Abstract
Introduction
Cancer patients are at risk for acute kidney injury (AKI) due to anticancer agents, tumor lysis syndrome, sepsis, and contrast-induced nephropathy. Renal replacement therapy (RRT) is associated with longer hospital stays, higher healthcare costs, and significant healthcare resource consumption. Therefore, early detection of severe AKI requiring RRT and the implementation of preventive measures are crucial. We comprehensively evaluated the associations with RRT by cancer site among Japanese cancer patients undergoing treatment through use of Hospital-Based Cancer Registry (HBCR) data linked to administrative data in a multicenter cohort study.
Methods
Patients were classified into three groups based on the timing of RRT after the initiation of cancer treatment: acute phase (≤30 days), post-acute phase (after 30 days), and no RRT. The hazard ratios (HRs) and subdistribution HRs for RRT incidence were evaluated by cancer site (colon, lung, stomach, liver, kidney, esophagus, pancreas, and hematologic malignancies) using Cox proportional hazards and Fine-Gray competing-risk models adjusted for patient characteristics, comorbidities, and diuretic use. The 1-year cumulative incidence of RRT was also calculated.
Results
A total of 230,850 patients were analyzed. Among them, 386 patients received RRT at ≤30 days after cancer treatment initiation (acute phase), 479 patients received RRT at >30 days (post-acute phase), and 229,985 patients did not receive RRT. In the Cox proportional hazards regression model, the HRs for RRT were significantly higher in the acute phase among patients with colorectal cancer (HR: 1.72, p < 0.001), liver cancer (HR: 1.80, p = 0.006), kidney cancer (HR: 2.49, p < 0.001), and hematologic malignancies (HR: 4.06, p < 0.001) when compared according to cancer site. In the post-acute phase, only hematologic malignancies showed a significantly increased HR for RRT (HR: 2.65, p < 0.001). The 1-year cumulative incidence of RRT among all patients was 0.413%.
Conclusion
The risk of RRT after cancer treatment was increased during the acute phase in patients with colorectal, liver, and kidney cancer and those with hematologic malignancies. Patients with hematologic malignancies always had a higher HR for RRT, but the HR was particularly high in the acute phase.
Keywords: Cancer site, Renal replacement therapy, Acute kidney injury
Introduction
Patients with cancer are known to be at risk for acute kidney injury (AKI), with a reported incidence ranging from approximately 7.5%–17.5% [1, 2]. Factors reported to cause AKI in cancer patients include anticancer agents [3, 4], tumor lysis syndrome (TLS) [5, 6], sepsis [7, 8], and contrast-induced nephropathy [9, 10].
It has also been reported that cancer patients with AKI have a 0.4%–1.16% probability of receiving renal replacement therapy (RRT) [1, 11]. In critically ill AKI patients undergoing cancer treatment in intensive care units, the need for RRT is even more pronounced and has been reported to influence the prognosis and treatment strategies [12, 13]. RRT is the only treatment option to prolong the lives of patients who have severe AKI with severe oliguria or anuria unresponsive to diuretics. Although RRT is effective, it leads to extended hospital stays, increased healthcare costs, and significant consumption of medical resources [2, 14, 15]. Therefore, it is extremely important to identify severe AKI that requires RRT and implement strategies to prevent severe AKI [16]. However, reports examining the association between cancer and RRT in cancer patients at potential risk for AKI are limited. One possible reason is that RRT is infrequently performed in cancer patients with AKI, which makes an evidence-based discussion challenging. In this study, we analyzed a large dataset to examine the incidence of RRT among cancer patients and explore its association with cancer site, aiming to provide a comprehensive understanding of the relationship between RRT and cancer. These new findings focusing on RRT may be useful for the development of renal protection strategies aimed at preventing severe AKI in cancer patients.
Studies have indicated that individuals with hypertension have a reduced nephron count, predisposing them to compromised renal function. Furthermore, evidence suggests that the Japanese population exhibits a lower nephron count relative to other ethnic groups [17–20]. This underscores the importance of investigating the incidence and risk of severe renal failure in Japanese cancer patients.
In this study, we used Hospital-Based Cancer Registry (HBCR) data linked to administrative data in a multicenter cohort study of cancer patients in Osaka Prefecture to determine the association between cancer site and RRT incidence. Understanding the risk of RRT in cancer patients is important for optimizing treatment planning and managing complications. To further explore this association, we classified the timing of RRT into two phases: acute (after the start of cancer treatment) and post-acute and evaluated differences in the association with RRT by cancer site in each phase. The purpose was to identify cancer sites that require particularly careful observation for severe AKI early after the start of cancer treatment and investigate cancer sites that require careful long-term observation for severe AKI.
Methods
Data Sources
The Osaka International Cancer Institute, in collaboration with the Council for Coordination of Designated Cancer Care Hospitals in Osaka, requested anonymized records from the HBCR and Diagnosis Procedure Combination (DPC) data from 66 hospitals across Osaka Prefecture. The DPC data included a comprehensive record of each patient’s comorbidities, medical history, treatments performed, and the dates of treatment at each hospital. Comorbidities were recorded using International Classification of Diseases, Tenth Edition (ICD-10) codes.
Comorbidities
Comorbidities in this study included hypertension, diabetes mellitus, chronic kidney disease (CKD), end-stage renal disease, coronary artery disease, and cerebrovascular disease based on QUAN reports [21].
Medications
Medication history extracted from the DPC database identified the use of angiotensin-converting enzyme (ACE) inhibitors, angiotensin II receptor blockers (ARBs), and diuretics.
Treatments
From the HBCR data, surgery, laparoscopic surgery, chemotherapy, endoscopic therapy, radiation therapy, and hormonal therapy were recorded as cancer treatments, while transcatheter arterial chemoembolization (TACE) was identified as a treatment in the DPC data. Moreover, the implementation of RRT and the frequency of contrast-enhanced computed tomography and magnetic resonance imaging examinations were identified from the DPC database.
Cancer Sites
Each patient’s cancer site was identified from the HBCR according to the International Classification of Diseases for Oncology, Third Edition (ICD-O-3) localization codes. In this study, colorectal (ICD-O-3: C18-C20), lung (C34), stomach (C16), liver (C22), kidney (C64,C65), esophagus (C15), pancreas (C25), and hematologic malignancies (C42,C77) were included in the analysis because they were identified as being associated with a high risk of AKI [22].
Study Population
The study included 234,795 patients who were newly diagnosed with cancer between 2019 and 2022 and who underwent surgery, laparoscopic surgery, chemotherapy, endoscopic therapy, radiation therapy, hormone therapy, or TACE. A total of 2,761 patients without information on the date of cancer treatment, RRT, or last DPC follow-up date were excluded. Additionally, 1,184 patients with end-stage renal disease at the time of the cancer diagnosis were excluded. These patients were excluded because an outcome assessment was not possible. Ultimately, 230,850 patients with newly diagnosed cancer were included in the final analysis. If more than one cancer site was registered, the first registered cancer site was used. If the registration dates were the same, the more advanced cancer was used.
