Clinical Epidemiology

Quantifying Duration of Proteinuria Remission and Association with Clinical Outcome in IgA Nephropathy

Mark Canney, Sean J. Barbour, Yuyan Zheng, Rosanna Coppo, Hong Zhang, Zhi-Hong Liu, Keiichi Matsuzaki, Yusuke Suzuki, Ritsuko Katafuchi, Heather N. Reich, Daniel Cattran and for the International IgA Nephropathy Network
JASN February 2021, 32 (2) 436-447; DOI: https://doi.org/10.1681/ASN.2020030349

Significance Statement

Recent studies support the use of reduction in proteinuria as a surrogate endpoint in IgA nephropathy for a treatment’s effect on progression to ESKD. However, the duration of proteinuria reduction required to mitigate this risk is unknown. This study evaluated adult patients with biopsy-proven IgA nephropathy at high risk of disease progression. Among 1864 patients who achieved a proteinuria remission (≥25% reduction in proteinuria and a reduction in proteinuria to an absolute value <1 g/d), each 3 months in remission was associated with an additional 9% reduction in the risk of ESKD, or a 50% decline in kidney function over a median follow-up of 3.9 years. These results demonstrate the need to consider both the duration and magnitude of proteinuria reduction when quantifying the risk of disease progression in IgA nephropathy.

Abstract

Background On the basis of findings of observational studies and a meta-analysis, proteinuria reduction has been proposed as a surrogate outcome in IgA nephropathy. How long a reduction in proteinuria needs to be maintained to mitigate the long-term risk of disease progression is unknown.

Methods In this retrospective multiethnic cohort of adult patients with IgA nephropathy, we defined proteinuria remission as a ≥25% reduction in proteinuria from the peak value after biopsy, and an absolute reduction in proteinuria to <1 g/d. The exposure of interest was the total duration of first remission, treated as a time-varying covariate using longitudinal proteinuria measurements. We used time-dependent Cox proportional hazards regression models to quantify the association between the duration of remission and the primary outcome (ESKD or a 50% reduction in eGFR).

Results During a median follow-up of 3.9 years, 274 of 1864 patients (14.7%) experienced the primary outcome. The relationship between duration of proteinuria remission and outcome was nonlinear. Each 3 months in sustained remission up to approximately 4 years was associated with an additional 9% reduction in the risk of disease progression (hazard ratio [HR], 0.91; 95% confidence interval [95% CI], 0.89 to 0.93). Thereafter, each additional 3 months in remission was associated with a smaller, nonsignificant risk reduction (HR, 0.99; 95% CI, 0.96 to 1.03). These findings were robust to multivariable adjustment and consistent across clinical and histologic subgroups.

Conclusions Our findings support the use of proteinuria as a surrogate outcome in IgA nephropathy, but additionally demonstrate the value of quantifying the duration of proteinuria remission when estimating the risk of hard clinical endpoints.

Although IgA nephropathy is the most common form of glomerular disease worldwide,1 treatment options have historically been limited. For the majority of patients, corticosteroids are the only immunosuppression therapy with sufficient evidence to justify a guideline recommendation.2 There are significant challenges unique to IgA nephropathy that have substantially hampered the design of clinical trials capable of evaluating new therapies,3 the most important of which is the considerable variation in the time lag between diagnosis and the development of a hard clinical endpoint such as ESKD, or a 50% reduction in kidney function.4⇓–6 As a result, clinical trials require long follow-up over many years to observe a treatment effect on the hard outcomes traditionally required by regulatory agencies for drug approval. This increases the cost and complexity of clinical trials in IgA nephropathy, thereby severely limiting the evaluation of new therapies. Consequently, there is recent interest in identifying earlier surrogate measures of kidney outcome to encourage the conduct and improve the feasibility of future clinical trials in IgA nephropathy.

Proteinuria has emerged as the most likely candidate surrogate endpoint in IgA nephropathy, given extensive literature demonstrating it is a strong and independent predictor of long-term kidney outcomes.4,5,7 In 2018, the Kidney Health Initiative (KHI) in collaboration with the Food and Drug Administration (FDA) proposed that, in IgA nephropathy, a short-term reduction in proteinuria was a reasonably likely surrogate endpoint for a treatment’s effect on progression to ESKD.8 This conclusion was on the basis of two types of evidence: first, observational studies have shown that long-term sustained reductions in proteinuria are associated with lower risk of disease progression5,9; second, a meta-analysis of 13 randomized clinical trials demonstrated that a short-term treatment effect on proteinuria was highly correlated with a treatment effect on long-term kidney endpoints.8,10 The decision to accept proteinuria as a surrogate outcome has resulted in a substantial increase in the evaluation of new therapeutic options that target multiple novel disease pathways in IgA nephropathy.11 Indeed, recent clinical trials have already adopted a relative reduction in proteinuria as their primary outcome, for example the Targeted-release budesonide versus placebo in patients with IgA nephropathy (NEFIGAN) study.12

Therapies in IgA nephropathy that demonstrate a short-term reduction in proteinuria in clinical trials can apply for accelerated FDA drug approval8; however, the clinical benefit of products approved under this program must be verified in a postmarketing confirmatory trial using a hard clinical outcome. A major reason for this requirement is that it is unknown how long a reduction in proteinuria needs to be maintained to mitigate the risk of disease progression, and thus constitute a valid surrogate outcome measure.8 To address this question, we evaluated the association between duration of proteinuria remission and the subsequent risk of disease progression in a large and multiethnic international cohort of adult patients with biopsy-proven IgA nephropathy.

