Case Series The efficacy of optic nerve ultrasonography for differentiating papilloedema from pseudopapilloedema in eyes with swollen optic discs Meira Neudorfer, Maytal Siegman Ben-Haim, Igal Leibovitch and Anat Kesler Department of Ophthalmology, Tel Aviv-Sourasky Medical Center, Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel ABSTRACT. Purpose: To evaluate the diagnostic yield of optic nerve ultrasonography (US) in distinguishing between papilloedema (swollen discs owing to raised intracranial pressure) and pseudopapilloedema. Methods: We prospectively evaluated all patients with bilateral optic disc swelling who underwent a complete neuro-ophthalmological examination. Suitable patients were referred for neuroimaging (computerized tomography or magnetic resonance imaging) and lumbar puncture. They underwent optic nerve US (A-mode and B-mode), and the findings were compared with the final clinical assessment. Sensitivity, specificity and predictive values for US distinction between true papilloedema and pseudopapilloedema were calculated and compared with those of the other imaging tests. Results: Forty-four patients were enrolled. Ultrasonography detected papilloedema with a high degree of sensitivity (85%) when the normal optic nerve width (ONW) was set at 3.3 mm, and with an even higher degree of sensitivity (95%) when the normal ONW was set at 3.0 mm. Ultrasonography had a high negative predictive value for detecting papilloedema: 83% when the normal ONW was set at 3.3 mm and 93% when it was set at 3.0 mm. There was a significant correlation between the US findings and the final diagnosis (p < 0.001) when the upper limit of the normal ONW was set at 3.0 mm. Conclusions: Ultrasonography findings of the ONW correlated well with the final diagnosis of papilloedema or pseudopapilloedema, especially when the upper limit of the normal ONW was set at 3.0 mm. Ultrasonography could be a useful non-invasive technique for differentiating papilloedema from other causes for swollen discs, such as pseudopapilloedema. Key words: papilloedema pseudopapilloedema swollen discs ultrasonography Acta Ophthalmol. 2011: 91: 376 380 ª 2011 The Authors Acta Ophthalmologica ª 2011 Acta Ophthalmologica Scandinavica Foundation doi: 10.1111/j.1755-3768.2011.02253.x [Correction added after EarlyView publication on 27 September 2011: The name of the author Maytal S. Ben-Haim has been corrected to Maytal Siegman Ben-Haim]. Introduction Funduscopy is generally performed to examine the optic nerve head and may identify a swollen disc secondary to papilloedema or pseudopapilloedema usually owing to optic nerve head drusen or congenital crowded discs (Kline et al. 2005). Optic nerve ultrasonography (US) allows clinicians to distinguish between an increase in the volume of the subarachnoid fluid and an increase in the width of the optic nerve or its sheath. The reference value of the upper limit of a normal optic nerve width (ONW) as seen on US has been set at 3.3 mm (Byrne & Green 2002). Optic nerve US has been used in the past to detect optic nerve swelling (Newman et al. 2002). However, its diagnostic yield in differentiating papilloedema from pseudopapilloedema in eyes with swollen optic discs has not been studied before. In this study, we investigated whether optic nerve US could contribute to the differential diagnosis of papilloedema and other conditions associated with swollen optic discs. We also determined the sensitivity, specificity and predictive values of US in comparison with other techniques commonly employed in this clinical setting. 376
Materials and Methods We carried out a combined historical (earlier test findings) and prospective (optic nerve US) study on all consecutive patients who presented to our department between 2003 and 2007 with bilateral optic disc swelling on funduscopy. Inclusion criteria were compliance to undergo the full procedure of ocular US, the availability of a comprehensive medical anamnesis and long-term follow-up data (mean ± SD 1.5 ± 1.6 years). Exclusion criteria were a history of any central nervous system or ocular disease, previous ocular surgery or trauma. Ultrasonography was performed in a standardized fashion (Byrne & Green 2002) after obtaining informed consent from the patients. Approval to conduct the study was obtained from the institutional review board. The retrieved data included the patient s medical history and the findings of a physical