Endpoint and Statistical Analysis
The primary endpoint of this study was the initiation of RRT after cancer treatment. Time to RRT, a secondary endpoint, was defined as the interval between the initiation of cancer treatment and the administration of the first RRT. RRT events were identified using DPC procedure codes indicating temporary vascular access catheter placement followed by CRRT or other RRT procedures, including “artificial kidney [other].” Procedure codes corresponding to chronic maintenance dialysis were excluded from the outcome definition. The first recorded RRT procedure date after the initiation of cancer treatment was used as the event date. Because laboratory data such as serum creatinine levels and urine output were not available in the database, non-RRT AKI episodes and AKI severity according to the KDIGO classification could not be assessed. For baseline characteristics, categorical variables were presented as the number (frequency) and were compared using the chi-squared test. Continuous variables are presented as the median (interquartile range) and compared using the Kruskal-Wallis rank sum test. In patients with multiple cancer treatment dates (e.g., surgery and chemotherapy performed on different dates), the earliest treatment date was used as the baseline. The termination date for those who did not receive RRT was the date of the last DPC follow-up. Missing values for height and weight were imputed using the median values of the study population. No missing data were present for the remaining covariates included in the multivariable analyses; therefore, complete-case exclusion was not required. The 1-year cumulative incidence of RRT was estimated for each cancer site individually as well as all cancer sites combined. The 1-year cumulative incidence of RRT was estimated using the Fine-Gray method, with death treated as a competing event. To obtain an overall picture of the timing of RRT implementation in cancer patients, we divided the period after the start of cancer treatment into two phases. Based on previous studies on AKI [23, 24], the follow-up period was segmented into two distinct phases: (1) the initial 30 days (acute phase) and (2) the subsequent phase from 30 days (post-acute phase). The earliest cancer treatment date was defined as the baseline (Day 0). Acute phase RRT was defined as RRT occurring within 30 days after the initiation of treatment, whereas post-acute phase RRT was defined as events occurring more than 30 days after the initiation of treatment without an upper time limit. Although cumulative incidence curves were presented up to 1 year, patients were followed until the initiation of RRT, death, or the end of available follow-up data, with a maximum follow-up period of 1,359 days. The overall study design is shown in Figure 1. When multiple treatment modalities were administered, the earliest treatment date was used as the baseline (Day 0) for all analyses. Consequently, acute and post-acute phase RRT classifications were determined relative to the first treatment date, regardless of the timing of subsequent treatments. The effect of cancer site on the incidence of RRT was assessed across two consecutive time periods, using multivariable Cox proportional hazards models and Fine-Gray subdistribution hazard models adjusted for age, sex, comorbidities, and diuretic use. In the Cox models, patients were censored at the time of death. To account for death as a competing risk, Fine-Gray models were additionally performed as a sensitivity analysis, treating death as a competing event, and adjusted hazard ratios (HRs) and subdistribution hazard ratios (sHRs) were estimated accordingly. In the Cox and Fine-Gray analyses, the “other cancer sites” group, which included cancer types other than colorectal, lung, stomach, liver, kidney, esophagus, pancreas, and hematologic malignancies, was used as the reference category for comparison. These were adjusted as potential confounding factors based on prior literature. Patients who received RRT in the acute phase were compared with other patients by Cox proportional hazards regression. We also compared patients who received RRT in the post-acute phase with those who did not receive RRT. To ensure transparency of the analysis, cancer sites were entered into the multivariable models as dummy variables, with the “other cancer sites” group used as the reference category for comparison. Furthermore, stratified analyses according to treatment modality (surgery/laparoscopic surgery and chemotherapy) were performed to evaluate whether the association between cancer site and RRT differed by treatment type. Statistical analyses were performed using R (ver. 4.4.1). All statistical tests were two-sided, and p values of <0.05 were considered statistically significant.
Fig. 1.
Study design and definition of acute and post-acute phase renal replacement therapy (RRT). a Overall study design. The earliest date of cancer treatment was defined as the baseline (Day 0). Acute phase RRT was defined as RRT occurring within 30 days after the initiation of treatment, whereas post-acute phase RRT was defined as RRT occurring more than 30 days after the initiation of treatment without an upper time limit. The cumulative incidence analysis was performed over a 1-year period, while patients were followed until the initiation of RRT, death, or the end of available follow-up data (maximum follow-up: 1,359 days). b Example patient classification. In this example, RRT initiated on Day 60 after the earliest treatment date was classified as post-acute phase RRT, despite chemotherapy being administered on Day 50.
Results
Baseline Characteristics
Of 230,850 patients, 386 received RRT during the acute phase after cancer treatment, while 479 received RRT during the post-acute phase (Fig. 2). The baseline clinical characteristics are shown in Table 1. There were no missing data for the main variables of interest in the analyzed cohort because patients with missing key variables were excluded before analysis. Detailed baseline characteristics, including missing data rates, are provided in online supplementary Table S1 (for all online suppl. material, see https://doi.org/10.1159/000553501). Baseline characteristics were compared among the three groups – the acute phase RRT group, post-acute phase RRT group, and non-RRT group – with corresponding p values calculated for group comparisons. Baseline CKD was defined based on the presence of ICD-10 code N18.9 and included as a comorbidity. In the comparison between the three groups, a significant difference was observed in males (acute phase: 72.5%, post-acute phase: 72.9%, non-RRT group: 53.8%). Furthermore, significant differences were observed in hypertension (acute phase, 25.4%; post-acute phase, 28.0%; non-RRT, 17.2%), diabetes (acute phase, 29.0%; post-acute phase, 35.1%; non-RRT group, 14.7%), and CKD (acute phase, 7.8%; post-acute phase, 6.7%; non-RRT group, 1.0%). With regard to medication history, there was a significant difference in the prevalence of ACE inhibitors or ARBs among the three groups (acute phase, 37.8%; post-acute phase, 50.9%; non-RRT, 26.6%), as were diuretics (acute phase, 86.3%; post-acute phase, 87.3%; non-RRT, 21.3%). A significant difference was also observed in the number of contrast agents used. Regarding treatment modalities, significant differences were observed among the three groups. Chemotherapy (acute phase, 46.9%; post-acute phase, 54.1%; non-RRT group, 35.9%) and surgery (acute phase, 38.1%; post-acute phase, 23.8%; non-RRT group, 31.2%) differed significantly among the groups. In contrast, laparoscopic surgery (acute phase, 18.4%; post-acute phase, 26.5%; non-RRT group, 27.1%), endoscopic therapy (acute phase, 4.7%; post-acute phase, 18.0%; non-RRT group, 18.2%), radiation therapy (acute phase, 3.4%; post-acute phase, 12.1%; non-RRT group, 13.3%), and hormone therapy (acute phase, 1.3%; post-acute phase, 5.6%; non-RRT group, 12.0%) also showed significant differences among the groups. No significant difference was observed in the use of TACE (acute phase, 1.8%; post-acute phase, 1.5%; non-RRT group, 1.1%). Significant differences in cancer site distribution were observed among the three groups. Colorectal cancer (acute phase, 21.0%; post-acute phase, 12.9%; non-RRT group, 16.7%), liver (acute phase, 7.3%; post-acute phase, 4.6%; non-RRT group, 2.2%), kidney (acute phase, 5.7%; post-acute phase, 4.4%; non-RRT group, 2.5%), esophagus (acute phase, 3.9%; post-acute phase, 6.1%; non-RRT group, 3.1%), and hematologic (acute phase, 17.1%; post-acute phase, 13.6%; non-RRT group, 2.7%) malignancies, while lung (acute phase, 5.4%; post-acute phase, 6.7%; non-RRT group, 11.7%) malignancies were significantly more common in the non-RRT group.
