Introduction: We examined the association of proteinuria with the risk for heart failure (HF) and other cardiovascular disease (CVD) events in patients with prior history of breast, colorectal, or stomach cancer using a nationwide population-based database. Methods: We conducted this retrospective observation study using the JMDC Claims Database and analyzed 55,191 patients with prior history of breast, colorectal, or stomach cancer. The median age was 54 (48–60) years, and 20,665 participants (37.4%) were men. Using urine dipstick data at baseline, 3,945 and 1,521 participants were categorized as having trace and positive proteinuria, respectively. Using Cox proportional hazards models, we examined the relationship of proteinuria with the incidence of HF and other CVD events. Results: Over a mean follow-up of 2.8 ± 2.2 years, 1,597 HF, 124 myocardial infarction, 1,342 angina pectoris, 719 stroke, and 361 atrial fibrillation events were recorded. Kaplan-Meier curves showed that the cumulative incidence for HF increased with proteinuria category (log-rank p < 0.001). After multivariable adjustment, hazard ratios of trace and positive proteinuria for HF were 1.24 (95% CI, 1.04–1.47) and 1.62 (95% CI, 1.30–2.02), respectively. The presence of proteinuria was also associated with a higher risk for angina pectoris and atrial fibrillation. Discussion: Proteinuria was associated with a greater risk of developing HF and other CVD events in patients with prior history of cancer. The optimal management strategy for patients with proteinuria and cancer needs to be established for the prevention of HF in cancer patients.

© 2022 The Author(s). Published by S. Karger AG, Basel

Introduction

In the USA, there are approximately 17 million cancer survivors. Furthermore, the number of cancer survivors is predicted to increase to >22 million by 2030 [1], and progress in cancer treatment (e.g., molecular targeted therapy) has improved the survival of a variety of cancer patients. Accordingly, cardiovascular disease (CVD), which can occur in the chronic phase of cancer patients, has been clinically recognized as an important issue. In particular, heart failure (HF) and cancer coexist frequently [2], and therefore, risk stratification of HF development in cancer patients is currently important. Proteinuria is not only an indicator of chronic kidney disease (CKD) [3-7] but also a risk factor for the development of CVD, including HF [8-11]. There have been previous studies on the relationship between renal insufficiency and outcomes in cancer patients [12-14]. However, most of these studies examined the association of reduced estimated glomerular filtration rate (eGFR) with all-cause (or cancer related) mortality, and clinical data on the association of proteinuria with incident CVD among cancer patients have been scarce. In this study, we selected three cancer types, namely, breast, colorectal, and stomach cancers, based on the incidence of cancers among the Japanese population (https://ganjoho.jp/reg_stat/statistics/stat/summary.html). The relationship between proteinuria and the risk of developing HF was analyzed using a nationwide population-based dataset. Using this large-scale database, it was also examined whether the association of proteinuria with incident HF would vary according to baseline characteristics (e.g., age, sex, presence of active treatment, cancer site, eGFR). We believe that it is important to uncover the clinical significance of proteinuria in cancer patients at this time when the importance of CVD (particularly HF) in patients with cancer is attracting increasing attention.

Materials and Methods

Anyone who purchases the JMDC Claims Database from JMDC Inc. (https://www.jmdc.co.jp/en/) can use this database.

Study Design and Data Source

This is a retrospective observational study using the JMDC Claims Database (JMDC; Tokyo, Japan) between January 2005 and April 2021 [15, 16]. This dataset includes individual health insurance records enrolled from more than 60 insurers and health checkup data (e.g., urine dipstick test). Claims data were recorded using the International Classification of Diseases, 10th Revision (ICD-10) codes. We extracted information on 68,070 people who were diagnosed with breast cancer (ICD-10: C50), colorectal cancer (ICD-10: C18-20), or stomach cancer (ICD-10: C16) and underwent a health checkup after the diagnosis of cancer more than 1 year after the date of health insurance enrollment (1-year lookback period). The following records were excluded: 6,227 individuals with a prior history of CVD or renal failure; 1,531 with missing data on cigarette smoking; and 5,121 with missing data on alcohol consumption. Finally, 55,191 participants were included in this study (online suppl. Fig. 1; for all online suppl. material, see www.karger.com/doi/10.1159/000527703).

Ethics

This study was performed in accordance with the ethical guidelines of The University of Tokyo (approval by the Ethical Committee of The University of Tokyo: 2018-10862) and in compliance with the Declaration of Helsinki. The requirement for informed consent was waived because all data in this dataset were anonymized and de-identified.

