Secular trends in the incidence of end-stage renal disease and its risk factors in Japanese patients with immunoglobulin A nephropathy

ABSTRACT

Background

There are limited data on secular trends in the incidence of end-stage renal disease (ESRD) and frequencies of its risk factors or treatment modalities in patients with immunoglobulin A nephropathy (IgAN).

Methods

This study divided 1255 patients with IgAN into three groups according to the timing of renal biopsy: 1979–89 (n = 232), 1990–99 (n = 574) and 2000–10 (n = 449). The age-adjusted incidence rates, incidence rate ratios and 95% confidence intervals (CIs) for ESRD were calculated by the person-year method and compared using Poisson regression analysis.

Results

A total of 63 patients (5.0%) developed ESRD. The age-adjusted incidence of ESRD decreased significantly over time, i.e. 11.5 per 1000 person-years (95% CI 5.4–24.6) in 1979–89, 6.5 per 1000 person-years (95% CI 1.0–25.2) in 1990–99 and 4.2 per 1000 person-years (95% CI 1.0–17.7) in 2000–10. The proportions of patients with preserved renal function and acute-stage inflammatory histologic changes (i.e. endocapillary hypercellularity and extracapillary proliferation) at the timing of biopsy increased over time, as did the rates of prescriptions of renin–angiotensin system blockers and corticosteroids (all P for trend <0.05). The effect of acute inflammatory histologic lesions on renal prognosis was drastically reduced over time.

Conclusions

These findings suggest that early diagnosis in the acute inflammatory phase and subsequent aggressive treatment may have contributed to the significant downward trend in the incidence of ESRD in patients with IgAN over three decades.

INTRODUCTION

The number of individuals with chronic kidney disease (CKD) is increasing rapidly, and CKD is emerging as a major worldwide public health concern. Following diabetic nephropathy, glomerulonephritis is the second most frequent cause of end-stage renal disease (ESRD) among patients with CKD [1]. Immunoglobulin A nephropathy (IgAN) is the most common type of primary glomerulonephritis occurring worldwide [2, 3]. Renal prognosis in the majority of patients with mild IgAN is generally good, although it is relatively poor in some progressive cases. For example, a study analyzing Japanese patients with IgAN found that the 5-, 10-, 15- and 20-year renal survival rates were 96, 85, 75 and 61%, respectively [4]. Long-term prognosis depends largely on early risk stratification at the time of diagnosis, suggesting the need for strategies that prevent disease progression in patients with IgAN.

The concept of CKD has emphasized the importance of early detection of kidney disease and appropriate referral to nephrologists [5]. Changes in lifestyles and treatment modalities over the past three decades may have affected the clinical prognosis of patients with IgAN. However, presently, there is limited information on secular changes in the incidence of and risk factors for ESRD. Increased knowledge of these trends may improve therapeutic strategies for patients with IgAN. In the present study, we therefore retrospectively examined long-term renal outcomes in a cohort of 1255 patients with IgAN. The aims of this study were to assess trends in the incidence of ESRD, and the frequency of its risk factors and treatment modalities over the last three decades, and to evaluate the impact of these secular changes on renal prognosis.

MATERIALS AND METHODS

Study population

A flow chart of participants is shown in Figure 1. A total of 1543 biopsy-proven IgAN patients were eligible to participate in the study. Participants underwent renal biopsy in seven participating institutions (Kyushu University Hospital, Japanese Red Cross Fukuoka Hospital, Hamanomachi Hospital, Munakata Medical Association Hospital, Japan Seamen’s Relief Association Moji Hospital, Karatsu Red Cross Hospital and Hakujyuji Hospital) between October 1979 and December 2010. Patients with biopsy specimens containing <10 glomeruli (n = 80), extreme outlier body mass index (BMI) value (>40 kg/m²) (n = 1), loss of follow-up (n = 26) or monotherapy with intravenous methylprednisolone (n = 4) were excluded. Additionally, patients lacking data on one or more of the following clinical parameters (n = 177) were excluded: serum creatinine (n = 2), proteinuria (n = 3), blood pressure (n = 2), BMI (n = 48), total cholesterol (n = 19), histology (n = 3), use of renin–angiotensin system blockers (RASBs) (n = 9) and tonsillectomy (n = 91). Finally, 1255 patients with primary IgAN were enrolled. These patients were divided into three cohorts according to the timing of renal biopsy: 1979–89 (n = 232), 1990–99 (n = 574) and 2000–10 (n = 449). These cohorts were followed until 31 December 1989, 31 December 1999 and 31 December 2011, respectively. The study protocol was approved by the Clinical Research Ethics Committees of the Institutional Review Boards of Kyushu University and all participating institutions (approval number 469-06).

