Abstract
Background/Aim: This study aimed to identify potential drug associations between proteinuria and nephrotic syndrome (NS) using the Japanese Adverse Drug Event Report (JADER) database.
Patients and Methods: We extracted data reported in JADER between April 2004 and May 2023, and conducted a comprehensive disproportionality analysis of spontaneous adverse event reports to identify drugs potentially linked to proteinuria and NS.
Results: Our analysis identified 20 and 32 drugs associated with proteinuria and NS, respectively. Notably, anti-vascular endothelial growth factor (anti-VEGF) agents represented 45% (9/20) of proteinuria-associated drugs and 34% (11/32) of NS-associated drugs. Furthermore, the association between anti-VEGF agents and these adverse events appeared to be independent of the route of administration, sex, or clinical background.
Conclusion: These findings suggest that anti-VEGF agents play a significant role in the development of proteinuria and NS. Given the widespread use of anti-VEGF therapy, heightened vigilance, routine monitoring, and timely interventions are crucial to mitigate these risks and improve patient outcomes.
Introduction
Nephrotic syndrome (NS) is a syndrome that causes an increase in urinary protein levels due to an increase in protein permeability resulting from renal glomerular loop failure and hypoproteinemia (1). NS is classified as primary NS of unknown pathology and secondary NS caused by drugs (1). Non-steroidal anti-inflammatory drugs (NSAIDs) and D-penicillamine are typical agents that cause NS (2). Despite advances in anticoagulant and immunosuppressive therapies, the average number of years of life lost due to NS is comparable to that of malignancy, suggesting that NS is a refractory disease (3).
Proteinuria is an indispensable index for the diagnosis of NS (2). Anticancer drugs targeting vascular endothelial growth factor (VEGF)-related molecules, such as multikinase inhibitors, increase the risk of this condition (4-6). Aggravation may lead to NS as well as chronic kidney disease and heart failure (7, 8). Even low levels of urinary protein excretion may lead to a high mortality risk (9). However, there is no fundamental treatment available, and patients are often managed by dose reduction or suspension of the offending drug (10).
Early detection and appropriate management may prevent the onset of NS and deterioration of the patient’s prognosis. Although there have been studies on drugs that may be pharmacologically associated with its development, there are limited examples of comprehensive analyses of clinically used drugs (11). Few studies have examined drugs associated with the development of NS (12, 13).
The spontaneously reported adverse event (AE) database is a valuable tool for identifying potential associations between specific drugs and the occurrence of AEs, establishing drug-associated AE hypotheses. The purpose of this study was to identify drugs that may be associated with proteinuria and NS using the Japanese Adverse Drug Event Report database (JADER). We also examined whether sex, age, body mass index (BMI), and concomitant medications were associated with the spontaneous reporting of these events.
Patients and Methods
Data source. The JADER is publicly accessible through the Pharmaceuticals and Medical Devices Agency (PMDA) website (https://www.pmda.go.jp). JADER has been freely available and has contained information on drug AEs and patients in Japan since April 1, 2004. JADER comprises four datasets: patient demographics (DEMO), drug details (DRUG), AEs (REAC), and medical history (HIST). DEMO includes basic patient data, such as sex, age, and year of report; DRUG provides drug details, such as name, route of administration, and date of use; REAC lists AEs with name, date of onset, and outcome; and HIST records the patient’s primary disease. AEs listed in the REAC are coded using the Medical Dictionary for Regulatory Activities classification version 24.1 and indicated by the Preferred Term. The DRUG category includes three subcategories: “suspected drugs”, “concomitant drugs”, and “drug interactions”. We focused our disproportionality analysis only on drugs classified as “suspected drugs” to concentrate on drug-related AEs.
In the DRUG table of the JADER database, the same drug may be registered multiple times within a single case ID when there are differences in dosage or administration date. Similarly, in the REAC table, identical adverse event terms may be registered multiple times for the same case ID if the onset dates differ. Such duplications can result in overestimation and introduce bias in the analysis. Therefore, we removed duplicate entries in both DRUG and REAC tables to ensure that each drug name and adverse event term appeared only once per case ID.
Definitions. In this study, NS was defined as a standardized MedDRA query report for nephrotic syndrome (code: 10029164), and proteinuria was defined as a standardized MedDRA query report for proteinuria (code: 10037032).