Methods

Design, Setting, and Participants

The study population included patients with biopsy-proven idiopathic IgA nephropathy and available MEST-C scores from the following international cohorts as previously described6,13⇓⇓⇓–17: the European Validation Study of the Oxford Classification of IgA Nephropathy (VALIGA), Oxford Derivation (Europe, Asia, North and South America), Oxford Validation (North America), Tokyo (Japan), and Beijing and Nanjing (China). We included 3335 adult patients (age ≥18 years) who did not have ESKD at the time of biopsy, and who had available data for eGFR and proteinuria during follow-up. The sample was then further restricted to those patients who met the definition for disease remission on the basis of proteinuria during longitudinal follow-up after kidney biopsy according to three criteria: (1) peak proteinuria value before remission ≥1 g/d; (2) a ≥25% reduction in proteinuria from the peak value after biopsy; and (3) an absolute reduction in proteinuria to <1 g/d. This definition could therefore include individuals who achieved a partial remission, maintaining residual proteinuria below 1 g/d, and those with a complete remission of proteinuria into the normal range. A 25% relative reduction in proteinuria was chosen on the basis of a meta-analysis of therapeutic clinical trials in IgA nephropathy, which showed that a 24% reduction in proteinuria between the treatment and control arms in these studies was associated with a 63% relative risk reduction in the composite endpoint of ESKD, doubling of serum creatinine, or death.8,10 This project was approved by the University of British Columbia research ethics board, with waived patient consent.

Variable Definitions

Proteinuria was taken from 24-hour urine collections. When these were not available, daily excretion of urinary protein was estimated from spot urine protein-to-creatinine ratios.18 Baseline mean arterial BP and creatinine were taken as the closest values within 6 months of the date of remission. GFR was estimated from creatinine using the Chronic Kidney Disease Epidemiology Collaboration formula.19 Use of medications that block the renin-angiotensin aldosterone system (RAAS), including angiotensin-converting enzyme inhibitors and angiotensin receptor blockers, was determined at the remission date. Use of immunosuppression was determined at or before the remission date. Race or ethnicity was self-reported by the patient as White, Chinese, Japanese, or other. Age was determined at the time of remission. The Oxford MEST score was on the basis of established criteria14,20,21 as follows: M0/M1 as a mesangial score ≤0.5 or >0.5 (≤50% or >50% of glomeruli with ≥4 mesangial cells per mesangial area); E1/E0 as the presence or absence of endocapillary hypercellularity; S1/S0 as the presence or absence of segmental sclerosis or tuft adhesions; and T0/T1/T2 as the degree of tubular atrophy or interstitial fibrosis (<25%, 25%–50%, and >50%, respectively). Crescents were considered as the presence or absence of cellular or fibrocellular crescents (C0 versus C1/C2). The percentage of glomeruli with crescents was not available.

Outcome

The primary outcome was a composite of the first occurrence of either ESKD (eGFR <15 ml/min per 1.73 m2, dialysis, or transplantation) or a reduction in eGFR to below 50% of the value at the time of remission, which persisted across all subsequent eGFR measurements.22,23

Exposure

The exposure of interest was the total duration of first remission (in months) and was treated as a time-varying covariate. All patients were in remission at the beginning of the observation period, and therefore their duration of remission was zero at the start of their time at risk. According to sequential proteinuria measurements during longitudinal follow-up, each patient’s remission status was iteratively re-evaluated. If they continued to satisfy the criteria for remission, their duration of remission represented the cumulative time in months from their remission date to the date of proteinuria measurement. At the first assessment date at which the proteinuria value no longer satisfied the definition of remission, the patient was considered to have experienced a relapse and their duration of first remission stopped increasing. Patients continued to be followed beyond that time for the risk of the primary outcome, and therefore were followed after the end of their first remission until they experienced an event or were censored. For patients who were not observed to have experienced a relapse, their duration of remission continued to increase until they experienced the primary outcome or they were censored. An example of how the exposure variable was dynamically evaluated is provided in Supplemental Appendix 1.

Statistical Analyses

Time from remission to the primary outcome, censored at the end of follow-up, was modeled using time-dependent Cox proportional hazards regression models to account for time-varying duration of first remission as the main exposure variable. To accomplish the time-dependent analysis, the data were formatted using a counting process in SAS. In this type of analysis, the duration of first remission is dynamically updated at each event in the cohort to estimate the hazard function. Further details are provided in Supplemental Appendix 1. To explore nonlinearity, the functional form of the unadjusted association between duration of remission, as a continuous variable, and the primary outcome was evaluated using restricted cubic splines with four knots at the fifth, 35th, 65th, and 95th percentiles. Because of a significant nonlinear relationship, and to simplify the presentation of the results, duration of remission was subsequently modeled using piece-wise regression, with two segments of linear relationships (on the natural log scale) before and after the observed change in the risk relationship. On the basis of the best model fit (lowest Akaike Information Criterion), the optimal cut-point between these two piece-wise linear relationships occurred at 51 months. This resulted in two estimates for the effects of each unit of time in remission: (1) during shorter remission duration up to a maximum of approximately 4 years, and (2) during longer remission duration beyond approximately 4 years.

The association between duration of remission and the primary outcome was then assessed using a multivariable regression model with covariates chosen on the basis of prior literature that included age, sex, race or ethnicity, mean arterial BP, eGFR, and proteinuria at remission; peak proteinuria before remission; time from biopsy to remission; MEST-C score; and use of RAAS blockade at remission or immunosuppression before remission. The proportional hazards assumption was evaluated using log-log plots and Martingale residuals. For each continuous covariate, the functional form of its association with the primary outcome (and potential need for variable transformation) was visually assessed using restricted cubic splines. Consistent with prior observations,17 the hazard ratio for Chinese (versus White) race or ethnicity was not proportional over time. Consequently, the multivariable model included an interaction between time and Chinese race or ethnicity.

