Rheumatology Advance Access originally published online on April 4, 2008
Rheumatology 2008 47(5):702-707; doi:10.1093/rheumatology/ken019
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Revised classification of lupus nephritis is valuable in predicting renal outcome with an indication of the proportion of glomeruli affected by chronic lesions
Department of Medicine and Clinical Science, Gunma University Graduate School of Medicine, Maebashi, Gunma, Japan.
Correspondence to: T. Kuroiwa, 3-39-22 Showa, Maebashi, Gunma 371-8511, Japan. E-mail: tkuroiwa{at}med.gunma-u.ac.jp
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Objectives. To determine if the International Society of Nephrology/Renal Pathology Society (ISN/RPS) 2003 classification of lupus nephritis (LN) is helpful in predicting renal outcome.
Methods. A total of 92 patients with LN who underwent renal biopsy in our hospital were re-classified according to the ISN/RPS 2003 criteria.
Results. The mean patient age was 36.8 yrs and the median observation period was 65 months. The relative frequency for each class was as follows: Class I (minimal mesangial LN) 0%, Class II (mesangial proliferative LN) 13%, Class III (focal LN) 17%, Class IV (diffuse LN) 60% and Class V (membranous LN) 10%. Within Class IV, diffuse segmental (Class IV-S) was 25% and diffuse global (Class IV-G) 75%. During the observation period, renal function was more likely to deteriorate in Class IV-G cases than in Class IV-S cases. Importantly, when Class IV-G was subdivided into cases involving active lesion alone [IV-G (A)] or chronic lesion [IV-G (A/C)], the majority of cases in IV-G (A) was nephrotic, but responded well to therapy. In contrast, renal function declined only in IV-G (A/C) cases. Patients with Class IV-G (A/C) had persistent proteinuria in spite of intensified therapies. Moreover, the higher proportion of chronic lesions was related with the deterioration of renal function.
Conclusions. This study showed that in Class IV-G cases, renal outcome differed in the presence of chronicity. Chronicity could be a critical factor in predicting outcome. Thus, the revised classification of LN is clinically valuable in identifying different renal outcomes among patients with diffuse LN.
KEY WORDS: SLE, Renal, Histopathology, Immunosuppressants
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Lupus nephritis (LN) is a common manifestation in SLE. In this disease, autoantibodies are thought to participate in the formation of glomerular immune deposits, initiating inflammatory responses. The morphological expression can vary considerably among patients or within an individual over time, which is related to the location, quantity and pro-inflammatory properties of immune deposits. The original World Health Organization (WHO) classification of LN was introduced in 1974 [1], defined in 1982 [2] and revised in 1995 [3]. This classification scheme has been used for histological diagnosis and has been shown to be valuable in predicting renal outcome and choosing appropriate therapies. However, its different categories and diagnostic terminology needed clarification.
Recently, the International Society of Nephrology/Renal Pathology Society (ISN/RPS) 2003 classification of LN was proposed [4]. In this classification, diagnostic terminology has been more clearly defined. As a result, histological diagnosis is expected to become more reproducible and differences among renal pathologists are to be decreased. The basic structure is similar to that of the WHO classification; Classes I and II are used for pure mesangial involvement (I, mesangial immune deposits without mesangial hypercellularity; II, mesangial immune deposits with mesangial hypercellularity); Class III for focal glomerulonephritis (involving <50% of total number of glomeruli); Class IV for diffuse glomerulonephritis (involving 
50% of total number of glomeruli); Class V for membranous lupus nephritis; and Class VI for advanced sclerosing lesions.
