Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
They certainly give very strange names to diseases
Plato (427–347 BC)
Seon-Kyu Lee and a formidable panel of authors undertook the laudable task of writing standards and guidelines for the management of an often poorly understood neurovascular entity which they chose to describe as intracranial ‘dural arteriovenous shunts’ (DAVS). I start and end this editorial by commending them on a clear, well-balanced, and concise review. The conclusions and recommendations are sensible, pragmatic and should be widely accepted and applied.
Between these book ends, an editorial always calls for candid critique, and I am in turn honored to be asked to comment on these standards and guidelines. I do so mostly regarding methodology and depth. But first, let me get out of the way the question of nomenclature. Yes, the long and widely used term ‘dural arteriovenous malformation’ (DAVM) may evoke inaccurate comparisons with the brain parenchymal arteriovenous malformations, fundamentally different lesions in terms of etiology, natural history, hemodynamics, pathobiology, and treatment strategies. And the word ‘malformation’ may wrongly imply to some a congenital anomaly, although it need not to. A more trendy term, ‘dural arteriovenous fistula’ (DAVF), unfortunately evokes the view of a single direct arteriovenous connection, belying the plexiform fistulous connections involving a pathologic dural leaflet in this disease. So who would disagree with the suggested term DAVS? If Lee and his broad panel of colleagues are for it, I won't argue with it.
The main critique I wish to bring forth relates to the methodology and rigor of the standards and guidelines, without questioning the veracity or gravitas of opinions by the very accomplished authors. They certainly adopted widely accepted and relevant American Stroke Association endorsed levels of evidence, including size of treatment effect (Classes I–III) and estimates of certainty (Levels A–C). But it is unclear if these were based on a systematic review of the literature, all the literature, or the arbitrary choice of evidence favored by authors of different sections of the manuscript, supporting their pre-existing opinions. What search terms were used, engine, covering what dates, and what were the criteria for considering or excluding a particular article? Understanding that the practicalities of modern publication limit manuscripts to a certain number of references (in this case 60 were cited), there should be an expressed methodology for limiting cited papers to certain types of studies meeting pre-articulated criteria of rigor, such as the number of patients, follow-up or content, or citation impact. Even case studies can be prioritized to cite the first, most recent, or other systematic reviews of the particular aspect of the disease. Instead, we see a hodge-podge of references dating back to the 1970s, but missing major papers on the topic such as those by Lasjaunias in the 1980s. Only one paper by each of Mullan and Djindjian are quoted, despite their acknowledged prolific, highly original, and fundamental contributions, while numerous historically contemporaneous references by others are included. The paper seems also to selectively ignore more recent formidable case series, novel concepts, and experimental models published in the past decade, such as those by Lawton. This reviewer cannot resist pointing out the lack of reference to a paper published in 1990, cited more than 600 times in other peer-reviewed publications, including the first systematic review and meta-analysis of all prior published reports of DAVMs.1 The paper presented for the first time actual data and statistical analyses demonstrating that lesions with leptomeningeal venous drainage, and no other features, correlated with an aggressive clinical course, which remains the most relevant principle governing the management of this pathology more than 25 years later. Even the first and—to my knowledge—only published monograph on this entity2 is not mentioned or cited. I am sure there were good objective reasons for citing other literature but not these. Such reasons would have been uncontestable if a methodology of literature search and citation had been articulated and followed, such as the preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA-statement.org).
With a potentially biased literature foundation, another question arises regarding the consensus methodology. How were different sections of the manuscript assembled and who drove the classification of evidence? Did everyone agree with every opinion? How was overt or covert dissent with a particular recommendation (or nuance thereof) handled? Here again, there are published methodologies such as the Delphi process which introduce rigor and minimize bias and the propagation of opinions weighed by one expert over another. These have recently been applied beautifully to complex neurovascular topics,3 and could have been followed here.
Finally, some relevant scientific and clinical knowledge gaps are not presented or discussed. The pathobiology of this entity, including enhanced angiogenesis activity in the setting of dural venous channel thrombosis, is not thoroughly reviewed. Nor are questions asked about the role of hypercoagulability in lesion genesis and recurrence. These have clinical relevance, as it remains unclear and the standards and guidelines are silent on whether a systematic search for hypercoagulability or associated infection or malignancy is important or useful when facing a DAVS diagnosis.
The field has been influenced by several named classifications; some are summarized in the paper (table 3) and there are others. Yet there is little questioning of the relative weight or implications of various features of each classification, and little is presented to indicate that any aspect of these elaborate schemes is superior to a more simplistic concept. In fact, the recommendations by the authors do not endorse anything more complex than symptomatic (yes/no) and leptomeningeal venous drainage (yes/no) classification as primary factors in the natural history and in the decision to treat.
In the end, notwithstanding the above critiques, the authors are commended on six common sense recommendations, which are supported by the best available evidence, and these should guide the current management of these lesions. Articulating these recommendations and simple conclusions in one concise document, endorsed by august experts, is a valuable service to the clinical community.
Competing interests None declared.
Provenance and peer review Not commissioned; internally peer reviewed.