Fig. 2.
Study population. TACE, transcatheter arterial chemoembolization; RRT, renal replacement therapy; DPC, Diagnosis Procedure Combination.
Table 1.
Baseline characteristics, treatment, and cancer sites
| | RRT acute phase 3861 | RRT post-acute phase 4791 | non-RRT 2299851 | p value2 |
|---|---|---|---|---|
| Baseline characteristics | ||||
| Age, years | 74.0 (68.0, 79.0) | 72.0 (65.0, 78.0) | 71.0 (61.0, 78.0) | <0.001 |
| Male | 280 (72.5) | 349 (72.9) | 123,839 (53.8) | <0.001 |
| Height, cm | 162.0 (155.0, 167.0) | 163.0 (156.0, 169.0) | 160.0 (154.0, 166.0) | <0.001 |
| Body weight, kg | 59.0 (50.8, 68.0) | 59.0 (51.0, 68.0) | 58.0 (50.6, 66.0) | 0.022 |
| Comorbidities | ||||
| cHypertension | 98 (25.4) | 134 (28.0) | 39,651 (17.2) | <0.001 |
| Diabetes mellitus | 112 (29.0) | 168 (35.1) | 33,753 (14.7) | <0.001 |
| Chronic kidney disease | 30 (7.8) | 32 (6.7) | 2,295 (1.0) | <0.001 |
| Coronary artery disease | 6 (1.6) | 7 (1.5) | 2,075 (0.9) | 0.175 |
| Cerebrovascular disease | 2 (0.5) | 6 (1.3) | 1,451 (0.6) | 0.221 |
| Medication | ||||
| ACE inhibitors or ARBs | 146 (37.8) | 244 (50.9) | 61,201 (26.6) | <0.001 |
| Diuretic | 333 (86.3) | 418 (87.3) | 48,961 (21.3) | <0.001 |
| Number of contrast media used | 1.0 (0.0, 3.0) | 2.0 (0.0, 4.0) | 1.0 (0.0, 3.0) | <0.001 |
| Treatment | ||||
| Surgery | 147 (38.1) | 114 (23.8) | 71,710 (31.2) | <0.001 |
| Laparoscopic surgery | 71 (18.4) | 127 (26.5) | 62,405 (27.1) | <0.001 |
| Endoscopic therapy | 18 (4.7) | 86 (18.0) | 41,808 (18.2) | <0.001 |
| Chemotherapy | 181 (46.9) | 259 (54.1) | 82,451 (35.9) | <0.001 |
| Radiation therapy | 13 (3.4) | 58 (12.1) | 30,699 (13.3) | <0.001 |
| Hormone therapy | 5 (1.3) | 27 (5.6) | 27,679 (12.0) | <0.001 |
| TACE | 7 (1.8) | 7 (1.5) | 2,518 (1.1) | 0.298 |
| Cancer site | ||||
| Colorectal | 81 (21.0) | 62 (12.9) | 38,332 (16.7) | 0.007 |
| Lung | 21 (5.4) | 32 (6.7) | 26,836 (11.7) | <0.001 |
| Stomach | 34 (8.8) | 55 (11.5) | 25,694 (11.2) | 0.330 |
| Liver | 28 (7.3) | 22 (4.6) | 4,989 (2.2) | <0.001 |
| Kidney | 22 (5.7) | 21 (4.4) | 5,700 (2.5) | <0.001 |
| Esophagus | 15 (3.9) | 29 (6.1) | 7,197 (3.1) | <0.001 |
| Pancreas | 7 (1.8) | 20 (4.2) | 7,271 (3.2) | 0.142 |
| Hematologic malignancies | 66 (17.1) | 65 (13.6) | 6,190 (2.7) | <0.001 |
| Other | 112 (29.0) | 173 (36.1) | 107,776 (46.9) | <0.001 |
Values are presented as the median (IQR) for continuous variables and n (%) for categorical variables.
The acute phase was defined as the period ≤30 days after the initiation of cancer treatment.
The post-acute phase was defined as >30 days of the initiation of cancer treatment.
p values were calculated using the Kruskal-Wallis test for continuous variables and the chi-squared for categorical variables, as appropriate.
1 Median (Q1, Q3); n (%).
2 Kruskal-Wallis rank sum test; Pearson’s chi-squared.
ACE, angiotensin-converting enzyme; ARB, angiotensin II receptor blocker; RRT, renal replacement therapy.
Clinical Endpoint
Figure 3 shows the 1-year cumulative incidence of RRT following the initiation of cancer treatment across all cancer sites. The incidence of RRT increased steeply until approximately 1.5 months. Thereafter, the incidence of RRT increased gradually. The 1-year cumulative incidence of RRT after cancer treatment was 0.413% across all cancer sites. Figure 4 shows the 1-year cumulative incidence of RRT following the initiation of cancer treatment for each cancer site. These 1-year cumulative incidence rates are shown in Table 2. The 1-year cumulative incidence of RRT varied across cancer sites, with the highest incidence observed in hematologic malignancies (2.237%), followed by liver (1.012%), kidney (0.878%), and esophageal (0.685%) cancer. The cumulative incidence was lower in pancreatic (0.439%), colorectal (0.414%), stomach (0.398%), and lung (0.225%) cancer. The median follow-up period was 9.6 months (interquartile range, 5.8–14.3), with a total follow-up of 202,891 person-years. Table 3 presents the HRs for RRT after cancer treatment according to cancer site and phase. Corresponding sHRs obtained from the Fine-Gray models are provided in online supplementary Table S2. Figures 5–7 present forest plots of HRs and sHRs for RRT after cancer treatment according to cancer site in the all, acute, and post-acute phases. Online supplementary Table S3 presents the HRs derived from multivariable Cox proportional hazards models, including all covariates entered into the analysis, with cancer sites incorporated as dummy variables and the “other cancer sites” group used as the reference category. In the acute phase after cancer treatment, the HRs for RRT by cancer site were significantly higher for colorectal (HR, 1.72; 95% confidence interval [CI]: 1.29–2.29, p < 0.001), liver (HR, 1.80; 95% CI: 1.18–2.74, p = 0.006), kidney (HR, 2.49; 95% CI: 1.57–3.95, p < 0.001), and hematologic (HR, 4.06; 95% CI: 2.98–5.53, p < 0.001) malignancies and were significantly lower for the lung (HR: 0.54, CI: 0.34–0.86, p = 0.009). Hematologic malignancies were the site with the highest significant HR. In the post-acute phase after cancer treatment, HRs by cancer site were significantly higher for hematologic malignancies (HR, 2.65; 95% CI: 1.98–3.55; p < 0.001) and significantly lower for the lung (HR: 0.56, CI: 0.38–0.82, p = 0.0027). In the post-acute phase, hematologic malignancy was the only significant risk factor. Moreover, among cancer sites, hematologic malignancy, the nature of which may differ from other cancer sites, was the only significant risk factor identified in both the acute and post-acute phases. In this registry-based retrospective cohort study, the risk of RRT differed substantially by cancer site.
