Structural modeling for Oxford histological classifications of immunoglobulin A nephropathy.

In immunoglobulin A nephropathy (IgAN), Cox regression analysis can select independent prognostic variables for renal functional decline (RFD). However, the correlation of the selected histological variables with clinical and/or treatment variables is unknown, thereby making histology-based treatment decisions unreliable. We prospectively followed 946 Japanese patients with IgAN for a median of 66 mo. and applied structural equation modeling (SEM) to identify direct and indirect effects of histological variables on RFD as a regression line of estimated glomerular filtration rate (eGFR) via clinical variables including amount of proteinuria, eGFR, mean arterial pressure (MAP) at biopsy, and treatment variables such as steroid therapy with/without tonsillectomy (ST) and renin-angiotensin system blocker (RASB). Multi-layered correlations between the variables and RFD were identified by multivariate linear regression analysis and the model's goodness of fit was confirmed. Only tubular atrophy/interstitial fibrosis (T) had an accelerative direct effect on RFD, while endocapillary hypercellularity and active crescent (C) had an attenuating indirect effect via ST. Segmental sclerosis (S) had an attenuating indirect effect via eGFR and mesangial hypercellularity (M) had accelerative indirect effect for RFD via proteinuria. Moreover, M and C had accelerative indirect effect via proteinuria, which can be controlled by ST. However, both T and S had additional indirect accelerative effects via eGFR or MAP at biopsy, which cannot be controlled by ST. SEM identified a systemic path links between histological variables and RFD via dependent clinical and/or treatment variables. These findings lead to clinically applicable novel methodologies that can contribute to predict treatment outcomes using the Oxford classifications.

Introduction

The international histological classification for IgA nephropathy (IgAN) known as the Oxford classification was developed by defining and selecting relevant pathological variables for renal functional decline (RFD). This classification utilizes five pathological variables assessed individually: mesangial hypercellularity (M), endocapillary hypercellularity (E), segmental glomerulosclerosis (S), tubular atrophy/interstitial fibrosis (T), and cellular or fibrocellular (active) crescent (C) [1-3]. Almost all previous studies used traditional Cox regression analysis, which selects independent prognostic variables to propose high-risk patients. However, the correlation of the selected histological variables with clinical and/or treatment variables is unknown, and this is important information when choosing therapy. As a consequence, high-risk patients, irrespective of either active or chronic histological variables, equally receive immunosuppressive treatment at the time of biopsy [4, 5]. Due to a lack of sufficient evidence for histology-based decision making, the Kidney Disease: Improving Global Outcomes (KDIGO) guidelines recommend the choice of immunosuppressant therapy be made, not based on histology, but mostly based on clinical features indicating more than 1 g/d proteinuria around the time of biopsy and during the two years after biopsy [6]. Therefore, risk stratification and treatment decisions currently still rely on inaccurate categorization of these risk factors.

Recently, a new International Risk-Prediction Tool for IgAN has been proposed as a more accurate tool to predict disease progression based on both histological and clinical risk factors, including treatment choice [7-10]. This is a personalized prediction model using an equation composed not only of clinical variables including urine protein excretion (UPE), estimated glomerular filtration rate (eGFR), and mean arterial pressure (MAP) at biopsy and treatment choice such as steroid therapy (ST) and renin–angiotensin system blocker (RASB) but also the histological variables MEST. This personalized equation predicts the probability of developing 50% decrease in eGFR or end-stage renal disease in 5 years [8]. However, the formula is composed of evenly evaluated histological, clinical, and treatment variables as a simple summation in the exponential function [8, 11]. Therefore, it is still not known how each histological variable will respond to treatment choice after renal biopsy.

RFD is associated not only directly with histological, clinical, and/or treatment variables, but also indirectly and unequally with clinical and/or treatment variables. Structural equation modeling (SEM) is a method which can estimate these complex interactions by adjusting for measurement errors in dependent variables using the error term “ε,” reducing bias in correlation estimates [12, 13].

In this prospective study, we therefore aimed to use SEM with multivariate linear regression analysis to find structural paths of correlation between each histological variable and slope as a regression line of eGFR (SLOPE) via clinical variables and treatment choice. The appropriately fitting model, including direct and indirect effects on RFD of Oxford classification variables via clinical and/or treatment variables, can then be implemented clinically.

Materials and methods

Objectives

This was a prospective clinical study. Patients with IgAN were recruited and their clinical data and renal biopsy materials collected at 44 kidney centers across Japan. This clinical research project was produced by the ethical committee of The Research Group on Progressive Renal Diseases organized by the Ministry of Health, Labour and Welfare in Japan. The study protocol was in accordance with the standards of the ethics committee at each center, and each patient consented to participate after being informed of the purpose and procedure of the study.

Patients with IgAN registered between April 2005 and August 2015 in the Japan IgA Nephropathy Prospective Cohort Study (JIGACS), which is a prospective observational study conducted at facilities throughout Japan, were included in the study. The protocol was approved by the Jikei University School of Medicine’s Institutional Review Board on Human Research, which acted as the main ethics committee for this study (No. 16–174 [4402]). Written consent was obtained from each patient and the date of consent acquisition was recorded in the designated column of the consent form. If the patient was below the age of 18 years, written consent was obtained from his/her guardian. Of 1130 patients, we excluded those with more than one of the following variables missing: MESTC score, baseline UPE, eGFR, or MAP, and SLOPE measurements, or if they had no follow-up data after renal biopsy. Secondary cases that showed mesangial IgA deposits, although with a predominant combined disease such as diabetes mellitus, were excluded. After exclusions, 946 patients were included in the study.

