors had more complications compared to patients with soft tumors (50.00% vs 12.50%) and also had longer operative times.

The most common complication was a postoperative CSF leak, which occurred in 4 patients (10%), and which required either a reoperation or a lumbar drain placement for management. The next most common complication was a new pituitary deficit postoperatively, which occurred in 3 patients (7.9%). One patient (2.6%) developed a transient cranial nerve (CN) III palsy. From the 6 patients with intermediate and firm tumors, 3 (50%) had a complication, compared to 4 (12.50%) in the soft tumor group.

The authors conclude that this study reveals that MRE is a valuable tool for preoperative prediction of tumor consistency in pituitary macroadenomas. Unfortunately, conventional T1- and T2- weighted signal intensity, even with the high resolution of 3T MRI, could not accurately predict tumor consistency, thus limiting its utility as a surgical planning tool. Nevertheless, knowledge of tumor consistency in the preoperative setting can have vast clinical applications like improving surgical planning, counseling the patient about potential surgical risks, and estimating the length of operation.

2 figures, 7 tables, with MR

6. Catapano JS, Rutledge C, Rumalla K, et al. External validation of the Lawton brainstem cavernous malformation grading system in a cohort of 277 microsurgical patients. J Neurosurg 2021 Oct;1–9. Available from: https://thejns.org/view/journals/j-neurosurg/aop/article-10.3171-2021.3.JNS204291/article-10.3171-2021.3.JNS204291.xml

In 2015, a BSCM grading system was proposed to help predict which patients might have favorable neurological outcomes after microsurgical resection. The so-called Lawton or Lawton-Garcia BSCM grading system consists of 5 variables with a total of 7 points assigned for size (maximum axial diameter ≤ 2 cm vs > 2 cm), lesion crossing the brainstem midpoint, the presence of a venous malformation or developmental venous anomaly (DVA), patient age, and time from the most recent hemorrhage to surgery. A total of 7 points can be assigned because age was weighted by coefficients in the multivariable model, resulting in 2 points for patients older than 40 years, and the time between hemorrhage and surgery has 3 intervals: acute (< 3 weeks), subacute (3–8 weeks), and chronic (> 8 weeks) or unruptured. BSCM grades ranged from 0 to VII, with good outcomes in patients with low-grade BSCMs (grades 0–II), poor outcomes in patients with high-grade BSCMs (grades VI and VII), and decreasing proportions of favorable outcomes for those with intermediate- grade BSCMs (grades III–V). The BSCM grading system was previously developed and internally validated from a retrospective analysis of 104 patients. Risk prediction models should be internally and externally validated before integration into patient care. The purpose of the present study was to externally validate the BSCM grading system in a larger cohort of 277 patients.

Most BSCMs were located in the pons (160/277, 57.8%), 77 (27.8%) were in the midbrain, and 40 (14.4%) were in the medulla. The mean (SD) lesion diameter was 17mm. Overall, 85 BSCMs (31%) were > 2 cm in diameter, and 157 BSCMs (57%) crossed the axial midpoint. Venous malformations or DVAs were identified radiographically in 155 patients (56%). Patients were evenly divided according to age, with 143 (52%) 40 years of age or younger and 134 (48%) older than 40 years. The interval from the last hemorrhage to surgery was acute in 54 patients (19%), subacute in 37 (13%), and chronic in 186 (67%).

A binary logistic regression model was fit with the variables from the original BSCM grading system. The model accurately and significantly predicted whether a patient would have an unfavorable functional outcome. The ROC analysis demonstrated acceptable discrimination for predicting unfavorable functional outcomes (mRS score > 2) with an AUROC of 0.74 (95% CI 0.68–0.80).

This study validates the Lawton BSCM grading system in a large cohort of patients from two high-volume neurosurgery centers. The BSCM grade predicted neurological outcomes with accuracy comparable to that of other grading systems in widespread use.

5 figures, 3 tables, no imaging

7. Cluceru J, Interian Y, Phillips JJ, et al. Improving the noninvasive classification of glioma genetic subtype with deep learning and diffusion-weighted imaging. Neuro Oncol 2022;24:639–52. Available from: https://academic.oup.com/neuro-oncology/article/24/4/639/6398212

Since the restructuring of the categorization of gliomas by the World Health Organization (WHO) in 2016 to include variations in underlying genetic and epigenetic alterations,1 the consortium that informs the WHO has begun to place even greater emphasis on the delineation of glioma categories by a mutation in isocitrate dehydrogenase 1 and/or 2 (IDH1 and/or 2) and codeletion of 1p and 19q chromosomal arms, prioritizing these features over grade. In contrast to the WHO 2016 guidelines that first stratify by grade and then use genetic alterations to further differentiate patients within a designated grade, the new 2021 WHO guidelines now recommend that the first diagnostic delineation relies on IDH-mutation, followed by 1p19q-codeletion status, as supported by evidence that these distinct genetic subtypes indicate drastic differences in overall survival and response to therapy. Due to this increasing emphasis on genetic alterations as a diagnostic tool, it has become a clinical standard to perform genetic testing on tissue acquired during surgery to decide subsequent treatment. Because genetic testing requires an invasive process, an alternative noninvasive approach is attractive, particularly if resection is not recommended.