Fig. 3.
Cumulative 1-year incidence of patients who received RRT across all cancer sites. The vertical axis shows the cumulative incidence (%) and the horizontal axis shows the disease duration (months).
Fig. 4.
Cumulative 1-year incidence of patients who received RRT for each cancer site. The vertical axis shows the cumulative incidence (%) and the horizontal axis shows the disease duration (months).
Table 2.
The 1-year cumulative incidence of RRT according to cancer site
| | 1-year cumulative incidence, % |
|---|---|
| All cancers | 0.413 |
| Cancer sites | |
| Colorectal | 0.414 |
| Lung | 0.225 |
| Stomach | 0.398 |
| Liver | 1.012 |
| Kidney | 0.878 |
| Esophagus | 0.685 |
| Pancreas | 0.439 |
| Hematologic malignancies | 2.237 |
The 1-year cumulative incidence of RRT was estimated using the Fine-Gray method, with death treated as a competing event. Cumulative incidence was evaluated from the date of the first cancer treatment.
Table 3.
Multivariable Cox proportional hazards analysis of RRT according to cancer site across all, acute, and post-acute phases
| Cancer sites | All phase | Acute phase | Post-acute phase | |||
|---|---|---|---|---|---|---|
| HR (95% CI) | p value | HR (95% CI) | p value | HR (95% CI) | p value | |
| Colorectal | 1.20 (0.98–1.47) | 0.081 | 1.72 (1.29–2.30) | <0.001 | 0.85 (0.64–1.14) | 0.291 |
| Lung | 0.55 (0.41–0.74) | <0.001 | 0.54 (0.34–0.86) | 0.009 | 0.56 (0.38–0.82) | 0.003 |
| Stomach | 0.97 (0.76–1.23) | 0.802 | 0.91 (0.62–1.33) | 0.617 | 1.02 (0.75–1.38) | 0.911 |
| Liver | 1.23 (0.91–1.67) | 0.181 | 1.80 (1.18–2.74) | 0.006 | 0.88 (0.56–1.37) | 0.564 |
| Kidney | 1.87 (1.35–2.58) | <0.001 | 2.49 (1.57–3.95) | <0.001 | 1.47 (0.93–2.32) | 0.098 |
| Esophagus | 0.89 (0.65–1.23) | 0.494 | 0.77 (0.45–1.32) | 0.337 | 0.98 (0.66–1.46) | 0.937 |
| Pancreas | 0.60 (0.40–0.89) | 0.011 | 0.39 (0.18–0.84) | 0.016 | 0.74 (0.46–1.18) | 0.203 |
| Hematologic malignancies | 3.23 (2.62–3.99) | <0.001 | 4.06 (2.98–5.53) | <0.001 | 2.65 (1.98–3.55) | <0.001 |
Hazard ratios (HRs) and 95% confidence intervals (95% CIs) were estimated using multivariable adjusted Cox proportional hazards models adjusted for age, sex, chronic kidney disease, diabetes mellitus, hypertension, coronary heart disease, and diuretic use. The “other cancer sites” group was used as the reference category. Acute phase renal replacement therapy (RRT) was defined as RRT occurring within 30 days after initiation of cancer treatment, and post-acute phase RRT as RRT occurring more than 30 days after treatment initiation.
Fig. 5.
Forest plots of adjusted hazard ratios and subdistribution hazard ratios for renal replacement therapy (RRT) across all phases according to cancer site. Forest plots showing adjusted hazard ratios derived from multivariable Cox proportional hazards models (a) and subdistribution hazard ratios derived from Fine-Gray competing-risk models (b) for RRT according to cancer site across all phases. Estimates were adjusted for age, sex, comorbidities, and diuretic use. The “other cancer sites” group was used as the reference category. Error bars represent 95% confidence intervals (CIs).
Fig. 7.
Forest plots of adjusted hazard ratios and subdistribution hazard ratios for post-acute phase renal replacement therapy (RRT) according to cancer site. Forest plots showing adjusted hazard ratios derived from multivariable Cox proportional hazards models (a) and subdistribution hazard ratios derived from Fine-Gray competing-risk models (b) for post-acute phase RRT according to cancer site. Post-acute phase RRT was defined as RRT occurring more than 30 days after the initiation of cancer treatment. Estimates were adjusted for age, sex, comorbidities, and diuretic use. The “other cancer sites” group was used as the reference category. Error bars represent 95% confidence intervals (CIs).
Fig. 6.
Forest plots of adjusted hazard ratios and subdistribution hazard ratios for acute phase renal replacement therapy (RRT) according to cancer site. Forest plots showing adjusted hazard ratios derived from multivariable Cox proportional hazards models (a) and subdistribution hazard ratios derived from Fine-Gray competing-risk models (b) for acute phase RRT according to cancer site. Acute phase RRT was defined as RRT occurring within 30 days after initiation of cancer treatment. Estimates were adjusted for age, sex, comorbidities, and diuretic use. The “other cancer sites” group was used as the reference category. Error bars represent 95% confidence intervals (CIs).
The results of the stratified analyses are shown in Table 4, and the corresponding analyses, including sHRs, are presented in online supplementary Table S4A and B. In the subgroup of patients who underwent surgery or laparoscopic surgery, colorectal cancer (HR 1.41, 95% CI: 1.03–1.93), particularly during the acute phase (HR 2.14, 95% CI: 1.42–3.21), was associated with an increased risk of RRT. Esophageal cancer also showed an elevated risk during the acute phase (HR 2.64, 95% CI: 1.23–5.70). In contrast, lung cancer was associated with a lower risk of RRT across all phases. In the chemotherapy subgroup, hematologic malignancies were consistently associated with an increased risk of RRT across all phases (all phase: HR 1.91, 95% CI: 1.43–2.54; acute phase: HR 1.98, 95% CI: 1.35–2.89; post-acute phase: HR 1.80, 95% CI: 1.16–2.79). In contrast, colorectal, lung, stomach, and pancreatic cancers were associated with lower risks of RRT in several phases.
Table 4.
