Journal Logo

Article

Renal Sarcoidosis

Clinical, Laboratory, and Histologic Presentation and Outcome in 47 Patients

Mahévas, Matthieu MD; Lescure, Francois Xavier MD; Boffa, Jean-Jacques MD, PhD; Delastour, Victoire MD; Belenfant, Xavier MD; Chapelon, Catherine MD; Cordonnier, Carole MD; Makdassi, Raifat MD; Piette, Jean-Charles MD; Naccache, Jean-Marc MD; Cadranel, Jacques MD, PhD; Duhaut, Pierre MD, PhD; Choukroun, Gabriel MD, PhD; Ducroix, Jean Pierre MD; Valeyre, Dominique MD

Author Information
doi: 10.1097/MD.0b013e31819de50f
  • Free

Abstract

INTRODUCTION

Sarcoidosis is a systemic disease of unknown cause characterized by the formation of granulomatous lesions in various organs, particularly the lower respiratory tract.35 Autopsy findings in sarcoidosis patients reveal up to 13% of interstitial granulomatous lesions in the renal parenchyma.7,23 Sarcoidosis renal manifestations also include abnormal calcium metabolism, nephrocalcinosis, and nephrolithiasis.27 Some studies have described clinical, laboratory, or histologic features in renal sarcoidosis (RS) and demonstrated that the granulomatous tubulointerstitial nephritis was associated with acute renal failure, and could lead to hemodialysis.4 Renal failure commonly ranges from 0.7% to 4.3% of cases in previous clinical series on sarcoidosis.10,25,28,32,37 Corticosteroid treatment is the cornerstone of treatment. Its efficacy has been demonstrated in advanced sarcoid tubulointerstitial nephritis.29 Immunosuppressive therapy may be required in some cases.9

Only a few studies have addressed the specific issues of renal involvement in sarcoidosis. Little is known about its prognostic factors and its relationship with other systemic manifestations and with calcium metabolism. We present here what is to our knowledge the largest retrospective cooperative series of patients with biopsy-proven renal involvement in sarcoidosis (47 patients). The results were collected by the French Sarcoidosis Group. We conducted the current study 1) to describe the clinical, laboratory, and histologic features at presentation, 2) to point out the clinical phenotype of sarcoidosis, and 3) to assess prognostic and therapeutic challenges.

METHODS

The current retrospective multicenter study received institutional review board approval; the need for informed consent was waived. Patients met the following inclusion criteria: 1) confirmed sarcoidosis according to the statement of the American Thoracic Society (ATS), the European Respiratory Society (ERS), and the World Association of Sarcoidosis and Other Granulomatous Disorders (WASOG);35 2) evidence of renal failure (estimated glomerular filtration rate [eGFR] <90 mL/min per 1.73 m2) or morphologic imaging abnormalities of kidney; 3) renal histology compatible with the diagnosis of RS (interstitial nephritis with or without granuloma); and 4) exclusion of any alternative diagnosis.

Patients were recruited between 1988 and 2007 from 6 departments in 3 different French hospitals. All were large urban referral hospitals (2 university hospitals: Assistance publique-hôpitaux de Paris and Centre hospitalier universitaire d'Amiens, and 1 community hospital: Centre hospitalier de Montreuil). The departments involved were the pulmonary (n = 1), nephrology (n = 3), and internal medicine (n = 2) departments. Recruitment was made through file research of patients diagnosed with sarcoidosis and renal involvement.

All data were collected in each center by 1 author (MM), and were systematically reviewed by the same person and another author (DV). Follow-up was investigated until December 2007. Initial analysis concerned 53 patients; 6 of them who did not meet the inclusion criteria and were excluded from the final analysis because of the absence of renal biopsy (n = 5) or the absence of sufficient criteria to diagnose sarcoidosis (isolated renal granulomatosis, n = 1). Forty-six patients presented with renal failure, while 1 presented with pseudotumoral infiltration of the kidney on computer tomography (CT) scan.

Clinical and Laboratory Data at Presentation

For all cases, history, medical examination, chest X-ray, and renal laboratory tests were compiled. Hypertension was defined according to the World Health Organization criteria. Thoracic sarcoidosis staging was made according to the ATS/ERS/WASOG statement of sarcoidosis.35 Renal function was evaluated using the following parameters: serum creatinine level (μmol/L), proteinuria defined as >0.3 g/24 h, hematuria as >104 red blood cells/mL, leukocyturia as >104 leukocytes/mL in the urine. The eGFR was calculated according to the Modification of Diet in Renal Disease (MDRD) renal disease equation, and all patients were classified according to the chronic kidney disease score (CKD).22 Calcemia was defined as total plasma calcium >2.75 mmol/L and hypercalciuria as urine calcium excretion >0.10 mmol/kg per day.

Results of spirometry (n = 39) and bronchoalveolar lavage (n = 39) were collected when available. Airflow obstruction was defined by forced expiratory volume in 1 second/forced vital capacity <70%, and restrictive syndrome by total lung capacity <80% predicted value. Serum 25OH-vitamin D3 (25OHD3) (n = 11), parathormone (PTH) (n = 10), and calciuria (n = 15) were measured in patients with hypercalcemia. Serum angiotensin-converting enzyme (n = 41), serum ionogram (n = 47), blood cell count (n = 47), and C-reactive protein (n = 35) were recorded. Anemia was defined as hemoglobin <13 g/dL for male patients and <12 g/dL for female patients.