Clinical data

Normally distributed variables, as determined using the Shapiro–Wilk test, non-parametric variables and categorical variables were expressed as mean ± standard deviation, as median and range, and as frequency and percentage, respectively. The data were analyzed using SPSS version 24 (IBM Corp., USA).

Clinical variables, which were collected within 1 mo of biopsy and during follow-up, were as follows. MAP was defined as diastolic pressure plus a third of the pulse pressure. The eGFR was calculated as per the Japanese-based equation: eGFR (ml/min/1.73 m2) = 194 × serum creatinine (sCr)−1.094 × age−0.287 (if female, ×0.739) [14]. In patients who were younger than 20 years, the eGFR was calculated incorporating polynomial formulas for sCr and body length [15]. The Oxford study excluded initial GFR <30 ml/min and initial proteinuria <0.5 mg/d, but no such exclusion criteria were applied in this study. Treatments taken within a year after biopsy were recognized as selected treatments: immunosuppressive treatment including ST with/without tonsillectomy (reported as an intended treatment regardless of the type or duration of therapy) or RASB indicating any exposure to angiotensin-converting enzyme inhibitor, angiotensin receptor blockers, or both. If the patients had already received RASB at the time of renal biopsy, this was also recorded as an initial treatment. Weight, height, sCr, and amount of proteinuria (g/d) were also recorded.

Outcomes

In patients where more than two eGFR measurements were available, SLOPE was calculated as a regression line and used as an outcome, as an indicator of RFD.

Pathological data

All cases were proven to be IgAN by biopsy. IgAN was defined by dominant mesangial depositions of IgA and its presence in more than 10 glomeruli. Pathological variables used in the present study included M0, M1, E0, E1, S0, S1, T0, T1, or T2, and C0, C1, or C2, which were defined according to the Oxford classification [1-3]. Briefly, M0 and M1 indicated the percentages of glomeruli with a mesangial hypercellularity of 0%–50% and <50%, respectively according to a simplification of the mesangial hypercellularity score as shown in the Oxford study [1, 2]. S0 or S1 and E0 or E1 indicated the absence or presence of S and E, respectively. T0, T1, and T2 indicated interstitial fibrosis at 0%–25%, 26%–50%, and >50%, respectively. C0, C1, and C2 indicated the percentage of glomeruli with either a cellular or fibrocellular crescent at 0%, 1%–24%, and ≥25%, respectively. Renal biopsies were scored for pathological variables according to the Oxford study by five renal pathologists (KJ, AH, AS, SH, RK) blinded to the clinical data. When disagreement occurred between observers, scoring was repeated by all five observers to reach a consensus. The agreement rate was good or moderate in our previous study; the intraclass correlation coefficient of M, E, S, T1/T2, and C1/C2 were 0.54, 0.57, 0.64, 0.72, and 0.57, respectively, among the five pathologists [16].

SEM

SEM model building and estimation was done using STATA/SE version 15 (Light Stone, USA). A two-sided P <0.05 was considered significant.

SEM included analysis of the direct paths between histological variables and SLOPE and indirect paths between histological variables and SLOPE via clinical factors and treatment choice, and linear regression analysis was used to find statistically significant direct and indirect effects. Linear regression was used instead of Cox regression to find indices of the model’s statistical fit, as goodness of fit could not be applied for the survival analysis model based on Cox multivariate analysis.

The candidate variables were selected according to a previous study by Barbour et al. [8] We first created a hypothetical model consisting of all histological, clinical, and treatment variables and SLOPE as a marker of RFD (Fig 1). Considering normality, clinical variables were modified as follows: The baseline UPE0, baseline square root eGFR0 (SReGFR0), and MAP at biopsy were centralized as each patient’s UPE0c, SReGFR0c, and MAPc, which was the value subtracted by an average of 946 patients in each category. Direct paths are indicated by arrows from each independent histological variable (M, E, S, T, and C), clinical variable (centralized UPE [UPEc0], centralized square root eGFR [SReGFR0c], or centralized MAP dichotomized with negative MAPc as MAPc0 and positive MAPc as MAPc1 [MAPc01]), and treatment variable (ST and RASB) after renal biopsy leading to SLOPE as an endpoint (Fig 1a–1c). Indirect paths are indicated by arrows from each independent histological variable to each mediating variable (clinical and treatment; Fig 1d and 1e) and from each dependent clinical variable to each dependent treatment variable (Fig 1f). As MAP and eGFR can affect the amount of proteinuria [6], the correlations between UPE0c and MAPc01 or SReGFR0c were added as clinical variables. The correlations of a–f were estimated using a multiple linear regression model. We reported standardized path coefficients (SC), P values, and confidence intervals (CI). In addition, error terms (ε1–ε6), which account for measurement error, were used in addition to latent clinical variables and SLOPE.

Fig 1

Hypothetical model incorporating all histological, clinical, and treatment variables and the slope of change in estimated glomerular filtration rate (SLOPE), as a marker of renal functional decline (RFD).