Stratified multivariable Cox proportional hazards analyses of RRT according to cancer site by treatment modality
| Cancer sites | All phase | Acute phase | Post-acute phase | |||
|---|---|---|---|---|---|---|
| HR (95% CI) | p value | HR (95% CI) | p value | HR (95% CI) | p value | |
| Surgery and laparoscopic surgery | ||||||
| Colorectal | 1.41 (1.03–1.93) | 0.031 | 2.14 (1.42–3.21) | <0.001 | 0.72 (0.43–1.21) | 0.216 |
| Lung | 0.56 (0.35–0.91) | 0.019 | 0.43 (0.20–0.92) | 0.030 | 0.68 (0.36–1.28) | 0.237 |
| Stomach | 1.31 (0.90–1.92) | 0.163 | 1.69 (1.02–2.78) | 0.040 | 0.95 (0.52–1.73) | 0.860 |
| Liver | 1.38 (0.89–2.14) | 0.152 | 1.88 (1.06–3.32) | 0.030 | 0.92 (0.45–1.86) | 0.806 |
| Kidney | 1.48 (0.94–2.34) | 0.091 | 1.85 (1.01–3.37) | 0.045 | 1.11 (0.55–2.24) | 0.779 |
| Esophagus | 1.85 (0.98–3.49) | 0.058 | 2.64 (1.23–5.70) | 0.013 | 1.06 (0.32–3.44) | 0.928 |
| Pancreas | 1.41 (0.73–2.72) | 0.311 | 1.46 (0.57–3.72) | 0.427 | 1.35 (0.53–3.44) | 0.529 |
| Chemotherapy | ||||||
| Colorectal | 0.37 (0.15–0.92) | 0.032 | 0.27 (0.06–1.09) | 0.066 | 0.51 (0.16–1.68) | 0.270 |
| Lung | 0.45 (0.27–0.73) | 0.001 | 0.51 (0.28–0.95) | 0.035 | 0.36 (0.16–0.82) | 0.014 |
| Stomach | 0.41 (0.22–0.77) | 0.005 | 0.19 (0.06–0.62) | 0.006 | 0.71 (0.33–1.53) | 0.382 |
| Liver | 0.46 (0.15–1.47) | 0.191 | 0.26 (0.04–1.86) | 0.177 | 0.78 (0.19–3.28) | 0.739 |
| Kidney | 1.86 (0.93–3.72) | 0.078 | 1.39 (0.50–3.87) | 0.527 | 2.49 (0.97–6.38) | 0.058 |
| Esophagus | 0.95 (0.45–1.97) | 0.882 | 0.77 (0.28–2.16) | 0.621 | 1.23 (0.43–3.50) | 0.694 |
| Pancreas | 0.26 (0.13–0.53) | <0.001 | 0.05 (0.01–0.36) | 0.003 | 0.59 (0.27–1.27) | 0.174 |
| Hematologic malignancies | 1.91 (1.43–2.54) | <0.001 | 1.98 (1.35–2.89) | <0.001 | 1.80 (1.16–2.79) | 0.008 |
Hazard ratios (HRs) were estimated using multivariable Cox proportional hazards models adjusted for age, sex, chronic kidney disease, diabetes mellitus, hypertension, coronary heart disease, and diuretic use. Acute phase renal replacement therapy (RRT) was defined as RRT occurring within 30 days after initiation of cancer treatment, and post-acute phase RRT as RRT occurring more than 30 days after treatment initiation. For the chemotherapy subgroup, the “other cancer sites” group was used as the reference category. For the surgery/laparoscopic surgery subgroup, patients with hematologic malignancies were included in the “other cancer sites” group because no events occurred in this subgroup, resulting in model instability. The “other cancer sites” group was used as the reference category.
CI, confidence interval; HR, hazard ratio.
Discussion
In this study, we utilized a comprehensive database integrating the HBCR and DPC system and obtained the following key findings: (1) the 1-year cumulative incidence of patients requiring RRT after cancer treatment was 0.413%; (2) in the acute phase after cancer treatment, cancers such as colorectal, liver, kidney, and hematologic malignancies were identified as important risk factors for the initiation of RRT; and (3) in the post-acute phase, hematologic malignancy was the only significant risk factor for RRT, but the risk was greater in the acute phase.
Clinically meaningful differences were also observed across the groups. Patients requiring RRT were older (median age: 74.0 years in the acute phase and 72.0 years in the post-acute phase vs. 71.0 years in the non-RRT group; p < 0.001) and more likely to be male (72.5% and 72.9% vs. 53.8%; p < 0.001). A higher burden of CKD and comorbidities in the RRT groups suggests that baseline renal vulnerability contributes to severe AKI requiring RRT. Differences in treatment modalities and medication use further indicate that treatment-related factors and underlying clinical status may influence both the development of AKI and decisions regarding RRT initiation.
Although previous studies have investigated the incidence of AKI in cancer patients, data specific to the patient population requiring RRT after cancer treatment remain lacking [1, 2]. This is mainly due to the low incidence, as evidenced by the 1-year cumulative incidence observed in this study (0.413%), which is consistent with the results of previous studies. The rarity of such events may be one reason why studies examining the correlation between specific cancer sites and the subsequent need for RRT are limited. To our knowledge, this study is the first comprehensive analysis to compare the risk of RRT among different cancer sites, specifically in cancer patients, using an extensive database integrating an HBCR and DPC.
Reports investigating the association between AKI and cancer site in patients with cancer alone are rare; Kitchlu et al. [25] calculated adjusted hazard ratios (aHRs) for the incidence of AKI by cancer site in cancer patients who received systemic therapy, with breast cancer as the reference group. Their analysis showed that multiple myeloma patients had the highest aHR (aHR, 4.30; 95% CI: 3.83–4.82), followed by patients with bladder cancer (aHR, 3.69; 95% CI: 3.28–4.16) and cervical cancer (aHR = 3.47; 95% CI: 2.90–4.14). Leukemia patients had the fourth highest hazard ratio (aHR, 3.39; 95% CI: 3.02–3.80), indicating that hematologic malignancies may confer a greater risk of AKI in comparison to other cancer sites.
Our study similarly evaluated HRs for the incidence of RRT by cancer site in a population restricted to cancer patients. Among patients with hematologic malignancies, the HR for RRT was 4.06 in the acute phase and 2.65 in the post-acute phase. These values parallel the elevated aHRs for AKI and suggest a comparatively greater likelihood of requiring RRT relative to other cancer sites. For kidney cancer, the acute phase HR for RRT was 2.48, which is comparable to the AKI HR of 2.43 reported by Kitchlu et al. [25]. Our results showed a similar pattern. To our knowledge, our study is the first to report on the incidence of RRT and cancer site in a population restricted to cancer patients, and our findings support previous studies that focused on the incidence of AKI and cancer site in a population restricted to cancer patients.
RRT in Cancer Patients
A recent population-based retrospective cohort study conducted in Korea by Kim et al. [26] examined the risk of renal failure among cancer patients and identified a significant association between cancer site and the initiation of RRT. This investigation was notable as it was the only study that utilized a large-scale population database. The findings indicated that multiple myeloma, leukemia, lymphoma, kidney cancer, ovarian cancer, and liver cancer were significantly correlated with an increased risk of requiring RRT, with multiple myeloma presenting the highest risk, irrespective of age or sex. In our study, hematologic malignancies, colorectal cancer, kidney cancer, and liver cancer were associated with an increased risk of RRT. However, differences were observed in the aHRs. These differences may be partially attributed to variations in the population databases employed. Prior studies have relied on population-based data, facilitating comparisons between cancer and non-cancer patients. In contrast, this study utilized a hospital-based database that included only patients who had undergone cancer treatment and compared the HRs for the incidence of RRT by cancer site. When comparing cancer patients with non-cancer patients, accounting for the effects of cancer treatment presents a substantial challenge. The comparison group in our study consisted only of individuals who had received cancer treatment. Although the analysis was not stratified by the type of cancer treatment, the large sample size may have mitigated the variability due to treatment differences, potentially leading to a uniform effect. Furthermore, given the limited applicability of comparing cancer and non-cancer patients in routine clinical practice, our findings may better represent the real-world clinical circumstances and include many aspects that are likely to be applicable to cancer treatment in clinical settings.