Renal Pathology

In each center, all renal biopsies were viewed by a pathologist who specialized in kidney pathology and were independently reviewed by a single pathologist (C. Cordonnier) unaware of clinical features. For light microscopic studies, biopsy specimens were fixed in formalin-alcohol, embedded in paraffin, and 2μ sections were cut. The sections were stained with hematoxylin-eosin, periodic acid-Schiff, Masson trichrome, and Ziehl-Neelsen. Interstitial fibrosis was scored based on the degree of lesion in the cortical sample as follows: 0-5% = F0, 6%-25% = F1, 26%-50% = F2, >50% = F3. Immunofluorescence-labeling studies were performed in sections cut from frozen specimens with antiserum direct against C3, C1q, IgA, IgG, IgM, light κ, and λ chains.

Follow-Up

We evaluated the outcome and treatment for each patient. Clinical, laboratory (serum creatinine level, eGFR, analysis of urinary sediment, calcemia) evaluation at 1 month (n = 38), at 1 year (n = 41), and at the end of follow-up (n = 47) were recorded. The RS outcome was assessed according to the degree of eGFR variation (Δ eGFR). Patients were classified in 3 groups: 1) complete response to therapy, defined as a degree of Δ eGFR ≥50%; 2) partial response to therapy, defined as a degree of Δ eGFR between 0 and 50%; and 3) unfavorable response to therapy, defined as a negative Δ eGFR. During the follow-up, treatments (respectively prednisone, methylprednisolone pulses, or immunosuppressive therapy) and relapses of sarcoidosis at a renal or extrarenal level were systematically recorded for each patient. The need for dialysis was specified.

Study Design

The study was designed to analyze retrospectively the epidemiologic characteristics and the clinical, laboratory, and renal pathological features of RS at presentation. The response of renal function to therapy was assessed at 1 month, 1 year, and at the end of follow-up. Relapses of sarcoidosis, either renal or extrarenal, were investigated. Eventually, response to therapy was correlated to renal pathology.

Statistical Analysis

Statistical analyses were performed with SAS version 8.0 (SAS Institute Inc, Cary, NC) and SPSS version 14.0 (SPSS Inc, Chicago, IL) software. A descriptive analysis was used to study the distribution of continuous variables (that is, means, medians, standard deviations, ranges) and the frequencies of categorical variables (N and percentages). Means were compared using the Student t test or, when appropriate, Wilcoxon or Mann-Whitney nonparametric test. Percentages were compared with the Pearson chi-squared test, or, when appropriate, the Fisher exact test. A p value of less than.05 was considered to denote statistical significance. Independent variables with a p value of.20 or less after univariate analysis were included in a multiple logistic regression model based on a stepwise ascending method. The adequacy of the model was studied using the maximum probability method. The stability of the model was then tested by the Hosmer and Lemeshow test.

RESULTS

Epidemiology

Forty-seven adult patients were assessed (30 male and 17 female patients, M/F ratio: 1.76). The estimated prevalence of RS in the pulmonary department and in both internal medicine departments was 1.5%. Thirty-two (68%) patients were white and 15 (32%) were black African or native Caribbean. Median age at onset of RS was 47 years old (range, 21-76 yr). The mean age at onset for women (50.6 yr) tended to be higher than for men (45 yr), but not significantly higher (Figure 1). The age and sex ratios were not different between white and black patients. There was no case of familial sarcoidosis. No one had occupational or environmental exposure known to induce granulomatous disease. Some of the patients had prior or concomitant comorbidities, including diabetes mellitus type II (n = 5), Biermer anemia (n = 1), multiple sclerosis treated with IFN-β (n = 1), primary antiphospholipid syndrome (n = 1), breast carcinoma (n = 1), and renal carcinoma (n = 1). Ten patients were active smokers.

F1-4
FIGURE 1:
Sex and age at presentation of renal sarcoidosis.

Clinical, Laboratory, and Radiologic Features at Presentation (Table 1 and Table 2)

T1-4
TABLE 1:
Clinical, Laboratory, Initial Chest X-Ray, and Renal Biopsy Features at Renal Sarcoidosis Presentation
T2-4
TABLE 2:
Extrarenal Manifestations of Sarcoidosis

RS occurred in the course of previously diagnosed sarcoidosis in 9 (19%) patients, with a median of 42 months after the initial diagnosis of sarcoidosis (range, 2-168 mo). For the other 38 (81%) patients, RS occurred concomitantly with at least 1 other localization of the disease, which facilitated the sarcoidosis diagnosis. Among these cases, renal failure revealed sarcoidosis in 23 patients leading to the discovery of other localizations. There was no case of an isolated renal localization at presentation preceding an additional localization of the disease.

Forty-two of 47 (90%) patients had thoracic involvement disclosed on chest X-ray (stage 1: n = 14, stage 2: n = 17, stage 3: n = 10, stage 4: n = 1). Twenty of 39 (51%) patients had abnormal pulmonary function tests (airway obstruction: n = 6, restrictive syndrome: n = 10, mixed syndrome: n = 4). Bronchoalveolar lavage revealed lymphocytosis higher than 20% in 21 of 39 (54%) patients, and the lymphocyte T CD4/CD8 ratio was higher than 3.5 in 9 of 17 (53%) patients.

