Hettinga YM, Scheerlinck LME, Lilien MR, Rothova A, de Boer JH. The Value of Measuring Urinary β2-Microglobulin and Serum Creatinine for Detecting Tubulointerstitial Nephritis and Uveitis Syndrome in Young Patients With Uveitis. JAMA Ophthalmol. 2015;133(2):140–145. doi:10.1001/jamaophthalmol.2014.4301
Tubulointerstitial nephritis and uveitis (TINU) syndrome is characterized by tubulointerstitial and ocular inflammation. Thus far, the value of noninvasive diagnostic tests is not known.
To determine whether urinary β2-microglobulin (β2M), urinary protein, and serum creatinine have predictive value for detecting TINU syndrome in young patients with uveitis.
Design, Setting, and Participants
This prospective cohort study was conducted July 2010 through February 2013 at a tertiary care referral center in Utrecht, the Netherlands. Forty-five consecutive new patients with uveitis aged 22 years or younger were enrolled.
Urinary β2M, urinary protein, and serum creatinine were measured prospectively, and the estimated glomerular filtration rate was calculated.
Main Outcomes and Measures
A post hoc analysis was performed to determine whether urinary β2M, urinary protein, serum creatinine, estimated glomerular filtration rate, and/or pyuria were correlated with definitive and probable cases of TINU syndrome.
Eighteen of the 45 patients (40%) in our cohort had elevated urinary β2M levels, and 10 patients (22%) had elevated serum creatinine levels. Twenty of 43 patients (47%) had proteinuria. Eight of the 45 patients were diagnosed by a pediatric nephrologist as having renal dysfunction that suggested acute interstitial nephritis. Of these 8 patients, 2 were definitively diagnosed as having TINU syndrome (confirmed by renal biopsy). After excluding other causes of renal dysfunction, the remaining 6 patients with uveitis and renal dysfunction fulfilled the criteria of probable TINU syndrome. The 8 patients with definitive or probable TINU syndrome had higher urinary β2M levels than patients with normal renal function (median β2M, 1.95 mg/L; 95% CI, 1.26-5.16 mg/L vs 0.20 mg/L; 95% CI, 0.19-0.21 mg/L; P < .001; Mann-Whitney U test). Our analysis revealed that the positive predictive value of increased β2M combined with increased serum creatinine was 100% for detecting definitive and/or probable TINU syndrome.
Conclusions and Relevance
These data suggest that urinary β2M and serum creatinine levels are sensitive and relatively simple diagnostic screening tools for detecting renal dysfunction to diagnose TINU syndrome in young patients with uveitis similar to those evaluated in this study.
The presence of acute tubulointerstitial nephritis (AIN) and uveitis is probably underdiagnosed among young patients. Acute tubulointerstitial nephritis and uveitis comorbidity in children has several causes including tubulointerstitial nephritis and uveitis (TINU) syndrome, systemic sarcoidosis, and systemic lupus erythematosus. Acute tubulointerstitial nephritis can lead to renal dysfunction, and identifying renal function is important because AIN is often asymptomatic and can lead to severe renal damage.
Tubulointerstitial nephritis and uveitis syndrome is a condition characterized by tubulointerstitial and ocular inflammation. It was first described in 1975 by Dobrin et al.1 It primarily affects young patients; the prevalence of TINU syndrome peaks at 14 years of age.2 Although a diagnosis of TINU syndrome accounts for only 1% to 2% of all patients at specialized uveitis centers,3,4 this number is likely a gross underestimation as the nephritis component in TINU is often self-limiting and therefore not recognized clinically. In general, following a diagnosis of AIN, a renal biopsy is needed to definitively diagnose TINU syndrome. However, a renal biopsy is highly invasive and is generally not tolerated well by children. On the other hand, an increased level of β2-microglobulin (β2M) in the urine has been proposed as a potential screening tool for detecting AIN in TINU syndrome.5,6 β2-microglobulin is a small (12-kDa) protein that is excreted by the glomeruli into the ultrafiltrate and then resorbed by the tubules. In tubulointerstitial nephritis, β2M is excreted but not resorbed, leading to elevated levels of β2M in the urine. However, the value of measuring urinary β2M for diagnosing TINU syndrome is currently unknown.