and neurological examination. The comprehensive ophthalmologic examination included funduscopy, brain and orbit imaging by computerized tomography (CT), CT angiography, magnetic resonance imaging (MRI) and an MR venogram. According to the results of the clinical assessment, suitable patients were referred for neuroimaging (CT or MRI) and lumbar puncture (LP), during which the cerebrospinal opening pressure value was measured. All enrolled patients underwent optic nerve US (A-mode and B-mode) that was performed with an I 3 system ABD (Diagnostic Ophthalmic Ultrasound; I 3 Innovative Imaging Inc., Sacramento, CA, USA) equipped with a 7.5-MHz transducer. The B-scan provides a bidimensional anatomic image of the optic nerve head, and the A-scan enables the cross-sectional measurement of the ONW. Detection of the doughnut or crescent sign in the B-Mode, an ONW of >3.3 mm in the retrobulbar area as observed in the A-Mode, or a positive 30-degree test was considered positive US findings. The 30-degree test is based on the assumption that when the eye turns, the optic nerve and its sheaths are stretched, thus distributing the increased subarachnoid fluid over a greater area. The ONW was measured twice: in primary position and 30 towards the probe. A reduction of more than 10% was considered positive. The analysed data were comprised of the results of imaging studies, LP values and long-term follow-up evaluations. We calculated the 95% confidence interval for the sensitivity, specificity and predictive values of optic nerve US and other techniques commonly employed in the clinical setting for the purpose of comparison. The calculated correlation coefficient (u) represents the association between the results of each technique and the final diagnosis. A correlation coefficient between 0.5 and 1.0 represents a high correlation. Results Forty-four patients (23 male, 21 female) were recruited to the study, of whom 39 (89%) were under 18 years of age when they were initially examined. The average age of the study participants was 12.7 years (range, 2 30 years). The patients were followed for an average duration of 18 ± 20 months, ranged from 2 days to 83 months. The 12 patients who did not undergo LP were followed for an average of 22 ± 23 months. The final diagnosis of the cause of bilateral optic disc swelling was papilloedema in 20 patients and pseudopapilloedema in 24 patients. All patients underwent complete neuro-ophthalmological examinations and suitable ones were referred for neuroimaging (CT or MRI) and LP. The availability of earlier test results allowed each patient to serve as his her own control; values from earlier tests were compared with optic nerve US results. The most common symptom among the patients was headache (n = 24, 55%). Headache was present in 54% of patients who were diagnosed with papilloedema (Tables 1 and 2). Headaches were occasionally accompanied by tinnitus, vomiting and or stomach ache. Eleven (25%) patients complained of visual disturbances, including either transient visual disturbances or temporary blurred vision (Tables 1 and 2). One patient who complained of temporary double vision was found to have sixth cranial nerve palsy. Retrospective data on multiple diagnostic tests that had been performed earlier were available for 39 patients (89%). These included pupillary reaction and eye movement tests, visual acuity tests, visual field tests, colour vision tests, the Amsler grid test and neurological tests. Thirty-one of these patients had normal findings, and there were no differences between them and the eight patients with abnormal findings in terms of being subsequently diagnosed as having papilloedema or pseudopapilloedema on US. Forty patients underwent funduscopy, which suggested the diagnosis of papilloedema in 18 and pseudopapilloedema in 22 (Tables 1 and 2). Of these 40 patients, 33 underwent CT imaging, 26 underwent MRI imaging and 22 underwent both. The Table 1. Clinical data and results of investigations on patients with final diagnosis of papilloedema or pseudopapilloedema. Papilloedema Pseudopapilloedema Total Headache Symptomatic* 13 11 24 Non-symptomatic 7 13 20 Visual disturbances With 6 5 11 Without 14 19 33 Funduscopy suggestive of Papilloedema 12 6 18 Pseudopapilloedema 7 15 22 Total 19 21 40 Presence of elevated intracranial pressure (ICP) by final diagnosis Increased ICP 8 2 10 No increased ICP 12 15 27 Total 20 17 37 * Headaches with or without vomiting and tinnitus. 377