One of the most important changes of the revised classification is that Class IV was further subdivided into diffuse glomerulonephritis with predominantly segmental (Class IV-S) or predominantly global (Class IV-G) involvement. Class IV-S is defined as Class IV with <50% capillary tuft involvement and Class IV-G as 50%. This change was prompted by differences in prognosis reported by Schwartz et al. [5] and Najafi et al. [6] that patients with predominantly segmental but diffuse involvement have a worse prognosis than those with a diffuse global endocapillary proliferative pattern of injury. Another important change is that in Classes III and IV, the proportion of glomeruli affected by active (A) and chronic (C) lesions should be indicated. For example, Class IV-S is subdivided into IV-S (A) and IV-S (C) when all of the involved glomeruli have active and chronic lesions, respectively, or into IV-S (A/C) when involved glomeruli have both active and chronic lesions. Considering previous studies showing the importance of chronic lesions in predicting renal outcome, this subcategorization can be very important [7, 8]. Patients with Class IV having chronic lesions may have a worse outcome than those with Class IV with active lesions.
In the present study, we sought to examine whether the ISN/RPS 2003 classification of LN is valuable in predicting renal outcome in patients with LN. We were particularly interested in whether the changes described above, subcategorization of Class IV into Class IV-S or IV-G, and the indication of activity or chronicity in Classes III and IV, are helpful. To this end, we histologically re-classified patients with LN who underwent renal biopsy in our hospital according to the ISN/RPS 2003 criteria. Clinical records were reviewed and information regarding renal function, serological activity and urinary protein was collected and compared across histological classes. Our data indicate that although renal function is more likely to deteriorate in Class IV-G, renal outcome differs in the presence of chronicity.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Patients
Patients with LN who underwent renal biopsy between 1986 and 2003 in our hospital were included in this study. All patients met the 1982 ACR revised criteria for the classification of SLE. For inclusion, patients had to have a follow-up period of >1 yr and adequate renal biopsy samples for histological diagnosis, including >10 glomeruli. For the patients who had repeated biopsies, samples and clinical records at the time of the first biopsy were used. Consequently, data of 92 patients were available in the study. For included patients, clinical records at biopsy and last observation were collected.
Histological diagnosis
Renal biopsy samples were histologically re-classified according to the ISN/RPS 2003 criteria. Data regarding immunofluorescence findings were available for 90% of the patients and data regarding electron microscopic findings were available for 40% of the patients. Histological diagnosis was determined by experienced nephrologists T.K., K.H. and K.U., who were blinded to clinical information.
Clinical data
Collection of clinical data was performed in December 2004. The following clinical parameters were evaluated at the time of biopsy and last observation: blood pressure, CH50, C3, C4, anti-DNA antibodies, serum creatinine and proteinuria. For proteinuria, the ratio of urinary protein to urinary creatinine (g/gCr) from a spot urine sample was used. Nephrotic syndrome was defined as an increase in the urinary protein ratio of 3.5 g/gCr and a decrease in serum total protein to 
6.0 g/dl or serum albumin to 3.0 g/dl. At the last examination, renal function was categorized by the level of serum creatinine (mg/dl) as follows: normal-mild renal insufficiency (<1.5), moderate renal insufficiency (1.5–3.0) and advanced renal insufficiency (3.0). Renal flare was defined as an increase of urinary protein of >1 g/gCr associated with an increase in serological activity.
Statistical analysis
For statistical testing of normality, the 
2 test for goodness of fit was used. Data with non-normal distribution were then summarized as median and range; data with normal distribution were summarized as mean and S.D. Statistical analysis was performed using the Kruskal–Wallis non-parametric analysis of variance test and the Wilcoxon rank-sum test for continuous and categorical data, and the Kaplan–Meier life-table method for renal survival. A P-value of <0.05 was considered statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Patients
Patient profiles in this study are listed in Table 1. A total of 92 patients were included in this study. The mean patient age at renal biopsy was 36.8 yrs and the median observation period was 65 months. Of the 92 patients, 78 (85%) were female. Data at biopsy are as follows: median serum creatinine, 0.7 (0.4–2.4) mg/dl; median urinary protein, 3.2 (0–42.6) g/gCr and mean serum albumin, 3.0 ± 0.8 g/dl. Cases were serologically active, with a median CH50 of 13.3 (<5–50.7) U/ml and a median anti-DNA antibody titre of 37 (0–838) IU. Hypertension was present in 50% of the cases.
