Immunophenotyping in Uveitis PDF

Ocular Immunology and Inflammation ISSN: 0927-3948 (Print) 1744-5078 (Online) Journal homepage: https://www.tandfonline.com/loi/ioii20 Lymphocyte Immunophenotyping and CD4/CD8 Ratio in Cerebrospinal Fluid for the Diagnosis of Sarcoidosis-related Uveitis Romain Paule, Laure Denis, Nico...

Ocular Immunology and Inflammation ISSN: 0927-3948 (Print) 1744-5078 (Online) Journal homepage: https://www.tandfonline.com/loi/ioii20 Lymphocyte Immunophenotyping and CD4/CD8 Ratio in Cerebrospinal Fluid for the Diagnosis of Sarcoidosis-related Uveitis Romain Paule, Laure Denis, Nicolas Chapuis, Julien Rohmer, Jérôme Hadjadj, Jonathan London, Anthony Chauvin, Clémence Bonnet, Luc Mouthon, Claire Le Jeunne, Dominique Monnet, Philippe Blanche, Antoine Brezin & Benjamin Terrier To cite this article: Romain Paule, Laure Denis, Nicolas Chapuis, Julien Rohmer, Jérôme Hadjadj, Jonathan London, Anthony Chauvin, Clémence Bonnet, Luc Mouthon, Claire Le Jeunne, Dominique Monnet, Philippe Blanche, Antoine Brezin & Benjamin Terrier (2019): Lymphocyte Immunophenotyping and CD4/CD8 Ratio in Cerebrospinal Fluid for the Diagnosis of Sarcoidosis- related Uveitis, Ocular Immunology and Inflammation, DOI: 10.1080/09273948.2019.1678647 To link to this article: https://doi.org/10.1080/09273948.2019.1678647 Published online: 31 Oct 2019. Submit your article to this journal View related articles View Crossmark data Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=ioii20 Ocular Immunology & Inflammation, 2019; 00(00): 1–9 © 2019 Taylor & Francis Group, LLC ISSN: 0927-3948 print / 1744-5078 online DOI: 10.1080/09273948.2019.1678647 ORIGINAL ARTICLE Lymphocyte Immunophenotyping and CD4/CD8 Ratio in Cerebrospinal Fluid for the Diagnosis of Sarcoidosis-related Uveitis Romain Paule, MD 1,2, Laure Denis, MD1, Nicolas Chapuis, MD3, Julien Rohmer, MD1, Jérôme Hadjadj, MD1, Jonathan London, MD1,2, Anthony Chauvin, MD4, Clémence Bonnet, MD5, Luc Mouthon, MD, PhD1,2, Claire Le Jeunne, MD, PhD 1,2, Dominique Monnet, MD, PhD2,5, Philippe Blanche, MD1, Antoine Brezin, MD, PhD2,5, and Benjamin Terrier, MD, PhD1,2 1 Department of Internal Medicine, Cochin Hospital, Paris, France, 2Paris Descartes University, Sorbonne Paris Cité, Paris, France, 3Hematology Laboratory, Cochin Hospital, Paris, France, 4Centre of Research in Epidemiology and Statistics Sorbonne Paris Cité (CRESS-UMR1153), Paris, France, and 5Department of Ophthalmology, Cochin Hospital, Paris, France ABSTRACT Background: This study aimed to assess the diagnostic relevance of CD4/CD8 ratio in cerebrospinal fluid (CSF) for the etiological diagnosis work-up of uveitis. Methods: We consecutively included patients who were referred to our department for the diagnostic workup of intermediate and/or posterior uveitis. Etiological diagnoses were established in a blind manner regarding CD4/CD8 ratio. Results: Fifty-two patients were included. A diagnosis of ocular sarcoidosis was made in 15 (29%) patients, 21% had another determined diagnosis while 50% remained of undetermined origin. Median CD4/CD8 ratio in CSF was 4.57 (IQR 3.39–5.47) in ocular sarcoidosis, 1.74 (1.60–3.18) in uveitis due to other determined cause (P =.008), and 2.83 (2.34–3.54) in those with uveitis of undetermined origin (P =.007). CD4/CD8 ratio >3.23 was associated with a diagnosis of ocular sarcoidosis. Conclusion: Determination of CD4/CD8 ratio in CSF can be useful for diagnosis work-up since a CD4/CD8 ratio >3.23 in CSF is associated with ocular sarcoidosis. Keywords: CD4/CD8 ratio, lymphocyte immunophenotyping, sarcoidosis, uveitis Uveitis is defined as an intraocular inflammation causes are multiple and include three major etiolo- that affects both the uveal tract and adjacent struc- gical frameworks, i.e. pure ophthalmological enti- tures. Its incidence ranges from 17 to 52/ ties, infectious diseases, and inflammatory diseases 100,000 person-years with an estimated prevalence such as sarcoidosis. The prevalence of sarcoidosis- of 38–204 cases per 100,000 individuals.1–4 Because related uveitis has not been well established but is of recurrent intraocular inflammation, uveitis can probably underestimated, since uveitis may pre- lead to transient or permanent visual impairment cede extraocular manifestations by more than and ocular complications. Its complications one year in