Extrathoracic localizations (renal excluded) were found in 29/47 (62%), of which 19 cases were associated with thoracic localization and 5 were not.

Constitutional symptoms (fatigue, weight loss >5 kg, or fever >38°C) were present in 20 (42.5%) cases. Fever >38°C was present in 8 of 47 (17%). Eleven of 47 (23%) patients had high blood pressure. One patient presented with lumbar pain, which was revealed by CT scan to be a pseudotumoral infiltration in the left kidney (Figure 2).

F2-4
FIGURE 2:
Pseudotumoral infiltration in the left kidney as seen on CT scan.

All patients but 1 presented with acute renal failure at diagnosis. The patient without renal failure displayed renal infiltration on CT. Renal function was usually severely impaired with a median serum creatinine level of 263.5 μmol/L (range, 88-1130 μmol/L). Median eGFR was 20.5 mL/min per 1.73 m2 (range, 4-93 mL/min per 1.73 m2). In 42 of 47 (90%) cases, eGFR was below 60 mL/min per 1.73 m2. Moderate proteinuria was found in 31 (66%) patients (median, 0.7 g/24 h; range, 0-2.7 g/24 h), microscopic hematuria in 11 (21.7%), and aseptic leukocyturia in 13 (28.7%) patients. Fourteen (30%) patients had no urinary abnormalities.

Fifteen of 47 (32%) patients had hypercalcemia (>2.75 mmol/L; median, 2.47 mmol/L; range, 2-3.91 mmol/L), 13 of whom had calcemia >3 mmol/L. Hypercalcemia was significantly more common in white patients (14 of 32 cases, 44%) than in black patients (1 of 14 cases, 6.5%) (odds ratio [OR], 8; 95% confidence interval [CI], 1.75-85; p < 0.001), and was more common in men than in women (12 men compared with 3 women). Eleven of the 22 (50%) patients diagnosed between June and September had hypercalcemia, compared with only 4 of the 25 (16%) patients diagnosed during the other months (OR, 7.8; 95% CI, 1.6-44.8; p < 0.001) (Figure 3). Hypercalciuria was constant in patients with hypercalcemia and was complicated by nephrolithiasis in 3 cases and nephrocalcinosis in 1 case, revealed by radiography. In patients with hypercalcemia, PTH levels were lower than the normal range in 7 of 10 patients (PTH <10 ng/L; normal range, 10-65 ng/L) and were normal in the other cases, while the serum 25OHD3 level was low in 7 of 9 patients (25OHD3 <10 nmol/L; normal range, 10-40 nmol/L). One patient had a tubular dysfunction that included hyperchloremic metabolic acidosis, glycosuria, aminoaciduria, and hyperphosphaturia. Serum angiotensin-converting enzyme was increased in 22 of 41 (55%) cases (median, 1.5 N). C-reactive protein was increased (>5 mg/L) in 9 of 30 patients (range, 3-100 mg/L). Anemia was found in 22 of 47 cases (range, 7-16 g/dL).

F3-4
FIGURE 3:
Seasonal variation of calcemia.

Histologic Findings

An informative percutaneous renal biopsy containing more than 6 glomeruli was performed on all patients. Thirty-seven patients presented with noncaseating granulomatous interstitial nephritis (GIN), and 10 with interstitial nephritis without granuloma. Epithelioid granulomas varied in intensity from extensive to isolated lesions as illustrated in Figure 4A and 4B. Giant cells were associated with GIN in 22 cases (Figure 4C). Interstitial nephritis was observed in all patients with acute (interstitial inflammatory cell infiltrate, and tubular inflammation) and chronic (tubular atrophy, fibrosis) lesions. Interstitial fibrosis was present in all cases to varying degrees. Interstitial calcifications were observed in 8 patients, all with hypercalcemia. Glomeruli lesions were rare. Immunofluorescence revealed that 2 patients had IgA mesangial deposits. Immunofluorescence also showed rare interstitial deposits of C3 (n = 3).

F4-4
FIGURE 4:
Granuloma infiltration of kidney. A. Low-power view showing a single granuloma composed of epithelioid cells, tubular atrophy, and interstitial inflammation. (Hematoxylin and eosin, magnification × 200.) B. Diffuse granuloma infiltration through the interstitium. (Hematoxylin and eosin, magnification × 200.) C. Giant cell and lymphocytic infiltrate through the interstitium. (Hematoxylin and eosin, magnification × 400).

Treatment and Evolution

All patients initially received prednisone at 1 mg/kg per day (n = 39), 0.7 mg/kg per day (n = 6), or 0.5 mg/kg per day (n = 2). The median duration of treatment was 18 months (range, 1-48 mo). Intravenous pulse methylprednisolone (MP) was administered in 8 cases at diagnosis (respectively 500 mg to 1 g for 3 days in a row for 7 patients, and 240 mg in 1 pulse for 1 patient) and in 2 other cases at 500 mg for 3 days in a row, respectively 3 and 6 months after initiation of prednisone. The patient with pseudotumoral renal infiltration displayed no renal failure and was not included in the following renal outcome analysis. In his case, the pseudotumoral infiltration was controlled by steroid therapy.