|
Histology
The relative frequency of each class was as follows: Class I (minimal mesangial LN) 0%, Class II (mesangial proliferative LN) 13%, Class III (focal LN) 17%, Class IV (diffuse LN) 60%, Class V (membranous LN) 10% and Class VI (advanced sclerosing LN) 0% (Table 2). Within Class IV, diffuse segmental (Class IV-S) was 25% and diffuse global (Class IV-G) 75%. Cases of III + V were considered to be Class III and cases of IV + V were considered to be Class IV.
|
Classes III and IV were further subdivided into (A), (A/C) and (C). Within Class III, (A) was 65%, (A/C) was 35% and (C) was 0%. Within Class IV-S, IV-S (A) was 71%, IV-S (A/C) was 29%, and IV-S (C) was 0%. Within Class IV-G, IV-G (A) was 51%, IV-G (A/C) was 49% and IV-G (C) was 0% (Table 2).
Clinical parameters and treatments among each class
Clinical data at renal biopsy by class are summarized in Table 3. In Class II, renal function was well preserved: serum creatinine was 0.7 mg/dl. Urinary protein excretion was mild: average urinary protein was 0.5 g/gCr and the percentage of patients with nephrotic syndrome was 8%. Cases of Class II were not serologically active. Data were similar between Classes III and IV-S. In both, renal function was still well preserved. The frequency of nephrotic syndrome and the levels of anti-DNA antibodies were comparable. As predicted, patients with Class IV-G had the most severe nephritis. These patients had the highest serum creatinine (0.8 mg/dl). Indeed, 81% of the patients in this class had nephrotic syndrome. Additionally, in Class V, 67% of patients had nephritic syndrome, as predicted.
|
In a comparison of Class IV-G with and without chronic lesions, the incidence of severe nephritis was similar. However, interestingly, serological activity was lower in Class IV-G with chronic lesions [IV-G (A/C)] with significantly lower anti-DNA antibodies (P = 0.048). In Classes III and IV-S, data of (A) were not significantly different from that of (A/C).
Treatments during follow-up are summarized in Table 4. Patients were treated with either steroid alone or a combination of steroid and immunosuppressants. Steroid pulse therapy was given to 17% of the patients in Class II, 25% in Class III, 36% in Class IV-S, 51% in Class IV-G [52% in Class IV-G (A), 50% in Class IV-G (A/C)] and 22% in Class V. Immunosuppressants were administered to 17% of the patients in Class II, 56% in Class III, 71% in Class IV-S, 78% in Class IV-G [81% in Class IV-G (A), 75% in Class IV-G (A/C)] and 33% in Class V. Among immunosuppressants, cyclophosphamide, administered either orally or intravenously, was most frequently used. Oral cyclophosphamide was prescribed to 17% in Class II, 38% in Class III, 36% in Class IV-S, 34% in Class IV-G [38% in Class IV-G (A), 30% in Class IV-G (A/C)] and 33% in Class V. Use of i.v. cyclophosphamide was limited to patients in Class IV: 7.1% in Class IV-S and 27% in Class IV-G [29% in (A) and 25% in (A/C)].
|
Renal function and outcome
Renal function at the last follow-up is shown in Table 5. Renal function was categorized by the level of serum creatinine (mg/dl) as follows: normal-mild renal insufficiency (<1.5), moderate renal insufficiency (1.5–3.0) and advanced renal insufficiency (
3.0). Overall, renal function was well preserved in our follow-up. After an average observation period of 7 yrs, renal function remained at normal-mild renal insufficiency in 93% of the patients. Only four patients (4%) reached advanced renal insufficiency. Three patients (3%) had moderate renal insufficiency. All patients in Classes II and V had normal-mild renal insufficiency. Even in Classes III and IV-S, renal function was well preserved; only one patient reached advanced renal insufficiency in Class III and one had moderate renal insufficiency in Class IV-S. In Class IV-G, renal outcome was the worst: three patients had advanced renal insufficiency and two had moderate renal insufficiency, although a statistically significant difference was not observed (P = 0.562). Importantly, when Class IV-G was subdivided into active lesion alone [Class IV-G (A)] or chronic lesion [Class IV-G (A/C)], renal function was significantly worse in Class IV-G (A/C) than in Class IV-G (A) (P = 0.016). Renal function remained at normal-mild renal insufficiency in all patients with Class IV-G (A).