up to 30% of patients.6,7 In Japan, account for nearly 10–15% of preventable causes sarcoidosis accounts for 10–15% of cases of of blindness and represent the fourth most com- uveitis8,9 whereas in a French retrospective study, mon cause of blindness among the working age sarcoidosis represented 7% of posterior uveitis or population in developed countries.5 Underlying panuveitis.10 This heterogeneity may be explained Received 2 April 2019; revised 1 October 2019; accepted 7 October 2019 Correspondence: Romain Paule, Department of Internal Medicine, Hôpital Cochin, 27, rue du Faubourg Saint-Jacques, Paris Cedex 14 75679, France. E-mail: [email protected] Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/ioii. 1 2 R. Paule et al. by patients’ characteristics, definition of sarcoido- to the department of Internal Medicine of Cochin sis, recruitment method but also with the use of Hospital (Paris, France), for the diagnostic work-up of variable paraclinical investigations. However, intermediate and/or posterior uveitis and who under- despite an extensive work-up, one third of uveitis went lumbar puncture. Lumbar puncture was assessed remains of undetermined origin or idiopathic.4,10 for etiological diagnosis of uveitis as described above, Due to its heterogeneous presentation and multiplicity and lymphocyte immunophenotyping was added to clas- of the underlying causes, uveitis represents a diagnostic sical analysis. No lumbar puncture was performed only and economic challenge,4,10 especially due to the lack of to assess lymphocyte immunophenotyping. a well-codified diagnostic procedure.11 Recently, several French law does not require patients’ informed studies aimed to improve the diagnosis work-up of uvei- consent for medical research studies involving data tis but the place of neurological explorations, and espe- collected from retrospective chart reviews. The cially lumbar puncture, is still unclear.11,12 When study was approved by a French ethics committee sarcoidosis is considered as a cause of uveitis, diagnostic (COCHIN Hospital, Paris). investigations can include labial salivary gland biopsy (LSGB), chest computed tomography (CT) scan, serum angiotensin converting enzyme (ACE), and bronchoal- Ocular Evaluation and Diagnostic veolar lavage fluid (BALF) analysis.11,13 Lumbar punc- Investigations ture with analysis of cerebrospinal fluid (CSF) can also be included, especially in intermediate and/or posterior All patients had complete ophthalmologic examina- uveitis since etiology can include some inflammatory tion, clinical examination and biological tests, i.e. and/or systemic disorders.13,14 In a recent study from blood count, serum creatinine, liver tests, serum our group, abnormal CSF analysis was found in 14% of C-reactive protein, ACE, protein electrophoresis, cal- patients.11 CSF analysis is recommended for the diagno- cium test, and an interferon-gamma (IFN-γ) release sis of primary oculocerebral lymphoma15 or neurological assay (QuantiFERON-TB Gold®) to search for latent Lyme disease.13 Furthermore, some authors recommend tuberculosis. Additional investigations included in to perform CSF analysis when uveitis is associated with most patients chest CT-scan, cerebral magnetic reso- optic neuritis and/or when the following diagnoses are nance imaging (MRI), bronchoscopy with BALF ana- suspected, i.e. Behçet disease, sarcoidosis, multiple lysis and/or LSGB when appropriate, according to sclerosis, Vogt-Koyanagi-Harada disease, syphilis, bilat- the physician’s discretion. eral papillar edema and sympathetic ophthalmia,16–18 By definition, lumbar puncture was performed for since it could have clinical or therapeutic all patients, and was considered abnormal in case of implications.19,20 cell count ≥10 cells/mm3. Lymphopenia was defined Yet, CSF analysis usually includes only classic cyto- by a cell count 1.5 the upper limit of normal immunophenotyping and especially CD4/CD8 ratio in (ULN) (>78 U/L). Hypergammaglobulinemia was CSF could be useful for the diagnostic workup of uveitis, defined by polyclonal gammaglobulins >15 g/L. but it has never been evaluated. On contrary, measure- Chest CT-scan was considered contributive in the ment of CD4/CD8 ratio is already performed in BALF as presence of findings suggestive of sarcoidosis (i.e. in vitreous specimens21,22 for diagnosis of