Clinical parameters

Demographic and clinical parameters, including patient age, sex, blood pressure, BMI, total cholesterol and serum creatinine concentrations, and 24-h urinary protein excretion or urinary protein–creatinine ratio were collected from kidney biopsy records. Hypertension was defined as blood pressure ≥140/90 mmHg. Hypercholesterolemia was defined as serum total cholesterol ≥220 mg/dL. Obesity was defined as BMI ≥25 kg/m². Total cholesterol concentration was determined enzymatically. Serum creatinine was measured using Jaffe’s method until April 1988, and enzymatically from May 1988 at Kyushu University. At the other participating institutions, serum creatinine was determined by Jaffe’s method until December 2000, and enzymatically from January 2001. Serum creatinine concentrations measured by Jaffe’s method were converted to enzymatically determined concentrations by subtracting 0.207 mg/dL [6]. Estimated glomerular filtration rate (eGFR) was calculated using the Schwartz formula in patients <18 years old, and using the following formula in patients >18 years old: eGFR (mL/min/1.73 m²) = 194 × Cr^−1.094 × age^−0.287 (if female, ×0.739) [7–9]. Impaired kidney function was defined as eGFR ≥60 mL/min/1.73 m². High-grade proteinuria was defined as protein excretion ≥1.0 g/24 h at the time of biopsy [10]. Patients treated with RASBs for at least 6 months at any time during the follow-up period were defined as having been treated with RASBs. We adopted the definition of RASB use according to previous investigations of IgAN patients [11, 12]. The majority of patients received steroid therapy for almost 2 years according to our treatment protocol. However, we did not have accurate information on the start, end or interruption dates of drug administration.

Pathological parameters including Oxford classification criteria

Pathological findings were evaluated using Oxford classifications [13]. Mesangial hypercellularity (M) was scored as M0 or M1, if more or less than 50% of glomeruli, respectively, showed hypercellularity (defined as more than four mesangial cells/mesangial area). Endocapillary hypercellularity (E) and segmental glomerulosclerosis (S) were defined as E0 and S0, respectively, if absent, and as E1 and S1, respectively, if present. Tuft adhesions were classified as S1 lesions. Tubular atrophy/interstitial fibrosis (T) was semiquantitatively classified according to the percentage of cortical area with tubular atrophy or interstitial fibrosis; i.e. as T0 if 0‒25%, T1 if 26‒50% and T2 if >50%.

Extracapillary proliferative lesions

Extracapillary proliferation (Ex) was defined as Ex0 if a cellular or fibrocellular crescent (excluding fibrous) was absent, and as Ex1 if present [13, 14].

Outcome

The primary outcome was ESRD, defined as the initiation of renal replacement therapy, including hemodialysis, peritoneal dialysis and kidney transplantation. Participants were followed-up by reviewing their medical records or by telephone consultation with the clinics and hospitals they attended, or with the patients themselves. Patients were censored at their date of death or at the end of follow-up. The detailed number of deaths and information on follow-up renal function or proteinuria, and doubling of serum creatinine could not be obtained because of the lack of medical record data.

Statistical analysis

The frequency of plausible risk factors was adjusted for age by the direct method, based on the World Health Organization standard population distribution. The age-adjusted mean values of each risk factor were calculated using analysis of covariance. Trends in the frequency and mean values of each risk factor were compared using a logistic or linear regression model. The age-adjusted incidence rates, incidence rate ratios and 95% confidence intervals (CIs) for ESRD were calculated by the person-year method and compared using Poisson regression analysis. Age- and sex-adjusted hazard ratios (HRs) with 95% CI of potential risk factors for the development of ESRD were calculated using a Cox proportional hazards model. Survival was assessed by the Kaplan–Meier method and compared by log-rank tests. Survival time of the cohort in the Cox regression analysis was assessed beginning from the date of biopsy. The differences in the effect of relevant risk factors with time were tested by adding an interaction term between the relevant risk factor (presence versus absence) and ordinal cohort number (1, 2 and 3) to the relevant Cox model. All statistical analyses were performed using SAS software package version 9.2 (SAS Institute, Cary, NC, USA) and STATA version 14 (Stata, College Station, TX, USA). A two-tailed P < 0.05 was considered statistically significant.