Direct paths are indicated as arrows from each independent histological variable (M, E, S, T, or C), clinical variable (UPEc0, SReGFR0c, and MAPc01), and treatment variable (ST with/without tonsillectomy and renin–angiotensin system blocker [RASB]) after renal biopsy leading to SLOPE (Fig 1a–1c). Indirect paths are indicated as each histological variable pointing to each mediating variable (clinical and treatment, Fig 1d and 1e) and pointing to each dependent treatment variable from each dependent clinical variable (Fig 1f).

These calculations using the hypothetical model identified significant correlations among histological, clinical, and treatment variables and SLOPE, allowing a new model to be drawn. Only paths with correlations of P <0.1 were drawn as arrows in this revised model. To determine any elimination bias, it was checked whether link paths with correlations of P <0.1 were the same as those with correlations of P <0.05. An appropriately fitting model was proven using population error (root mean square error of approximation [RMSEA] < 0.05 with 90% CI), baseline comparison (comparative fit index [CFI] > 0.90), and size of residuals (standardized root mean square residual [SRMSR] < 0.05) [17-19]. The sample size of 946 observations for the present analysis consisting of 19 variables including M0, M1, E0, E1, S0, S1, T0, T1 or T2, C0, C1 or C2, UPE0c, SReGFR0c, MAPc0, MAPc1, ST0, ST1, RASB0, RASB1, and SLOPE was appropriate because an ideal sample size-to-parameters ratio would be 20:1, indicating that the minimum sample size should be 19 × 20 [20].

Results

Clinical profile

Clinical profiles of 946 Japanese with IgAN are shown in Table 1. Patients had a median age of 37.1 years old (2.8–87.5 years old) and were followed up for a median of 66 mo (1–174 mo). There was an equal distribution of males (49%) and females (51%) indicating no gender-based skewing of the results. MAP at renal biopsy was 90.0 ± 13.7 mmHg and dichotomized and centralized MAPc1 was 45%. eGFR0 was 75.6 ± 28.9 ml/min and UPE0 was 1.1±2.2 mg/dl. Body mass index (BMI) was 22.1 ± 3.8. The SLOPE was −0.10 ± 0.51 ml/min/1.73 m2/y.

Table 1

Clinical, treatment, and histological profiles of the IgAN patients in this study.

Clinical profile		Histological profile
			n (%)
Cohort	946	M
Male, n (%)	463 (49%)	M0	671 (71%)
Female, n (%)	483 (51%)	M1	275 (29%)
Age	37.1 (2.8–87.5)	E
Ethnicity	Japanese	E0	611 (65%)
MAP at biopsy (mmHg)	90.0 ± 13.7	E1	335 (35%)
MAPc01, n (%)		S
Yes	430 (45%)	S0	247 (26%)
No	516 (55%)	S1	699 (74%)
eGFR (ml/min/1.73 m 2 )	75.6 ± 28.9	T
UPE0 (at biopsy)	1.1 ± 2.2	T0	739 (78%)
BMI	22.1 ± 3.8	T1	170 (18%)
eGFR slope (ml/min/1.73m 2 /y)	−0.10 ± 0.51	T2	37 (4%)
Period of follow-up (mo)	66 (1–174)	C
		C0	580 (61%)
Treatment choice	n (%)	C1	357 (38%)
ST		C2	9 (1%)
Yes	605 (64%)
No	341 (36%)
RASB
Yes	539 (57%)
No	407 (43%)

BMI: body mass index; eGFR: estimated glomerular filtration rate; MAP: mean arterial pressure; MAPc01: centralized dichotomized MAP; RASB: renin-angiotensin system blocker; ST: steroid therapy with/without tonsillectomy; UPE0: baseline urine protein excretion. M0 and M1 indicated the percentages of glomeruli with a mesangial hypercellularity of 0%–50% and <50%, respectively. S0 or S1 and E0 or E1 indicated the absence or presence of S and E, respectively. T0, T1, and T2 indicated interstitial fibrosis at 0%–25%, 26%–50%, and >50%, respectively. C0, C1, and C2 indicated the percentage of glomeruli with either a cellular or fibrocellular crescent at 0%, 1%–24%, and ≥25%, respectively.

Treatment

Table 1 shows an overview of treatment choices for the study participants. ST was used to treat 64%, which included steroid pulse therapy in 20%, steroid pulse with tonsillectomy in 38%, oral steroid therapy in 5%, and oral steroids with tonsillectomy in 1%. RAS blockade was used to treat 57%. A small proportion (5%) was treated only with tonsillectomy, and 34% were not treated at all.

Pathological profile

Histological results from renal biopsy samples are shown in Table 1. The proportion of the patients with M1, E1, S1, T1, T2, C1, and C2 was 29%, 35%, 74%, 18%, 4%, 38%, and 1%, respectively.

A hypothetical model was drawn as described in the Methods and tested with a multivariate linear regression model using SEM (Fig 1). After removing non-significant paths (P >0.1), SEM was performed again to find an appropriately fitting model. The significant paths were the same whether the significance threshold was set at P <0.05 or P <0.1 as shown in Fig 2. Therefore, no elimination bias was apparent. As indices of the models’ statistical fit, RMSEA, CFI, and SRMSR were 0.05 as same as 0.05, 0.93 as more than 0.90, and 0.03 as less than 0.05, respectively. Therefore, these results indicate that the model was an appropriate fit according to a likelihood estimation of appropriate fit as shown in the Methods [17, 18]. An analyzing process was shown in a flow chart (S1 File).

