Eye (2013) 27, 1–13 & 2013 Macmillan Publishers Limited All rights reserved 0950-222X/13 www.nature.com/eye 1;2 1 2 Systematic review of AIM Miguel , F Henriques , LFR Azevedo , REVIEW AJR Loureiro1 and DAL Maberley3 Purtscher’s and Purtscher-like retinopathies Abstract Purpose (1) To describe the clinical necessary, but may not be feasible to characteristics of Purtscher and Purtscher-like conduct. retinopathies, including etiologies, Eye (2013) 27, 1–13; doi:10.1038/eye.2012.222; fundoscopic signs, results of complementary published online 23 November 2012 investigation, treatments, and outcomes. (2) To compare visual acuity (VA) of patients Keywords: corticosteroids; Purtscher’s receiving corticosteroids for PuR compared retinopathy; Purtscher-like retinopathy; with observation. systematic review Methods Systematic review of several databases (1980–2010): Medline, EMBASE, ISI, EBSCO, Science Direct and Google Scholar. Study selection criteria included: (A) Studies of PuR with ophthalmology Z Introduction assessments; (B) 3 of 5 diagnostic criteria 1Ophthalmology of PuR; (C) Quantified VA at presentation. Purtscher’s retinopathy is an occlusive Department, Central For quantitative assessment (purpose 2), microvasculopathy associated with trauma. It University Hospital of we selected only studies that reported was first described in 1910 by Otmar Purtscher1 Coimbra, Coimbra, Portugal whether corticosteroids were administered, in a middle-aged man who fell off a tree and 2 and with vision assessments after at least suffered cranial trauma. There is some doubt Department of Health Information and Decision 1 month. about the incidence of this pathology; authors Sciences, Center for Results (1) From 670 studies initially found, have estimated its incidence at 0.24 persons per Research in Health 40 were included (68 cases, 110 eyes): 1 million per year (including Purtscher and Technologies and prospective, 5 case series, and 34 case reports. Purtscher-like retinopathies)2 while others Information Systems Mean VA at presentation was 1.3 logMAR believe that the incidence is higher because the (CINTESIS), Faculty of 3 Medicine, University of (logarithm of the minimum angle of condition may frequently be asymptomatic. Porto, Porto, Portugal resolution) (o20/200; range: 20/20-light Bilaterality is present in up to 60% of cases. perception). Purtscher flecken were Purtscher’s retinopathy generally occurs as a 3Retinal Division, underreported. Trauma and acute pancreatitis result of cranial trauma or thoracic compression. Department of were the most frequent etiologies. There When there is a non-traumatic etiology, the Ophthalmology and Visual were six deaths, all with systemic correct designation is Purtscher-like Sciences, University of British Columbia, associations. (2) There was no statistically retinopathy. Frequent causes of Purtscher-like Vancouver, British significant difference between VA retinopathy include acute pancreatitis, renal Columbia, Canada improvement for patients treated with failure, and autoimmune disease. Purtscher-like corticosteroids compared with observation. retinopathy in acute pancreatitis is an indicator Correspondence: Visual prognostic factors identified included of multiorgan failure and is often associated AIM Miguel, Rua Quinta 1 etiology (pancreatitis and trauma were with a fatal outcome.3 do Sardoal, VE 3, n 10, 4430-182 Vila Nova de associated with higher probability of visual The diagnosis of Purtscher’s and Purtscher- Gaia, Portugal. improvement) and male gender. like retinopathies (PuR) is clinical, with a Tel: þ 351 932482477. Conclusion Due to limitations of case presentation that usually includes sudden E-mail: myworld_ana@ reports and series, the presented data are vision loss of variable severity, hours to days hotmail.com only useful as broad characterizations of the after the causal pathology. Fundoscopic signs clinical course of PuR. Further studies, include cotton-wool spots and intraretinal Received: 7 March 2012 Accepted in revised form: possibly including trials to assess the hemorrhages, described in 83–92% of a series of 2 October 2012 4 effectiveness of corticosteroids use, and cases. Purtscher flecken are considered to be Published online: larger prospective cohort studies, are pathognomonic,4 but occur in only 50% of cases. 