|
Decline of renal function during the follow-up period was further examined by Kaplan–Meier analysis with doubling serum creatinine as the end point. First, we determined overall renal outcome. Of the patients, 81% did not have doubling of serum creatinine at 120 months (Fig. 1A). Second, we determined which class had the worst renal outcome. Renal function deteriorated most in Class IV: 15% of patients reached the end point after 120 months (Fig. 1B). Third, we determined as to which is worse in renal outcome—Class IV-S or IV-G. A statistically significant difference did not exist between Classes IV-S and IV-G; however, there seemed to be some trend towards a worse outcome in Class IV-G (P = 0.68) (Fig. 1C). Fourth, we determined which is worse in renal outcome—Class IV-G (A) or IV-G (A/C). As shown in Fig. 1D, renal outcome of Class IV-G was substantively different when chronic lesions were present. After 120 months, 50% in Class IV-G (A/C) reached the end point, while no patients in Class IV-G (A) reached the end point (Fig. 1D). Fifth, we determined if renal outcome could be related with the degree of chronic lesions. Here, the degree of glomerular chronic lesions was defined as the ratio of the number of glomeruli with chronic lesions alone to the total number of glomeruli that have lesions, either active or chronic. Patients in Class IV-G (A/C) were then divided into two groups according to the degree of glomerular chronic lesion: >50% and <50%. Figure 1E shows that patients in whom the degree of glomerular chronic lesions was >50% have a much higher risk for deterioration of renal function than those having the degree <50%. Austin et al. [7, 8] have shown that chronicity index has prognostic value in predicting renal outcome. As shown in Fig. 2, the correlation between chronicity index and the degree of chronic lesions was highly significant. Thus, the risk of doubling creatinine increases in proportion to increasing chronic lesions and the degree of chronic lesions could have prognostic value in predicting renal outcome.
|
|
Urinary protein
The level of urinary protein at the last follow-up is shown in Table 6. In Class IV-G, 71% of the patients had urinary protein <0.5 g/gCr. Consistent with better renal outcome, >91% in Class IV-G (A) had urinary protein levels <0.5 g/gCr, compared with only 50% in Class IV-G (A/C). Accordingly, response to the therapies differed among these classes. The majority of patients in IV-G (A) had nephrotic syndrome and highly active serological examination results, but they responded well to therapy, resulting in remission. In contrast, patients in Class IV-G (A/C) had persistent proteinuria in spite of intensified therapies.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
The new classification scheme for LN has clearer definitions of diagnostic terminology and the subsequent histological diagnosis is expected to be more reproducible. One of the most notable changes is that diffuse LN (Class IV) was subdivided into diffuse LN with segmental (Class IV-S) or global (Class IV-G) involvement and that in Classes III, IV-S and IV-G, the proportion of glomeruli affected by active and chronic lesions should be indicated. The question we asked in this study is whether such changes in the new classification are useful in predicting the renal outcome of LN. We first found that Class IV-G had the highest frequency at 45%. These patients also had the most severe nephritis: 80% presented with nephrotic syndrome at the time of renal biopsy. Renal function at the last follow-up was the worst in Class IV-G. This was corroborated by the result of Kaplan–Meier analysis with doubling serum creatinine as the end point. However, the most important finding in this study was that within Class IV-G, renal outcome is primarily determined by the degree of glomerular chronicity. Patients with Class IV-G with active lesions alone [Class IV-G (A)] had very good renal outcome. Indeed, no patients in this class had renal insufficiency or end-stage renal disease at the last follow-up. They also responded very well to the therapy. The level of urinary protein was <0.5 g/gCr in >90% of this class, suggesting that these patients have a very low risk of deterioration of renal function. Collectively, these data suggest that patients with Class IV-G (A), despite their very severe nephritis, have a very low risk for deterioration of renal function under the influence of effective therapy.