RESULTS

Baseline characteristics in the three cohorts

The age-adjusted frequencies of the risk factors for ESRD among the three patient cohorts are shown in Table 1. During the period from 1979 to 2010, the mean age of the overall patient population increased by 4.3 years. No clear trend was observed in the age-adjusted frequency of female patients over time. The mean follow-up period became significantly longer over time. The age-adjusted frequencies of hypertension and mean arterial pressure decreased significantly across the three cohorts. The age-adjusted frequency of obesity increased significantly, whereas the frequency of hypercholesterolemia had no clear trend. The age-adjusted mean urinary protein excretion was unchanged, whereas the mean eGFR increased significantly over time. The age-adjusted frequencies of endocapillary hypercellularity, segmental glomerulosclerosis and extracapillary proliferation increased significantly over the study period (P for trend <0.001). As shown in Figure 2, the age-adjusted proportions of patients treated with RASBs, oral prednisolone and intravenous methylprednisolone increased significantly (P for trend <0.001). Notably, the proportion of patients receiving combined therapy with intravenous methylprednisolone and oral corticosteroids increased steeply from 7.4% in the second cohort to 36.7% in the third cohort.

Trends in the incidence of ESRD

Among the 1255 patients, 63 (5.0%) developed ESRD. Kaplan–Meier curves for the development of ESRD are shown in Figure 3. The 5-year renal survival rates among the three groups increased significantly over time. They were 86.0% in 1979‒89, 94.3% in 1990‒99 and 95.1% in 2000‒10 (log-rank = 7.94, P = 0.02). The age-adjusted incidence rate (per 1000 person-years) of ESRD decreased dramatically over time, and was 63% lower in 2000‒10 compared with 1979‒89 (P = 0.006; Table 2). Throughout the three time periods, the age-adjusted incidence rate of ESRD was drastically reduced in patients with acute inflammatory histologic lesions (P < 0.001), whereas it remained unchanged in those with chronic histologic lesions (Figure 4).

Trends in the effects of risk factors on ESRD

The effects of potential risk factors on the incidence of ESRD were subsequently analyzed (Table 3). Hypertension was not significant in the first (1979‒89) or second (1990‒99) patient cohorts, whereas it was a strong risk factor for ESRD in the third cohort (2000‒10) (age- and sex-adjusted HR 3.27, 95% CI 1.12‒9.56). Obesity did not contribute to renal outcomes, despite its increased trend over time. Hypercholesterolemia was a powerful risk factor in all cohorts. High grade proteinuria and CKD had substantially greater initial effects on the incidence of ESRD than other risk factors over time. Nevertheless, there was no significant heterogeneity in the effects of these factors on the development of ESRD among the three cohorts. Mesangial hypercellularity score was not significant in the first patient cohort. However, in the second and third cohorts, it was a significant risk factor for ESRD, with its effect increasing gradually over time. Endocapillary hypercellularity was a significant risk factor in the first cohort (HR 3.93, 95% CI 1.49‒10.5) but not in the second (HR 1.13, 95% CI 0.52‒2.47) or third cohort (HR 0.86, 95% CI 0.34‒2.18). The effect of endocapillary hypercellularity decreased significantly over the three cohorts (P for difference over time = 0.04). Segmental sclerosis was a significant risk factor for ESRD in the second (HR 8.27, 95% CI 1.12‒60.9), but not in the third cohort, with no reliable evidence of it affecting the occurrence of ESRD in the first cohort, likely because of the small number of events. The effect of mild tubular/interstitial lesions (T1) on the development of ESRD was not significant in any of the cohorts, whereas severe tubular/interstitial lesions (T2) were a significant risk factor for ESRD across all three cohorts, with the HR increasing steeply from 5.52 (95% CI 2.53‒12.1) in the second cohort to 18.4 (95% CI 6.66‒50.8) in the third cohort (P for difference with time = 0.03). Extracapillary proliferation was associated with increased risk for ESRD in the first (HR 8.03, 95% CI 2.30‒28.1) and second (HR 4.49, 95% CI 1.55‒13.0) cohorts, but its effect was reduced drastically in the third cohort (HR 1.43, 95% CI 0.54‒3.83) (Table 3). Additionally, multivariate analysis of the entire cohort showed that hypercholesterolemia, advanced renal insufficiency, severe proteinuria and chronic histological lesions (mesangial hypercellularity score, segmental glomerulosclerosis and tubular atrophy/interstitial fibrosis) at the time of biopsy remained independent prognostic factors, whereas the influence of acute inflammatory findings (endocapillary hypercellularity, extracapillary proliferation) was attenuated (Supplementary data, Table S3).