In this study, we prospectively investigated the predictive value of urinary β2M and protein for detecting AIN to diagnose TINU syndrome in a cohort of young patients with uveitis. In addition, we measured serum creatinine levels, administered urinalyses, and calculated estimated glomerular filtration rate (eGFR) as an indicator of renal function. We then determined the positive predictive value (PPV) of these diagnostic procedures for detecting TINU syndrome in young patients with uveitis.
We performed a prospective study from July 2010 through February 2013 at the Department of Ophthalmology, University Medical Center Utrecht, Utrecht, the Netherlands. We recruited 45 consecutive new patients who presented at our outpatient clinic with uveitis with an age at onset of uveitis of 23 years or younger (range, 4-22 years). Because the University Medical Center Utrecht is a tertiary referral center, most patients were referred to our hospital by a referring ophthalmologist. Uveitis was diagnosed in accordance with the current diagnostic criteria of the International Uveitis Study Group7 and was based on a combination of clinical, laboratory, and imaging findings.
All patients were screened using our local diagnostic protocol (established in collaboration with pediatricians) to identify underlying uveitis-related disorders of unknown origin. In brief, we measured the following in each patient: complete blood cell count, erythrocyte sedimentation rate, liver function, renal function (ie, serum creatinine), potassium, sodium, calcium, uric acid, angiotensin-converting enzyme, C-reactive protein, antinuclear antibody (ANA) screening, human leukocyte antigen B27, and rheumatoid factor. In the event that another specific disease was suspected, additional tests (eg, immunological and/or virological tests) were performed.
This study was evaluated and approved by the institutional ethical committee of the University Medical Center Utrecht, which concluded that the Dutch Medical Research Involving Human Subjects Act did not apply and written informed consent of patients and or parents/guardians was not needed.
In addition to the previously mentioned screening protocol, urinary β2M, protein levels, and white blood cell count and/or cellular casts were rated in each patient on a urine sample at the first evaluation of renal function. A Siemens BN ProSpec nephelometer was used to measure urinary β2M. In our laboratory, the cutoff value for a positive urinary β2M result is 0.20 mg/L. Protein loss was measured using a urine dipstick (Multistix). This is a semiquantitative measurement and 0.10 g/L or greater is the cutoff for a positive protein loss result.
In 41 patients, white blood cell count was measured using urine dipstick (Multistix) and/or microscopic urinalysis was done. Pyuria was defined as more than 25 leucocytes/µL or if cellular casts were seen.
Serum creatinine was measured in 43 patients using a DxC800 analyzer (Beckman Coulter). The cutoff values for children 15 years or younger and 15 years or older were 0.74 and 1.17 mg/dL, respectively (to convert serum creatinine to micromoles per liter, multiply by 88.4). After May 2012, an enzymatic analyzer (Beckman Coulter) was used to measure serum creatinine (with the same cutoff values as just described). Patients with elevated test results were referred to a pediatric nephrologist for further evaluation. We also calculated eGFR in these 43 patients. Glomerular filtration rate (GFR) is currently the most reliable indicator of renal function in children. However, because standard GFR measurement techniques are costly, cumbersome, and often not available, we used eGFR, which is based on serum creatinine concentration. In patients younger than 18 years, we used the Pottel equation8; for patients 18 years and older, we used the chronic kidney disease epidemiology collaboration equation.9 The cutoff value for normal eGFR was 90 mL/min/1.73 m2.
Two patients with severe renal dysfunction underwent a renal biopsy by a pediatric nephrologist to confirm AIN.
We used the diagnostic criteria proposed by Mandeville et al2 to diagnose TINU syndrome. In brief, a diagnosis of TINU syndrome requires the presence of both AIN and uveitis with no other known systemic diseases that can cause AIN or uveitis. Cases of TINU syndrome can be categorized further as either definitive or probable TINU. With definitive TINU syndrome, AIN is diagnosed histopathologically (via a renal biopsy) and uveitis is present. In probable TINU syndrome, in addition to the presence of uveitis, AIN is diagnosed clinically (without a renal biopsy). The clinical diagnostic criteria for AIN in probable TINU syndrome are a combination of the presence of all 3 factors: (1) abnormal renal function (elevated serum creatinine or a decreased clearance), (2) abnormal urinalysis results (elevated urinary β2M, proteinuria, leucocyturia, and/or pyuria), and (3) systemic illness (fever, weight loss, anorexia, malaise, fatigue, pain, arthralgia, and/or myalgia).2
We calculated the PPV and the negative predictive value (NPV) for urinary β2M, serum creatinine, eGFR, proteinuria, and pyuria as screening methods for detecting definitive and/or probable TINU syndrome. The 95% CIs were calculated in R (version 3.1.0) using the adjusted Wald method (Agresti-Coull method).10 To calculate mean β2M levels, we conservatively used our cutoff point of 0.20 mg/L in patients without detectable urinary β2M levels.