In contrast, patients with Class IV-G having chronic lesions [IV-G (A/C)] had much higher risk for the deterioration of renal function. At the last follow-up, 25% in this class had worsening of renal function, including renal insufficiency or end-stage renal disease. They also had significantly higher risk for having doubling serum creatinine compared with those with Class IV-G (A). It is noteworthy that the risk is amplified as the degree of chronic lesions increases. Concordantly, 45% of the patients had >1.0 g/gCr of urinary protein at their last examination. Prolonged urinary protein is a risk factor for the deterioration of renal function. Therefore, more patients in this class may have worsening of renal function over time. These observations led us to conclude that the revised classification ‘with a description of the degree of glomerular chronicity’ is helpful in predicting renal outcome in LN.
Austin et al. [7, 8] have shown that the National Institutes of Health (NIH) chronicity index, but not the activity index, has diagnostic value in predicting renal outcome. In our cohort, the NIH chronicity index was well correlated with the proportion of chronic lesions. In addition, patients who reached doubling serum creatinine also had a high chronicity index. In contrast, patients with Class IV-G (A), in general, had very high activity index; however, this was not associated with deterioration of renal function. Revised criteria, by indicating the proportion of chronic lesions, have the advantage to be able to indicate the magnitude of the NIH chronicity index.
Before the ISN/RPS 2003 classification system was proposed, Najafi et al. [6] analysed 85 patients with lupus glomerulonephritis and revealed the poor outcome of lupus nephritis with segmental necrosis involving over 50% of glomeruli as compared with diffuse proliferative lupus nephritis. They concluded that histopathological categorization among patients with severe lupus glomerulonephritis does provide information relevant to their long-term outcome [6]. After the proposal of the ISN/RPS 2003 classification system, three groups reported the relation between ISN/RPS 2003 and renal outcome [9–11]. All of them, however, have mainly focused on the difference in renal outcome between Classes IV-S and IV-G, and the importance of chronicity has been sparsely discussed. Compared with these studies, our study is unique in emphasizing the importance of the indication of chronicity as well as the class. In terms of controversy between Classes IV-S and IV-G, we did not detect a significant difference in the rates of doubling serum creatinine between the groups. Yokoyama et al. [9] examined 60 cases of LN with 187 months of mean follow-up by defining primary outcome as end-stage renal disease. Although they found a trend towards a worse prognosis for Class IV-S compared with Class IV-G, it was not statistically significant. Mittal et al. [10] directly compared clinical, histological and outcome differences between IV-S and IV-G by performing a retrospective analysis of 70 patients with lupus nephritis. However, no significant difference was detected in renal outcome in the two groups after average follow-ups of 38 months for Class IV-S and 55 months for Class IV-G. Similarly, Hill et al. [11] examined the clinical and morphological differences between Class IV-S and Class IV-G. In their analysis, survivals from doubling creatinine at 10 yrs did not differ between the two classes. Collectively, these results including ours suggest there are no conclusive differences in renal outcome between Class IV-S and Class IV-G.
Because this is an observational study, before concluding that there is a difference in renal outcomes, careful evaluation of factors affecting renal outcome is important. Among others, treatment is an important variable. In general, patients with IV-G were more intensively treated than those with IV-S. For example, steroid pulse in induction therapy and immunosuppressants during the follow-up period were more frequently used in IV-G compared with IV-S in our cohort. It is unlikely that such intensive therapy itself was negatively correlated with renal survival. In contrast, treatment was similar between IV-G (A) and IV-G (A/C). In this study, although cases of Class III + V or IV + V were categorized to III or IV, respectively, the presence of Class V in either Class III or IV did not affect the renal outcome. Evaluation of other clinical factors potentially affecting renal outcome, such as hypertension and hyperlipidaemia, would be interesting, but remained undetermined in this study. That will be the issue to be examined in a future study.