Table 2. Comparison of the calculated sensitivity, specificity and predictive values of the different tests symptoms for the differential diagnosis of papilloedema and pseudopapilloedema. Test symptom Sensitivity (%) Specificity (%) PPV (%) NPV (%) Correlation coefficient u Headache 65 (41 84) 54 (33 74) 54 (33 74) 65 (41 84) +0.19 p > 0.1(NS) Visual disturbances 30 (13 54) 79 (57 92) 55 (25 82) 58 (39 74) +0.11 p > 0.3 (NS) General examination 22 (7 48) 81 (57 94) 50 (17 83) 55 (36 72) +0.04 p > 0.5 (NS) Funduscopy 63 (39 83) 71 (48 89) 67 (41 86) 68 (45 85) +0.35 p > 0.05 (NS) Imaging tests 40 (20 64) 88 (62 98) 80 (44 96) 56 (36 74) +0.32 p > 0.05 (NS) US (ONW upper limit of 3.3 mm) 85 (61 96) 63 (41 80) 65 (44 82) 83 (58 96) +0.48 p < 0.005 US (ONW upper limit of 3.0 mm) 95 (73 99.7) 58 (37 77) 66 (46 81) 93 (66 99.7) +0.56 p < 0.001 TN = True negative, TP = True positive, FN = False negative, FP = False positive, NS = non-significant, NPV = Negative predictive value, PPV = Positive predictive value, ONW = optic nerve width, US = Ultrasonography. The results of calculated sensitivity, specificity and predictive values are expressed as percentages (%), with corresponding 95% confidence intervals in parentheses. Correlation coefficients are presented with their corresponding statistical significance expressed as p-values. Sensitivity = TP (TP + FN). Specificity = TN (TN + FP). NPV = TN (TN + FN). PPV = TP (TP + FP). Table 3. Association between optic nerve width and the diagnosis of papilloedema or pseudopapilloedema in two conditions of normal optic nerve width. radiological data, including revisions, were thoroughly reviewed to trace the documentation of imaging indicators of increased intracranial pressure (ICP) and thickening or curling of the optic nerves. All patients who were Papilloedema Pseudopapilloedema Total Upper limit of normal optic nerve width of 3.3 mm Increased optic nerve width 17 9 26 Normal optic nerve width 3 15 18 Upper limit of normal optic nerve width of 3.0 mm Increased optic nerve width 19 10 29 Normal optic nerve width 1 14 15 Fig. 1. A 10-year-old boy presented with headache and visual disturbances. Funduscopy shows bilateral swollen discs. Ultrasonography reveals bilateral optic nerve head drusen. Right: colour photograph of right eye fundus. Left: B-mode scan of the right eye. eventually diagnosed with papilloedema underwent at least one imaging test (Tables 1 and 2). Lumbar puncture was performed in 32 patients, of whom 20 were found to have an increased opening cerebral spinal fluid (CSF) pressure. All the patients who were eventually diagnosed with papilloedema underwent LP. Opening CSF pressures of 20 cm H 2 O and 25 cm H 2 O were considered the upper normal limits for children and adults, respectively. Lumbar puncture results were highly correlated with the diagnosis of papilloedema. Specific bilateral US measurements of the ONW were available for 39 patients (89%). The association between the ONW and the diagnosis of papilloedema and pseudopapilloedema is presented in Table 3, where the data are based on the average ONW of both eyes. There were no patients with a unilateral increase in the ONW. Ultrasonography revealed an increased ONW in 26 patients and a normal ONW in 18 patients (Tables 2 and 3). Figures 1 and 2 illustrate the different echographic findings in papilloedema and pseudopapilloedema. When all the earlier and current test results were considered in combination, 20 patients were ultimately diagnosed as having papilloedema and the other 24 as having pseudopapilloedema. Nineteen of the 20 patients with papilloedema had a final diagnosis of pseudotumour cerebri (PTC). Brain imaging was normal in all 19 patients with PTC; the opening CSF 378