The goal of this study was to evaluate the effects of training strategy and incorporation of biologically relevant images on predicting genetic subtypes with deep learning. The authors dataset consisted of 384 patients with newly diagnosed gliomas who underwent preoperative MRI with standard anatomical and diffusion-weighted imaging, and 147 patients from an external cohort with anatomical imaging. Using tissue samples acquired during surgery, each glioma was classified into IDH-wildtype (IDHwt), IDH-mutant/1p19q-noncodeleted (IDHmut-intact), and IDH-mutant/1p19q-codeleted (IDHmut-codel) subgroups. After optimizing training parameters, top performing convolutional neural network (CNN) classifiers were trained, validated, and tested using combinations of anatomical and diffusion MRI with either a 3-class or tiered structure.

The best model used a 3-class CNN containing diffusion-weighted imaging as an input, achieving 85.7% (95% CI: [77.1, 100]) overall test accuracy and correctly classifying 95.2%, 88.9%, 60.0% of the IDHwt, IDHmut-intact, and IDHmut-codel tumors. In general, 3-class models outperformed tiered approaches by 13.5%-17.5%, and models that included diffusion-weighted imaging were 5%-8.8% more accurate than those that used only anatomical imaging.

They conclude that training a classifier to predict both IDH-mutation and 1p19q-codeletion status outperformed a tiered structure that first predicted IDH-mutation, then 1p19q-codeletion. Including apparent diffusion coefficient (ADC), a surrogate marker of cellularity, more accurately captured differences between subgroups.

4 figures, 2 tables, with MR images

8. van Solinge TS, Nieland L, Chiocca EA, et al. Advances in local therapy for glioblastoma — taking the fight to the tumour. Nat Rev Neurol 2022;18:221–36

Glioblastoma remains one of the most lethal neurological malignancies despite decades of unrelenting effort by the research and medical communities to combat this disease. Few new therapies have shown efficacy for mitigating glioblastoma since the introduction of temozolo­ mide as part of the Stupp standard of care protocol in 2005. Before the Stupp protocol was introduced, median survival was around 12 months, which has since increased to 16 months owing to various improvements in treatment, including optimization of the Stupp protocol, advances in imaging and radiotherapy, and gross total resection safeguarded by intraoperative mapping. In addition, tumor treating fields therapy, in which mitosis is hindered by alternating electric fields, has produced improvements in long-term overall survival (OS) in patients with primary or recurrent glioblastoma. Although these improvements are encouraging, the long-term prospects for patients with glioblastoma remain extremely poor. The absence of new treatment modalities for glioblastoma cannot be attributed to lack of effort: currently, 1,593 trials are registered under “glioblastoma” on ClinicalTrials.gov. The resistance of glioblastoma to treatment is widely known and can be explained by several distinctive characteristics of the tumor. Glioblastoma is notoriously heterogeneous, with an abundance of signaling pathways even within the same tumor mass, thereby limiting the options for targeted therapies. The tumor microenvironment strengthens the resistance of glioblastoma resistance to radiation and chemotherapy, and the low immunogenicity of glioblastoma hinders a strong immunological response. In addition, infiltration of glioma (stem) cells deep into the brain excludes effective treatment by resection alone. Moreover, the blood–brain barrier (BBB) prevents many systemically administered chemotherapeutics from reaching sufficient concentrations in the brain without serious adverse effects.

In this review the authors discuss current and future local therapies for glioblastoma, examining treatment of the tumor cavity and other direct approaches to the tumor. They highlight landmark studies to provide an overview of local therapies that have been— or are currently being — explored in patients with glioblastoma.

3 figures, 5 tables, no imaging

The American Society of Neuroradiology is accredited by the Accreditation Council for Continuing Medical Education (ACCME) to provide continuing medical education for physicians. Visit the ASNR Education Connection website to claim CME credit for this podcast.

Journal Scan – This Month in Other Journals, July 2022
Jeffrey Ross
Fatal error: Uncaught Error: Call to undefined function get_cimyFieldValue() in /home2/ajnrblog/public_html/wp-content/themes/ample-child/author-bio.php:13 Stack trace: #0 /home2/ajnrblog/public_html/wp-content/themes/ample-child/content-single.php(35): include() #1 /home2/ajnrblog/public_html/wp-includes/template.php(812): require('/home2/ajnrblog...') #2 /home2/ajnrblog/public_html/wp-includes/template.php(745): load_template('/home2/ajnrblog...', false, Array) #3 /home2/ajnrblog/public_html/wp-includes/general-template.php(206): locate_template(Array, true, false, Array) #4 /home2/ajnrblog/public_html/wp-content/themes/ample/single.php(21): get_template_part('content', 'single') #5 /home2/ajnrblog/public_html/wp-includes/template-loader.php(106): include('/home2/ajnrblog...') #6 /home2/ajnrblog/public_html/wp-blog-header.php(19): require_once('/home2/ajnrblog...') #7 /home2/ajnrblog/public_html/index.php(17): require('/home2/ajnrblog...') #8 {main} thrown in /home2/ajnrblog/public_html/wp-content/themes/ample-child/author-bio.php on line 13