Abstract
Background Torcular dural sinus malformations (tDSMs) are congenital complex vascular anomalies often referred as a single unit. Nevertheless, they possess distinct anatomical features, clinical diversity, and markedly different outcomes.
Objective On the basis of our institutional experience and analysis of published data, we propose a grading system. Methods We have identified 44 papers to which we added our four institutional cases for a total of 126 patients.
Eight predictor variables were studied. In order to assess their individual impact on mortality and possible correlations, a logistic regression model was constructed through a stepwise forward process.
Results Overall mortality was 22.1%. Mortality was higher in tDSM patients diagnosed postnatally, 40.7% versus a 15.6% in prenatally found cases (p = 0.007).
We divided the patients into four grades. Grade I comprised patients with no feeder evidence and possessed the best outcomes (mortality of 7.55%).
Mortality rose for grades II and III defined respectively by scarce and multiple feeders. Brain damage was the defining feature of grade IV.
A mortality of 75% could be observed within this grade. Grade IV was further divided into grades IVa (antenatal) and IVb (postnatal cases).
Furthermore, our logistic regression model found that brain damage (OR 11.3, p < 0.001, 95% CI 2.97-42.91) and patent feeders (OR 4, p = 0.03, 95% CI 1.15-13.86) were major determinants of poor outcome (area under ROC curve of 81.44%). Conclusion The grading system (tDSM-GS) streamlines classification into four different grades facilitating both diagnosis, clinical decision-making, and proper prognostication.