Fig 2

SEM was performed to find an appropriate fitting model after removing non-significant paths (P > 0.05).

Statistically significant paths between histological variables, clinical variables, or treatment variables and SLOPE are shown as arrows with standardized coefficients marked. T, besides ST, UPE0c, and SReGFR0c, showed direct correlations with SLOPE (P < 0.05) (blue arrows). All histological variables showed indirect effects on SLOPE via clinical variables or ST (red arrows).

Analyzing indirect contributors to SLOPE, ST, which correlated with S, E, C, UPE0c, SReGFR0c, and MAPc01, was correlated with SLOPE. RASB, which correlated with UPE0c, SReGFR0c, and MAPc01, was not correlated with SLOPE. The error terms ε1–ε6 of each of MAPc01, SReGFR0c, UPEc0, ST, RASB, and SLOPE were 0.93, 0.68, 0.89, 0.88, 0.83, and 0.95, respectively.

In the SEM analysis, statistically significant direct correlations with SLOPE (P <0.05) were independent accelerative histological variable T, independent accelerative clinical variables UPE0c and SReGFR0c, and independent attenuating treatment variable ST. As well as the direct correlations with SLOPE, correlations between clinical and histological variables and between treatment and histological variables or clinical variables were calculated and indicated indirect correlations with SLOPE (Fig 2 and Table 2). UPE0c positively correlated with M and C and negatively correlated with SReGFR0c. MAPc01 positively correlated with S and T. SReGFR0c negatively correlated with S and T. ST negatively correlated with MAP01c and positively correlated with UPE0c, SReGFR0c, S, E, and C. RASB, although correlated with UPE0c, MAPc01, and SReGFR0c, was not an independent variable for SLOPE.

Table 2

Direct and indirect correlations between histological variables and change in estimated glomerular filtration rate (SLOPE) via clinical variables and treatment variables.

	SC	SE	z	P>z	[95% Confidence Interval]
SLOPE
UPE0c	−0.073	0.033	−2.200	0.028	−0.138	−0.008
SReGFR0c	−0.224	0.039	−5.800	<0.001	−0.299	−0.148
ST	0.127	0.032	3.99	<0.001	0.065	0.19
T	−0.095	0.038	−2.500	0.013	−0.170	−0.020
_cons	−0.318	0.056	−5.630	<0.001	−0.429	−0.207
UPE0c
SReGFR0c	−0.205	0.031	−6.660	<0.001	−0.266	−0.145
M	0.194	0.031	6.26	<0.001	0.133	0.254
C	0.102	0.031	3.31	0.001	0.042	0.162
_cons	−0.205	0.042	−4.870	<0.001	−0.287	−0.122
MAPc01
S	0.091	0.032	2.84	0.004	0.028	0.152
T	0.237	0.03	7.79	<0.001	0.177	0.297
_cons	0.637	0.066	9.61	<0.001	0.507	0.766
SReGFR0c
S	−0.095	0.027	−3.52	<0.001	−0.148	−0.042
T	−0.545	0.021	−25.740	<0.001	−0.587	−0.504
_cons	0.448	0.051	8.79	<0.001	0.349	0.548
ST
UPE0c	0.108	0.032	3.41	0.001	0.046	0.17
MAPc01	−0.069	0.033	−2.100	0.036	−0.133	−0.004
SReGFR0c	0.109	0.034	3.24	0.001	0.043	0.175
S	0.127	0.032	3.97	<0.001	0.064	0.19
E	0.114	0.033	3.41	0.001	0.048	0.18
C	0.180	0.034	5.24	<0.001	0.112	0.247
_cons	0.956	0.076	12.53	<0.001	0.807	1.106
RASB
UPE0c	0.088	0.03	2.88	0.004	0.028	0.147
MAPc01	0.207	0.031	6.61	0	0.146	0.269
SReGFR0c	−0.300	0.031	−9.550	0	−0.361	−0.238
_cons	0.976	0.049	19.87	0	0.88	1.072

SC: standardized coefficient; SE: standard error; T: tubular atrophy/interstitial fibrosis; M: mesangial hypercellularity; C: active crescent; S: segmental glomerulosclerosis; E: endocapillary hypercellularity; ST: steroid therapy including tonsillectomy; RASB: renin-angiotensin system blocker; UPE0c: centralized base line urine protein excretion; SReGFR0c: centralized square root baseline eGFR; MAPc01: centralized dichotomized baseline mean arterial pressure; cons: constant.