Renal Outcome at 1 Month

Mean eGFR increased from 20 ± 19.1 mL/min per 1.73 m2 to 44 ± 24.7 mL/min per 1.73 m2 after 1 month of steroids (p < 0.001, n = 38). Response to therapy was observed in 32 of 38 (84%) patients with a complete response in 23 of 38 (60.5%) patients, a partial response in 9 of 38 (23.5%) patients, and unfavorable response in 6 of 38 (16%) patients. Transitory hemodialysis was necessary for 3 patients. Calcemia was normalized for all patients. The response at 1 month was not significantly correlated to age, sex, race, or multiorgan involvement.

Outcome at 1 Year

Mean eGFR was 47 ± 19.9 mL/min per 1.73 m2 at 1 year, not different from the value at 1 month. Response to therapy was observed in 34 of 41 (83%) patients, with a complete response in 27 of 41 (66%) patients, a partial response in 7 of 41 (17%), and unfavorable response in 7 of 41 (17%) patients. Hemodialysis was stopped for the 3 patients who had needed it. A complete response to therapy at 1 year was found to be strongly correlated with the complete response at 1 month (OR, 7; 95% CI, 2-41; p < 0.01). Positive response to therapy was observed in all patients receiving MP with a complete response in 8 of 10 (80%) cases, but the difference regarding the response to conventional steroid therapy was not significant. Hypercalcemia was strongly correlated with a complete response to therapy at 1 year (OR, 16; 95% CI, 1.8-137; p = 0.003). In a multivariate analysis (including treatment with MP [preselected variable], calcemia, and initial eGFR <30 mL/min per 1.73 m2), hypercalcemia was independently correlated with complete response (OR = 18.9; p = 0.001). There was no recurrence of hypercalcemia during the course of the disease. During the first year, clinical relapses of sarcoidosis occurred in 6 (13%) cases and concerned extrarenal but not renal localizations.

Outcome at End of Follow-Up

Follow-up data were available until 2007 for all patients except 8 patients whose follow-up ended before 2007. These cases were not remarkable according to the last information available. In them, the last visit was also considered to be "end of follow-up." After a median follow-up of 24 months (range, 3-180 mo), the mean eGFR was 49.1 ± 25 mL/min per 1.73 m2. Response to therapy was observed in 35 of 46 (76%) patients, with a complete response in 30 of 46 (65%), a partial response in 5 of 46 (11%), and unfavorable response in 11 of 46 (24%) patients. Complete response at the end of follow-up was significantly related to a complete response at 1 month (OR, 7.6; 95% CI, 2-41; p < 0.001). Thirty-one of 46 (66%) cases had an eGFR <60 mL/min per 1.73 m2. Hemodialysis was permanent in 2 patients (15 and 2 years after the beginning of the disease, respectively). The distribution and evolution of patients according to the CKD stage are shown in Table 3. Patient outcomes are summarized in Figure 5.

F5-4
FIGURE 5:
Outcome at the end of follow-up.
T3-4
TABLE 3:
Distribution and Evolution of Patients According to CKD Stage

In 6 of 46 (13%) cases, steroids were permanently stopped (median of follow-up, 28.5 mo; range, 12-52 mo): 1 patient experienced end-stage renal failure necessitating hemodialysis, and 5 patients were considered free of active disease, although the eGFR was abnormal for 4 of them.

In 41 (87%) cases, steroids were maintained at the end of follow-up. In 24 of 41 (59%) cases, no relapse occurred (after a mean therapy duration of 24 mo).

During the whole evolution, 17 relapses were observed (6 in the first year) after a mean steroid therapy duration of 24 months (range, 1-48 mo). Relapses occurred 13 times in 12 cases after 1 year. Relapses concerned the kidney 7 times and were either limited to the kidney (n = 3) or associated with relapse of other localizations or occurrence of new localizations (n = 4). Relapses were purely extrarenal in 6 cases. Relapses occurred after steroids were discontinued in 10 patients with a median of 5.5 months (range, 1-48 mo), and occurred during steroid therapy in 7 patients with a median dose of 10 mg/d (range, 3-40).

Recurrences were treated with prednisone in all cases, and with additional immunosuppressive therapy in 9 cases. The median duration of the relapse treatment was 48 months (range, 24-76 mo). The eGFR at the end of follow-up of patients with renal relapse was not different from that of patients who had not relapsed (median, 47 mL/min per 1.73 m2 vs. 49 mL/min per 1.73 m2, respectively). Unfavorable response to therapy at 1 month was associated with relapse during the disease progression up to the end of follow-up (p = 0.049). Characteristics of patients who had relapse are presented in Table 4.

T4-4
TABLE 4:
Patients With Relapse

One patient died from myocardial infarction. One patient developed multiple myeloma 2 years after the diagnosis, and another developed primary hyperparathyroidism 6 months after the diagnosis. The only systematically recorded adverse side effect of corticosteroid therapy was type 2 diabetes mellitus, which was observed in 6 patients. No other occurrence of severe adverse event was mentioned. In particular, no symptomatic osteoporosis was noticed. Because of the retrospective nature of the study, minor adverse events secondary to corticosteroid therapy could not be systematically investigated.