23 November 2012 Purtscher’s retinopathy—a systematic review AIM Miguel et al 2 It is widely assumed that the flecken are the result of Controlled Trials (in www.controlled-trials.com). occlusion of the precapillary arterioles. The characteristic The search query for Pubmed was ‘(Purtscher[tiab] OR finding is intraretinal whitening but with a clear zone Purtscher-like[tiab] OR angiopat*[tiab]) AND (within 50 mm) on either side of the retinal arterioles, retinopath*[tiab]’; similar searches were adjusted to each venules, and precapillary arterioles. This is in contrast to database. All searches were made from 1 to 4 May 2010. cotton-wool spots, which have ill-defined edges,4,5 and No publication date or language exclusions were made. are located superficially over vessels. Fluorescein We also hand-searched reference books, journals, and angiography may show leakage of dye from retinal bibliographies from relevant papers. We mailed (or sent arterioles, capillaries, and venules in patients with letters) to experts in the field for help in identifying the relatively mild Purtscher’s retinopathy. Arteriolar unpublished data and gray literature to avoid obstruction is noted in more severe cases. There are many publication bias. theories about pathogenesis but typically Purtscher’s microangiopathy is thought to be due to precapillary Selection criteria occlusion, due to fat emboli, leukoaggregation, or other mechanisms. Inclusion criteria were Treatment is variable with some physicians observing expectantly and others primarily administering - Case reports, case series or trials meeting predeter- corticosteroids. There have been non-systematic reviews mined quality criteria, based on the Cochrane Collabora- of Purtscher’s retinopathy;6,7 however, no recent tion,8 STARD,9 and STROBE.10 systematic review has been performed. There is a need - Patients that matched at least three of five diagnostic for an updated systematic review of Purtscher’s PuR criteria (adapted from several papers4,7,8): retinopathy given ongoing controversies regarding 1. Purtscher flecken treatment, and due to further published cases since prior 2. Retinal hemorrhages, low-to-moderate number (1–10) reviews. 3. Cotton-wool spots (typically restricted to posterior We present a systematic review of Purtscher’s pole) retinopathy and Purtscher-like retinopathies. Part one of 4. Probable or plausible explanatory etiology this review will describe the clinical characteristics of PuR, namely: clinical presentation, causal/associated 5. Complementary investigation compatible with pathologies, results of diagnostic testing, treatments diagnosis performed, and clinical outcomes. - Quantified VA at presentation. In part two of this review, our goal is to assess visual - Follow-up by an ophthalmologist with fundoscopic acuity (VA) changes of patients receiving corticosteroids evaluation, at least one month after initial diagnosis. compared with those without treatment (main outcome). Secondary comparisons between groups include an For the assessment of corticosteroid efficacy in PuR, we assessment of clinical changes and prognostic factors. only included studies that fulfilled another two inclusion criteria: studies with follow-up at 1 month (with all assessments and VAs registered), and studies reporting Materials and methods whether corticosteroids were administered (if studies did not explicitly note corticosteroids use for each patient, Search methods then they were excluded). The search methodology was guided by the Cochrane Exclusion criteria included: duplicate reports, studies Collaboration’s recommendations.8 Two independent without follow-up, and studies without examinations reviewers, AM and FH, examined each title and abstract performed by an ophthalmologist. To avoid language to exclude obviously irrelevant reports. They then bias, publication language was not an exclusion criteria. independently examined the full text of each article to determine eligibility according to the noted selection Data collection and analysis criteria below. Disagreements were solved by consensus and registered to allow assessment of reliability using the Two reviewers, AM and FH, independently collected kappa statistics. data and assessed quality. A form was developed Electronic searches included the following databases: specifically for this project to assist in consistent data Medline (via Pubmed)—1950 to May 2010, EMBASE— collection from each study. This form included: year of 1974 to May 2010; ISI—2000 to May 2010, and other publication, author’s name, as well as each subject’s age, databases via EBSCO, Science Direct, and Google Scholar. gender, retinal pathology, etiology of retinopathy, and We also searched clinicaltrials.gov and Current laterality, VA at presentation (VA0); VA at 1 month (VA1); Eye Purtscher’s retinopathy—a systematic review AIM Miguel et al 3 VA at 2 months(VA2); VA at 6 months(VA6), diagnostic analysis of each involved eye (‘all-eyes’), comparison of criteria of PuR, signs and results of tests, and treatment eyes with best VA (‘best-eye’), and worst VA (‘worst- (including dosing and administration route). The eye’). Best eye and worst eye analyses included cases of investigators had clear coding instructions.