H.S. Lee’s grading system developed in 1987 by incorporating mesangial proliferation, segmental lesion, crescent, and interstitial fibrosis showed good correlation with proteinuria, hypertension, and impaired renal function at the time of biopsy. Interestingly, episodes of macroscopic hematuria which is a favorable factor for renal progression in Korea were less frequent in patients with higher grade of this system [28].

Some modification of H.S. Lee’s grading system with particular emphasis on crescent, segmental sclerosis, and global sclerosis which implicate, respectively, active necrotizing glomerular inflammation, podocyte depletion, and resultant irreversible glomerular damage correlated well with patients’ age, CCr, 24-h urinary protein, and the prevalence of hypertension in dose-dependent manner. Moreover, this grading system predicted renal progression independent of clinical risk findings such as initial renal function and proteinuria, whereas the Hass classification did not [32].

Class IV/V lesions of WHO classification were a prognostic factor of renal progression independent of renal insufficiency (serum creatinine ≥1.4 mg/dL) and heavy proteinuria [33].

These studies suggest that specific histological features such as mesangial proliferation, crescent, segmental sclerosis, and advanced tubulointerstitial lesions rather than the whole system of histological classification were suggested to be significant prognostic factors independent of clinical features even before the introduction of Oxford classification.

After the Oxford classification was introduced as a new histological classification system for IgAN in 2009, several studies were performed to validate the usefulness of this classification in Korean patients.

All studies identified T1 and T2 lesions, namely, advanced tubulointerstitial lesions, as a predictor of renal progression independent of clinical findings in multivariable analysis [34–36]. T1 and T2 were also predictors of renal progression in posttransplant IgAN [37]. These findings confirmed the well-established prognostic importance of tubulointerstitial lesion in essentially all glomerular diseases including IgAN.

E1 was associated with more frequent prescription of steroid in studies where steroid was infrequently used, i.e., 18 % and 11 % of subjects of study, respectively [35, 36]. But in one study in which 38 % of patients had received steroid, S1 but not E1 was associated with the more frequent use of steroid [34]. RAS blockade treatment was more frequent in patients with M1 [34, 36]. Thus, at least in current situation, where the prescription of immunosuppressive drugs is largely dependent on the decision of individual physician without agreed guideline for immunosuppressive treatment, the reported relationship of specific Oxford lesions with steroid treatment be viewed with the consideration for criteria by which steroid was used in a specific study.

The frequency of M1, S1, E1, T1, and T2 increased along with the increasing grade of WHO classification [34] and also showed correlation with Hass classification in patients with posttransplant IgAN [37]. All MEST variables, especially S and E, correlated with activity index in semiquantitative classification, but only S and T showed relationship with chronic index [34]. Hass and Oxford classifications were comparable in providing additive predictive value for renal progression to known clinical parameters such as proteinuria and eGFR [36], but Oxford classification was superior to Haas classification in posttransplant IgAN [37]. Thus, pre-existing pathological grading systems of IgAN such as WHO and Hass classification could provide useful prognostic information as much as Oxford classification, but Oxford classification is superior to old systems in the convenience in clinical application and consistency in interpretation of pathologic findings by different pathologists.

Similar to conclusion from Oxford group, a retrospective analysis of 430 patients in which 18.8 % of patients had crescent showed that crescent was not a prognostic factor for renal progression independent of clinical findings and T lesion although patients with crescent had higher UPCR, lower eGFR, lower serum albumin level, and more prescription of RAS blockade and glucocorticoid treatment during follow-up. Because the patients in this study had a relatively good basal eGFR of 80.5 ± 24.1 ml/min/1.73 m², this study could not give the answer about the prognostic significance of crescent in patients having worse eGFR <30 ml/min/1.73 m² as Oxford classification could not [38]. But in posttransplant IgAN, 4-year graft survival after biopsy was 30.0 % in patients having crescents compared with 70.8 % patients without crescents despite the enhanced immunosuppression in the former group. Moreover, IgAN was the cause of graft failure in 66.7 % of patients with crescents, whereas IgAN was the cause of graft failure in only 13.6 % of patients without crescents [39]. Thus, the prognostic significance of crescent in IgAN remained to be defined especially in patients with low basal eGFR and/or patients on the immunosuppressive therapy at the time of renal biopsy.

Although the excellent prognosis is taken for granted for patients presenting with normotension, normal renal function, and normal or low-grade proteinuria, this does not apply to all such patients.