Urine Dipstick Test and Other Measurements

Urine dipstick tests were performed as part of a health checkup using fresh midstream urine samples. Protein concentrations were estimated as negative, trace (10–20 mg/dL), or positive (≥30 mg/dL) [11]. Obesity was defined as a body mass index ≥25 kg/m2 [17]. Hypertension was defined as systolic blood pressure ≥140 mm Hg, diastolic blood pressure ≥90 mm Hg, or blood pressure-lowering medication use. Diabetes mellitus was defined as fasting glucose ≥126 mg/dL or the use of glucose-lowering medications. Dyslipidemia was defined as low-density lipoprotein cholesterol ≥140 mg/dL, high-density lipoprotein cholesterol <40 mg/dL, triglycerides ≥150 mg/dL, or the use of lipid-lowering medications. Information on cigarette smoking (current or noncurrent) and alcohol consumption (every day or not every day) was obtained from self-reported questionnaires at the health checkup.

Outcomes

Clinical outcomes were collected between January 2005 and April 2021. The primary outcome of the present study was defined as HF (ICD-10: I500, I501, I509, and I110). The secondary outcomes included myocardial infarction (ICD-10: I210, I211, I212, I213, I214, I219), angina pectoris (AP) (ICD-10: I200, I201, I208, I209), stroke (ICD-10: I630, I631, I632, I633, I634, I635, I636, I638, I639, I600, I601, I602, I603, I604, I605, I606, I607, I608, I609, I610, I611, I613, I614, I615, I616, I619, I629, G459), and atrial fibrillation (AF) (ICD-10: I480, I481, I482, I483, I484, I489).

Statistical Analysis

Data are presented as numbers (percentages) for categorical variables or medians (interquartile ranges) for continuous variables. Summary statistics were calculated for the characteristics of participants between proteinuria categories (negative, trace, and positive). The statistical significance of differences among the three proteinuria categories was determined using analysis of variance for continuous data and χ2 tests for categorical data. The cumulative incidence of HF events was calculated using the Kaplan-Meier method and compared between the three proteinuria categories using the log-rank test. Cox regression analyses were performed to assess the association of proteinuria (negative proteinuria as a reference) with a risk of HF or other CVD events. Hazard ratios (HRs) were calculated in an unadjusted model (model 1), age-sex-adjusted model (model 2), and after adjustment for age, sex, obesity, hypertension, diabetes mellitus, dyslipidemia, cigarette smoking, alcohol consumption, and cancer site (model 3).

Seven sensitivity analyses were performed, and the first was multiple imputations for missing data [16, 18]. On the assumption of missing data at random, missing data were imputed for covariates using the chained equation method with 20 iterations as described by Aloisio et al. [19], and HRs and standard errors were obtained using Rubin’s rules [20]. Second, death could be considered as a competing risk with HF events. Since the purpose of the present study was to clarify the association of proteinuria with incident HF, a competing risk analysis was performed using the Fine-Gray model [21]. Third, the predicted albumin-to-creatinine ratio was calculated using the following conversion equation for urine dipstick protein adjusted for sex, hypertension, and diabetes mellitus: predicted albumin-to-creatinine ratio = exp (2.0373 + 0.7270 [if trace] + 1.6775 [if +] + 3.2622 [if ++] + 4.5435 [if >++] + 0.0822 [if female] + 0.27249 [if diabetes mellitus] + 0.33627 [if hypertension]) [22]. The study participants were divided into three groups according to the predicted urine albumin-to-creatinine ratio (<30 mg/g, 30–299 mg/g, ≥300 mg/g). Thereafter, the relationship between this category and incident HF was analyzed. Fourth, the induction period was set as 1 year and participants with a follow-up period of ≥365 days were included. Fifth, subgroup analyses were conducted based on age (age ≥50 years vs. age <50 years), sex (men vs. women), the presence of active treatment for cancer (surgical treatment, chemotherapy, radiation therapy within 6 months or 12 months), cancer sites, and eGFR (eGFR ≥60 mL/min/1.73 m2 vs. eGFR <60 mL/min/1.73 m2). The p values for the interaction were calculated using a multivariable model. Sixth, we analyzed 28,733 patients with available data on eGFR and added eGFR to covariates in the multivariable Cox regression analysis. Seventh, we analyzed 30,393 participants who had the same results on dipstick tests at the initial checkup and at the checkup 1 year later. Statistical significance was set at p < 0.05. Statistical analyses were conducted using Stata (StataCorp LLC, College Station, TX, USA).