ocular sarcoi- mediastinal and/or hilar lymph nodes with short dosis. Furthermore, in comparison to classic cytomor- axis diameter >10 mm, and/or parenchymal lung phology, flow cytometry analysis is known to abnormalities suggestive of sarcoidosis), or findings considerably improve CSF analysis sensitivity for typical of healed tuberculosis (calcified granulomas patients with meningeal involvement during lymphoma, predominantly in the upper lobes, and apical scar- even in samples with very low cell counts or paucicellular ring). When bronchoscopy was performed, bronchial samples.23,24 biopsies were considered positive if they showed In the present study, we aimed to evaluate the diag- granuloma, while the BALF was considered contribu- nostic interest of lymphocyte immunophenotyping and tive in case of alveolar lymphocytosis suggestive of CD4/CD8 ratio in CSF for the etiological diagnosis and sarcoidosis (defined by a lymphocyte count >15% of classification of intermediate and/or posterior uveitis. the white cells count and a CD4/CD8 lymphocyte ratio >3.5).11 MATERIALS AND METHODS Uveitis Diagnostic Criteria Patients Uveitis were classified according to the recently From May 2016 to March 2018, we included consecutive revisited Standardization of Uveitis Nomenclature patients referred from the department of Ophthalmology criteria.25 Chronic uveitis was defined as ocular Ocular Immunology & Inflammation  CSF CD4/CD8 Ratio in Uveitis 3 inflammation lasting >3 months. Granulomatous distributed variables are expressed as mean ± standard uveitis was defined by mutton-fat keratic precipi- deviation (SD), and non-normally distributed variables tates (large or small) and/or anterior chamber iris as median [interquartile (IQR)] for quantitative variables. nodules. Snowballs were defined by the presence of Mann-Whitney nonparametric tests were used for the vitreous opacities. Peripheral multifocal choroiditis comparison of medians. ANOVA was used to examine (PMC) was diagnosed on the basis of 3 criteria: (1) group differences for each of the dependent variables multiple (>10), small, round, punched-out lesions in with a Bonferroni correction. Receiver operating charac- the peripheral retina; (2) no central chorioretinal teristic (ROC) curve were used to determine the optimal lesions; and (3) intraocular inflammation.26 CD4/CD8 ratio. Data were analyzed using GraphPad Consistent with the revised criteria of International Prism 5, GraphPad Software CA, USA. For all analyses, Workshop on Ocular Sarcoidosis (IWOS), ocular sarcoi- a p value 0.4 g/L) in 8 cases (median 0.68 g/L, range 0.22–1.96). Only Statistical Analysis one patient had neurological symptoms, i.e. head- aches, and only two had an abnormal cerebral MRI. Descriptive analysis included the use of number and Four out the 10 patients with abnormal lumbar percentage for qualitative variables. Normally puncture had a diagnosis of ocular sarcoidosis, © 2019 Taylor & Francis Group, LLC 4 R. Paule et al. TABLE 1. Baseline characteristics of the 52 patients included in CD8 ratio in CSF = 6.41), and a 63-year-old woman with the study. uveitis of undetermined origin had an increased (CD4/ CD8 ratio in CSF = 8). The latter had unilateral inter- Demography n (%) mediate uveitis with no extraocular symptoms. All the Age, median (IQR) 50 (32.5–63) exams performed did not reveal any abnormality, and Male Sex 25 (48) lumbar puncture was normal (i.e. cell count 3.23 was also useful Clinical extra-ophthalmologic manifestations to distinguish ocular sarcoidosis from uveitis due Clinical extra-ophthalmologic manifestations 17 (33) Arthralgia 2 (12) to another inflammatory disease in case of bilateral uveitis (P =.02). In the case of granulomatous features on ophthalmologic examination, median while all the remaining patients were classified as CD4/CD8 ratio in CSF was 4.14 (3.33–4.86) for having uveitis of undetermined origin. patients with ocular sarcoidosis (n = 9), and 2.70 Median cell count in CSF according to uveitis diag- (2.33–3.03) for patients with uveitis of undeter- noses based on IWOS criteria were as followed: in mined origin (n = 5) (P =.04). In the absence of patients with definite sarcoidosis (n = 8), median CSF granulomatous features, median CD4/CD8 ratio in cell count was 6 (3–13.25) but only two patients had CSF was 5.46 (4.33–5.74) for patients with ocular abnormal CSF (62 and 127 cells/mm3, respectively); in sarcoidosis (n = 6), 1.74 (1.60–3.18) for those with patients with presumed sarcoidosis (n = 4), median CSF determined underlying cause (n = 11) (P =.02), cell count was 3.5 (2.5–9.25) with only one patient having and 2.90 (2.35–3.56) for those with uveitis of unde- abnormal CSF with count of 25/mm3; in patients with termined origin (n = 21) (P =.01). probable sarcoidosis (n = 3), median CSF cell count was 2 Finally, in patients with abnormal lumbar punc- (1.5–11.5) with only one patient having a abnormal CSF ture, median CD4/CD8 ratio in CSF was 5.65 (IQR with a count of 21/mm3. 