As shown in Table 3, direct and/or indirect effects of histological variables on SLOPE were calculated as total coefficients by combining the direct and indirect coefficients for each histological variable (Table 3). An analyzing process was shown in a flow chart (S1 File). The total coefficients were the sums of the direct and indirect coefficients, where each indirect coefficients of M, E, S, T, or C in Table 3 was the integration of each coefficient of the clinical and treatment variables in Table 2 according to the paths shown in Fig 2, which stood between histological variables and SLOPE (Fig 2). T showed an total accelerative effect composed of the direct accelerative coefficient and the indirect accelerative coefficients as via SReGFR0c and ST and via MAPc01 and ST. E showed an attenuating indirect effect via ST. C showed an attenuating indirect effect composed of an attenuating indirect effect via ST and attenuating indirect effect via UPE0c and ST. S showed an attenuating indirect effect via SReGFR0c and additionally an accelerative indirect effect via SReGFR0c and ST, as well as via MAPc01 and ST. M showed an accelerative effect composed of accelerative indirect effect via UPE0c and an attenuating indirect effect via UPE0c and ST (Table 3). In summary, only T showed an accelerative direct effect on RFD, whereas E and C were not independent variables for RFD, correlated significantly with ST, and showed attenuating effects on RFD via ST. S showed attenuating effects on RFD via SReGFR0c. Both C and M had additional accelerative effects via UPE0c, which can be controlled by ST, thus changing the accelerative effect to attenuating effect. On the other hand, both T and S had additional indirect accelerative effects on RFD via SReGFR0c or via MAPc01, which could not be controlled by ST, thus continuing to exhibit the accelerative effect. The error terms ε1–ε6 of each MAPc01, SReGFR0c, UPEc0, ST, RASB, and SLOPE were 0.93, 0.68, 0.89, 0.88, 0.83, and 0.95, respectively (Fig 2).

Table 3

Direct and indirect effects of Oxford histological variables on change in estimated glomerular filtration rate (SLOPE).

Histological	Direct	Indirect		Total
variable	coefficient	coefficient	via	coefficient
M		−0.014	UPE0c	−0.011
M		0.003	UPE0c, ST
E		0.014	ST	0.014
S		0.021	SReGFR0c	0.019
S		−0.001	SReGFR0c, ST
S		−0.001	MAPc01, ST
T	−0.095			−0.105
T		−0.008	SReGFR0c, ST
T		−0.002	MAPc01, ST
C		0.023	ST	0.024
C		0.001	UPE0c, ST

M: mesangial hypercellularity; E: endocapillary hypercellularity; S: segmental glomerulosclerosis; T: tubular atrophy/interstitial fibrosis; C: active crescent; SReGFR0c: centralized square root baseline eGFR; MAPc01: centralized dichotomized baseline mean arterial pressure; UPE0c: centralized baseline urine protein excretion; ST: steroid therapy including tonsillectomy.

Discussion

In this prospective multicenter study involving 946 Japanese IgAN patients, we applied SEM to evaluate structural correlations associated with the change in eGFR as RFD (SLOPE). Consequently, this analysis showed systemic path links between SLOPE and histological variables via clinical and/or treatment variables.

We identified the contributors to SLOPE as T (T1 or T2, independent accelerative histological variable), SReGFR0c and UPE0c (independent accelerative clinical variables), and ST (independent attenuating treatment variable). RASB was not an independent variable for SLOPE. Further, the selected contributors as well as the independent variables for SLOPE were connected via dependent histological, clinical, and/or treatment variables. Therefore, we investigated the statistically significant structural correlations among the histological, clinical, and treatment variables and SLOPE.

Using the coefficients of the direct and indirect correlations, we calculated the total effect of each histological variable on SLOPE (Table 3). T was the only histological variable with a direct (accelerative) effect on SLOPE. It had additional indirect effects (accelerative) via SReGFR0c or MAPc01, which could not be controlled by ST, thus maintaining the accelerative effect. Both E and C attenuated SLOPE via ST. M showed an overall accelerative effect on SLOPE, thereby incorporating an accelerative effect via UPE0c. Further, both M and C had additional accelerative effects on SLOPE via UPE0c, which was controlled by ST, thereby transforming the accelerative effect to attenuating effect. S had an attenuating indirect effect via eGFR. Moreover, both T and S had indirect accelerative effects via eGFR0c or MAP, which could not be controlled by ST, thus continuing with the accelerating effect. If S0 developed to S1, SReGFR0c decreased (Table 2). This negative correlation meant that if eGFR at biopsy was high, SLOPE declined strongly, while if it was low, there was a weaker decline. This is also true in normal kidneys [21].

The above findings suggest that ST was chosen for patients with E1 and C1 or C2 but not with T1 or T2 and that it effectively attenuated decline in eGFR. This is in partial agreement with another Japanese study, which found that patients with E1, S1, or C1 treated with ST had significantly better prognosis than the non-treatment group [22]. Similarly, in two other studies, the presence of E was strongly associated with subsequent ST, and there was a higher rate of decline in renal function in patients who were not treated with immunosuppressants [1] or corticosteroids [23]. Our study also showed that C and M had additional accelerating effects via UPE0c on SLOPE, which were controlled and thus led to the change from accelerative to attenuating effect by ST. Therefore, ST can be considered to be effective for the patients with C and M by diminishing the accelerative effect of UPE0c. Other studies have shown that patients had an increased risk of disease progression with extremely increasing C, even with immunosuppression [5, 24, 25]. T and S had additional accelerative effects via SReGFR0 or MAPc01, which could not be controlled by ST. This is consistent with the previous findings, which reported that S (as well as M and T) had predictive prognostic value for RFD [1, 5, 23]. However, the prognostic role of S1 for RFD can be small, when early ST prevented a progression from active crescent as C1/C2 to S. M appears to be a sensitive histological variable, as it can also be a risk factor for RFD via acceleration of proteinuria, when proteinuria cannot be controlled by ST [5, 26, 27]. However, the original Oxford study selected S1 and T1/T2 but not M1 for RFD in a multivariate linear regression model adjusted for initial eGFR, MAP, and UPE0 [1, 28]. The VALIGA study suggested also M1 was a steroid-responsive variable [26, 29, 30]. These conflicting results illustrate the benefit of analyzing both the direct and indirect effects.