ResultsCharacteristics of Study Population

The participants examined in this study totaled 55,191 patients. Among the study participants, 26,020 (47.1%) participants had a history of breast cancer, 18,873 (34.2%) had a history of colorectal cancer, 11,480 (20.8%) had a history of stomach cancer, 1,160 had a history of two cancers, and 11 had a history of three cancers. The median (interquartile range) age was 54 (48–60) years, and 20,665 participants (37.4%) were men. Participants were categorized according to urine dipstick test as negative (n = 49,725), trace (n = 3,945), or positive (n = 1,521). Participants in the trace proteinuria and positive proteinuria groups had a higher prevalence of hypertension, obesity, diabetes mellitus, dyslipidemia, cigarette smoking, and alcohol consumption than participants in the negative proteinuria group (Table 1).

Table 1.Proteinuria Category and HF Events

During a mean follow-up of 2.8 ± 2.2 years, 1,597 HF events were recorded. The cumulative incidence of HF events was lowest in the negative proteinuria group, followed by the trace proteinuria group, and then the positive proteinuria group (Fig. 1) (log-rank p < 0.001). The event rates for HF events were lowest in the negative proteinuria group (99.4 [94.2–104.8] per 10,000 person-years), followed by the trace proteinuria group (130.7 [111.0–154.0] per 10,000 person-years), and then the positive proteinuria group (213.5 [173.7–262.5] per 10,000 person-years). In an unadjusted model (model 1), trace proteinuria and positive proteinuria were associated with a higher risk of HF events compared with negative proteinuria. After multivariable adjustment (model 3), the HRs (95% CI) for HF events were 1.24 (95% CI, 1.04–1.47) and 1.62 (95% CI, 1.30–2.02) for trace proteinuria and positive proteinuria, respectively (Fig. 2).

Fig. 1.

Kaplan-Meier curves for HF. We compared the cumulative incidence of HF between the proteinuria categories using the log-rank test. Participants were categorized according to urine dipstick test as negative, trace, or positive (≥1+).

Fig. 2.

Proteinuria category and HF. The incidence rate was per 10,000 person-years. We performed the Cox regression analysis to examine the association between proteinuria category and the risk for HF. Model 1 is unadjusted. Model 2 includes adjustment for age and sex. Model 3 includes adjustment for age, sex, hypertension, obesity, diabetes mellitus, dyslipidemia, cigarette smoking, alcohol consumption, and cancer site. HRs (95% confidence interval) were presented.

Proteinuria Category and Other Cardiovascular Events

During the follow-up, 124 myocardial infarction, 1,342 AP, 719 stroke, and 361 AF events were recorded. In the unadjusted model (model 1), proteinuria was associated with an elevated risk of AP, stroke, and AF. In multivariable adjustment models (model 3), positive proteinuria was associated with a higher risk for AP (HR 1.51, 95% CI, 1.18–1.93) and AF (HR 1.89, 95% CI, 1.27–2.82) compared to negative proteinuria (Table 2).

Table 2.

Proteinuria category and other cardiovascular events

Sensitivity Analyses

First, multiple imputations for covariates were performed, and results with multiple imputations versus without were similar in terms of the association of proteinuria category with incident HF (online suppl. Table 1). Second, the relationship between proteinuria category and the risk of developing HF did not change in the competing risk analysis using the Fine-Gray model (online suppl. Table 2). Third, compared with people with predicted albumin-to-creatinine ratio of <30 mg/g, HRs for developing HF were 1.71 (95% CI, 1.36–2.15) for those with predicted albumin-to-creatinine ratio of 30–299 mg/g and 2.13 (95% CI, 1.23–3.69) for those with predicted albumin-to-creatinine ratio of ≥300 mg/g (online suppl. Table 3). Fourth, 41,869 participants were included with a follow-up period of ≥365 days for HF. Even in this population, the risk of developing HF increased with proteinuria (online suppl. Table 4). Fifth, the results of the subgroup analyses are summarized in Online Supplementary Figure 2. The association between proteinuria category and incident HF was consistent in subgroups stratified by age and active treatment for cancer. The risk of developing HF increased with proteinuria category in men and baseline eGFR ≥60 mL/min/1.73 m2, but not in women having a history of breast cancer and eGFR <60 mL/min/1.73 m2. The p value for the interaction between men and women was <0.001. Sixth, even after adjustment for eGFR, the risk of incident HF increased with proteinuria category (online suppl. Table 5). Seventh, in the population who had the same results on dipstick tests at the initial checkup and at the checkup 1 year later, positive proteinuria was associated with a higher risk of developing HF than negative proteinuria (HR 2.37, 95% CI, 1.63–3.46) (online suppl. Table 6).