5.15–5.82) for patients with ocular sarcoidosis (n = 4) Using TransFix®, median levels of CD4/CD8 ratio in and 2.54 (IQR 1.98–2.95) for those with uveitis of CSF according to uveitis diagnoses are shown in Table 2 undetermined origin (n = 6) (P =.002). In contrast, and in Figure 1. Median CD4/CD8 ratio in CSF was 4.57 in those with normal lumbar puncture, median CD4/ (IQR 3.39–5.47) in ocular sarcoidosis, 1.74 (1.60–3.18) in CD8 ratio in CSF was 3.96 (IQR 3.02–5.14) for patients uveitis due to another determined cause (P =.008), and with ocular sarcoidosis (n = 11), 1.74 (1.60–3.18) for 2.83 (2.34–3.54) in those with uveitis of undetermined those with determined underlying cause (n = 11) origin (P =.007). Two non-sarcoidosis patients had an (P =.04), and 2.99 (IQR 2.35–3.55) for those with increased CD4/CD8 ratio in CSF: a 22-year-old woman uveitis of undetermined origin (n = 6) (P =.14). with uveitis related to Bartonella henselae infection (CD4/ These data are summarized in Table 3. TABLE 2. Determination of CD4/CD8 ratio in CSF according to uveitis diagnoses. Patients with abnormal CSF Diagnoses n (%) n (%) Median CD4/CD8 ratio (IQR) Definite sarcoidosis 8 (15) 2 (25) 4.50 (3.80–5.54) Presumed sarcoidosis 4 (8) 1 (25) 3.08 (2.55–3.96) Probable sarcoidosis 3 (6) 1 (34) 5.43 (5.00–5.63) Uveitis of determined origin 11 (21) 0 1.74 (1.60–3.18) Uveitis of undetermined origin 26 (50) 6 (23) 2.83 (2.34–3.54) Ocular Immunology & Inflammation  CSF CD4/CD8 Ratio in Uveitis 5 Features Associated with an Increased CD4/ CD8 Ratio in CSF ROC curve analysis showed that the CD4/CD8 ratio threshold with the best performance was >3.23 for the diagnostic of ocular sarcoidosis with a 80.0% sensitivity, a 67.6% specificity, a 50.0% positive pre- dictive value and a 89.3% negative predictive value, an area under the curve of 0.77 (0.62–0.92) and a Youden index at 0.48 (Figure 2). A CD4/CD8 ratio >1.73 had a 100% sensitivity but a poor speci- ficity of 20%. By analogy with the cutoff used in BALF, a CD4/CD8 ratio >3.5 had a 60.0% sensitiv- ity, a 77.8% specificity, a 60.0% positive predictive value, and a 77.8% negative predictive value for the diagnostic of ocular sarcoidosis. Among the whole study population, CD4/CD8 ratio >3.23 was associated with a diagnosis of ocular sarcoi- dosis [OR 7.38, 95% CI (1.76–30.96), P <.01], with find- ings suggestive of sarcoidosis on chest CT-scan [OR 5.13, 95% CI (1.27–20.8), P =.049] and with evidence of gran- uloma on tissue biopsy [OR 8,2, 95% CI (1.2–56.3), P =.046], but not with a contributive BALF [OR 1.26, 95% CI (0.26–11.13), P =.66]. In contrast, a CD4/CD8 ratio >3.23 was neither associated with lymphopenia [OR 2.24, 95% CI (0.7–7.2), P =.28], ACE >1.5 ULN FIGURE 1. Comparisons of CD4/CD8 ratio in CSF between [OR 0.71, 95% CI (0.14–3.58), P = 1] nor hypergamma- ocular sarcoidosis (definite, presumed, or probable), other globulinemia [OR 0.67, 95% CI (0.16–2.72), P =.73]. Also, determined causes of uveitis and uveitis of undetermined CD4/CD8 ratio >3.23 was not associated with character- origin. Levels of CD4/CD8 ratios in CSF were higher in patients with ocular sarcoidosis as compared to that of istics of uveitis (anterior, intermediate and/or posterior) patients with uveitis due to other determined causes of nor specific ocular findings (granulomatous features, uveitis and uveitis of undetermined origin. snowballs, and/or PMC). TABLE 3. Median (IQR) CD4/CD8 ratios according subgroup analyses in patients with uveitis. A: Ocular C: Uveitis of sarcoidosis B: Uveitis of determined origin undetermined origin (n = 15) (n = 11) (n = 26) A vs. A vs. Subgroup n (%) Median (IQR) Median (IQR) Median (IQR) P value B C Uveitis features