DISCUSSION

The present study showed that the incidence of ESRD decreased significantly over the past three decades in patients with IgAN. Our results clearly demonstrated the downward trend in the frequency of patients with impaired kidney function at the time of biopsy, and the upward trends in those with acute inflammatory findings (i.e. endocapillary hypercellularity and extracapillary proliferation) and those receiving treatment with RASBs and aggressive immunosuppressive therapy. Notably, the effect of acute inflammatory histologic lesions on the development of ESRD was drastically reduced over time. These findings suggest that patients with IgAN are gradually being identified during the acute inflammatory phase of the disease, and subsequent aggressive treatment with immunosuppressive agents likely improves their renal survival.

Several previous reports have assessed secular changes of renal prognosis in patients with IgAN. Longitudinal data from 304 patients with IgAN indicated that patients who were diagnosed between 1996 and 2006 had a favorable renal outcome compared with those diagnosed between 1981 and 1995; the 10-year renal survival rates were 95.7 and 75.2%, respectively [15]. Another retrospective cohort analysis showed that the cumulative renal survival rate was significantly higher in patients diagnosed within the last 10 years compared with those diagnosed during the previous 10 years (86.6% versus 79.1%) [16]. These results were consistent with our findings, although none of the identified studies addressed trends of risk factors and treatment modalities over the analyzed periods of time. To our knowledge, the present study is the first investigation to assess the secular changes of risk factors and treatment modalities and their impact on renal prognosis in patients with IgAN.

Corticosteroid therapy has been reported to be beneficial for inhibiting the progression to ESRD in patients with IgAN [17–21], although its definitive clinical efficacy remains inconclusive. The present study showed that increased use of corticosteroid therapy likely reduced the incidence of ESRD, and attenuated the prognostic relevance of acute inflammatory lesions in the development of ESRD throughout the analyzed time periods. This inverse association was particularly apparent in the combination with intravenous methylprednisolone and oral corticosteroids. Recently, several reports have shown that the significance of endocapillary hypercellularity and extracapillary proliferation on renal outcome is influenced by immunosuppressive therapy. A study of a large pooled cohort by Haas et al. [22] showed that the effect of crescents is only observed in patients not undergoing immunosuppressive therapy, and not in immunosuppressed patients. Furthermore, Chakera et al. [23] reported that endocapillary hypercellularity was a significant independent predictor of time to ESRD in a cohort of patients with IgAN receiving no immunosuppressive therapy. These findings are consistent with our results, and support the hypothesis that the prognostic relevance of acute inflammatory lesions is affected by immunosuppression-related bias. Another important finding from our study was that there was a significant downward trend in the incidence rate of ESRD in patients with acute inflammatory lesions, but not in those with chronic lesions. A previous rebiopsy study from China also demonstrated that immunosuppressive therapy reduced acute inflammation, but not the chronic lesions (i.e. mesangial proliferation and tubular atrophy/interstitial fibrosis) [24]. These results supported the evidence that the renal protective effect of corticosteroids is mediated primarily through suppression of the acute inflammatory process in glomerulonephritis.

The proportion of patients who were treated with RASBs increased significantly over time. RASBs are considered to play a central role in the treatment of IgAN, given that they are administered continuously over a considerable number of years [17, 18]. Surprisingly, our survey showed that their rate of administration was <50% in the observed cohorts, even in the most recent one (2000‒10), suggesting that in addition to the use of RASBs, comprehensive management including diagnostic modalities and immunosuppressive intervention can result in improved renal prognosis in patients with IgAN.

Hypertension has been reported to be an important risk factor affecting the prognosis of patients with IgAN [2, 25]. The present study showed that hypertension was a significant influencing factor for ESRD in the third cohort (2000‒10). However, we could not take into consideration the influence of patients taking antihypertensive medication at the time of biopsy because of the lack of relevant information in medical records. The percentage of patients treated for hypertension likely increased over time, resulting that the increasing role of hypertension observed in the current study might be weakened over the period. Future studies are required to examine the significance of blood pressure in patients with IgAN treated for hypertension.