Fig. 2. A 28-year-old woman presented with headache. Funduscopy shows bilateral swollen discs. Ultrasonography reveals the doughnut sign, indicating increased optic nerve width (ONW). Lumbar puncture demonstrates elevated intracranial pressure = 320 mmh 2 O. Diagnosis: pseudotumour cerebri. Right: colour photograph of right eye fundus. Middle: B-mode scan of the right eye showing the ring of fluid around the nerve (doughnut sign) (68 db). Left: A-mode scan of the right eye, ONW = 4 mm (red line). pressure was elevated in all 20 patients with papilloedema. Optic nerve head drusen were detected in 11 patients who had been diagnosed with pseudopapilloedema. Ten other patients were diagnosed with structural pseudopapilloedema. In one case, the ONW was increased, but the final diagnosis was pseudopapilloedema: this patient developed PTC with increased ICP 2 years later. We presume that the increase in ONW in that patient was an early sign of an otherwise asymptomatic PTC. Comparisons of the calculated sensitivity, specificity and predictive values of the tests are presented in Table 2. Ultrasonography of the optic nerve had the highest sensitivity and the highest negative predictive value. Ultrasonography also had the best correlation with the final diagnosis and proved to be the only test whose correlation to the final diagnosis reached a level of significance. Defining the upper limit of the normal ONW at 3.0 mm rather than at 3.3 mm decreased the calculated specificity of the test but improved the calculated sensitivity and predictive values, as well as the correlation between the US results and the final diagnosis, resulting in an increased level of significance of the correlation. Discussion Our study showed that when comparing the various non-invasive procedures in current use for examining bilateral optic nerve swelling, optic nerve US was the most sensitive tool for differentiating papilloedema from pseudopapilloedema and it had the best correlation with the final diagnosis. Procedures that emerged as being more specific than US included the general ophthalmologic examination (including funduscopy) and imaging tests (CT and MRI). However, the calculated sensitivities of these tests were lower than those of US for the purposes of papilloedema screening. While diagnostic tools of high specificity could be useful to confirm the diagnosis of papilloedema, those with high sensitivity rather than high specificity could be useful screening tools for indicating which patients require additional invasive tests to confirm the final diagnosis. Papilloedema is often associated with symptoms such as headache, nausea, vomiting and visual disturbances (Soler et al. 1998; Newman et al. 2002; Piovesan et al. 2002; Ophir et al. 2005; Sa nchez-tocino et al. 2006; Mishra et al. 2007). Headache was the most frequent symptom of increased ICP in our study (65% of patients), followed by vision disturbances (30%), vomiting (15%) and tinnitus (10%). Funduscopy is not considered sufficiently sensitive for the diagnosis of papilloedema (Newman et al. 2002), and this was also confirmed in our study (sensitivity of only 63%). Swollen optic discs on funduscopy do not differentiate between papilloedema and pseudopapilloedema, especially when optic nerve drusen are present. Moreover, optic disc drusen in the paediatric population are usually buried and undetectable by funduscopy, further complicating the diagnosis (Auw-Haedrich et al. 2002). Brain imaging techniques, such as MRI and CT, are essential tools in the diagnosis of papilloedema with a suspected cerebral mass (Soler et al. 1998). Computerized tomography may show widening of the optic disc and is considered to be of some value in diagnosing drusen, especially in adults with calcified drusen. Magnetic resonance imaging may reveal elevation of the optic disc in patients with high ICP (Jinkins et al. 1996) and has also been shown to produce a constellation of specific signs that can assist in establishing the diagnosis of PTC with a sensitivity of 90% and a specificity of 100% (Brodsky & Vaphiades 1998). However, this requires a detailed analysis of the neuroimaging signs by a very experienced examiner. Both CT and MRI are limited by their high cost and availability, as well as the need for anaesthesia in young children who undergo these procedures. Computerized tomography also requires a relatively increased exposure to radiation. Several other non-invasive techniques, such as optic coherence tomography (Ophir et al. 2005) and confocal scanning laser ophthalmoscopy, have been tested for their value in detecting papilloedema (Tamburrelli et al. 2000). However, their role in the differential diagnosis of pseudopapilloedema and papilloedema must be investigated in greater depth. Lumbar puncture is a well-established method for determining the CSF opening pressure and fluid content. However, to reduce the risk of cerebral herniation, this test should only be performed after neuroimaging. A high CSF opening pressure is an essential criterion for establishing the diagnosis of papilloedema, while an increased ICP with normal CSF contents (cells and chemistry) is required for the diagnosis of PTC (Newman et al. 2002). We found a strong correlation between the CSF opening pressure 379