Discussion

The present study included 55,191 people with a history of breast, colorectal, or stomach cancer, all of which are common cancer sites in Japan, and demonstrated that the incidence of developing HF increased with proteinuria. Those with trace or positive proteinuria had more compromised clinical characteristics compared with those with no proteinuria at baseline. However, patients with both trace and positive proteinuria were associated with a greater risk of developing HF than those with no proteinuria, even after adjustment for covariates. In addition, proteinuria increased the subsequent risk of AP and AF. This relationship between proteinuria and incident HF was also observed in patients receiving active treatment for cancer or with preserved eGFR. At present, this study is the first large-scale epidemiological study to uncover the association between proteinuria and incident HF and other CVDs in patients with cancer.

CKD is commonly observed in cancer patients, and previous epidemiological studies showed that the presence of CKD increased cancer mortality [12-14]. For example, an analysis of 8,223 cancer patients showed that the HRs of cancer-specific mortality were 1.12 (95% CI, 1.01–1.26) for patients with eGFR of 30–59 mL/min/1.73 m2 and 1.75 (95% CI, 1.32–2.32) for patients with eGFR of <30 mL/min/1.73 m2 compared with those with eGFR of ≥60 mL/min/1.73 m2 [13]. Similarly, analysis of an Australian population-based cohort showed that the adjusted HR for cancer-specific mortality for those with eGFR <60 mL/min/1.73 m2 was 1.27 (95% CI, 1.00–1.60) compared with those with eGFR ≥60 mL/min/1.73 m2 [14]. However, little is known about the relationship between CKD assessed using proteinuria and incident CVD in cancer patients.

This study is distinguishable from previous studies in that an association of semiquantitatively assessed proteinuria using the dipstick urine test was found to have a greater risk of developing HF and other CVD events. A multitude of sensitivity analyses strengthened this primary finding, and several points should be noted. First, the association of proteinuria with a greater risk of developing HF was more pronounced in men than in women, and the association of proteinuria with incident HF was attenuated in patients with a history of breast cancer. The potential sex difference in the relationship between proteinuria and HF development requires further investigation. Second, proteinuria was associated with a higher risk of developing HF irrespective of the presence of active treatment for cancer (surgical treatment, chemotherapy, and radiation therapy); therefore, the assessment of proteinuria using the urine dipstick test would be useful for the risk stratification of HF events in cancer patients in both the active cancer treatment phase and the chronic (treatment free) phase. Third, a stepwise increase in the risk of developing HF with proteinuria was present in patients with eGFR ≥60 mL/min/1.73 m2. Therefore, the results highlight the importance of considering patients with cancer and proteinuria as a high-risk subset for developing HF even if their eGFR values were preserved and point out the need to revisit the importance of proteinuria assessment in cancer patients. We previously reported that the presence of proteinuria was associated with a greater risk of developing HF among a general population [11], and therefore, the assessment of proteinuria was important for the risk stratification of HF in both people with and without cancer. However, given that cancer itself, cancer-associated comorbidities, or cancer treatment would cause proteinuria (or kidney damage), the detailed assessment and evaluation of proteinuria would be clinically more important in cancer patients. As a next step, methods to prevent the development of HF and other CVD events in cancer patients with proteinuria should be further investigated. In this study, a robust relationship was found between proteinuria and incident HF. However, whether pharmacological intervention (e.g., renin-angiotensin system inhibitor, sodium-glucose cotransporter 2 inhibitor, nonsteroidal mineralocorticoid receptor antagonist) or nonpharmacological intervention (e.g., salt restriction, maintaining physical activity, body weight loss) for proteinuria could reduce the future risk of HF remains unclear, and further investigations are warranted.

The present study had several strengths. This large-scale, population-based study included more than 50,000 patients with a history of cancer. There is also a high retention of study participants in the JMDC Claims Database. The administrative claims records are included in this database, and therefore, all CVD events can theoretically be tracked even if a study participant sees different medical providers. Cancer patients frequently see different physicians, which leads to a loss of observation; hence, this is another advantage of using the JMDC Claims Database.