Bilateral (n = 26, 50%) 4.05 (3.36–4.79) 1.72 (1.07–2.04) 3.23 (2.56–4.18) 0.02 0.02 0.13 Unilateral (n = 26, 50%) 5.49 (5.45–5.82) 1.85 (1.62–4.21) 2.50 (1.56–3.09) 0.08 0.06 0.06 Posterior involvement (= 35, 4.57 (3.33–5.44) 1.72 (1.55–4.00) 3.03 (2.12–3.53) 0.03 0.07 0.001 67%) No posterior involvement 4.63 (3.42–5.85) 1.85 (1.79–1.90) 2.70 (2.34–3.84) 0.12 0.17 0.06 (n = 17, 33%) Granulomatous features (n = 8, 4.14 (3.33–4.86) - 2.70 (2.33–3.03) - - 0.04 15%) No granulomatous features 5.46 (4.33–5.74) 1.74 (1.60–3.18) 2.90 (2.35–3.56) 0.02 0.02 0.01 (n = 44, 85%) Lumbar puncture Contributive (n = 10, 19%) 5.65 (5.15–5.82) - 2.54 (1.98–2.95) - - 0.002 Non-contributive (n = 42, 81%) 3.96 (3.02–5.14) 1.74 (1.60–3.18) 2.99 (2.35–3.55) 0.06 0.04 0.14 © 2019 Taylor & Francis Group, LLC 6 R. Paule et al. DISCUSSION includes some inflammatory disorders, especially Behçet disease, sarcoidosis, multiple sclerosis, Vogt- In the present study, we evaluated the diagnostic Koyanagi-Harada disease and sympathetic interest of lymphocyte immunophenotyping and ophthalmia.34 This hypothesis is supported by CD4/CD8 ratio in the CSF of patients with intermedi- encouraging results in vitreous analysis of patients ate and/or posterior uveitis. We showed that the with uveitis, where ocular sarcoidosis showed higher diagnosis of ocular sarcoidosis was significantly asso- CD4/CD8 ratio than other etiologies,21,22 but at the ciated with an increased CD4/CD8 ratio in CSF, espe- cost of a more invasive procedure. However, no data cially when the CD4/CD8 ratio is >3.23, a threshold was available so far in CSF from patients with uveitis. which is quite similar to the one used in BALF in the In the setting of intermediate and/or posterior uvei- setting of pulmonary sarcoidosis. tis, we found that an increased CD4/CD8 ratio in CSF We recently evaluated the relevance of investiga- was significantly associated with a diagnosis of ocular tions for the etiological diagnosis of uveitis in 300 sarcoidosis, in comparison with uveitis due to other patients.11 In this study, lumbar puncture, performed determined cause and with uveitis of undetermined without lymphocyte immunophenotyping, was origin. This finding was especially true in case of uveitis abnormal in only 14% of cases, and ACE level in with posterior involvement and in those with abnormal CSF was not associated with the diagnosis of sarcoi- lumbar puncture. This stronger association could be dosis. Also, other studies showed that ACE levels in explained by more frequent granulomatous manifesta- CSF had a poor sensitivity and specificity for the tions and by contiguous inflammation of CSF in case of diagnosis of neurosarcoidosis.30,31 Overall, diagnostic posterior involvement. However, lymphocyte immuno- interest of CSF analysis was very low and thus phenotyping tended to be useful even in low-cellularity remained unclear. Lymphocyte immunophenotyping CSF. We showed previously that the main feature asso- was shown to greatly improve the diagnosis of central ciated with an abnormal lumbar puncture (without nervous system localizations of hematologic malig- Transfix® analysis) was an abnormal cerebral MRI.11 nancies in comparison with classic cytomorphology. In the present study, when both cerebral MRI and lum- In hematologic malignancies, the diagnostic value of bar puncture with classic cytomorphology were normal flow cytometry was more than twice that of cytomor- (n = 32, i.e. 62% of patients), lymphocyte immunophe- phology, particularly in low-cellularity samples.32,33 notyping using TransFix® was associated to diagnosis In uveitis, improving the detection of inflammatory of ocular sarcoidosis in 10 cases, that is, 2/3 of all cases cells in CSF could be useful to guide the diagnosis of sarcoidosis in this study. Also, lumbar puncture was toward the spectrum of uveomeningitis, which abnormal in only 10 cases, while immunophenotyping FIGURE 2. Receiver Operating Characteristic (ROC) curves for CD4/CD8 ratio in CSF for the diagnosis of ocular sarcoidosis. ROC curve analysis showed that CD4/CD8 ratio threshold with the best performance was >3.23 for the diagnostic of ocular sarcoidosis with a 80.0% sensitivity, 67.6% specificity, 50.0% positive predictive value and 89.3% negative predictive value, an area under the curve of 0.77 (0.62–0.92) and a Youden index at 