In our study, the earliest recorded date of any cancer therapy was designated as the baseline, and a survival analysis was conducted to evaluate the time to the initiation of RRT. Consequently, we examined both the acute and post-acute phases following the initiation of cancer treatment. These findings provide valuable insights into clinical practice. Notably, the HRs for the incidence of RRT differed between the acute and post-acute phases according to the cancer site. For colorectal, kidney, and liver cancers, the risk of requiring RRT was significantly elevated during the acute phase only, and no statistically significant differences were observed thereafter. Nevertheless, because the earliest treatment date was used as the baseline, some post-acute RRT events may have occurred in temporal proximity to subsequent systemic therapy or staged surgical interventions. An examination of the cumulative incidence of RRT (Fig. 3, 4) revealed a steep increase within approximately 1 month after the initiation of cancer treatment, followed by a steady, linear rise. Similar patterns have been documented in previous studies. In such studies, the cumulative incidence of postoperative complications after liver resection was evaluated, and it was found that complications increased rapidly until approximately the 20th postoperative day and then reached a plateau around the 30th day [27]. Thus, the observation that the HR for RRT was significantly increased only during the acute phase in patients with colorectal, kidney, and liver cancers may partly reflect postoperative AKI and other treatment-related complications occurring shortly after the initiation of cancer treatment. However, because detailed laboratory data, cancer stage, performance status, and mechanism-specific causes of AKI were not available, alternative explanations cannot be excluded. Differences in treatment composition, referral patterns, institutional practices regarding initiation of RRT, and residual confounding may also have contributed to the observed phase-specific associations.
The observed temporal patterns of RRT risk may partly reflect differences in treatment-related nephrotoxicity across cancer types. Platinum-based chemotherapy, which is commonly used for gastrointestinal, lung, and esophageal cancers, is a well-established cause of AKI [28]. In addition, immune checkpoint inhibitors, which are increasingly used in lung, liver, kidney, and gastrointestinal cancers, may induce delayed immune-mediated kidney injury [29]. In contrast, patients with hematologic malignancies are more likely to receive intensive chemotherapy, including high-dose methotrexate and hematopoietic stem cell transplantation, and are at particularly high risk for tumor lysis syndrome, severe infections, and sepsis. These mechanisms may partly explain why hematologic malignancies showed persistently elevated RRT risk during both the acute and post-acute phases in the present study.
The KDIGO guidelines state that, in the absence of emergency indications such as hyperkalemia, pulmonary edema, or severe metabolic acidosis, the initiation of RRT should be deferred until AKI has persisted for more than 72 h [30]. Based on the findings obtained in this study, it was revealed that patients with colon, liver, kidney, and hematologic malignancies had significantly higher HRs when undergoing RRT relative to patients with other types of cancer. When AKI occurs in these cancer types, closer monitoring and careful clinical judgment are needed to determine whether RRT should be initiated. Especially in the acute phase immediately after the initiation of cancer treatment, it is necessary to carefully deal with changes in renal function. Furthermore, in hematologic malignancies, it is important to manage AKI from a long-term perspective.
Hematologic Malignancies
The rate of RRT in patients with hematologic malignancies was significantly higher in both the acute and post-acute settings. This distinct pattern, different from that observed in other cancer sites, highlights the inherent tendency of hematologic malignancies to induce severe renal complications that require RRT. Previous extensive studies have consistently demonstrated a higher risk of AKI in cohorts of patients with hematologic malignancies. Furthermore, the present study revealed that RRT is more likely to be initiated because of severe AKI. Notably, the HR for RRT was higher in the acute phase (HR, 4.06) than in the post-acute phase (HR, 2.65), confirming that the time immediately after therapeutic intervention is an extremely high-risk period for RRT.
Patients with hematologic malignancies frequently undergo intensive treatment regimens, including high-dose methotrexate-based chemotherapy and hematopoietic stem cell transplantation, both of which are associated with substantial nephrotoxicity and AKI requiring RRT. High-dose methotrexate is a well-established cause of AKI due to crystal-induced tubular obstruction and delayed drug clearance, particularly in patients with hematologic malignancies [31]. In particular, AKI is a well-recognized complication after hematopoietic stem cell transplantation and has been associated with increased mortality and poor clinical outcomes [31, 32]. These treatment-related factors may partly explain the persistently elevated HRs observed in hematologic malignancies across both the acute and post-acute phases. In addition, disease-related immunosuppression and treatment-induced neutropenia increase susceptibility to severe infections and sepsis, both of which are major causes of AKI requiring RRT in patients with hematologic malignancies [33].
TLS is a major contributor to the increased risk of RRT in the acute phase after cancer treatment in patients with hematopoietic malignancies [34, 35]. TLS usually develops within 48–72 h after the start of chemotherapy and is characterized by a series of metabolic disturbances, including hyperuricemia, hyperkalemia, hyperphosphatemia, hypocalcemia, and metabolic acidosis. In parallel with advances in anticancer agents, prophylactic strategies have been developed to reduce nephrotoxicity associated with TLS.
Despite existing preventive strategies, our results suggest that the incidence of AKI requiring RRT remains disproportionately higher immediately after the start of cancer treatment than in later periods. This study underscores that continuous monitoring and optimization of TLS prophylaxis are essential to further reduce the risk of acute renal complications in this vulnerable patient population.
This study was based on a database of approximately 230,000 cancer patients collected over a relatively short period of time (3 years) for cases with low incidence and difficult data collection. Thus, it may accurately reflect contemporary clinical practice.
The study population was derived from 66 hospitals in Osaka, Japan, within a uniform healthcare system setting. Therefore, the findings are likely generalizable to cancer patients treated in similar clinical settings in Japan. However, caution is warranted when extrapolating these results to other countries with different healthcare systems, patient characteristics, or cancer treatment practices.
Prevention
These findings may help identify patients who require intensified renal monitoring after the initiation of cancer treatment. In the overall cohort, colorectal, liver, kidney, and hematologic malignancies were associated with higher acute phase RRT risk, suggesting the need for early monitoring of serum creatinine, urine output, volume status, and avoidance of nephrotoxic agents during the first 30 days. Among surgical patients, the elevated acute phase risk observed in colorectal, liver, kidney, stomach, and esophageal cancers highlights the importance of perioperative kidney-protective strategies, including hemodynamic optimization, adequate hydration, and careful use of nephrotoxic drugs. In contrast, hematologic malignancies showed persistently elevated risk in both the acute and post-acute phases, particularly among patients receiving chemotherapy, suggesting the need for prolonged monitoring and the prevention of tumor lysis syndrome, infection-related AKI, and chemotherapy-associated nephrotoxicity.