measurements and the diagnosis of papilloedema. Although CSF opening pressure is considered the gold standard in the diagnosis of papilloedema (Kline et al. 2005), our study results may indicate that US can be used as a screening tool in predicting papilloedema. In view of the untoward adverse effects of LP (severe headaches, back pain and an increased risk of cerebral herniation) and, on the other hand, the high negative predictive value of US in detecting papilloedema, asymptomatic patients with optic disc swelling might be adequately monitored solely by repeated US, and the use of LP could be deferred or even avoided. Defining the upper limit of the normal ONW at 3.3 mm resulted in a sensitivity of 85% and a specificity of 63% for the detection of papilloedema in patients with swollen discs. Some patients with an ONW lower than this threshold were diagnosed later on as having papilloedema. Setting the ONW upper limit at 3.0 mm increases the sensitivity of US as a screening test for papilloedema while decreasing its specificity. Both predictive values are increased, and the correlation with the final diagnosis is improved. These results suggest considering the adoption of 3.0 mm as the normal upper limit of the ONW in US screening for papilloedema. The limitations of our study stem mainly from the relatively small sample size. Furthermore, the results we report reflect the experience of a single and highly experienced operator (M.N.). In addition, we used a single instrument, the I 3 System-ABD Diagnostic Ophthalmic Ultrasound (Innovative Imaging), for all the tests, and extrapolating our results to other US machines should be done with caution. In conclusion, we demonstrated that US, which is a cheap and noninvasive modality, identified papilloedema with a high sensitivity and a moderate specificity. Our study supports the use of US as a screening tool in patients with bilateral swelling of the optic nerve head when the results are taken together with a complete ophthalmological examination. References Auw-Haedrich C, Staubach F & Witschel H (2002): Optic disk drusen. Surv Ophthalmol 47: 515 532. Brodsky MC & Vaphiades M (1998): Magnetic resonance imaging in pseudotumour cerebri. Ophthalmology 105: 1686 1693. Byrne SF & Green RL (2002): Ultrasound of the eye and orbit, 2nd edn. St Lewis: Mosby. Jinkins JR, Athale S, Xiong L, Yuh WT, Rothman MI & Nguyen PT (1996): MR of optic papilla protrusion in patients with high intracranial pressure. AJNR Am J Neuroradiol 17: 665 668. Kline LB, Arnold AC, Eggenberger E et al. (2005): Neuro-ophthalmology: Section 5. Basic and clinical science course. USA: American Academy of Ophthalmology. Mishra A, Mordekar SR, Rennie IG & Baxter PS (2007): False diagnosis of papilloedema and idiopathic intracranial hypertension. Eur J Paediatr Neurol 11: 39 42. Newman WD, Hollman AS, Dutton GN & Carachi R (2002): Measurement of optic nerve sheath diameter by ultrasound: a means of detecting acute raised intracranial pressure in hydrocephalus. Br J Ophthalmol 86: 1109 1113. Ophir A, Karatas M, Ramirez JA & Inzelberg R (2005): OCT and chronic papilledema. Ophthalmology 112: 2238. Piovesan EJ, Lange MC, Piovesan Ldo R, de Almeida SM, Kowacs PA & Werneck LC (2002): Long-term evolution of papilledema in idiopathic intracranial hypertension: observations concerning two cases. Arq Neuropsiquiatr 60: 453 457. Sánchez-Tocino H, Bringas R, Iglesias D, González-Pérez A & Del Villar-Galán R (2006): Utility of optic coherence tomography (OCT) in the follow-up of idiopathic intracranial hypertension in childhood. Arch Soc Esp Oftalmol 81: 383 389. Soler D, Cox T, Bullock P, Calver DM & Robinson RO (1998): Diagnosis and management of benign intracranial hypertension. Arch Dis Child 78: 89 94. Tamburrelli C, Salgarello T, Caputo CG, Giudiceandrea A & Scullica L (2000): Ultrasonographic evaluation of optic disk swelling: comparison with CSLO in idiopathic intracranial hypertension. Invest Ophthalmol Vis Sci 41: 2960 2966. Received on January 29th, 2011. Accepted on July 15th, 2011. Correspondence: Meira Neudorfer, MD Department of Ophthalmology Tel Aviv-Sourasky Medical Center 6 Weizman Street Tel Aviv 64239, Israel Tel: + 972 3 6973408 Fax: + 972 3 6973867 Email: meiraneu@netvision.net.il 380