We performed three sensitivity analyses. First, we repeated our primary analysis in the subgroup of patients with a minimum of three proteinuria measurements after the remission date. Second, we repeated our primary analysis using different definitions of the end of remission (i.e., relapse). Finally, to confirm the primary analysis results for short periods of remission duration, we performed a landmark analysis at 3 and 6 months after the onset of first remission (see Supplemental Appendix 2 for further details). Descriptive variables are presented as mean (SD) or median (interquartile range [IQR]) as appropriate. Categorical variables are presented as frequency (percentage). Two tailed P values <0.05 were considered statistically significant. Missing predictor variables were imputed using multiple imputations by chained equations as done previously.17,24⇓–26 Analyses were performed using SAS Version 9.4 (SAS Institute, Cary, NC).

Results

Baseline Characteristics

A total of 1864 patients were included in the remission cohort (Figure 1). Patients had a median age of 36.8 years (IQR, 29.4–46.6) and 56% were male (Table 1). The majority of patients had well-preserved kidney function with median eGFR of 78 ml/min per 1.73 m2 (IQR, 55–104) and had a peak proteinuria before remission of 1.8 g/d (IQR, 1.3–2.8). Most patients were receiving RAAS blockade and 42% of patients had received immunosuppression at or before remission. The median number of proteinuria measurements after entering remission was 7 (IQR, 3–11) (Supplemental Figure 1), and the median time between sequential proteinuria measurements was 3.5 months (IQR, 2.1–6.7). The median duration of first remission was 22 months (IQR, 8–48) and the distribution of time in remission is shown in Figure 2. A description of missing data for covariates is provided in Supplemental Table 1.

Figure 1.
Figure 1.

Flowchart of case ascertainment.

View this table:
Table 1.

Characteristics of 1864 patients who achieved a proteinuria remission

Figure 2.
Figure 2.

Histogram of the distribution of duration of first remission.

Association Between Remission Duration and Outcome

A total of 274 patients (14.7%) within the remission cohort experienced ESKD or a 50% decline in eGFR during a median follow-up time of 3.9 years (IQR, 2.1–6.3), with a 5-year risk of 14.9% (95% confidence interval (95% CI), 13% to 17.1%) (Supplemental Figure 2). Among patients who experienced the primary outcome, 224 (82%) experienced the event after the end of their first remission, with a median (IQR) time between the end of first remission and the primary outcome of 2.6 years (1.1–4.4). The functional form of the relationship between duration of remission, unadjusted for other covariates, and the primary outcome is illustrated as a restricted cubic spline in Figure 3. This analysis suggested a nonlinear relationship, with an initial steep reduction in risk associated with progressively longer duration of remission up to approximately 4 years, followed by an attenuation in benefit for additional remission duration thereafter. We therefore generated two continuous linear exposure variables for inclusion in a regression model to account for this nonlinearity, one for the steep reduction in risk that occurs during periods of remission lasting up to approximately 4 years, and a second for the more attenuated risk reduction that occurs during extended duration of remission beyond 4 years.

Figure 3.
Figure 3.

Restricted cubic spline illustrating the functional form of the unadjusted association between duration of first remission and the primary outcome (ESKD or a 50% decline in eGFR). The y-axis represents the hazard ratio of the primary outcome on the logarithmic scale. The graph shows an initial steep reduction in risk associated with longer duration of remission up to approximately 4 years, followed by an attenuation in further risk reduction.

The Cox proportional hazards models estimating the unadjusted and multivariable-adjusted risk of the primary outcome associated with duration of remission are shown in Table 2. Output from the full multivariable model is provided in Supplemental Table 2. Each 3 months during shorter remission, up to a maximum of approximately 4 years, was associated with a 9% relative-risk reduction in the primary outcome (hazard ratio, 0.91; 95% CI, 0.89 to 0.93). Beyond 4 years, the effect of additional time in remission was substantially attenuated in that each additional 3 months in remission was associated with a nonsignificant 1% further reduction in relative risk (hazard ratio, 0.99; 95% CI, 0.96 to 1.03). Adjusting for demographic and clinical characteristics did not appreciably change the results. On the basis of these model estimates, Figure 4 shows the hazard ratio associated with cumulative duration of remission. For example, 4 years in remission was associated with a 79% (95% CI, 68% to 86%) reduction in the risk of the primary outcome, whereas extending this remission duration to 6 years resulted in only a slight further reduction in risk to 82% (95% CI, 73% to 87%).

View this table:
Table 2.

Association between duration of remission and the primary outcome (composite of ESKD or a 50% decline in eGFR)

Figure 4.
Figure 4.

Smoothed plot of the hazard ratio (gray line) and associated 95% confidence interval (shaded area) for the risk of the primary outcome associated with the cumulative duration of remission.

To explore if patient characteristics might explain the attenuated benefit of additional time in remission beyond approximately 4 years, the baseline characteristics of patients with shorter compared with longer remission duration are shown in Table 3. Clinical factors including use of RAAS blockade, immunosuppression exposure, BP, and eGFR were similar between the two groups. Although there were several differences between the groups, the most striking were that patients with longer remission duration beyond approximately 4 years had a lower nadir proteinuria (median 0.13 versus 0.36 g/d), and were much more likely to achieve proteinuria values of <0.3 g/d compared with those shorter remission duration (84% versus 43%). They were also more likely to be White, and less likely to have mesangial hypercellularity, crescents, or interstitial fibrosis/tubular atrophy at the time of biopsy.