Because this study was based on a Japanese cancer patient population, caution is warranted when generalizing the findings to other populations. Ethnic and regional differences in nephron number, body size, baseline kidney function, prevalence of comorbidities, distribution of cancer types, treatment strategies, and criteria for initiating dialysis may have influenced the absolute incidence of RRT [36, 37]. Previous studies have suggested that nephron endowment and susceptibility to kidney disease may vary across ethnic groups and geographic regions, potentially affecting vulnerability to AKI and progression to severe kidney injury requiring RRT [38]. Nevertheless, the temporal pattern of increased RRT risk shortly after the initiation of treatment and the relative differences observed across cancer sites may have broader clinical relevance beyond Japan, although the absolute incidence of RRT may vary among populations.
The biological mechanisms underlying cancer-specific differences in RRT risk remain incompletely understood. Metabolomic approaches may provide further insights into the metabolic and inflammatory pathways associated with kidney injury in patients with cancer. Previous studies have suggested that metabolomics can identify biomarkers and metabolic signatures associated with kidney dysfunction and the progression of renal disease [39]. Such approaches may help clarify why certain malignancies, particularly hematologic malignancies, are associated with a markedly elevated risk of RRT following cancer treatment.
Limitations
This study had several limitations. First, sex, age, comorbidities, and medication history were extracted from the database as risk factors for the incidence of RRT and used as adjustment factors in the Cox proportional hazards model. However, there may be other factors that should be adjusted. Second, the database used in this study did not include blood biochemistry test data. The severity of AKI is often diagnosed using the KDIGO [30] or RIFLE [40] criteria, and simple comparisons with clinical studies using these diagnostic criteria may not be appropriate. Because laboratory measurements were not available, residual confounding related to baseline kidney function and AKI severity could not be fully addressed, despite adjustment for CKD and exclusion of patients with CKD stage 5. Third, prophylactic RRT before the start of cancer treatment was not included in the primary endpoint, which may have led to underestimation. The indications for initiating RRT may vary across hospitals, clinical departments, and cancer types, which might have introduced bias. In addition, the database did not contain hospital-level identifiers and information regarding the site of RRT initiation (e.g., intensive care unit versus general ward). We were unable to evaluate inter-hospital variation in RRT rates or perform analyses incorporating hospital as a random effect or stratification factor. Therefore, differences in institutional practices and thresholds for the initiation of RRT may have contributed to the observed variation in the incidence of RRT. Patients with hematologic malignancies are at higher risk of AKI due to conditions such as tumor lysis syndrome and may therefore be more likely to receive RRT than those with solid tumors. In older patients with severe AKI and multiple organ failure, RRT may be withheld because of poor overall prognosis. In contrast, patients with hematologic malignancies tend to be younger and may be considered more likely to recover from AKI, potentially leading to more aggressive use of RRT. All participating hospitals in this registry had intensive care unit capability or equivalent resources, enabling the provision of RRT for severe AKI. However, nephrology service structures and emergency care systems were not standardized across institutions, which may have introduced inter-institutional variability. In addition, because RRT events were identified based on administrative DPC codes used for reimbursement, misclassification due to coding errors or omissions is possible. The use of administrative coding data may underestimate the prevalence of comorbidities because only coded diagnoses could be captured.
Conclusion
The risk of RRT after cancer treatment differed according to cancer site. Colorectal, liver, kidney, and hematologic malignancies were associated with increased acute phase RRT risk, whereas hematologic malignancies remained at increased risk beyond 30 days. These findings may support risk stratification and kidney-protective strategies in patients undergoing cancer treatment.
Acknowledgments
We are deeply grateful to Dr. Nariaki Matsuura, President of the Osaka International Cancer Institute, for his invaluable support, without which this research would not have been possible. We also wish to express our sincere appreciation to Mizuki Shimadzu Kato, Toshitaka Morishima, and Isao Miyashiro of the Cancer Control Center at the Osaka International Cancer Institute for providing the essential data that made this study possible. The authors thank ThinkSCIENCE K.K. for English language editing. ChatGPT (OpenAI) was used solely to refine the language and assist with formatting. The authors take full responsibility for the scientific content of this paper.
Statement of Ethics
This retrospective observational study was approved by the Institutional Review Board of the Osaka International Cancer Institute (Approval No. 21065–6). As anonymized data from a population-based cancer registry were used, the requirement for individual informed consent was waived in accordance with institutional and national ethical guidelines.
Conflict of Interest Statement
The authors have no conflicts of interest to declare.
Funding Sources
This study was supported by an internal research grant from the Osaka International Cancer Institute. The funding body had no role in the study design, data collection, analysis, interpretation, or writing of the manuscript.
Author Contributions
E.U. served as the corresponding author, conceived and designed the study, performed data cleaning and statistical analyses, and drafted the manuscript. H.K. contributed to the study conception and provided methodological and statistical advice. M.K., as an intensive care physician, contributed to the selection of comorbidity variables used in the analysis and revised the manuscript for important intellectual content. All authors reviewed and approved the final version of the manuscript and agree to be accountable for all aspects of the work.
Funding Statement
This study was supported by an internal research grant from the Osaka International Cancer Institute. The funding body had no role in the study design, data collection, analysis, interpretation, or writing of the manuscript.
Data Availability Statement
The data that support the findings of this study were provided by the Cancer Control Center, Osaka International Cancer Institute. These data are not publicly available due to institutional and privacy restrictions but may be available from the corresponding author upon reasonable request and with permission from the data provider.
Supplementary Material.
Supplementary Material.
Supplementary Material.