We were able to stratify the Oxford histological classifications from the viewpoint of treatment response in predicting future outcomes. Several validation studies of the Oxford classification using Cox analysis showed merely different independent histological, clinical, and treatment variables without suggesting their concrete clinical use, which may depend on differences in the cohorts [4, 5, 31, 32]. SEM can clarify these differences by analyzing the correlations between Oxford histological variables and RFD via clinical and treatment variables.

In conclusion, SEM identified a systemic path links between histological variables and RFD via dependent clinical and/or treatment variables. We focused not only on direct effect but also on indirect effect of histological variables on RFD, where direct effects of clinical variables on RFD, such as SReGFR0c, UPE0c and ST, were influenced by histological variables. To the best of our knowledge, there has been no such systematic research previously to propose the correlation among histological, clinical, and treatment variables in depth. These findings lead to clinically applicable novel methodologies that can contribute to predict treatment outcomes using the Oxford classifications and improve care for IgAN patients using personalized medicine.

As a limitation, this prospective study using SEM cannot predict the prognosis of renal function by Cox regression analysis because goodness of fit could not be applied for the survival analysis model. Obtaining a more apparent eGFR slope needs longer period of follow-up.

(DOCX)

Click here for additional data file.

9 Jun 2022

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Zhanjun Jia

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please note that PLOS ONE has specific guidelines on code sharing for submissions in which author-generated code underpins the findings in the manuscript. In these cases, all author-generated code must be made available without restrictions upon publication of the work. Please review our guidelines at https://journals.plos.org/plosone/s/materials-and-software-sharing#loc-sharing-code and ensure that your code is shared in a way that follows best practice and facilitates reproducibility and reuse.

3. Please provide additional details regarding participant consent. In the ethics statement in the Methods and online submission information, please ensure that you have specified (1) whether consent was informed and (2) what type you obtained (for instance, written or verbal, and if verbal, how it was documented and witnessed). If your study included minors, state whether you obtained consent from parents or guardians. If the need for consent was waived by the ethics committee, please include this information.

If you are reporting a retrospective study of medical records or archived samples, please ensure that you have discussed whether all data were fully anonymized before you accessed them and/or whether the IRB or ethics committee waived the requirement for informed consent. If patients provided informed written consent to have data from their medical records used in research, please include this information.

4. Thank you for stating the following financial disclosure:

"This study was supported in part by a Grant-in-Aid for Progressive Renal Diseases Research, Research on Rare and Intractable Disease, from the Ministry of Health, Labour and Welfare of Japan. This research was supported by AMED under Grant Number JP19ek0109261."

Please state what role the funders took in the study.  If the funders had no role, please state: "The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript."

If this statement is not correct you must amend it as needed.

Please include this amended Role of Funder statement in your cover letter; we will change the online submission form on your behalf.

5. In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available. For more information about our data policy, please see http://journals.plos.org/plosone/s/data-availability.

Upon re-submitting your revised manuscript, please upload your study’s minimal underlying data set as either Supporting Information files or to a stable, public repository and include the relevant URLs, DOIs, or accession numbers within your revised cover letter. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories. Any potentially identifying patient information must be fully anonymized.

Important: If there are ethical or legal restrictions to sharing your data publicly, please explain these restrictions in detail. Please see our guidelines for more information on what we consider unacceptable restrictions to publicly sharing data: http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions. Note that it is not acceptable for the authors to be the sole named individuals responsible for ensuring data access.

We will update your Data Availability statement to reflect the information you provide in your cover letter.

Additional Editor Comments:

Some comments from the experts need to be well addressed.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Partly

Reviewer #3: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: N/A

Reviewer #3: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: No

Reviewer #3: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: No

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Previous studies have identified risk factors for IgA nephropathy progression: eGFR and macrohematuria at the time of renal biopsy (Shu, D., Xu, F., Su, Z. et al. Risk factors of progressive IgA nephropathy which progress to end stage renal disease within ten years: a case–control study. BMC Nephrol 18, 11 (2017).

This study from Kensuke Joh et al evaluates the diagnostic approach for the treatment of IgA Nephropathy. The number of renal biopsies, the duration of follow-up and the quality of the statistical analysis permit the novelty of these useful and interesting conclusions:

Only tubular atrophy/interstitial fibrosis had an accelerative direct effect on renal function decline, while endocapillary hypercellularity and active crescents had an attenuating indirect effect via steroid treatment. This message is very important. A renal biopsy must be performed as soon as possible in order to initiate treatment for endocapillary hypercellularity and active crescent.

This study satisfies all the criteria to be accepted by PLOS One

I have only very few comments to this article: there are some acronyms in the abstract that are not identified and difficult the complete understanding of the message.

line 43 S and M

line 44 UPE

line 46 eGFR0c

Reviewer #2: The present study was a prospective clinical study for patients with IgAN among 44 kidney centers across Japan. By using the Oxford classification, the pathological variables were defined in the research and analysis. The authors applied structural equation modeling (SEM) to evaluate structural correlations associated with the change in eGFR, predicting the systemic path links between renal functional decline (RFD) and histological variables vis clinical or treatment variables. This is the only systematic research that proposes the correlation between histological, clinical, and treatment variables. Overall, it is an exciting study, and most of the conclusions may provide some information for IgAN patients using personalized medicine. However, I have some concerns about this study.

