xRead - September 2022

Cerebrospinal fluid rhinorrhea diagnosis

approaches have become the standard of care because of their high success rate and lower morbidity profile. In or der for these endoscopic CSF leak repairs to be successful, accurate diagnosis and localization are vital preoperative steps. It has been shown that failure to accurately localize the site of a leak can lead to surgical failure. 8 Given that rhinorrhea is a nonspecific finding, the pres ence of CSF in a sample must be confirmed before costly and even invasive testing is pursued. Once confirmed, the site of CSF fistula must be localized prior to repair. Commonly used modalities for CSF rhinorrhea confirmation include use of the ring sign, glucose testing, beta-2 transferrin, beta trace protein, and radionuclide cisternography. The site of CSF fistula can be localized using methods such as high resolution computed tomography (HRCT), magnetic res onance cisternography (MRC), CT cisternography (CTC), and intrathecal fluorescein (IF). Of note, with the exception of HRCT, these imaging and endoscopic modalities are ca pable of simultaneous confirmation and localization of CSF fistula. These options vary widely in availability, cost, and invasiveness. The purpose of this study is to thoroughly review the lit erature on the diagnosis of CSF leaks and provide evidence based recommendations through a structured and system atic review process. Our second objective was to define a diagnosis flowchart to help physicians diagnose and localize CSF rhinorrhea based on the current best evidence. Materials and methods The methodology described by Rudmik and Smith 9 for the development of an evidence-based review with recommen dations were followed in preparing and writing this article. Nine diagnostic and localization modalities commonly used in the setting of CSF leak were selected for inclusion in the analysis: ring sign, glucose testing, beta-2 transferrin, beta trace protein, RNC, HRCT, MRC, CTC, and IF. A systematic review of the literature was performed us ing PubMed, EMBASE, and Cochrane Review Databases from January 1, 1990 through September 30, 2014 using the terms “cerebrospinal fluid” and/or “cerebrospinal fluid rhinorrhea” in combination with each of the 9 diagnos tic and localization modalities (the search terms used are listed in Supporting material 1). This resulted in a total of 1313 abstracts, which were evaluated for relevance. The ex clusion criteria applied were language other than English, nonhuman studies, nonrhinologic source of the CSF leak, and case reports. The remaining 68 relevant studies were in cluded for evaluation and graded for level of evidence using reported research methodology for diagnostic studies. 10 Of note, many of these were cross-sectional studies, indicating level 2 evidence; however, they were downgraded because of lack of consistently applied reference standards and blinding. Following analysis of each article, summary tables were developed, as well as an aggregate grade of evidence, benefit-harm assessments, and value judgment for each

modality. The costs of these diagnostic methods were de termined based on average nationwide reimbursement from Centers for Medicare and Medicaid Services (CMS) for the given Current Procedural Terminology (CPT) codes. 11 When the available literature was sufficient, recommen dations were made based on published evidence or biblio graphic search. The literature was reviewed and the initial manuscript was prepared by 3 authors (G.M.O., J.A.A., and R.R.O). Additional authors were then asked to criti cally evaluate the recommendations based on their review of the literature (R.S. and R.H.) per the protocol for the above-mentioned online iterative process. CSF combined with blood can leave a “ring” or a “halo” on bed sheets or similar white media. This “ring sign” has been discussed as the first clue of a CSF leak, especially in a trauma patient. Although such an immediate indica tor would be convenient, the reliability of this sign is poor. Only 1 study examined this phenomenon. Dula and Fales 12 presented 1 drop combinations of blood and either CSF, saline, tap water, or nasal fluid to 2 blinded Emergency Department physicians for analysis of ring sign. They also tested different concentrations of CSF and blood and var ious white media, including standard filter paper, coffee filters, bed linens, and paper towels for quality of ring sign. They found that, although a mixture of blood and CSF does reliably produce a ring sign in CSF concentrations of 30% to 90% on any of the tested filter media, so does blood mixed with any of the other clear fluids. The ring sign was similar among any of these combinations, giving it very poor specificity for identifying CSF. No other studies have corroborated or challenged these conclusions (Supporting material 2). 12 Aggregate grade of evidence: N/A; Benefit: Rapid bedside test; Harm: High potential for misdiagnosis; Cost: None; Benefits-harm assessment: High potential for misdiagnosis outweighs any potential benefit of early information from this test; Value judgments: None; Recommendation level: Recommend against; Intervention: Avoiding the ring sign in diagnosing a CSF leak. Glucose testing Testing clear rhinorrhea for glucose was traditionally con sidered an option for quick diagnosis of CSF leak, due to the belief that only the presence of CSF would lead to a positive glucose result. However, this assumption has been challenged over the last few decades. Results Confirmation Ring sign

International Forum of Allergy & Rhinology, Vol. 6, No. 1, January 2016

9