Twenty-two of the patients in our cohort (49%) were boys and 23 patients (51%) were girls. The mean age at uveitis diagnosis was 10 years (range, 4-22 years). The anatomic classification of uveitis was as follows: 22 patients (49%) had intermediate uveitis, 13 patients (29%) had anterior uveitis, 9 patients (20%) had panuveitis, and 1 patient (2%) had posterior uveitis.
The average interval between the onset of uveitis and subsequent β2M testing was 10 months (range, 1 week-5 years) and the median was 4 months. Thirty-eight patients (84%) had bilateral uveitis, and 7 patients (16%) had unilateral uveitis.
Forty-four patients had ANA screening, and the results of 9 of these patients (20%) were positive. Three of these 9 patients were ultimately diagnosed as having silent chronic anterior uveitis typical of juvenile idiopathic arthritis, although they did not meet the International League of Associations for Rheumatology criteria.11 Two patients were diagnosed as having juvenile idiopathic arthritis–associated uveitis, and 1 patient was diagnosed as having psoriasis. The remaining 3 ANA-positive patients had panuveitis or intermediate uveitis of unknown cause (Table).
Eighteen of the 45 patients (40%) in our cohort had elevated urinary β2M levels (mean, 6.52 mg/L; range, 0.22-82.7 mg/L). Ten of the 45 children (22%) had increased serum creatinine levels (mean, 1.15 mg/dL; range, 0.41-1.72 mg/dL). Eight of these 10 patients were diagnosed by a pediatric nephrologist as having either definitive or probable TINU syndrome; the remaining 2 patients had no other clinical signs of renal pathology and were diagnosed as having uveitis of unknown cause (Table).
Of the 43 patients for whom protein levels were measured, 20 (47%) had proteinuria. We then calculated eGFR in these 43 patients and found an eGFR of less than 90 mL/min/1.73 m2 in 11 patients. Two of these patients had definitive TINU syndrome, 4 had probable TINU syndrome, 1 patient had uveitis secondary to herpes simplex virus, and the remaining 4 patients had uveitis of unknown cause (Table).
Five of the 8 patients with either definitive or probable TINU syndrome were male (62.5%), and 3 (37.5%) were female. The mean age of these 8 patients was 14 years (range, 12-20 years). A renal biopsy was performed in 2 of these patients, and both biopsies contained a focal tubulointerstitial inflammatory infiltrate containing lymphocytes and nonspecific histiocytes. The results of 1 of these 2 patients was negative for urinary β2M; however, a renal biopsy confirmed the diagnosis of definitive TINU syndrome. The 6 other patients with definitive or probable TINU syndrome did not receive a renal biopsy because their postevaluation renal dysfunction was relatively mild and therefore did not justify this invasive procedure. All 8 patients with probable TINU syndrome had systemic signs and symptoms such as malaise, weight loss, or anorexia. After excluding other possible causes of renal disease, all 6 patients were diagnosed as having probable TINU syndrome based on the 3 criteria of Mandeville et al.2 Seven of the 8 patients with TINU syndrome had intermediate uveitis and 1 had panuveitis.