Sparing Agents and Immunosuppressive Therapy

Several immunosuppressive agents (low-dose methotrexate, azathioprine, and mycophenolate mofetil) were added as steroid-sparing agents in 12 patients (9 cases for extrarenal relapses and 3 for renal relapses). In the 3 cases with renal relapse, mycophenolate mofetil (n = 2) and azathioprine (n = 1) were used and allowed a sustainable control of the renal function while reducing the steroid level below 5 mg/d for 2 patients and 12 mg/d for 1 patient. No severe adverse event relative to the immunosuppressive therapy was observed.

Interstitial Fibrosis, a Predictive Factor of Renal Function Outcome (Figure 6)

F6-4
FIGURE 6:
eGFR at presentation, at 1 year, and at the end of follow-up as a function of renal fibrosis score (F1, F2, F3 as described in Methods section).

We found an inverse relationship between the response to steroids treatment and the initial degree of interstitial fibrosis. In patients with F1, initial eGFR increased significantly after treatment, from 25.3 ± 12 to 53.7 ± 27.9 mL/min per 1.73 m2 at 1 year (p < 0.05), and to 52.8 ± 28.8 mL/min per 1.73 m2 at the end of follow-up (p < 0.05). In patients with F2, initial eGFR increased significantly from 28.6 ± 23.7 to 39.3 ± 12.2 mL/min per 1.73 m2 at 1 year (p < 0.05), and to 45.8 ± 19.6 mL/min per 1.73 m2 at the end of follow-up (p < 0.05). In contrast, in patients with F3, initial eGFR did not improve with treatment: from 16 ± 12.7 to 18 ± 11.5 mL/min per 1.73 m2 at 1 year, and to 15.2 ± 8.9 mL/min per 1.73 m2 at the end of follow-up. So, patients with F3 did not respond to therapy. There was no relationship between eGFR at 1 month, 1 year, and at the end of follow-up for any other histologic parameter, including the presence of epithelioid granulomas, of giant cells, tubular atrophy, interstitial inflammatory cell infiltrates, vascular intimal sclerosis, or calcium deposits.

DISCUSSION

In the current study we present the largest (to our knowledge) clinical series on RS with data on renal function and histology, but as well as the associated clinical manifetations and long-term outcome. All medical departments involved in this study were experienced with sarcoidosis and used similar procedures for investigation and treatment. This study makes 2 strong points. The first concerns the sarcoidosis clinical phenotype in case of renal localization. Our population of patients with RS was unique regarding the high prevalence of hypercalcemia. Hypercalcemia occurred mostly in white males during summer months. We reported a high prevalence of fever and M/F ratio, unusual features in common forms of sarcoidosis. The other point is the high responsiveness of renal insufficiency to steroids, which was sustained with time. More importantly, we demonstrate 2 predictive factors of long-term renal function outcome: 1) renal function outcome was inversely related to the degree of interstitial renal fibrosis; and 2) long-term renal response to treatment was related to the response at 1 month.

Most published series have focused on renal function.5,29 In our experience, RS mostly occurs at the presentation of sarcoidosis. More rarely (19%), it occurs during the course of previous sarcoidosis with a long time interval. We did not observe any case of isolated granulomatous interstitial nephritis preceding the diagnosis of sarcoidosis, as described in another report;5 and RS was different from ocular, skin, neurologic, or nasosinusal localizations that may remain isolated for a long time before further localizations. Thus, RS should be readily suspected in a patient with renal failure and with an evident sarcoidosis context. Taking into account the possibility of delayed cases of RS in patients with sarcoidosis, one of the take-home recommendations is to assess systematically the baseline renal function, every 6 months during the survey, and in case of outbreak or relapse of sarcoidosis.

The clinical phenotype of RS did not differ significantly from that of patients without renal involvement according to the prevalence of thoracic and extrathoracic localizations, which were similar to the ACCESS study.32 However, fever was unusually frequent in our population, considering the absence of Löfgren or Heerfordt syndromes, which are well-known causes of fever in sarcoidosis. We also observed a male predominance in RS (M/F ratio = 1.76 in our series) as reported in the literature on RS,5 but in contrast with common forms of sarcoidosis, in which an increased prevalence of 10%-33% is usually observed in female patients.31

The high prevalence of hypercalcemia (32%), which concerned white patients almost exclusively, was probably one of the most striking findings. Berliner et al5 previously reported a prevalence of 19.5% (Ca >2.75 mmol/L) in RS. In sarcoidosis without renal involvement, hypercalcemia was less common and was noted in only 3.7% cases in the prospective ACCESS study.32 In that study, hypercalcemia was also observed mostly in white patients.32 In the current series, the prevalence of hypercalcemia reached 50% in white patients during summer months. Thus, the association between RS and hypercalcemia deserves discussion. All our cases with hypercalcemia had hypercalciuria. Insufficient excretion of calcium does not seem to explain hypercalcemia, any more than renal involvement does. Activated monocytes/macrophages in granulomas are capable of producing calcitriol and an excessive production of calcitriol is considered to be the main mechanism of abnormal calcium metabolism in sarcoidosis.1 Except for renal localization, sarcoidosis was very similar in our cases when nonrenal localizations were concerned to what was described in the ACCESS study, in which hypercalcemia was rare. Thus, a first hypothesis accounting for the high frequency of hypercalcemia in RS is that granulomas in the kidney could be an important source of calcitriol, either because of a large amount of granulomas in the kidney or because of an increased propensity of the 1α OH vitamin D hydroxylase gene to be over-expressed in renal granulomas. However, we did not find any correlation between the score of granulomas in renal biopsy and the calcemia level. Since some haplotype-phenotype correlations have been demonstrated in sarcoidosis, one can suppose that some haplotypes could predispose both to the presence of RS and to abnormal calcium metabolism, perhaps HLA DPB1*0101, which has been associated with hypercalcemia in white patients.30