Obesity did not contribute to renal outcomes despite its increased trend over time. However, the frequency of obesity in the present study was lower than in previous reports [26, 27]. The westernization of lifestyle in Japan is expected to increase the prevalence and effects of obesity. Therefore, it is important to continue to monitor the association between obesity and disease prognosis in Japanese patients with IgAN. Hypercholesterolemia remained a significant risk factor for ESRD throughout the study period. Sensitivity analysis showed that this trend was unchanged after excluding patients with nephrotic-range proteinuria (Supplementary data, Table S1), suggesting that lipotoxicity itself may have an unfavorable effect on renal prognosis [28, 29]. These findings may highlight the importance of managing lipid metabolism in patients with IgAN.

This study had several limitations. First, the relatively small number of events weakened the statistical power for assessing associations between risk factors and ESRD. A multivariable-adjusted model incorporating many covariates was unable to determine reliable associations between risk factors and ESRD. Second, a single measurement of risk factors may have been a potential source of misclassification of the included patients. Such misclassification would have weakened the associations observed in this study, which may have biased the results toward the null hypothesis. Third, the difference in the follow-up period of each cohort may have been a potential source of bias in directly comparing the incidence of ESRD among the three cohorts. However, although the median follow-up period increased significantly over time, the incidence of ESRD decreased secularly. This suggests that the heterogeneity in the follow-up period of each cohort had minimal effect on our main conclusions. Fourth, this was a retrospective study conducted over different periods of time, and therefore the association between immunosuppressive treatment and outcome is presumed to be modified by other factors that cannot be readily measured (e.g. differences in screening programs for microscopic hematuria or threshold of renal biopsy over the study periods). Indeed, the proportion of cases with severe grades of microscopic hematuria at the time of biopsy increased over time (Figure 5). Moreover, stratified analysis combined with urinary protein excretion indicated that the proportion of patients with both mild proteinuria (<1 g/day) and high-grade hematuria (more than 2+ by dipstick urinalysis) had increased to approximately 50% in the most recent cohort (Figure 6). These findings suggest that the concept of CKD, which emphasizes the importance of early detection of kidney disease and appropriate referral to nephrologists, has recently been introduced to family physicians and specialists in Japan. Finally, it is possible that advances in medical technology have improved the safety of renal biopsy procedures, resulting in overdiagnosis of nonprogressive mild cases throughout the time periods. However, our sensitivity analysis revealed that the downward trend in the incidence of ESRD in patients with preserved kidney function (eGFR ≥60 mL/min/1.73 m²) did not differ substantially from that in the overall population (Supplementary data, Table S2). Therefore, we believe this limitation affects our conclusion to a lesser degree.

In conclusion, this study demonstrated that the incidence of ESRD decreased significantly over the past three decades in patients with IgAN. These results indicated that early diagnosis by renal biopsy during an acute inflammatory phase, and aggressive treatment likely contributed to the significant downward trend in the incidence of ESRD.

AUTHORS’ CONTRIBUTIONS

S.T. contributed to the study design, acquisition of data, statistical analysis, interpretation of data and drafting of the manuscript. T.N. contributed to the study design, statistical analysis, interpretation of data and drafting of the manuscript. K.T. and M.T. contributed to the funding, acquisition of data and critical revision of the manuscript. R.K. contributed to the acquisition of data, pathological evaluation of kidney biopsies and critical revision of the manuscript. K.M. and A.T. contributed to the pathological evaluation of kidney biopsies and critical revision of the manuscript. H.H., H.O. and T.K. contributed to the critical revision of the manuscript and study supervision. All authors provided critical reviews of the draft and approved the final version.

ACKNOWLEDGEMENTS

The authors would like to thank the investigators at the participating institutions: Tetsuhiko Yoshida, M.D., Hirofumi Ikeda, M.D., Ph.D., Takashi Inenaga, M.D., Akinori Nagashima, M.D., Ph.D., Tadashi Hirano, M.D. and Hideko Noguchi.

FUNDING

This work was supported by the Grants-in-Aid for Scientific Research [No. 15H06800] from the Ministry of Education, Culture, Sports, Science and Technology of Japan.

SUPPLEMENTARY DATA

Supplementary data are available online at http://ndt.oxfordjournals.org.

CONFLICT OF INTEREST STATEMENT

None declared.

REFERENCES