0.48. A threshold of 1.73 had a 100% sensitivity but a poor specificity of 20%. By analogy with the cutoff used in bronchoalveolar lavage fluid, CD4/CD8 ratio >3.5 had a 60.0% sensitivity, a 77.8% specificity, a 60.0% positive predictive value and a 77.8% negative predictive value for the diagnostic of ocular sarcoidosis. Ocular Immunology & Inflammation  CSF CD4/CD8 Ratio in Uveitis 7 analysis of CSF was available in all cases. These results patients as it is not without risk, possibly reducing the suggest that determination of lymphocyte immunophe- clinical relevance of the study. notyping could be useful even in patients without any Second, the small number of patients with each diag- abnormal findings on MRI or abnormal lumbar punc- nosis can limit the significance of the study and deserves ture based on classic cytomorphology. validation of our results in other cohorts of patients. Using ROC curve analysis, a threshold >3.23 had the Overall, these findings suggest that lymphocyte immu- best performance to distinguish ocular sarcoidosis from nophenotyping and CD4/CD8 ratio in CSF could be useful other causes of uveitis. A CD4/CD8 ratio >3.23 was especially in the absence of biological features and/or of associated with the diagnosis of ocular sarcoidosis, specific ocular findings suggestive of sarcoidosis, since whether granulomatous features were present on uveitis diagnosis is particularly challenging in these situa- ophthalmologic examination or not. To support the rele- tions. This is particularly relevant since identifying an vance of this threshold, a CD4/CD8 ratio >3.23 was underlying condition associated with uveitis could have significantly associated with the most robust criteria for clinical (especially cardiac) or therapeutic implications.19,20 the diagnosis of sarcoidosis, that is, granuloma on biopsy In contrast, in the presence of biological features and/or of and/or contributive chest CT-scan. In contrast, a CSF specific ocular characteristics suggestive of sarcoidosis, CD4/CD8 ratio >3.23 was not associated with lymphocyte immunophenotyping could be probably not a contributive BALF and with neither biological features necessary since the diagnosis can be made without inva- suggestive of sarcoidosis, such as lymphopenia, ACE sive procedures. >1.5 ULN and hypergammaglobulinemia, nor specific In conclusion, lymphocyte immunophenotyping characteristics of uveitis including granulomatous fea- and CD4/CD8 ratio in CSF can be useful in the etio- tures, snowballs, and/or peripheral multifocal choroidi- logical workup of patients with intermediate and/or tis. Increased ACE level is commonly used for the posterior uveitis, since a CD4/CD8 ratio >3.23 is asso- diagnosis of ocular sarcoidosis,14 and a recent study ciated with ocular sarcoidosis. showed a high negative predictive value for the diagno- sis of sarcoidosis in the setting of uveitis.35 On the con- CONTRIBUTORSHIP STATEMENT trary, a recent study showed that ACE was positive in only 62% of patients with a biopsy-proven diagnosis of All the authors took part on the acquisition, analy- sarcoidosis, supporting the need for additional biomar- sis, or interpretation of data for the work; AND kers in patients with normal ACE. drafting the work or revising it critically for impor- We could have expected to have a higher CD4/ tant intellectual content; AND final approval of the CD8 ratio in definite sarcoidosis group compared version to be published; AND agreement to be to presumed and probable sarcoidosis. The absence accountable for all aspects of the work in ensuring of difference between these groups could be that questions related to the accuracy or integrity related to the classification criteria we used. of any part of the work are appropriately investi- Distinction between definite and presumed sarcoi- gated and resolved. dosis is based on the IWOS criteria, which require histological evidence of sarcoidosis to consider definite sarcoidosis, what can be difficult to obtain DECLARATION OF INTEREST in ocular sarcoidosis. Bronchial biopsies are rarely contributive and bronchial echo-endoscopy The authors report no conflicts of interest. The remains an invasive procedure with general authors alone are responsible for the content and anesthesia, much more than lumbar puncture. writing of the article. Also, cases of presumed