View this table:
Table 3.

Characteristics of patients with duration of remission up to approximately 4 yr compared with those with extended duration of remission beyond 4 yr

Subgroup Analysis

The relationship between each additional 3 months of time in first remission and the primary outcome was examined within subgroups on the basis of individual components of the MEST-C score, RAAS blockade use, immunosuppression exposure, peak proteinuria before remission, proteinuria at remission, and time to remission. This subgroup analysis was done using only interaction terms with the effect of additional time in remission during shorter remissions up to approximately 4 years, and not during longer remissions thereafter, because the latter result was NS in the primary analysis. The results are provided in Supplemental Table 3 and summarized in Figure 5. For each subgroup, the upper bound of the 95% CI did not cross 1 and the interaction term P values were NS (except for endocapillary hypercellularity), indicating a consistent association between remission duration and outcome across all subgroups. The interaction between endocapillary cellularity and remission duration was statistically significant (P=0.01), although the difference in risk estimates between E1 and E0 was of marginal clinical significance (hazard ratio, 0.95; 95% CI, 0.91 to 0.98 versus 0.91; 95% CI, 0.88 to 0.93, see Supplemental Table 3).

Figure 5.
Figure 5.

Association between each additional 3 months in remission and the primary outcome in predefined subgroups. Continuous variables such as time to remission, peak proteinuria before remission, and proteinuria at remission were categorized on the basis of their median values. The results are presented for the effect of additional time in remission during shorter remissions of up to approximately 4 years. The multivariable model is adjusted for age, sex, race/ethnicity, MEST score, crescents, mean arterial pressure, eGFR (log-transformed), peak proteinuria before remission, proteinuria at remission, time from biopsy to remission, RAAS blockade at remission, and immunosuppression at or before remission.

Sensitivity Analysis

We performed three sensitivity analyses. First, we repeated the restricted cubic spline analysis in the 1471 patients who had a minimum of three proteinuria measurements beyond their remission date. The relationship between duration of remission and the primary outcome was consistent in this subpopulation (Supplemental Figure 3). Second, we repeated the analysis using two more conservative definitions of relapse: a 50% increase in proteinuria to an absolute value of ≥1 g/d (Supplemental Figure 4), or a 100% increase in proteinuria to an absolute value of ≥1 g/d (Supplemental Figure 5). In both cases, the results were similar to the primary analysis. Finally, we used a landmark analysis to explore the long-term risk of the primary outcome associated with experiencing a relapse before 3 months or before 6 months after the onset of first remission. Patients who relapsed before the 3-month and 6-month landmark had significantly higher risk of the primary outcome compared with patients who were still in remission at the specified landmark (Supplemental Figure 6). This was consistent with the primary analysis results demonstrating a reduced risk associated with each 3 months of additional remission duration.

Discussion

In this large international cohort of adult patients with biopsy-proven IgA nephropathy, we observed a nonlinear dose-response relationship between the duration of first proteinuria remission and the risk of disease progression. Each 3 months of additional time in remission was associated with a further 9% reduction in the risk of ESKD or a 50% decline in eGFR. This finding was robust to adjustment for recognized demographic, clinical, and histologic risk factors. Beyond approximately 4 years of remission duration, there was a substantial attenuation in the benefit accrued from additional time spent in remission. Rather than there being a minimum duration of proteinuria remission in IgA nephropathy, our results demonstrate that even short durations in remission are associated with significant reductions in the risk of disease progression. This suggests that when proteinuria is used as a surrogate outcome in IgA nephropathy, there is a need to consider the duration of remission in addition to the magnitude of proteinuria reduction when evaluating the anticipated effect on long-term clinical endpoints.

Quantifying the association between duration of proteinuria remission and the long-term risk of an established GFR-based kidney outcome has immediate significance to the current clinical trial development programs in IgA nephropathy that utilize proteinuria as a surrogate outcome. The KHI/FDA have proposed that a clinical trial demonstrating a short-term reduction in proteinuria can be used for accelerated drug approval, provided a longer-term postmarketing trial is conducted to confirm a therapeutic benefit on hard clinical outcomes. As such, the KHI/FDA workgroup emphasized the anticipated treatment effect on proteinuria must be of sufficient duration to ensure the ability to demonstrate a reduction in the risk of a clinical endpoint within the limited timeframe available for a postmarketing trial. In this context, the KHI/FDA workgroup highlighted that the use of proteinuria as a surrogate outcome in this type of accelerated drug approval program is severely limited by the absence of contemporary data on the duration of proteinuria reduction necessary to confer a protective effect on kidney function.8 This study addresses this key knowledge gap identified by the KHI/FDA, and can be used to inform future clinical trials investigating new therapies in IgA nephropathy. The results from the meta-analysis by Inker et al.10 on the magnitude of short-term reduction in proteinuria can be combined with our results on the duration of proteinuria remission to generate a more accurate assessment of the anticipated long-term effect on a hard clinical outcome (ESKD or a 50% decline in eGFR). This could be used to optimize the design of a short-term trial by identifying the duration of proteinuria remission necessary to feasibly demonstrate an effect on a GFR-based outcome during a postmarketing follow-up trial. Importantly, our results were independent of how patients achieved a remission in proteinuria, meaning these data can be used to inform clinical trials for not only immunosuppression treatments, but also for supportive therapies or alternative drug interventions.