This study also had several limitations. Although the latest clinical guidelines recommend the use of quantitative evaluation for proteinuria (e.g., the urine protein-to-creatinine ratio or the urine albumin-to-creatinine ratio) [23, 24], the JMDC Claims Database included results from the urine dipstick test (semiquantitative evaluation) alone. However, the primary result did not change when study participants were categorized using the predicted albumin-to-creatinine ratio, as shown in online supplementary Table 3. Furthermore, given its feasibility in clinical practice, the urine dipstick test is inexpensive and easy to conduct; therefore, the results of the present study demonstrate the robust relationship of proteinuria, assessed using the urine dipstick test, with a greater risk for HF, which is practically important. Multivariate analysis was conducted in this study; however, there could be unmeasured confounders and residual biases. For example, socioeconomic status and sodium intake could have affected the relationship between proteinuria and incident CVD (including HF), but these data are not available for this dataset. The incidence of CVD events in the JMDC Claims Database is comparable to that in other epidemiological datasets in Japan [25, 26], and the specificity of recorded diagnoses (including CVD) in an administrative claims database was reported to be high [27]. Therefore, our data could reflect real-world clinical settings in Japan, although recorded diagnoses in the administrative claims database should generally be considered less well-validated and there remains uncertainty regarding the accuracy of the diagnoses for CVD. Since people aged >75 years were not included in this dataset, whether the findings of this study could be expanded to elderly patients should be confirmed using other datasets. Detailed information on cancer (i.e., cancer stage) is not available in the JMDC Claims Database, and the association between proteinuria and incident HF did not change in a competing risk model. However, the severity of cancer may influence this association. Because we analyzed patients with three cancer sites in this study, it remains unclear whether our primary findings could be applicable to people having a history of other cancer sites. The JMDC Claims Database lacks data on CVD death or etiology of HF (e.g., HF with reduced or preserved ejection fraction).

In conclusion, not only positive proteinuria but also trace proteinuria was associated with a greater risk of developing HF in patients with a history of breast, colorectal, or stomach cancer. These results are helpful for the risk stratification of future HF events in cancer survivors and suggest the importance of multidisciplinary team management (e.g., oncologists, nephrologists, cardiologists) for further improvement in the prognosis of cancer survivors.

Statement of Ethics

This study was performed in accordance with the ethical guidelines of The University of Tokyo (approval by the Ethical Committee of The University of Tokyo: 2018-10862) and in accordance with the principles of the Declaration of Helsinki. The requirement for informed consent was waived because all data in this dataset were anonymized and de-identified.

Conflict of Interest Statement

Research funding and scholarship funds (Hidehiro Kaneko and Katsuhito Fujiu) from Medtronic Japan Co., Ltd, Boston Scientific Japan Co., Ltd, Biotronik Japan, Simplex QUANTUM Co., Ltd, and Fukuda Denshi, Central Tokyo Co., Ltd.

Funding Sources

This work was supported by grants from the Ministry of Health, Labour and Welfare, Japan (19AA2007 and H30-Policy-Designated-004) and the Ministry of Education, Culture, Sports, Science and Technology, Japan (17H04141).

Author Contributions

Conception and design: H. Kaneko, Y. Suzuki, A. Okada, K. Node, and I. Komuro. Analysis of data: Y. Suzuki, A. Okada, H. Itoh, K. Fujiu, N. Michihata, T. Jo, and H. Yasunaga. Interpretation of data: H. Kaneko, A. Okada, K. Fujiu, K. Kamiya, A. Matsunaga, J. Ako, H. Morita, A. Nishiyama, K. Node, Yasunaga, and I Komuro. Drafting of the manuscript: H. Kaneko, A. Okada, Y. Suzuki, N. Takeda, and H. Morita. Critical revision for important intellectual content: N. Takeda, H. Morita, H. Yasunaga, and I. Komuro. Final approval of the submitted manuscript: M. Nangaku, H. Yasunaga, and I. Komuro.

Data Availability Statement

This database is available for anyone who purchases it from the JMDC Inc. (https://www.jmdc.co.jp/en/). Hidehiro Kaneko, as the corresponding author, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

This article is licensed under the Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC). Usage and distribution for commercial purposes requires written permission. Drug Dosage: The authors and the publisher have exerted every effort to ensure that drug selection and dosage set forth in this text are in accord with current recommendations and practice at the time of publication. However, in view of ongoing research, changes in government regulations, and the constant flow of information relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any changes in indications and dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new and/or infrequently employed drug. Disclaimer: The statements, opinions and data contained in this publication are solely those of the individual authors and contributors and not of the publishers and the editor(s). The appearance of advertisements or/and product references in the publication is not a warranty, endorsement, or approval of the products or services advertised or of their effectiveness, quality or safety. The publisher and the editor(s) disclaim responsibility for any injury to persons or property resulting from any ideas, methods, instructions or products referred to in the content or advertisements.