sarcoidosis, despite the lack of histological evidence of sarcoidosis, exhib- ited clinical and paraclinical features for which the ORCID diagnosis of sarcoidosis could be clearly retained. Therefore, our level of certainties for the diagnosis Romain Paule http://orcid.org/0000-0003-4360- of sarcoidosis was very high, even in patients with 4157 presumed sarcoidosis. Claire Le Jeunne http://orcid.org/0000-0002-0116- Finally, our study has some limitations. First, ana- 2277 lysis of CSF using lumber puncture may be consid- ered as an invasive procedure, not widely performed in the diagnostic work-up of uveitis. The issue of REFERENCES considering the risk of an invasive procedure when contemplating lumbar puncture should be consid- 1. Prete M, Dammacco R, Fatone MC, Racanelli V. Autoimmune uveitis: clinical, pathogenetic, and therapeu- ered. Outside the French attitude, a large number of tic features. Clin Exp Med. 2016 May;16(2):125–136. centers do not conduct lumbar punctures in uveitis doi:10.1007/s10238-015-0345-6. © 2019 Taylor & Francis Group, LLC 8 R. Paule et al. 2. Gritz DC, Wong IG. Incidence and prevalence of uveitis NCBI [Internet]. https://www-ncbi-nlm-nih-gov.gate2. in Northern California; the Northern California epide- inist.fr/pubmed/27266701. Accessed March 5, 2019. miology of uveitis study. Ophthalmology. 2004 Mar;111 18. Reekie I, Reddy Y. Use of lumbar punctures in the manage- (3):491–500. discussion 500. doi:10.1016/j. ment of ocular syphilis. Semin Ophthalmol. 2018;33 ophtha.2003.06.014. (2):271–274. doi:10.1080/08820538.2016.1228986. 3. Rim TH, Kim SS, Ham D-I, et al. Incidence and prevalence of 19. Choi RY, Rivera-Grana E, Rosenbaum JT. Reclassifying uveitis in South Korea: a nationwide cohort study. Br J Ophthalmol. idiopathic uveitis: lessons from a tertiary uveitis center. 2018;102(1):79–83. doi:10.1136/bjophthalmol-2016-309829. Am J Ophthalmol. 2018 Oct 20;198:193–199. 4. Barisani-Asenbauer T, Maca SM, Mejdoubi L, Emminger W, 20. Han YS, Rivera-Grana E, Salek S, Rosenbaum JT. Machold K, Auer H. Uveitis- a rare disease often associated Distinguishing uveitis secondary to sarcoidosis from idio- with systemic diseases and infections - a systematic review of pathic disease: cardiac implications. JAMA Ophthalmol. 2018 2619 patients. Orphanet J Rare Dis. 2012 Aug 29;7:57. Feb 1;136(2):109–115. doi:10.1001/jamaophthalmol.2017.5466. doi:10.1186/1750-1172-7-47. 21. Maruyama K, Inaba T, Sugita S, et al. Comprehensive 5. Suttorp-Schulten MS, Rothova A. The possible impact of analysis of vitreous specimens for uveitis classification: uveitis in blindness: a literature survey. Br J Ophthalmol. a prospective multicentre observational study. BMJ Open. 1996 Sep;80(9):844–848. doi:10.1136/bjo.80.9.844. 2017 Nov 16;7(11):e014549. 6. Blaise P, Fardeau C, Chapelon C, Bodaghi B, Le Hoang P. 22. Maruyama K, Inaba T, Tamada T, Nakazawa T. Vitreous Minor salivary gland biopsy in diagnosing ocular lavage fluid and bronchoalveolar lavage fluid have equal sarcoidosis. Br J Ophthalmol. 2011 Dec;95(12):1731–1734. diagnostic value in sarcoidosis. Medicine (Baltimore). 2016 doi:10.1136/bjophthalmol-2011-300129. Dec;95(49):e5531. doi:10.1097/MD.0000000000004864. 7. Ma S, Rogers SL, Hall AJ, et al. Sarcoidosis related uveitis: 23. Alvarez R, Dupuis J, Plonquet A, et al. Clinical relevance of clinical presentations, disease course and rates of systemic flow cytometric immunophenotyping of the cerebrospinal disease progression after uveitis diagnosis. Am fluid in patients with diffuse large B-cell lymphoma. Ann J Ophthalmol. 2018 Sep 20;198:30–36. Oncol. 2012 May;23(5):1274–1279. doi:10.1093/annonc/ 8. Iwata D, Mizuuchi K, Aoki K, et al. Serial frequencies and mdr436. clinical features of uveitis in Hokkaido, Japan. Ocul 24. Subirá D, Castañón S, Aceituno E, et al. Flow cytometric Immunol Inflamm. 2017 Aug 24;25(sup1):S15–8. analysis of cerebrospinal fluid samples and its usefulness in doi:10.1080/09273948.2016.1184286. routine clinical practice. Am J Clin Pathol. 2002 Jun;117 9. Kitamei H, Kitaichi N, Namba K, et al. Clinical features of (6):952–958. doi:10.1309/123P-CE6V-WYAK-BB1F. intraocular inflammation in Hokkaido, Japan. Acta 25. Mochizuki M, Smith JR, Takase H, et al. Revised criteria Ophthalmol. 