The findings of this study also have significant clinical implications for physicians in the management of their patients with IgA nephropathy. The results demonstrate that virtually any duration of proteinuria reduction confers a clinical benefit to patients with IgA nephropathy, and the longer an individual maintains a proteinuria remission, the better their long-term outcome. This finding was consistent across a variety of primary clinical, pathologic, and therapeutic criteria at the time of remission, including MEST score, the use of RAAS blockade or immunosuppression to achieve proteinuria remission, and the magnitude of the peak proteinuria that preceded a remission. This information is also valuable to the individual patient with IgA nephropathy who can be informed of the value of a reduction in proteinuria, even for seemingly short periods of time, for long-term benefit on their kidney survival. The results were consistent across the spectrum of time from biopsy to first remission, suggesting patients can experience a benefit from remission even if it occurs late in their disease course. The relationship between duration of proteinuria remission and the risk of ESKD or a 50% decline in eGFR was nonlinear, with an attenuation in risk reduction demonstrated after approximately 4 years in remission. This is not meant to infer there is no benefit to maintaining a remission beyond that time. Rather, remission duration beyond 4 years was associated with a >80% relative reduction in the risk of the primary outcome, presumably because (as shown in Table 3) these patients experienced near-complete remission of proteinuria during follow-up. This finding is consistent with previous studies showing the most favorable prognosis was associated with lower time-averaged proteinuria.4⇓–6 Importantly, as shown in Table 3, the group of patients with sustained long-term remission could not have been identified on the basis of their clinical characteristics at or before the onset of first remission.

The importance of proteinuria reduction for long-term kidney health has been demonstrated consistently in other glomerular diseases, including FSGS and membranous nephropathy, such that for both of these conditions a disease remission is defined by the magnitude of proteinuria reduction.27,28 Rarely, however, has the duration of proteinuria reduction been investigated. A study from the Toronto GN Registry group evaluated the association between duration of complete or partial proteinuria remission and the risk of disease progression in patients with membranous nephropathy.29 The authors used a landmark analysis to estimate the risk of ESKD or 50% reduction in eGFR, conditional on remission status at fixed time points after the date of first remission. Similar to our results, the authors observed that maintaining remission was associated with a significantly reduced risk of the primary outcome from as early as 3 months. The landmark method is a useful construct for estimating risk over time; however, it has a number of limitations related to censored events and changing remission status depending on the choice of the landmark times.30 In the present analysis, we used a time-varying approach in which remission status was defined during longitudinal follow-up, as opposed to being classified at a prespecified time point. Each patient thus contributed information to the analysis, and all events were used in the calculation of risk estimates. As a sensitivity analysis, we compared our primary findings to those from a landmark analysis at 3 and 6 months after the onset of first remission, and demonstrated consistent results. However, the time-varying approach also facilitated an interrogation of nonlinear associations between time in remission, as a continuous variable, and the risk of disease progression. This allowed us to use a data-driven approach to determine there is no minimum duration of proteinuria remission in IgA nephropathy that confers benefit.

Our study has a number of a limitations. We chose a 25% relative reduction in proteinuria to be consistent with the recent meta-analysis by Inker and colleagues.8,10 We acknowledge that our findings on the basis of this definition of proteinuria remission may not apply similarly to the full spectrum of the magnitude of proteinuria reduction. However, we repeated the analyses with two more conservative definitions of relapse, which did not appear to affect the results. Unlike the previous meta-analysis of randomized clinical trials,10 this study was conducted in observational cohorts and the results are therefore subject to residual unmeasured confounding. Although the results were robust to multivariable adjustment, including known histologic and clinical risk factors for disease progression, we encourage external validation of our findings. Although our subgroup analyses suggested the relationship between duration of remission and the primary outcome did not differ among those who received immunosuppression or RAAS blockade, we were not able to assess specific types, dosing, and duration of immunosuppression and/or RAAS blockade. This will need to be addressed in prospective clinical trials that report both the degree and duration of proteinuria reduction. Although >70% of proteinuria measurements were on the basis of 24-hour urine collections, the rest were estimated from spot urine protein-to-creatinine ratios, which may have contributed to measurement error in the exposure variable. Strengths of our study include the large and multiethnic sample of patients with biopsy-proven disease, substantial observation time, and available histology, and the use of a time-varying approach to accurately classify remission status from repeated measurements of proteinuria.

In conclusion, we observed a nonlinear dose-response relationship between the duration of proteinuria remission and the risk of ESKD or a 50% decline in eGFR in adult patients with biopsy-proven IgA nephropathy. This finding was independent of recognized risk factors, such as the MEST score and prior use of either RAAS blockade or immunosuppression. This study provides new data to support the use of proteinuria as a surrogate outcome measure in IgA nephropathy by demonstrating the duration of proteinuria remission needs to be considered when evaluating proteinuria reduction as a surrogate measure for the long-term benefit on kidney outcomes.

Disclosures

D. Cattran reports receiving grants from Genentech and personal fees from Calliditis Pharmaceutical, Chemocentryx, Dimerix, Novartis, and Principia Biopharma, outside the submitted work. H. Reich reports having consultancy agreements with Calliditas, Omeros - Academic; receiving research funding from clinical trial recruitment from Alnylam, Calliditas, Chemocentryx, and Omeros; receiving Honoraria from Novartis; being a member of the Editorial Board of Kidney International and a scientific advisor to Omeros; other interests/relationships from peer-reviewed funding from the Kidney Foundation of Canada, Canadian Institutes for Health Research, Fast Foundation, and the Pearson Foundation. S. Barbour is a Scholar with the Michael Smith Foundation for Health Research. All remaining authors have nothing to disclose.