Disproportionality analysis. Disproportionality analyses of drug-associated AEs are frequently performed using spontaneous reporting databases. The analyses included reports published between April 2004 and May 2023. To assess the association between specific drugs and the presence or absence of AEs (proteinuria and NS) in case reports, reporting odds ratios (ROR) were calculated using a 2-by-2 contingency table with four categories: (a) number of patients with proteinuria or NS after target drug use, (b) number of patients without proteinuria or NS after target drug use, (c) number of patients with proteinuria or NS after non-target drug use, and (d) number of patients without proteinuria or NS after non-target drug use. The ROR is the reporting rate of a specific AE attributable to a specific drug divided by the reporting rate of the same adverse reaction attributable to all other drugs in the database (14). Disproportionality was defined as the lower limit of the 95% confidence interval (CI) of ROR >1 (15, 16). The current analysis included drugs with at least 10 reported AEs. We excluded fluids and blood products that were unlikely to contribute to the development of proteinuria.
Calculation of time to onset. Time-to-onset analysis utilized information structured according to when the medication was first administered and when adverse reactions first appeared. The interval between these events was determined by calculating the difference between the initial dosage date and the date of onset of the first adverse reaction. In cases where a patient experienced multiple simultaneous adverse effects, these were tallied as a single occurrence for that individual. The study imposed a two-year (730-day) cap on the time-to-onset of adverse reactions. Histograms were used to visualize the distribution of the days until AEs manifested.
Stratified analysis of anti-VEGF agents. Anti-VEGF agents are defined as drugs that act on the VEGF pathway, including VEGF monoclonal antibodies and VEGF receptor tyrosine kinase inhibitors (17). Stratified analyses were conducted for anti-VEGF agents based on age, sex, BMI, and the presence or absence of concomitant medications. Furthermore, we calculated the ROR by collectively categorizing the target drugs into oral medications, injectable medications, and all target drugs.
Detailed age, weight, and height data are not publicly available in the JADER database. For instance, ages between 50 and 59 years are reported as “50s”. Consequently, in this study, we analyzed these cases using a representative age of 50 years. Similar data processing was applied to the height and weight measurements before analysis. BMI was calculated based on processed data. Subsequently, age stratification was performed by dividing patient groups into <70 and ≥70 years (18). Regarding BMI, the analysis was conducted by categorizing patients into three groups: <18.5 kg/m2, 18.5-24.9 kg/m2, and ≥25 kg/m2. We stratified bevacizumab, ramucirumab, and aflibercept as injectable, and the others as orally administered. We also analyzed the effects of concomitant use of renin-angiotensin system (RAS) inhibitors, calcium channel blockers (CCB), and diuretics.
Statistical analysis. Statistical analyses, including disproportionality and calculation of time to onset, were conducted using JMP Pro®15.2 (SAS Institute, Cary, NC, USA). Details of each method are described in their respective sections.
Ethical considerations. The research protocol adhered to the Ethical Guidelines for Epidemiological Studies of the Ministry of Health, Labor and Welfare and the principles outlined in the Declaration of Helsinki. As this was an observational study using anonymized data from the JADER database, without involving any therapeutic interventions or human sample collection, the Ethical Review Committee of Hyogo Medical University determined that formal ethical approval was unnecessary. Furthermore, because the study relied solely on publicly accessible data, individual patient consent was not required.
Results
We analyzed 2,200,140 cases registered in JADER between April 2004 and May 2023. Among the analyzed cases, proteinuria and NS were reported in 2,279 and 2,642 cases, respectively (Figure 1). Disproportionality was detected for 20 and 32 drugs for proteinuria and NS, respectively (Table I). Anti-VEGF agents accounted for 9 (45%) and 11 (34%) drugs detected as disproportionate in the proteinuria and NS groups, respectively (Table I). In addition to anti-VEGF agents, drugs with disproportionality in proteinuria or NS include antineoplastic agents, antimicrobials, NSAIDs, and immunosuppressants. Thirteen drugs (axitinib, atezolizumab, aflibercept beta, irinotecan, interferon beta, cabozantinib, sunitinib, pazopanib, bucillamine, bevacizumab, ramucirumab, regorafenib, and lenvatinib), including 9 anti-VEGF agents, showed disproportionality in both proteinuria and NS.