2009 Jun;87(4):424–428. doi:10.1111/j.1755- of International Workshop on Ocular Sarcoidosis 3768.2008.01282.x. (IWOS) for the diagnosis of ocular sarcoidosis. Br 10. Bodaghi B, Cassoux N, Wechsler B, et al. Chronic severe J Ophthalmol. 2019 Feb 23;103:1418–1422. doi:10.1136/ uveitis: etiology and visual outcome in 927 patients from bjophthalmol-2018-313356. a single center. Medicine (Baltimore). 2001 Jul;80(4):263–270. 26. Abad S, Meyssonier V, Allali J, et al. Association of per- doi:10.1097/00005792-200107000-00005. ipheral multifocal choroiditis with sarcoidosis: a study of 11. Hadjadj J, Dechartres A, Chapron T, et al. Relevance of thirty-seven patients. Arthritis Rheum. 2004 Dec 15;51 diagnostic investigations in patients with uveitis: retro- (6):974–982. doi:10.1002/art.20839. spective cohort study on 300 patients. Autoimmun Rev. 27. Bansal R, Gupta A, Gupta V, Dogra MR, Bambery P, Arora SK. 2017 May;16(5):504–511. doi:10.1016/j.autrev.2017.03.006. Role of anti-tubercular therapy in uveitis with latent/manifest 12. de Parisot A, Kodjikian L, Errera M-H, et al. Randomized tuberculosis. Am J Ophthalmol. 2008 Nov;146(5):772–779. controlled trial evaluating a standardized strategy for uvei- doi:10.1016/j.ajo.2008.06.011. tis etiologic diagnosis (ULISSE). Am J Ophthalmol. 2017 28. International Team for the Revision of the International Jun;178:176–185. doi:10.1016/j.ajo.2017.03.029. Criteria for Behçet’s Disease (ITR-ICBD). The 13. Sève P, Cacoub P, Bodaghi B, et al. Uveitis: diagnostic International Criteria for Behçet’s Disease (ICBD): work-up. A literature review and recommendations from a collaborative study of 27 countries on the sensitivity an expert committee. Autoimmun Rev. 2017 Dec;16 and specificity of the new criteria. J Eur Acad Dermatol (12):1254–1264. doi:10.1016/j.autrev.2017.10.010. Venereol. 2014 Mar;28(3):338–347. doi:10.1111/jdv.12107. 14. Grumet P, Kodjikian L, de Parisot A, et al. Contribution of 29. Satsangi J, Silverberg MS, Vermeire S, Colombel J-F. The diagnostic tests for the etiological assessment of uveitis, Montreal classification of inflammatory bowel disease: contro- data from the ULISSE study (Uveitis: clinical and medicoe- versies, consensus, and implications. Gut. 2006 Jun;55 conomic evaluation of a standardized strategy of the etio- (6):749–753. doi:10.1136/gut.2005.082909. logical diagnosis). Autoimmun Rev. 2018 Apr;17(4):331–343. 30. Khoury J, Wellik KE, Demaerschalk BM, Wingerchuk DM. doi:10.1016/j.autrev.2017.10.018. Cerebrospinal fluid angiotensin-converting enzyme for diag- 15. Gundlach E, Hoffmann MM, Prasse A, Heinzelmann S, nosis of central nervous system sarcoidosis. Neurologist. 2009 Ness T. Interleukin-2 receptor and angiotensin-converting Mar;15(2):108–111. doi:10.1097/NRL.0b013e31819bcf84. enzyme as markers for ocular sarcoidosis. PLoS One. 31. Bridel C, Courvoisier DS, Vuilleumier N, Lalive PH. 2016;11(1):e0147258. doi:10.1371/journal.pone.0147258. Cerebrospinal fluid angiotensin-converting enzyme for 16. Sakata VM, Da Silva FT, Hirata CE, de Carvalho JF, diagnosis of neurosarcoidosis. J Neuroimmunol. 2015 Yamamoto JH. Diagnosis and classification of Aug;15(285):1–3. doi:10.1016/j.jneuroim.2015.05.020. Vogt-Koyanagi-Harada disease. Autoimmun Rev. 2014 32. Bromberg JEC, Breems DA, Kraan J, et al. CSF flow cytometry May;13(4–5):550–555. doi:10.1016/j.autrev.2014.01.023. greatly improves diagnostic accuracy in CNS hematologic 17. Clinical manifestations and cerebrospinal fluid status in malignancies. Neurology. 2007 May 15;68(20):1674–1679. ocular syphilis in HIV-negative patients. - PubMed - doi:10.1212/01.wnl.0000261909.28915.83. Ocular Immunology & Inflammation  CSF CD4/CD8 Ratio in Uveitis 9 33. Bento LC, Correia RP, Alexandre AM, et al. Detection of central Rev Med Interne. 2016 Jan;37(1):25–34. doi:10.1016/j. nervous system infiltration by myeloid and lymphoid hema- revmed.2015.09.009. tologic neoplasms using flow cytometry analysis: diagnostic 35. Niederer RL, Al-Janabi A, Lightman SL, Tomkins- accuracy study. Front Med (Lausanne). 2018;5:70. doi:10.3389/ Netzer O. Serum angiotensin-converting enzyme has fmed.2018.00070. a high negative predictive value in the investigation 34. Abad S, Terrada C, Trad S, et al. [Management of uveomenin- for systemic sarcoidosis. Am J Ophthalmol. 2018 gitis in internal medicine: proposal for a diagnostic work-up]. Oct;194:82–87. doi:10.1016/j.ajo.2018.07.010. © 2019 Taylor & Francis Group, LLC