What is the basis for patients with Class IV-G having chronic lesions to have a worse prognosis? Chronic, sclerotic lesions in glomeruli are thought to be irreversible and refractory to therapy, while active lesions are reversible and can respond to therapy. Concordantly, persistent urinary protein was more often observed in chronic types than in active types, both after induction therapy and at the last follow-up. Recent studies have shown that urinary protein itself is an important risk factor for the deterioration of renal function, which is in part a result of direct interstitial inflammation. Such a feedback loop between sclerotic lesions and urinary protein can be a mechanism towards chronic renal injury. Another question is why sclerotic lesions mostly develop in patients with Class IV-G. One possibility is that patients with chronic lesions might have had renal biopsy only after their disease has already progressed. The initial delay before diagnosis and the start of therapy could result in a critical difference in renal outcome. Another possibility is that chronic lesions may be more prone to develop into sclerotic lesions from exaggerated inflammatory responses, due to a patient's genetic factors or quality of immune complexes.
In summary, our study indicates that the revised classification of LN is clinically valuable in identifying different renal outcomes among patients with diffuse LN. Our results suggest that chronicity may be a stronger determinant of outcome than class. Even in very active cases of Class IV-G, the patients’ renal outcome is favourable when the chronic lesions are absent. In sharp contrast, patients with IV-G (A/C) have a much higher risk. Our results indicate that the degree of chronic lesion has a very strong influence on renal outcome.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
The authors would like to thank Dr James E. Balow, National Institutes of Health, for critical review of the manuscript and Dr Yoshito Tsukada, Fujioka General Hospital, Dr Yukiyasu Watanabe, Watanabe Clinic and Dr Ryoji Wakamatsu, Nishikatagai Clinic, for providing data. The authors would also like to thank Ms Rumiko Koitabashi for technical assistance.
Disclosure statement: The authors have declared no conflicts of interest.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
- McCluskey RT. Lupus nephritis. In: Kidney pathology decennial 1966–1975.—Sommers SC, ed. (1975) East Norwalk, CT: Appleton-Century-Crofts.
- Churg J, Sobin LH. Renal disease: classification and atlas of glomerular disease. (1982) Tokyo: Igaku-Shoin.
- Churg J, Bernstein J, Glassock RJ. Renal disease: classification and atlas of glomerular disease. (1995) 2nd. New York: Igaku-Shoin.
- Weening JJ, D’Agati VD, Schwartz MM, et al. The classification of glomerulonephritis in systemic lupus erythematosus revisited. J Am Soc Nephrol (2004) 15:241–50.
[Abstract/Free Full Text] - Schwartz MM, Bernstein J, Hill GS, et al. Predictive value of renal pathology in diffuse proliferative lupus glomerulonephritis. Lupus Nephritis Collaborative Study Group. Kidney Int (1989) 36:891–6.
- Najafi CC, Korbet SM, Lewis EJ, et al. Significance of histologic patterns of glomerular injury upon long-term prognosis in severe lupus glomerulonephritis. Kidney Int (2001) 59:2156–63.[Web of Science][Medline]
- Austin HA III, Muenz LR, Joyce KM, et al. Prognostic factors in lupus nephritis. Contribution of renal histologic data. Am J Med (1983) 75:382–91.[CrossRef][Web of Science][Medline]
- Austin HA III, Muenz LR, Joyce KM, et al. Diffuse proliferative lupus nephritis: identification of specific pathologic features affecting renal outcome. Kidney Int (1984) 25:689–95.[Web of Science][Medline]
- Yokoyama H, Wada T, Hara A, et al. The outcome and a new ISN/RPS 2003 classification of lupus nephritis in Japanese. Kidney Int (2004) 66:2382–8.[CrossRef][Web of Science][Medline]
- Mittal B, Hurwitz S, Rennke H, et al. New subcategories of class IV lupus nephritis: are there clinical, histologic, and outcome differences? Am J Kidney Dis (2004) 44:1050–9.[CrossRef][Web of Science][Medline]
- Hill GS, Delahousse M, Nochy D, et al. Class IV-S versus class IV-G lupus nephritis: clinical and morphologic differences suggesting different pathogenesis. Kidney Int (2005) 68:2288–97.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||