Hypercalcemia is aggravated by sunlight, and its seasonal variation with higher rate during spring and summer has been reported before.1,2,36 We confirm here this seasonal hypercalcemia variation. Hypercalcemia responded dramatically to steroids, and no relapse was observed. Vitamin D3 metabolism was not comprehensively investigated in the current study, and the frequent hypercalcemia seen in our series of RS patients deserves further studies with a focus on vitamin D3 metabolism in the particular setting of RS with hypercalcemia. We found that PTH level tended to be low, by that excluding the role of hyperparathyroidism at the origin of hypercalcemia. Serum 25OHD3 level was low too, a finding already reported in sarcoidosis with abnormal calcium metabolism.27

In the current series, renal function was severely impaired, with a median serum creatinine level of 263.5 μmol/L; 90% of patients had an initial eGFR <60 mL/min per 1.73 m2. Renal function impairment was associated with mild proteinuria (in 64% of cases), and more rarely with aseptic leukocyturia and microscopic hematuria. Thus, the absence of proteinuria, hematuria, or leukocyturia did not exclude the presence of RS. Comparable findings have been described in other studies on interstitial granulomatous nephritis6,18 and on RS.4,8,11,13,15-17,20,26,29,34 In the 94 cases of sarcoid granulomatous interstitial nephritis compiled in the literature and described by Berliner et al,5 the mean creatinine level was 424 μmol/L. A large majority of patients had proteinuria of 1 g/d or less. So, in both series, most patients with RS had a severe renal disease. The patient with a pseudotumoral infiltration in the left kidney illustrates the polymorphism of RS clinical expression.24

GIN is the most typical histologic feature observed in patients with RS.3,4,6,8,11-13,15-21,29,33,34 In the current study, 10 patients had interstitial nephritis without granulomas in the renal biopsy specimen. This population did not differ from other patients with GIN. No other cause of interstitial nephropathy has been found. Thus, we considered them to be authentic cases of RS. In them, sparse granulomatous lesions could have been missed in renal biopsies.

Since the first report in 1955 of successful therapy for sarcoid GIN by Berger and Relman,3 steroids have been the gold standard treatment for RS. However, there is no standardized protocol for doses or duration. Most authors recommend doses of 0.5-1 mg/kg per day of oral prednisone as the initial regimen.5,29 Most patients in our series received high-dose prednisone at 1 mg/kg per day. We showed that most patients responded to steroids with a clear improvement of eGFR at 1 month, at 1 year, and at last follow-up. However, 2 major limits of the efficacy of the treatment have been demonstrated in our study. First, a few patients did not respond to treatment. Second, although steroids improved the renal function of a large majority of patients, only a minority of them recovered normal renal function. At the end of follow-up, only 4 patients had an eGFR >90 mL/min per 1.73 m2. More than 60% of patients had an eGFR <60 mL/min per 1.73 m2, and 2 patients had end-stage renal failure that necessitated hemodialysis. These results are in accordance with those obtained by Rajakariar et al29 in a series of 19 patients and with those obtained by Berliner et al5 in the compilation of 94 cases.

In the current study it is noteworthy that we showed that long-term renal function outcome in response to treatment was predictable at presentation. The renal function improvement was inversely related to the initial fibrosis score. Patients with an extensive interstitial fibrosis F3 score were unresponsive to treatment. This observation is in agreement with that of Rajakariar et al.29 Thus, the presence of interstitial renal fibrosis at presentation is certainly an important piece of data to consider for explaining the limitation of treatment efficacy on renal function. An unsolved question is why renal fibrosis is seen, even at a variable degree from case to case, in all patients. It could appear to be a paradox since in most cases, diagnosis of RS was established at presentation of sarcoidosis. A possible hypothesis could consist of a specific trend of renal granulomas to promote the development of fibrotic lesions very rapidly. An alternative explanation is the insidious nature of RS and its late diagnosis. Nevertheless, in order to contain renal fibrosis, steroids therapy should be initiated as early as possible.

We also demonstrated that the renal function response at 1 month was highly predictive of the long-term response. This observation suggests that renal granulomatous and inflammatory renal lesions are rapidly responsive to treatment and stay stable under treatment, thus without progression of the renal fibrotic lesions. In addition, we found that patients with an unfavorable response at 1 month were more susceptible to relapse at the end of follow-up. These findings emphasize the predictive value of eGFR response at 1 month of treatment. We believe this result is clinically relevant and may help clinicians monitor steroid therapy for duration and dose.