Funding

Funding support for this project was provided by grant funding from the Canadian Institutes of Health Research (Institute of Nutrition, Metabolism and Diabetes, grant PCG-155557). The European Validation Study of the Oxford Classification of IgA Nephropathy (VALIGA) was supported by a grant from the first research call and the Immunonephrology Working Group of the European Renal Association—European Dialysis and Transplant Association. The Oxford derivation and North American Validation studies were supported by the International IgA Nephropathy Network, the Toronto GN Registry, and the Toronto General Hospital Foundation (Sorenson and McCann Fund). The funders had no role in the design and conduct of the study, collection, management, analysis, and interpretation of the data, preparation, review, or approval of the manuscript, and decision to submit the manuscript for publication.

Acknowledgments

The authors would like to acknowledge all of the investigators that contributed to the datasets used in this analysis as part of the International IgA Nephropathy Network (see Supplemental Material for details).

Dr. S.J. Barbour, Dr. M. Canney, and Dr. D. Cattran designed the study and developed the analytic plan. All analyses were performed by Ms. Yuyan Zheng. Dr. M. Canney and Dr. S.J. Barbour drafted the manuscript. All authors reviewed the results, revised the manuscript and approved the final draft before submission.

Supplemental Material

This article contains the following supplemental material online at http://jasn.asnjournals.org/lookup/suppl/doi:10.1681/ASN.2020030349/-/DCSupplemental.

Supplemental Summary 1. Collaborator information.

Supplemental Appendix 1. Specification of the exposure variable.

Supplemental Appendix 2. Description of the landmark analysis.

Supplemental Table 1. Frequency of missing data for covariates.

Supplemental Table 2. Results of multivariable Cox proportional hazards model for primary analysis. Due to nonproportional hazards, the Chinese race variable was split into two periods of time.

Supplemental Table 3. Multivariable association between each additional 3 months in remission during shorter remissions up to approximately 4 years and the primary outcome within predefined subgroups. Continuous variables such as time to remission, peak proteinuria before remission, and proteinuria at remission were categorized on the basis of their median values.

Supplemental Figure 1. Distribution of the number of proteinuria measurements per patient after remission.

Supplemental Figure 2. Probability of not experiencing the primary outcome (ESKD or a 50% decline in eGFR).

Supplemental Figure 3. Restricted cubic spline of the unadjusted association between duration of remission and the primary outcome (ESKD or a 50% decline in eGFR) among patients with at least three proteinuria measurements beyond their date of remission. The y-axis represents the natural logarithm of the relative hazard of the primary outcome.

Supplemental Figure 4. Restricted cubic spline of the unadjusted association between duration of remission and the primary outcome (ESKD or a 50% decline in eGFR) when relapse was defined as a 50% increase in proteinuria to an absolute value of ≥1 g/d. The y-axis represents the natural logarithm of the relative hazard of the primary outcome.

Supplemental Figure 5. Restricted cubic spline of the unadjusted association between duration of remission and the primary outcome (ESKD or a 50% decline in eGFR) when relapse was defined as a 100% increase in proteinuria to an absolute value of ≥1 g/d. The y-axis represents the natural logarithm of the relative hazard of the primary outcome.

Supplemental Figure 6. Kaplan-Meier curves illustrating the risk of the primary outcome (ESKD or a 50% decline in eGFR) among patients who experienced a relapse before key landmark times versus those who stayed in remission. Results are presented for two landmark times: 3 months (left panel) and 6 months (right panel) after the onset of remission. Patients who relapsed at or before 3 months (hazard ratio 1.57; 95% confidence interval, 1.09 to 2.27) and patients who relapsed at or before 6 months (hazard ratio, 2.14; 95% confidence interval, 1.62 to 2.82) had a higher risk of experiencing the primary outcome compared with patients who stayed in remission beyond those landmark times.

Footnotes

  • M.C., S.J.B., and D.C. contributed equally to this work.

  • Published online ahead of print. Publication date available at www.jasn.org.