The 6 patients with probable TINU syndrome and 1 of the 2 patients with definitive TINU syndrome had a β2M level of 1.0 mg/L or greater; the other patient with definitive TINU syndrome did not have an elevated β2M value. In 10 of 11 patients with elevated β2M and normal renal function, the β2M levels were within the range of 0.2 to 1.0 mg/L. The 8 patients with definitive or probable TINU syndrome had higher urinary β2M levels than the patients with normal renal function (median β2M, 1.95 mg/L; 95% CI, 1.26-5.16 mg/L vs 0.20 mg/L; 95% CI, 0.19-0.21 mg/L; P < .001; Mann-Whitney U test). Two of the 11 patients with uveitis, elevated urinary β2M, and no other signs of AIN were diagnosed as having another clinical entity including herpes simplex virus infection (confirmed by an aqueous humor analysis) and ANA-positive chronic anterior uveitis with no signs of arthritis.12,13 The conditions of the other 9 patients remained undiagnosed despite comprehensive additional testing.
Next, we calculated the predictive value of the specific tests with respect to detecting definitive and/or probable TINU syndrome. The PPV of greater than 0.20 mg/L β2M in the urine for predicting TINU syndrome was 39% (7 of 18; 95% CI, 20-61) and the NPV was 96% (26 of 27; 95% CI, 80-100). With a cutoff of 1.0 mg/L, the PPV increased to 88% (7 of 8; 95% CI, 51-100) and the NPV to 97% (36 of 37; 95% CI, 85-100). The PPV of increased serum creatinine for predicting TINU syndrome was 80% (8 of 10; 95% CI, 48-95) and the NPV was 100% (33 of 33; 95% CI, 91-100). Combining elevated urinary β2M and elevated serum creatinine yielded an increased PPV of 100% (7 of 7; 95% CI, 68-100) and an NPV of 97% (35 of 36; 95% CI, 85-100).
In addition, the PPV and the NPV of proteinuria for predicting TINU syndrome were 30% (6 of 20; 95% CI, 14-52) and 91% (21 of 23; 95% CI, 72-99), respectively. The eGFR generally confirmed the existence of renal dysfunction in our patients. The PPV and NPV of eGFR for definitive or probable TINU syndrome were 55% (6 of 11; 95% CI, 28-79) and 94% (30 of 32; 95% CI, 79-99), respectively.
Pyuria was more frequently found in patients with proven or probable TINU syndrome (5 of 8 patients; 2 patients with TINU syndrome and 3 patients with probable TINU) compared with the group with other uveitis entities (7 of 31; P = .02; Fisher exact).
The results support our hypothesis that measuring urinary β2M and serum creatinine is a sensitive screening method for diagnosing definitive and/or probable TINU syndrome in young patients with uveitis. Moreover, combining these 2 measurements increases the PPV. Measuring β2M in the urine is relatively easy, and serum creatinine is routinely measured as part of the standard uveitis workup. The urinary β2M levels in our patients with uveitis with definitive and probable TINU syndrome were higher than in patients without suspected renal dysfunction. Consistent with previous reports,5 proteinuria was not specific for detecting renal dysfunction in our cohort. Given the risk for irreversible renal damage, we recommend that children with renal dysfunction receive additional follow-up care by a pediatric nephrologist.
Most studies regarding TINU syndrome were performed in children with clinically confirmed tubulointerstitial nephritis who were referred for uveitis screening. In these cases, the diagnosis of tubulointerstitial nephritis was confirmed by a nephrologist (usually via renal biopsy). Renal biopsies from patients with TINU syndrome typically reveal inflammatory cells (primarily mononuclear cells including lymphocytes, plasma cells, and histiocytes) and edema in the renal interstitium.2 In our prospective study of children with uveitis, we studied a different population of patients who initially presented with uveitis and were subsequently screened for renal disorders following a finding of increased urinary β2M. To our knowledge, this is the first study to screen renal function in patients with uveitis and then correlate those results with a diagnosis of TINU syndrome. Given that elevated urinary β2M and elevated serum creatinine levels are associated with the diagnosis of renal pathology, referral to a nephrologist is recommended in these cases. In patients who are suspected of having TINU syndrome, the main differential diagnoses include renal sarcoidosis, systemic lupus erythematosus, and Sjögren syndrome. It is important to note that, in particular, TINU syndrome and sarcoidosis can be accompanied by uveitis.