A challenging question is whether another treatment could have improved the results obtained with prednisone. As seen above, one hypothesis is that no treatment can further modify the renal function, since only fibrotic lesions were unresponsive to prednisone. However, since renal fibrosis could be a rapid process, the question of using more rapidly efficient treatments is raised. In this context, we scrutinized the effect of MP in 10 of our cases. The current series is, to our knowledge, the first to describe the outcome of patients receiving MP, since only a few cases have been reported before.14 In our study, MP seems to have been beneficial. Initial eGFR for patients receiving MP was significantly lower than for others. At the end of follow-up, 5 of 10 (50%) patients receiving MP had an eGFR >60 mL/min per 1.73 m2, compared to 10 of 36 (27.7%) patients treated with oral steroids. These results suggest a better response to intravenous high-dose steroids than to oral steroids, but need to be assessed in a prospective trial.

Important clinical considerations are how long to administer the treatment, and what should be the follow-up modalities. Some authors recommended maintaining the steroids indefinitely.29 Our results showed that relapses occur during reduction or after discontinuation of steroids with a median of 24 months of treatment. Relapses occurred in renal as well as extrarenal localizations in up to 36% of cases, some of whom had severe cardiac or central nervous system involvement. Altogether, these observations support the necessity for prolonged treatment and for checking patients during treatment and for a long time after its cessation, periodically assessing creatininemia. Because of the necessity of long-term treatment and the risk of corticosteroid-induced side effects such as diabetes mellitus, as occurred in 6 of our cases, the use of corticosteroid-sparing agents is an important issue. In the current study we observed a beneficial sparing effect from azathioprine and mycophenolate mofetil in 3 patients.

Limitations of the current study are in part a result of the retrospective collection of data and incomplete data monitoring. For example, the side effects of steroids were not systematically recorded. Calcitriol metabolism was not investigated extensively. Despite these limitations, the study has highlighted some important features of RS. We show here, for the first time to our knowledge, the predictive value of interstitial renal fibrosis and renal function response at 1 month on long-term renal function outcome. Our results confirm the efficacy of steroids on renal function in RS. Nevertheless, the incomplete response observed in most of our patients suggests that current treatments could be improved. The use of intravenous steroids and steroid-sparing agents needs to be tested in prospective trials.