1. It is hard to catch the data and analysis process in the manuscript. A flow chart will help in this regard.

2. Figure 1 shows those variables for the SEM. The authors should outline the possible relationships between those variables but not just indicate the paths in the single image.

3. After removing non-significant paths is not clear. How to remove it? What is the difference without removal?

4. Can the authors simply the models? So complicated with the variables and sub-variables in this study. Making a table for results and conclusions will be much easier to follow.

5. The description for figure 2 is confusing.

Reviewer #3: This an paper dealing with finding of independent prognostic variables for renal functional decline in IgA nephropathy. Although the subject is of great interest, overall the article is not clear, with to many abbreviation within the text. Therefore, the manuscript is difficult to read. Some abbreviations are unnecessary – eg. UPE could be replaced with proteinuria (1 word), ST- steroids etc . Some of the abbreviations found in abstract must be spelled out completely on initial appearance in text (S,M,C, MEST…)

In table 1 The period of follow up overlaps with Treatment choice. Please clarify.

The authors should explain the raison for choosing different parameters like square root eGFR0 instead of eGFR; furthermore, it is not clear what does represent the values UPE0c, SReGFR0c and MAPc- why not MAP0c? …and the confusion continues with C0, C1 etc ….

The square root GFR could not be considered a predictor of SLOPE since GFR slopes depends on the eGFR.

The authors should also emphasize on the clinical values of their findings. A limitation section should be added as well.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Teresa Adragao

Reviewer #2: No

Reviewer #3: No

**********

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

1 Aug 2022

Dear Editor-in-Chief, PLOS ONE

Thank you for the thoughtful and constructive feedback you provided regarding our manuscript, entitled “Structural modeling for Oxford histological classifications of immunoglobulin A nephropathy.” We agree with you and have incorporated these suggestions throughout our paper.

With these changes to our final manuscript, we hereby resubmit our manuscript for a secondary evaluation. Thank you once again for your consideration of our paper.

Answers for the reviewer’s comments

Reviewer #1: Previous studies have identified risk factors for IgA nephropathy progression: eGFR and macrohematuria at the time of renal biopsy (Shu, D., Xu, F., Su, Z. et al. Risk factors of progressive IgA nephropathy which progress to end stage renal disease within ten years: a case–control study. BMC Nephrol 18, 11 (2017).

Thank you for introducing an interesting paper. For the present study, we have selected clinical variables according to the study by Barbour et al., who selected amount of proteinuria, eGFR, and MAP but not macrohematuria (JAMA Intern Med. 2019;179: 942-52). The information on macrohematuria was unfortunately not collected in the national-wide prospective study and therefore not used in the present study.

This study from Kensuke Joh et al evaluates the diagnostic approach for the treatment of IgA Nephropathy. The number of renal biopsies, the duration of follow-up and the quality of the statistical analysis permit the novelty of these useful and interesting conclusions:

Only tubular atrophy/interstitial fibrosis had an accelerative direct effect on renal function decline, while endocapillary hypercellularity and active crescents had an attenuating indirect effect via steroid treatment. This message is very important. A renal biopsy must be performed as soon as possible in order to initiate treatment for endocapillary hypercellularity and active crescent.

This study satisfies all the criteria to be accepted by PLOS One

I have only very few comments to this article: there are some acronyms in the abstract that are not identified and difficult the complete understanding of the message.

Line 43 S and Line 44 M

line 44 UPE

line 46 eGFR0c

We have changed some acronyms to their expanded terms in the abstract section.

Reviewer #2: The present study was a prospective clinical study for patients with IgAN among 44 kidney centers across Japan. By using the Oxford classification, the pathological variables were defined in the research and analysis. The authors applied structural equation modeling (SEM) to evaluate structural correlations associated with the change in eGFR, predicting the systemic path links between renal functional decline (RFD) and histological variables vis clinical or treatment variables. This is the only systematic research that proposes the correlation between histological, clinical, and treatment variables. Overall, it is an exciting study, and most of the conclusions may provide some information for IgAN patients using personalized medicine. However, I have some concerns about this study.

1. It is hard to catch the data and analysis process in the manuscript. A flow chart will help in this regard.

Thank you for your valuable suggestion. We have created a flow chart and added the text as Supplementary Material 1 (Page 14; Line 262).

2. Figure 1 shows those variables for the SEM. The authors should outline the possible relationships between those variables but not just indicate the paths in the single image. After removing non-significant paths is not clear. How to remove it? What is the difference without removal?

The final relationship was illustrated in Fig. 2 corresponding with yellow as clinical variables and green as treatment variables, among which all significant relationships were connected by the arrows. Blue lines show a direct relationship with the slope. All histological variables showed additional indirect effects on SLOPE via clinical variables or ST (red arrows). An analyzing process was shown in a flow chart (Supplementary Material 1), indicating the results of linear regression analysis among nonsignificant histological, clinical, and treatment variables (see Step 3 in Supplementary Material 1). We selected the significant paths from the result of Step 3 by removing nonsignificant paths (P > 0.05). SEM was performed again to attest to the goodness of fit for an appropriately fitting model (see Step 4 in Supplementary Material 1).

2. Can the authors simply the models? So complicated with the variables and sub-variables in this study. Making a table for results and conclusions will be much easier to follow.