References

  1. ↵
    1. McGrogan A,
    2. Franssen CFM,
    3. de Vries CS
    : The incidence of primary glomerulonephritis worldwide: A systematic review of the literature. Nephrol Dial Transplant 26: 414–430, 2011pmid:21068142
    OpenUrlCrossRefPubMed
  2. ↵
    1. Cattran DC,
    2. Feehally J,
    3. Cook HT,
    4. Liu ZH,
    5. Fervenza FC,
    6. Mezzano SA, et al
    .: Kidney disease: Improving global outcomes (KDIGO) glomerulonephritis work group. KDIGO clinical practice guideline for glomerulonephritis. Kidney Int Suppl 2: 139–274, 2012
    OpenUrlCrossRef
  3. ↵
    1. Leaf DE,
    2. Appel GB,
    3. Radhakrishnan J
    : Glomerular disease: Why is there a dearth of high quality clinical trials? Kidney Int 78: 337–342, 2010 pmid:20520593
    OpenUrlCrossRefPubMed
  4. ↵
    1. Le W,
    2. Liang S,
    3. Hu Y,
    4. Deng K,
    5. Bao H,
    6. Zeng C, et al
    .: Long-term renal survival and related risk factors in patients with IgA nephropathy: Results from a cohort of 1155 cases in a Chinese adult population. Nephrol Dial Transplant 27: 1479–1485, 2012pmid:21965586
    OpenUrlCrossRefPubMed
  5. ↵
    1. Reich HN,
    2. Troyanov S,
    3. Scholey JW,
    4. Cattran DC
    ; Toronto Glomerulonephritis Registry: Remission of proteinuria improves prognosis in IgA nephropathy. J Am Soc Nephrol 18: 3177–3183, 2007pmid:17978307
    OpenUrlAbstract/FREE Full Text
  6. ↵
    1. Coppo R,
    2. Troyanov S,
    3. Bellur S,
    4. Cattran D,
    5. Cook HT,
    6. Feehally J, et al
    .; VALIGA study of the ERA-EDTA Immunonephrology Working Group: Validation of the Oxford classification of IgA nephropathy in cohorts with different presentations and treatments. Kidney Int 86: 828–836, 2014pmid:24694989
    OpenUrlCrossRefPubMed
  7. ↵
    1. Donadio JV,
    2. Bergstralh EJ,
    3. Grande JP,
    4. Rademcher DM
    : Proteinuria patterns and their association with subsequent end-stage renal disease in IgA nephropathy. Nephrol Dial Transplant 17: 1197–1203, 2002pmid:12105241
    OpenUrlCrossRefPubMed
  8. ↵
    1. Thompson A,
    2. Carroll K,
    3. A Inker L,
    4. Floege J,
    5. Perkovic V,
    6. Boyer-Suavet S, et al
    .: Proteinuria reduction as a surrogate end point in trials of IgA nephropathy. Clin J Am Soc Nephrol 14: 469–481, 2019pmid:30635299
    OpenUrlAbstract/FREE Full Text
  9. ↵
    1. Barbour SJ,
    2. Cattran DC,
    3. Espino-Hernandez G,
    4. Hladunewich MA,
    5. Reich HN
    : Identifying the ideal metric of proteinuria as a predictor of renal outcome in idiopathic glomerulonephritis. Kidney Int 88: 1392–1401, 2015pmid:26287314
    OpenUrlCrossRefPubMed
  10. ↵
    1. Inker LA,
    2. Mondal H,
    3. Greene T,
    4. Masaschi T,
    5. Locatelli F,
    6. Schena FP, et al
    .: Early change in urine protein as a surrogate end point in studies of IgA nephropathy: An individual-patient meta-analysis. Am J Kidney Dis 68: 392–401, 2016 pmid:27032886
    OpenUrlPubMed
  11. ↵
    1. Floege J,
    2. Barbour SJ,
    3. Cattran DC,
    4. Hogan JJ,
    5. Nachman PH,
    6. Tang SCW, et al
    .; Conference Participants: Management and treatment of glomerular diseases (part 1): Conclusions from a Kidney Disease: Improving Global Outcomes (KDIGO) Controversies Conference. Kidney Int 95: 268–280, 2019
    OpenUrlPubMed
  12. ↵
    1. Fellström BC,
    2. Barratt J,
    3. Cook H,
    4. Coppo R,
    5. Feehally J,
    6. de Fijter JW, et al
    .; NEFIGAN Trial Investigators: Targeted-release budesonide versus placebo in patients with IgA nephropathy (NEFIGAN): A double-blind, randomised, placebo-controlled phase 2b trial. Lancet 389: 2117–2127, 2017pmid:28363480
    OpenUrlCrossRefPubMed
  13. ↵
    1. Herzenberg AM,
    2. Fogo AB,
    3. Reich HN,
    4. Troyanov S,
    5. Bavbek N,
    6. Massat AE, et al
    .: Validation of the Oxford classification of IgA nephropathy. Kidney Int 80: 310–317, 2011pmid:21544062
    OpenUrlCrossRefPubMed
  14. ↵
    1. Cattran DC,
    2. Coppo R,
    3. Cook HT,
    4. Feehally J,
    5. Roberts ISD,
    6. Troyanov S, et al
    .; Working Group of the International IgA Nephropathy Network and the Renal Pathology Society: The Oxford classification of IgA nephropathy: Rationale, clinicopathological correlations, and classification. Kidney Int 76: 534–545, 2009pmid:19571791
    OpenUrlCrossRefPubMed
  15. ↵
    1. Zeng CH,
    2. Le W,
    3. Ni Z,
    4. Zhang M,
    5. Miao L,
    6. Luo P, et al
    .: A multicenter application and evaluation of the oxford classification of IgA nephropathy in adult Chinese patients [published correction appears in Am J Kidney Dis 61: 186, 2013]. Am J Kidney Dis 60: 812–820, 2012pmid:22819700
    OpenUrlCrossRefPubMed
  16. ↵
    1. Shi SF,
    2. Wang SX,
    3. Jiang L,
    4. Lv JC,
    5. Liu LJ,
    6. Chen YQ, et al
    .: Pathologic predictors of renal outcome and therapeutic efficacy in IgA nephropathy: Validation of the oxford classification. Clin J Am Soc Nephrol 6: 2175–2184, 2011pmid:21852672
    OpenUrlAbstract/FREE Full Text
  17. ↵