Several hypotheses have been suggested regarding the cause of TINU syndrome, and these include genetic predisposition, hypersensitivity to drugs, and infection.14- 16 For example, a very strong association between the HLA-DRB1 and HLA-DQA1 genes and biopsy-confirmed TINU syndrome has been confirmed by independent groups.17- 19 In addition, the use of antibiotics that are often prescribed to treat respiratory tract infections and/or the use of nonsteroidal anti-inflammatory drugs is associated with AIN but the role in pathogenesis of TINU syndrome is not elucidated yet.2 Lastly, because serological autoantibodies—including ANAs, anticardiolipin antibodies, rheumatoid factor, and cytoplasmic antineutrophil cytoplasmic antibodies—have been detected in patients, an autoimmune hypothesis has been suggested with respect to TINU syndrome pathogenesis.2 However, whether these findings are casually related to TINU syndrome or are merely coincidental remains unknown. In our study, none of the patients with definitive or probable TINU syndrome were ANA positive, and none were treated with antibiotics in the 4-month period before the onset of uveitis.
In our study, we included children and young adults only because the peak age at onset for TINU syndrome is approximately 14 years, although TINU syndrome can also occur in older patients.2 However, proteinuria in patients older than 45 years might be because of causes other than TINU syndrome including hypertension or diabetes mellitus.
Whether TINU syndrome is more prevalent in one sex than another is currently unknown. Some groups have reported a female predominance,20 whereas others found an apparent male predominance.2,5 In our relatively small and young cohort, 62.5% of patients with TINU syndrome were male. All of our patients with TINU syndrome had anterior segment inflammation and vitritis, whereas other studies reported anterior segment inflammation only.5,21
Previous reports have suggested that uveitis is caused by TINU syndrome in approximately 2% of all children with uveitis.4 However, the actual prevalence of TINU remains unclear because TINU syndrome is likely underdiagnosed and may be more common than currently appreciated.20 The relatively high prevalence of renal dysfunction in our cohort of children and young adults with uveitis might have been owing to a referral bias because our center serves as both a secondary and tertiary referral center and has a special interest in this particular disease. Because Goda et al5 found that the serum creatinine level was elevated in only 25% patients with biopsy-proven TINU syndrome, the number of patients with TINU syndrome might be underestimated in our series.
A limitation of our study was that some of the children with probable TINU syndrome did not receive a biopsy to confirm the diagnosis. In these cases, we used the Diagnostic Criteria for TINU Syndrome published by Mandeville et al2; these criteria were based on all published medical literature regarding TINU syndrome that was available at the time. All our patients with (probable) TINU fulfill these complete criteria of a minimum of 3 clinical factors previously listed combined with typical bilateral uveitis.
A relatively long delay from the onset of symptoms until referral was another limitation of our study. However, the median referral time was 4 months and urinary β2M and elevated serum creatinine levels are known to remain elevated for a long period.2 Nevertheless the number of patients with TINU syndrome in our series might be underestimated.
In summary, in agreement with a comprehensive review on TINU syndrome,20 we think that TINU syndrome is a probably underdiagnosed cause of uveitis in children. To identify the specific AIN in children with uveitis, thereby potentially leading to the diagnosis TINU syndrome, we recommend that all young patients with bilateral anterior uveitis with or without intermediate or posterior involvement of unknown origin should be tested for both urinary β2M and serum creatinine. These data suggest that urinary β2M and serum creatinine levels are a sensitive and relatively simple diagnostic screening tool for detecting renal dysfunction to diagnose TINU syndrome in young patients with uveitis similar to those evaluated in this study.
Corresponding Author: Ymkje M. Hettinga, MD, Department of Ophthalmology, University Medical Center Utrecht, E.03-136, Heidelberglaan 100, PO Box 85500, 3508 CX Utrecht, the Netherlands (firstname.lastname@example.org).
Submitted for Publication: May 27, 2014; final revision received September 2, 2014; accepted September 6, 2014.
Published Online: October 30, 2014. doi:10.1001/jamaophthalmol.2014.4301.
Author Contributions: Drs Hettinga and de Boer had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Hettinga, de Boer.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Hettinga, de Boer.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Hettinga, Scheerlinck.
Obtained funding: Hettinga, de Boer.
Study supervision: Rothova.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.
Funding/Support: This research was financially supported by Stichting Uitzicht, the Netherlands.
Role of the Funder/Sponsor: Stichting Uitzicht had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Contributions: We thank Rolf Groenwold, MD, PhD, of the Julius Center for Health Sciences for assistance in the statistical analysis. He did not receive compensation for his contribution.