In our study of this proliferative LN cohort, the mNIH C index from the 2018 revised classification showed a significant association of a 30% decrease of eGFR. The renal outcome was significantly associated with all C index components, including tubulointerstitial lesions not only glomerular lesions. On the other hand, there was no significant association between renal outcome and the activity and chronicity subdivisions based on the 2003 ISN/RPS classification. In terms of associations with renal prognosis, the activity and chronicity assessment based on the 2018 revised ISN/RPS classification was more useful compared to that of the 2003 ISN/RPS classification.
Tao et al had already studied the association of 2018 revised ISN/RPS classification with renal prognosis, and reported that fibrous crescent, tubular atrophy/interstitial fibrosis, and the C index were associated with poor renal prognosis (18). Our study also showed significant associations of the C index and its components including fibrous crescent, tubular atrophy, and interstitial fibrosis by a 30% decrease of eGFR, which supported the previous study. In this study, we further investigated the usefulness of 2018 revised ISN/RPS classification in terms of the association with renal prognosis compared to 2003 ISN/RPS classification by reclassifying the patients of LN using both classification criteria. Such reports had not been made until now, making this the first report of its kind.
In this study, the mNIH C index showed significant association by a 30% decrease of eGFR, although it was not shown in the activity and chronicity subdivisions in 2003. The most important difference in activity and chronicity assessment between the 2003 and 2018 ISN/RPS classifications is the inclusion of the evaluation for tubulointerstitial lesions. This is because tubulointerstitial lesions, not only glomerular lesions, has shown to be significantly associated with renal prognosis of LN in previous studies (8, 13, 19, 20). It is known that whatever the causes of Chronic Kidney disease (CKD) are, CKD gradually exacerbates tubulointerstitial hypoxia by multifactorial mechanisms such as loss of peritubular capillaries, decreased oxygen diffusion by fibrosis, or decreasing of blood flow by glomerulosclerosis. As a result of tubulointerstitial hypoxia, CKD progress, and it forms a malignant cycle. It is known as the ‘final common pathway’ (21). We should recognize tubulointerstitial lesions as the important prognostic factors even in a glomerular disease. Actually, in IgA nephropathy, a common form of glomerulonephritis, interstitial fibrosis and tubular atrophy are shown to be significant lesions that strongly associate with renal prognosis (22). In this context, adoption of evaluation of tubulointerstitial lesions in the classification of LN should improve the predictability of renal outcome.
As explanations for why chronicity assessment of the 2003 ISN/RPS classification showed no significant association with renal prognosis, in addition to the absence of the evaluation for tubulointerstitial lesions mentioned above, there is a problem in designation of subdivisions. Chronicity subdivisions in 2003 classification is not quantitative. Patients with a single C lesion and patients with diffuse C lesions are classified into the same C subdivision. In A/C subdivision, whether the active lesion is dominant, or the chronic lesion is dominant is not expressed. Subdivision of A/C occupies a large proportion (54.3% in our cohort), and they are formed as a heterogeneous group. Hiramatsu et al has reported that class IV-G (A/C) patients with diffuse C lesions are more likely to reach to the outcome of the doubling of sCr than class IV-G (A/C) patients with focal C lesions (23). This study showed the importance of quantitatively of chronic lesions in prediction of renal prognosis.
Meanwhile, the A index was not significantly associated with renal outcome in our study. Austin et al showed patients with high A index had a strong association with ESRD (10), but no significant associations have been shown in recent reports (11, 13, 19). In recent years, the treatment regimen of glucocorticoid with other immunosuppressants including CYC or MMF for the remission induction in recent years would sufficiently improve the some components of A index (24). A index may be less important as a renal prognostic factor in this era.
The mNIH index of the 2018 revised classification is proposed to be applied to all classes, unlike the activity and chronicity subdivision of the 2003 ISN/RPS classification which are limited to class III/IV ± V patients. Although many patients in class I, II, V had a low A index and C index, some patients had a relatively high C index. In multivariate analysis using cox proportional hazard models in all classes, C index was also shown to have significant associations with renal prognosis. Even in class II or pure class V, for cases with high C index, careful observation and follow-up on renal function may be desirable. Additionally, we set 30% GFR decline as the primary endpoint. Although arrival of doubling sCr, which is strongly associated with subsequent risk of ESRD, has been widely accepted as the renal outcome, it is a late event. In fact, in this study, a 30% GFR decline was observed in 36 patients, while the doubling of sCr and ESRD was observed only in 13 and 3 patients, respectively. It is shown that 30% eGFR decline, in replacement of sCr doubling, is effective as an early predictive marker of CKD progression (25).
There are several limitations. First, this study is a retrospective observational study. Second, the nationality and race are limited to mostly Japanese. Third, MMF or CYC as the current standard drug for proliferative LN was only used for 25% of patients in our cohort. This is because MMF or CYC was recently approved for the treatment of LN as health insurance treatment in Japan.
The strength of this study is that Compared to SLE patient cohorts of previous reports, there are more patients in this study. In addition, this is the first report that investigates the usefulness of 2018 revised ISN/RPS classification compared to 2003 ISN/RPS classification in terms of the association with renal prognosis.
While the activity and chronicity subdivision of the 2003 ISN/RPS classification showed no association with renal prognosis, the mNIH C index of the 2018 revised classification showed a significant association by a 30% eGFR decrease. Each category of the C index was independently associated with poor renal prognosis. In terms of associations with renal prognosis, the 2018 activity and chronicity assessment was more useful compared to the 2003 activity and chronicity assessment.