    1. Barbour SJ,
    2. Coppo R,
    3. Zhang H,
    4. Liu ZH,
    5. Suzuki Y,
    6. Matsuzaki K, et al
    .; International IgA Nephropathy Network: Evaluating a new international risk-prediction tool in IgA nephropathy [published correction appears in JAMA Intern Med 179: 1007, 2019 10.1001/jamainternmed.2019.2030]. JAMA Intern Med 179: 942–952, 2019pmid:30980653
    OpenUrlCrossRefPubMed
  18. ↵
    1. Hogan MC,
    2. Reich HN,
    3. Nelson PJ,
    4. Adler SG,
    5. Cattran DC,
    6. Appel GB, et al
    .: The relatively poor correlation between random and 24-hour urine protein excretion in patients with biopsy-proven glomerular diseases. Kidney Int 90: 1080–1089, 2016pmid:27528553
    OpenUrlPubMed
  19. ↵
    1. Levey AS,
    2. Stevens LA,
    3. Schmid CH,
    4. Zhang YL,
    5. Castro AF 3rd,
    6. Feldman HI, et al
    .; CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration): A new equation to estimate glomerular filtration rate [published correction appears in Ann Intern Med 155: 408, 2011]. Ann Intern Med 150: 604–612, 2009pmid:19414839
    OpenUrlCrossRefPubMed
  20. ↵
    1. Roberts ISD,
    2. Cook HT,
    3. Troyanov S,
    4. Alpers CE,
    5. Amore A,
    6. Barratt J, et al
    .; Working Group of the International IgA Nephropathy Network and the Renal Pathology Society: The Oxford classification of IgA nephropathy: Pathology definitions, correlations, and reproducibility. Kidney Int 76: 546–556, 2009pmid:19571790
    OpenUrlCrossRefPubMed
  21. ↵
    1. Trimarchi H,
    2. Barratt J,
    3. Cattran DC,
    4. Cook HT,
    5. Coppo R,
    6. Haas M, et al
    .; IgAN Classification Working Group of the International IgA Nephropathy Network and the Renal Pathology Society; Conference Participants: Oxford classification of IgA nephropathy 2016: An update from the IgA Nephropathy Classification Working Group. Kidney Int 91: 1014–1021, 2017pmid:28341274
    OpenUrlCrossRefPubMed
  22. ↵
    1. Inker LA,
    2. Lambers Heerspink HJ,
    3. Mondal H,
    4. Schmid CH,
    5. Tighiouart H,
    6. Noubary F, et al
    .: GFR decline as an alternative end point to kidney failure in clinical trials: A meta-analysis of treatment effects from 37 randomized trials. Am J Kidney Dis 64: 848–859, 2014pmid:25441438
    OpenUrlCrossRefPubMed
  23. ↵
    1. Levey AS,
    2. Inker LA,
    3. Matsushita K,
    4. Greene T,
    5. Willis K,
    6. Lewis E, et al
    .: GFR decline as an end point for clinical trials in CKD: A scientific workshop sponsored by the National Kidney Foundation and the US Food and Drug Administration. Am J Kidney Dis 64: 821–835, 2014pmid:25441437
    OpenUrlCrossRefPubMed
  24. ↵
    1. Graham JW,
    2. Olchowski AE,
    3. Gilreath TD
    : How many imputations are really needed? Some practical clarifications of multiple imputation theory. Prev Sci 8: 206–213, 2007pmid:17549635
    OpenUrlCrossRefPubMed
  25. ↵
    1. White IR,
    2. Royston P,
    3. Wood AM
    : Multiple imputation using chained equations: Issues and guidance for practice. Stat Med 30: 377–399, 2011pmid:21225900
    OpenUrlCrossRefPubMed
  26. ↵
    1. Moons KGM,
    2. Donders RART,
    3. Stijnen T,
    4. Harrell FE Jr.
    : Using the outcome for imputation of missing predictor values was preferred. J Clin Epidemiol 59: 1092–1101, 2006pmid:16980150
    OpenUrlCrossRefPubMed
  27. ↵
    1. Troyanov S,
    2. Wall CA,
    3. Miller JA,
    4. Scholey JW,
    5. Cattran DC
    ; Toronto Glomerulonephritis Registry Group: Idiopathic membranous nephropathy: Definition and relevance of a partial remission. Kidney Int 66: 1199–1205, 2004pmid:15327418
    OpenUrlCrossRefPubMed
  28. ↵
    1. Troyanov S,
    2. Wall CA,
    3. Miller JA,
    4. Scholey JW,
    5. Cattran DC
    ; Toronto Glomerulonephritis Registry Group: Focal and segmental glomerulosclerosis: Definition and relevance of a partial remission. J Am Soc Nephrol 16: 1061–1068, 2005pmid:15716334
    OpenUrlAbstract/FREE Full Text
  29. ↵
    1. Cattran DC,
    2. Kim ED,
    3. Reich H,
    4. Hladunewich M,
    5. Kim SJ
    ; Toronto Glomerulonephritis Registry group: Membranous nephropathy: Quantifying remission duration on outcome. J Am Soc Nephrol 28: 995–1003, 2017pmid:27756808
    OpenUrlAbstract/FREE Full Text
  30. ↵
    1. Dafni U
    : Landmark analysis at the 25-year landmark point. Circ Cardiovasc Qual Outcomes 4: 363–371, 2011pmid:21586725
    OpenUrlAbstract/FREE Full Text
Back to top

In this issue

View Selected Citations (0)
Print
Download PDF
Email Article
Citation Tools
Request Permissions
Quantifying Duration of Proteinuria Remission and Association with Clinical Outcome in IgA Nephropathy
Mark Canney, Sean J. Barbour, Yuyan Zheng, Rosanna Coppo, Hong Zhang, Zhi-Hong Liu, Keiichi Matsuzaki, Yusuke Suzuki, Ritsuko Katafuchi, Heather N. Reich, Daniel Cattran, for the International IgA Nephropathy Network
JASN Feb 2021, 32 (2) 436-447; DOI: 10.1681/ASN.2020030349
del.icio.us logo Digg logo Reddit logo Twitter logo Facebook logo Google logo Mendeley logo

Jump to section

Keywords