Process of constructing a data analysis table. Reports were extracted from the Japanese Adverse Drug Event Report (JADER) database (April 2004–May 2023). Four datasets (DEMO, DRUG, REAC, and HIST) were merged, and only records involving suspected drugs and adverse events coded as proteinuria or nephrotic syndrome (NS) were retained. Duplicate reports were removed if they shared the same case ID, drug name, and adverse event term. The final dataset was used to calculate the reporting odds ratios (RORs).
Number of cases, reporting odds ratio (ROR), and 95% confidence interval (CI) of proteinuria and nephrotic syndrome by drug.
We created box plots of the time to onset of proteinuria or NS from the start of administration for the drugs detected as disproportionate and classified them according to the class of drugs (Figure 2). Our analysis showed that 10 (50%) of the 20 drugs associated with proteinuria onset were reported within the first month of treatment. An analysis of 31 drugs associated with NS revealed that 11 (36%) were reported within the first month of treatment. Among the thirteen drugs with disproportions detected for both AEs, nine drugs caused proteinuria earlier than they caused NS.
Box plot for time to onset of proteinuria (A) and nephrotic syndrome (B) for drugs with significant disproportionality. The horizontal axis represents the number of days from drug administration to the onset of the adverse event. Drugs are grouped into three categories: (a) anti-VEGF agents, (b) other anticancer drugs, and (c) other drugs. Each box indicates the interquartile range (IQR), with the horizontal line representing the median. Whiskers extend to 1.5 times the IQR from the first and third quartiles. Dots represent outliers beyond the whisker range.
The disproportionality analysis of anti-VEGF agents among the predefined strata revealed disproportionality across all strata for both proteinuria and NS (Table II). Stratified analysis of intravenous and oral anti-VEGF agents revealed a disproportion in all predefined strata (Table III and Table IV).
Stratified analysis of the number of cases, reporting odds ratio (ROR), and 95% confidence interval (CI) of proteinuria and nephrotic syndrome using anti-VEGF agents.
Stratified analysis of the number of cases, reporting odds ratio (ROR), and 95% confidence interval (CI) of proteinuria and nephrotic syndrome using intravenous anti-VEGF agents.
Stratified analysis of the number of cases, reporting odds ratio (ROR), and 95% confidence interval (CI) of proteinuria and nephrotic syndrome using oral anti-VEGF agents.
Discussion
Disproportionality analysis revealed that 20 and 32 drugs were associated with proteinuria and NS, respectively. Among the 20 drugs with disproportionality in proteinuria, 9 (45%) were anti-VEGF agents, and among the 32 drugs with disproportionality in NS, 11 (34%) were anti-VEGF agents. Our findings support the previously reported association of proteinuria and NS with anti-VEGF agents (10).
Anti-VEGF agents induce minimal change disease (MCNS) and renal thrombotic microangiopathy (TMA) in the renal glomeruli (19). Both pathologies are reversible, and stopping anti-VEGF exposure resolves proteinuria (10). Therefore, monitoring and early detection of proteinuria may be useful for preventing the development of NS in patients receiving anti-VEGF agents.
In our analysis, among anti-VEGF agents, sorafenib and nintedanib were the only drugs associated with NS. Previous studies have reported a lower frequency of proteinuria with sorafenib than with other anti-VEGF agents (20). Nintedanib is a therapeutic agent for idiopathic pulmonary fibrosis. Drug treatment options are limited for idiopathic pulmonary fibrosis; therefore, proteinuria may be perceived by practitioners as a minor AE and may be underreported.
Our analysis showed that drugs with a disproportionality detected in both AEs caused proteinuria earlier than they caused NS. Some drugs showing disproportionate effects have not reported NS in clinical trials (21). The occurrence of proteinuria may shift to NS; therefore, renal function monitoring and treatment interruption should be considered. The NS occurrence reporting time for aflibercept was shorter than that for proteinuria. The indication for aflibercept is bevacizumab treatment for colorectal cancer (22). All patients were pretreated with bevacizumab; therefore, the possibility of bevacizumab-induced induction cannot be ruled out. Previous studies have reported that prior anti-VEGF use increases the risk of proteinuria, resulting in an earlier onset (23). In addition, proteinuria worsens in patients at the start of treatment (24). Thus, the possibility of the early reporting of NS was considered. It is necessary to evaluate drugs that showed disproportionality in our analysis, including prior treatment, in historical cohort studies.