An analyzing process was shown in a flow chart (Supplementary Material 1). The final result was shown in Table 3, showing the direct and indirect effects of Oxford histological variables (M, E, S, T, and C) on changes in the estimated glomerular filtration rate (SLOPE). To introduce Table 3, direct and/or indirect effects of histological variables on SLOPE were calculated as total coefficients, which were the total of direct and indirect coefficients, where each indirect coefficient of M, E, S, T, or C in Table 3 indicated the integration of each coefficient of the clinical and treatment variables in Table 2 according to the paths shown in Fig. 2.

The description for figure 2 is confusing.

Figure 2 showed an appropriately fitting structural modeling. Using this model, Table 3 was made as the conclusion showing direct and indirect effects of Oxford histological variables on changes in the estimated glomerular filtration rate (SLOPE).

Reviewer #3: This paper dealing with finding of independent prognostic variables for renal functional decline in IgA nephropathy. Although the subject is of great interest, overall, the article is not clear, with too many abbreviations within the text. Therefore, the manuscript is difficult to read. Some abbreviations are unnecessary – eg. UPE could be replaced with proteinuria (1 word), ST- steroids etc.

We agree that several abbreviations are found within the text; therefore, we should avoid some unnecessary abbreviations. However, each abbreviation has its original meaning, e.g., UPEc0 (centralized urine protein excretion (UPE), SReGFR0c centralized square root estimated glomerular filtration rate (eGFR), MAPc01: centralized mean arterial pressure (MAP) dichotomized with negative MAPc as MAPc0 and positive MAPc as MAPc1. ST not only means steroid therapy but also steroid therapy with/without tonsillectomy. Therefore, we believe that it is better to use abbreviations than to spell out on an initial appearance in the text. Please let us know if you still need the abbreviations to be defined at some places.

Some of the abbreviations found in abstract must be spelled out completely on initial appearance in text (S,M,C, MEST…)

In the abstract, we have spelled out the initial appearance of S and M. We used proteinuria instead of UPE in the text (Page 3; Lines 34–46).

In table 1 The period of follow up overlaps with Treatment choice. Please clarify.

In Table 1, the follow-up period was separately indicated from the treatment choice as ST and RASB.

The authors should explain the raison for choosing different parameters like square root eGFR0 instead of eGFR; furthermore, it is not clear what does represent the values UPE0c, SReGFR0c and MAPc- why not MAP0c? …and the confusion continues with C0, C1 etc ….

To obtain the goodness of fit for an appropriately fitting model, we changed the clinical variables respecting the variable normality, such as UPEc0 (centralized urine protein excretion (UPE), SReGFR0c centralized square root estimated glomerular filtration rate (eGFR), MAPc01: centralized mean arterial pressure (MAP) dichotomized with negative MAPc as MAPc0 and positive MAPc as MAPc1. The final choice of centralization (UPEc0), square root (centralized square root eGFR), or dichotomized MAPc01 was the best choice to obtain the most appropriate fitting model.

The square root GFR could not be considered a predictor of SLOPE since GFR slopes depends on the eGFR.

We believe that eGFR at biopsy was high, and subsequently, SLOPE declined strongly, whereas when it was low, there was a weaker decline. This is also true in normal kidneys. Therefore, according to our understanding, eGFR at the time of biopsy can influence SLOPE.

The authors should also emphasize on the clinical values of their findings. A limitation section should be added as well.

Thank you for your valuable comments. We added the following sentence in the conclusion (Page 21; Lines 390–393).

“We focused not only on direct effect but also on the indirect effect of histological variables on RFD, where direct effects of clinical variables on RFD, such as SReGFR0c, UPE0c, and ST, were influenced by histological variables.”

We added the following information to the limitation section (Page 21; Lines 398–401).

“As a limitation, this prospective study using SEM could not predict the prognosis of renal function via Cox regression analysis because the goodness of fit could not be applied to the survival analysis model. Obtaining a more apparent eGFR slope needs longer follow-up.”

Sincerely,

Kensuke Joh, MD, PhD

Department of Pathology, The Jikei University School of Medicine,

3-25-8 Nishi-Shinbashi, Minato-Ku, Tokyo, 105-8461, Japan

Tel: 81-3-3433-1111, Fax: 81-3-5472-0700

E-mail: johken@med.tohoku.ac.jp

Submitted filename: Response to reviewers.docx

Click here for additional data file.

17 Aug 2022

Structural modeling for Oxford histological classification s of immunoglobulin A nephropathy

PONE-D-22-13403R1

Dear Dr. Joh,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Zhanjun Jia

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: N/A

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This study from Kensuke Joh et al evaluates the diagnostic approach for the treatment of IgA Nephropathy. The number of renal biopsies, the duration of follow-up and the quality of the statistical analysis permit the novelty of useful and interesting conclusions. It is possible to change the decline of renal function based on the results of renal biopsy.

The final review of this article answers my previous comments

Reviewer #2: The authors have addressed all the questions, and the current version of the manuscript has dramatically improved and become more readable.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Teresa Adragao MD PhD

Reviewer #2: No

**********

31 Aug 2022

PONE-D-22-13403R1

Structural modeling for Oxford histological classifications of immunoglobulin A nephropathy

Dear Dr. Joh:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Zhanjun Jia

Academic Editor

PLOS ONE