REFERENCES

1. Adams JS, Sharma OP, Gacad MA, Singer FR. Metabolism of 25-hydroxyvitamin D3 by cultured pulmonary alveolar macrophages in sarcoidosis. J Clin Invest. 1983;72:1856-1860.
2. Adams JS, Singer FR, Gacad MA, Sharma OP, Hayes MJ, Vouros P, Holick MF. Isolation and structural identification of 1,25-dihydroxyvitamin D3 produced by cultured alveolar macrophages in sarcoidosis. J Clin Endocrinol Metab. 1985;60:960-966.
3. Berger KW, Relman AS. Renal impairment due to sarcoid infiltration of the kidney; report of a case proved by renal biopsies before and after treatment with cortisone. N Engl J Med. 1955;252:44-49.
4. Bergner R, Hoffmann M, Waldherr R, Uppenkamp M. Frequency of kidney disease in chronic sarcoidosis. Sarcoidosis Vasc Diffuse Lung Dis. 2003;20:126-132.
5. Berliner AR, Haas M, Choi MJ. Sarcoidosis: the nephrologist's perspective. Am J Kidney Dis. 2006;48:856-870.
6. Bijol V, Mendez GP, Nose V, Rennke HG. Granulomatous interstitial nephritis: a clinicopathologic study of 46 cases from a single institution. Int J Surg Pathol. 2006;14:57-63.
7. Branson JH, Park JH. Sarcoidosishepatic involvement: presentation of a case with fatal liver involvement; including autopsy findings and review of the evidence for sarcoid involvement of the liver as found in the literature. Ann Intern Med. 1954;40:111-145.
8. Brause M, Magnusson K, Degenhardt S, Helmchen U, Grabensee B. Renal involvement in sarcoidosis-a report of 6 cases. Clin Nephrol. 2002;57:142-148.
9. Carmichael P, O'Donnell JP. The protean face of renal sarcoid. J Nephrol. 2003;16:721-727.
10. Chapelon C, Uzzan B, Piette JC, Jacques C, Coche E, Godeau P. [Sarcoidosis in internal medicine. A cooperative study of 554 cases]. Ann Med Interne (Paris). 1984;135:125-131.
11. Coutant R, Leroy B, Niaudet P, Loirat C, Dommergues JP, Andre JL, Baculard A, Bensman A. Renal granulomatous sarcoidosis in childhood: a report of 11 cases and a review of the literature. Eur J Pediatr. 1999;158:154-159.
12. Dobrin RS, Vernier RL, Fish AL. Acute eosinophilic interstitial nephritis and renal failure with bone marrow-lymph node granulomas and anterior uveitis. A new syndrome. Am J Med. 1975;59:325-333.
13. Duvic C, Herody M, Rossignol P, Lecoules S, Didelot F, Nedelec G. [Renal manifestations of sarcoidosis. A report of nine cases]. Rev Med Interne. 1999;20:226-233.
14. Gottlieb JE, Israel HL, Steiner RM, Triolo J, Patrick H. Outcome in sarcoidosis. The relationship of relapse to corticosteroid therapy. Chest. 1997;111:623-631.
15. Grateau G, Hannedouche T, Dehaine V, Fournier G, Barbanel B, Grunfeld JP, Jungers P. [Renal failure disclosing interstitial granulomatous nephropathy of sarcoidosis. 4 cases]. Presse Med. 1988;17:1393-1395.
16. Guenel J, Chevet D. [Interstitial nephropathies in sarcoidosis. Effect of corticosteroid therapy and long-term evolution. Retrospective study of 22 cases]. Nephrologie. 1988;9:253-257.
17. Hannedouche T, Grateau G, Noel LH, Godin M, Fillastre JP, Grunfeld JP, Jungers P. Renal granulomatous sarcoidosis: report of six cases. Nephrol Dial Transplant. 1990;5:18-24.
18. Javaud N, Belenfant X, Stirnemann J, Laederich J, Ziol M, Callard P, Ronco P, Rondeau E, Fain O. Renal granulomatoses: a retrospective study of 40 cases and review of the literature. Medicine (Baltimore). 2007;86:170-180.
19. Kenouch S, Mery JP. [Involvement of the renal interstitium in systemic diseases (II)]. Nephrologie. 1988;9:117-124.
20. Kettritz R, Goebel U, Fiebeler A, Schneider W, Luft F. The protean face of sarcoidosis revisited. Nephrol Dial Transplant. 2006;21:2690-2694.
21. Laroche M, Ader JL, Mazieres B. [Sarcoidosic tubulopathy simulating hyperparathyroidism]. Presse Med. 1990;19:1548.
22. Levey AS, Eckardt KU, Tsukamoto Y, Levin A, Coresh J, Rossert J, De Zeeuw D, Hostetter TH, Lameire N, Eknoyan G. Definition and classification of chronic kidney disease: a position statement from Kidney Disease: Improving Global Outcomes (KDIGO). Kidney Int. 2005;67:2089-2100.
23. Longcope WT, Freiman DG. A study of sarcoidosis; based on a combined investigation of 160 cases including 30 autopsies from The Johns Hopkins Hospital and Massachusetts General Hospital. Medicine (Baltimore). 1952;31:1-132.
24. Lucci S, Rivolta R, Fazi M, Iorio L, Raschella GF, Merlino G, Giordano R, Redler A. [Sarcoidosis and clear cell carcinoma of the kidney: the sixth case?]. G Chir. 2002;23:75-78.
25. Mayock RL, Bertrand P, Morrison CE, Scott JH. Manifestations of sarcoidosis. Analysis of 145 patients, with a review of nine series selected from the literature. Am J Med. 1963;35:67-89.
26. McCurley T, Salter J, Glick A. Renal insufficiency in sarcoidosis. A clinical and pathologic study. Arch Pathol Lab Med. 1990;114:488-492.
27. Meyrier A, Valeyre D, Bouillon R, Paillard F, Battesti JP, Georges R. Resorptive versus absorptive hypercalciuria in sarcoidosis: correlations with 25-hydroxy vitamin D3 and 1,25-dihydroxy vitamin D3 and parameters of disease activity. Q J Med. 1985;54:269-281.
28. Milman N, Selroos O. Pulmonary sarcoidosis in the Nordic countries 1950-1982. Epidemiology and clinical picture. Sarcoidosis. 1990;7:50-57.
29. Rajakariar R, Sharples EJ, Raftery MJ, Sheaff M, Yaqoob MM. Sarcoid tubulo-interstitial nephritis: long-term outcome and response to corticosteroid therapy. Kidney Int. 2006;70:165-169.
30. Rossman MD, Thompson B, Frederick M, Maliarik M, Iannuzzi MC, Rybicki BA, Pandey JP, Newman LS, Magira E, Beznik-Cizman B, Monos D. HLA-DRB1*1101: a significant risk factor for sarcoidosis in blacks and whites. Am J Hum Genet. 2003;73:720-735.
31. Rybicki BA, Iannuzzi MC. Epidemiology of sarcoidosis: recent advances and future prospects. Semin Respir Crit Care Med. 2007;28:22-35.
32. Semenzato G. ACCESS: A Case Control Etiologic Study of Sarcoidosis. Sarcoidosis Vasc Diffuse Lung Dis. 2005;22:83-86.
33. Silverstein E, Brunswick J, Rao TK, Friedland J. Increased serum angiotensin-converting enzyme in chronic renal disease. Nephron. 1984;37:206-210.
34. Simonsen O, Thysell H. Sarcoidosis with normocalcemic granulomatous nephritis. Five case reports and a review of 24 cases in the literature. Nephron. 1985;40:411-417.
35. Statement on sarcoidosis. Joint Statement of the American Thoracic Society (ATS), the European Respiratory Society (ERS) and the World Association of Sarcoidosis and Other Granulomatous Disorders (WASOG) adopted by the ATS Board of Directors and by the ERS Executive Committee, February 1999. Am J Respir Crit Care Med. 1999;160:736-755.
36. Vidal M, Ramana CV, Dusso AS. Stat1-vitamin D receptor interactions antagonize 1,25-dihydroxyvitamin D transcriptional activity and enhance stat1-mediated transcription. Mol Cell Biol. 2002;22:2777-2787.
37. Wirnsberger RM, de Vries J, Wouters EF, Drent M. Clinical presentation of sarcoidosis in The Netherlands: an epidemiological study. Neth J Med. 1998;53:53-60.
© 2009 Lippincott Williams & Wilkins, Inc.