Our disproportionality analysis, stratified by anti-VEGF agents, suggested that anti-VEGF agents may be associated with the onset of proteinuria or NS, regardless of the route of administration, sex, or patient background. Previous studies have shown that obesity and old age increase the risk of proteinuria (25, 26). In addition, our analysis revealed an association between anti-VEGF agents and proteinuria or NS, even when used in combination with RAS inhibitors, CCB, and diuretics. RAS inhibitors are known to suppress proteinuria, and there have been reports showing a suppressive effect on proteinuria when anti-VEGF agents are used (24, 27). Analysis of a spontaneous reporting database may have detected drugs used to treat adverse events as signals.
Our findings revealed an association between the occurrence of NS and the use of rifampin, isoniazid, and etanercept. Rifampin and isoniazid are commonly prescribed to patients with tuberculosis (TB) or latent TB infection. Two case reports have implicated rifampin and isoniazid in the development of MCNS, which resolves following treatment discontinuation or corticosteroid therapy (28, 29). These reports, which also noted the presence of proteinuria, underscore the importance of monitoring renal function and proteinuria for several months after the initiation of treatment with these drugs.
Similarly, etanercept-induced MCNS has been reported to develop 2–6 months after the initiation of therapy, with both cases presenting with proteinuria (30, 31). While the disproportionality observed exclusively in NS in our study may reflect a tendency for serious AEs to be preferentially reported, the potential for rapid progression of NS and the risk of undetected proteinuria should not be overlooked.
While this study was conducted as a hypothesis-generating analysis using disproportionality methods on spontaneous reporting data, the clinical implications are nevertheless significant. Signal detection plays a key role in pharmacovigilance by identifying potential drug-event associations that may not have been previously recognized in clinical trials or cohort studies. In particular, the observed associations with anti-VEGF agents and the stratified findings across clinical subgroups provide valuable insights for clinical monitoring and early intervention. Although causality cannot be established, the findings underscore the need for vigilance in patient populations at risk and may inform further epidemiological and mechanistic studies.
Study limitations. Firstly, only AEs that were actually reported in the JADER were included. In addition, there is a possibility that serious AEs are preferentially reported or that the reports are duplicated. Secondly, information such as the grade of the AE, dose of the drug, and patient’s treatment history was not available. Thirdly, the data registered in the database were insufficient, and missing data were excluded from the analysis of the number of days until AE onset.
Our disproportionality analysis using the JADER database suggests that anti-VEGF drugs may play a role in the development of NS and proteinuria. Notably, these associations were observed regardless of the route of administration, sex, or patient characteristics, indicating a broad and clinically relevant risk. Furthermore, our findings revealed that drugs exhibiting disproportionality for both AEs tended to induce proteinuria earlier than NS. This temporal relationship highlights the potential for early detection and intervention at the onset of proteinuria to mitigate or delay progression to NS. These insights underscore the importance of vigilant monitoring and proactive management strategies for patients receiving anti-VEGF therapy.
Acknowledgements
We would like to thank Editage (www.editage.com) for English language editing.
Footnotes
Authors’ Contributions
Sho Masago: Investigation; Data curation; Visualization; Project administration; Writing – original draft. Kenta Yamaoka: Conceptualization; Supervision; Writing – review and editing. Mayako Uchida: Writing – review and editing. Yoshihiro Uesawa: Methodology; Writing – review and editing. Kennosuke Yorifuji: Writing – review and editing. Tadashi Shimizu: Project administration; Supervision; Writing – review and editing.
Data Availability Statement
The data supporting the findings of this study are available at https://www.info.pmda.go.jp/fukusayoudb/CsvDownload.jsp. All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding authors.
Conflicts of Interest
All Authors declare no conflicts of interest.
Artificial Intelligence (AI) Disclosure
During the preparation of this manuscript, a large language model (ChatGPT 4o, OpenAI) was used solely for language editing and stylistic improvements in select paragraphs. No sections involving the generation, analysis, or interpretation of research data were produced by generative AI. All scientific content was created and verified by the authors. Furthermore, no figures or visual data were generated or modified using generative AI or machine learning–based image enhancement tools.
- Received April 5, 2025.
- Revision received April 25, 2025.
- Accepted April 28, 2025.
- Copyright © 2025 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.