The article titled “Systematic Review of the Predictive Value of Negative Brain or Low Probability Brain MRIs in Patients with CSF Venous Fistulas” (American journal of neuroradiology) is a systematic review analyzing data from nine studies (898 patients) to assess how reliably negative brain MRI or low-probability MRI scores (Bern, Mayo) can rule out cerebrospinal fluid (CSF) venous fistulas, a surgically treatable cause of spontaneous intracranial hypotension (SIH) 1).
The authors found that although negative brain MRI has a high negative predictive value (89%), CVFs were still identified in a notable number of patients—especially when using the Mayo score (NPV only 65%).
The authors conclude that negative imaging does not exclude CVF, and invasive spinal imaging (e.g., DSM, dynamic CT myelography) should be considered in cases with strong clinical suspicion. This is especially relevant for neurosurgeons, as surgical ligation of CVFs may be necessary.
All authors are affiliated with Cedars-Sinai Medical Center (Los Angeles), with Dr. Wouter I. Schievink, a neurosurgeon, being a leading expert in the field of CSF leaks.
This so-called systematic review by Tay et al. (AJNR, June 2025) ambitiously sets out to clarify the diagnostic value of negative brain MRI in detecting cerebrospinal fluid venous fistulas (CVFs), yet ends up as a textbook example of how quantitative gloss can’t save a weak clinical premise.
❌ Flawed Foundation: Heterogeneity Buried Under Pooled Numbers
The authors admit up front that their review suffers from substantial heterogeneity in reference standards, imaging protocols, and scoring systems across studies. Yet, they proceed to pool data as if they were comparing apples to apples—when in reality, they’re juggling an orchard of inconsistencies. The Bern and Mayo scores are fundamentally different tools, and merging their results under the same meta-analytical umbrella is methodologically indefensible.
❌ Diagnostic Performance? Or Statistical Mirage?
An NPV of 89% for negative brain MRI is presented as clinically reassuring. But what does that mean in a low-prevalence population, where even random guesses produce high NPV? The authors make no serious attempt to contextualize this value against pre-test probabilities, clinical severity, or the dire consequences of missed CVFs.
Even worse: they downplay the real-world implication that up to 35% of patients with “low-probability” MRI scores still had a fistula. If your scoring system misses one in three patients, it is not a diagnostic aid — it’s a liability.
❌ What Is “Low Probability” Anyway?
The review’s use of the Bern and Mayo scoring systems is shaky. These tools were not designed to exclude CVFs—they were created to stratify general SIH suspicion, often with dural leaks in mind. Using them to estimate the likelihood of a specific leak subtype (venous fistulas) is a classic case of misapplied tools, with conclusions that border on tautology.
❌ The Neurosurgical Perspective: Where’s the Relevance?
Despite being authored by a team from a prestigious center (Cedars-Sinai), including Dr. Wouter Schievink—a neurosurgeon well-known in the field—the review fails to provide any actionable guidance for surgical decision-making. There is no discussion of:
How to select patients for surgical exploration despite negative imaging.
What intraoperative findings to expect.
How to interpret “false negatives” in the context of patient disability.
For a condition where the definitive solution is surgical ligation, this omission is not just a gap—it’s a gaping hole.
❌ Conclusion: A Review That Confirms What We Already Knew—And Still Gets It Wrong
That patients with CVFs can have negative brain MRIs is not new. What this review adds is a false sense of security about when to stop looking. Instead of advocating for clinical vigilance and targeted imaging, it wraps known facts in misleading statistics and undercuts its own relevance by drowning in weak methodology.
This is not a clarifying contribution—it’s diagnostic noise dressed up as evidence-based guidance.
Verdict:
Methodologically shaky, clinically misleading, and surgically irrelevant. A missed opportunity from a group that should have known better.