References
- 1. Christiansen CF, Johansen MB, Langeberg WJ, Fryzek JP, Sørensen HT. Incidence of acute kidney injury in cancer patients: a Danish population-based cohort study. Eur J Intern Med. 2011;22(4):399–406. [DOI] [PubMed] [Google Scholar]
- 2. Cheng Y, Nie S, Li L, Li Y, Liu D, Xiong M, et al. Epidemiology and outcomes of acute kidney injury in hospitalized cancer patients in China. Int J Cancer. 2019;144(11):2644–50. [DOI] [PubMed] [Google Scholar]
- 3. Lyrio R, Rocha BRA, Corrêa A, Mascarenhas MGS, Santos FL, Maia RDH, et al. Chemotherapy-induced acute kidney injury: epidemiology, pathophysiology, and therapeutic approaches. Front Nephrol. 2024;4:1436896. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Markowitz GS, Fine PL, Stack JI, Kunis CL, Radhakrishnan J, Palecki W, et al. Toxic acute tubular necrosis following treatment with zoledronate (Zometa). Kidney Int. 2003;64(1):281–9. [DOI] [PubMed] [Google Scholar]
- 5. Kunlayawutipong T, Rattanathammethee T, Punnachet T, Hantrakun N, Piriyakhuntorn P, Hantrakool S, et al. Prevalence and risk factors for hyperuricemia and hyperuricosuria in patients with hematologic malignancies. Front Med. 2024;11:1343000. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Abdel-Nabey M, Chaba A, Serre J, Lengliné E, Azoulay E, Darmon M, et al. Tumor lysis syndrome, acute kidney injury and disease-free survival in critically ill patients requiring urgent chemotherapy. Ann Intensive Care. 2022;12(1):15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Peerapornratana S, Manrique-Caballero CL, Gómez H, Kellum JA. Acute kidney injury from sepsis: current concepts, epidemiology, pathophysiology, prevention and treatment. Kidney Int. 2019;96(5):1083–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Williams MD, Braun LA, Cooper LM, Johnston J, Weiss RV, Qualy RL, et al. Hospitalized cancer patients with severe sepsis: analysis of incidence, mortality, and associated costs of care. Crit Care. 2004;8(5):R291–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Aspelin P, Aubry P, Fransson SG, Strasser R, Willenbrock R, Berg KJ, et al. Nephrotoxic effects in high-risk patients undergoing angiography. N Engl J Med. 2003;348(6):491–9. [DOI] [PubMed] [Google Scholar]
- 10. Waybill MM, Waybill PN. Contrast media-induced nephrotoxicity: identification of patients at risk and algorithms for prevention. J Vasc Interv Radiol. 2001;12(1):3–9. [DOI] [PubMed] [Google Scholar]
- 11. Salahudeen AK, Doshi SM, Pawar T, Nowshad G, Lahoti A, Shah P. Incidence rate, clinical correlates, and outcomes of AKI in patients admitted to a comprehensive cancer center. Clin J Am Soc Nephrol. 2013;8(3):347–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Benoit DD, Hoste EA. Acute kidney injury in critically ill patients with cancer. Crit Care Clin. 2010;26(1):151–79. [DOI] [PubMed] [Google Scholar]
- 13. Calças Marques R, Reis M, Pimenta G, Sala I, Chuva T, Coelho I, et al. Severe acute kidney injury in hospitalized cancer patients: epidemiology and predictive model of renal replacement therapy and in-hospital mortality. Cancers. 2024;16(3):561. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Xiong M, Wang L, Su L, Luo W, Li Y, Li L, et al. Acute kidney injury among hospitalized children with cancer. Pediatr Nephrol. 2021;36(1):171–9. [DOI] [PubMed] [Google Scholar]
- 15. Ng AP, Chervu N, Branche C, Bakhtiyar SS, Marzban M, Toste PA, et al. National clinical and financial outcomes associated with acute kidney injury following esophagectomy for cancer. PLoS One. 2024;19(3):e0300876. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Bagshaw SM, Wald R, Adhikari NKJ, Bellomo R, da Costa BR, Dreyfuss D, et al. Timing of initiation of renal-replacement therapy in acute kidney injury. N Engl J Med. 2020;383(3):240–51. [DOI] [PubMed] [Google Scholar]
- 17. Kanzaki G, Puelles VG, Cullen-McEwen LA, Hoy WE, Okabayashi Y, Tsuboi N, et al. New insights on glomerular hyperfiltration: a Japanese autopsy study. JCI Insight. 2017;2(19):e94334. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Keller G, Zimmer G, Mall G, Ritz E, Amann K. Nephron number in patients with primary hypertension. N Engl J Med. 2003;348(2):101–8. [DOI] [PubMed] [Google Scholar]
- 19. Kanzaki G, Tsuboi N, Shimizu A, Yokoo T. Human nephron number, hypertension, and renal pathology. Anat Rec. 2020;303(10):2537–43. [DOI] [PubMed] [Google Scholar]
- 20. Hughson MD, Gobe GC, Hoy WE, Manning RD Jr, Douglas-Denton R, Bertram JF. Associations of glomerular number and birth weight with clinicopathological features of African Americans and whites. Am J Kidney Dis. 2008;52(1):18–28. [DOI] [PubMed] [Google Scholar]
- 21. Quan H, Sundararajan V, Halfon P, Fong A, Burnand B, Luthi JC, et al. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care. 2005;43(11):1130–9. [DOI] [PubMed] [Google Scholar]
- 22. Munch PV, Nørgaard M, Jensen SK, Birn H, Schmidt H, Christiansen CF. Risk of and mortality after acute kidney injury following cancer treatment: a cohort study. Cancer Med. 2025;14(3):e70646. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Jiang W, Song L, Gong W, Li X, Shi K, Li L, et al. The role of phospholipid transfer protein in sepsis-associated acute kidney injury. Crit Care. 2025;29(1):33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Godi I, Feltracco P, Lorenzoni G, Antonelli A, Salvador R, Gregori D, et al. Incidence, risk factors, and consequences of acute kidney injury in patients undergoing esophageal cancer surgery: a historical cohort. Kidney Dial. 2024;4(2):93–104. [Google Scholar]
- 25. Kitchlu A, McArthur E, Amir E, Booth CM, Sutradhar R, Majeed H, et al. Acute kidney injury in patients receiving systemic treatment for cancer: a population-based cohort study. J Natl Cancer Inst. 2019;111(7):727–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Kim CS, Kim B, Suh SH, Oh TR, Kim M, Choi HS, et al. Risk of kidney failure in patients with cancer: a South Korean population-based cohort study. Am J Kidney Dis. 2022;79(4):507–17.e1. [DOI] [PubMed] [Google Scholar]
- 27. Watanabe G, Kawaguchi Y, Ichida A, Ishizawa T, Akamatsu N, Kaneko J, et al. Understanding conditional cumulative incidence of complications following liver resection to optimize hospital stay. HPB Oxf. 2022;24(2):226–33. [DOI] [PubMed] [Google Scholar]
- 28. Pabla N, Dong Z. Cisplatin nephrotoxicity: mechanisms and renoprotective strategies. Kidney Int. 2008;73(9):994–1007. [DOI] [PubMed] [Google Scholar]
- 29. Cortazar FB, Marrone KA, Troxell ML, Ralto KM, Hoenig MP, Brahmer JR, et al. Clinicopathological features of acute kidney injury associated with immune checkpoint inhibitors. Kidney Int. 2016;90(3):638–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Khwaja A. KDIGO clinical practice guidelines for acute kidney injury. Nephron Clin Pract. 2012;120(4):c179–84. [DOI] [PubMed] [Google Scholar]
- 31. Wiczer T, Dotson E, Tuten A, Phillips G, Maddocks K. Evaluation of incidence and risk factors for high-dose methotrexate-induced nephrotoxicity. J Oncol Pharm Pract. 2016;22(3):430–6. [DOI] [PubMed] [Google Scholar]
- 32. Raina R, Herrera N, Krishnappa V, Sethi SK, Deep A, Kao W-M, et al. Hematopoietic stem cell transplantation and acute kidney injury in children: a comprehensive review. Pediatr Transpl. 2017;21(4):e12935. [DOI] [PubMed] [Google Scholar]
- 33. Rosner MH, Perazella MA. Acute kidney injury in the patient with cancer. Kidney Res Clin Pract. 2019;38(3):295–308. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Matuszkiewicz-Rowinska J, Malyszko J. Prevention and treatment of tumor lysis syndrome in the era of onco-nephrology progress. Kidney Blood Press Res. 2020;45(5):645–60. [DOI] [PubMed] [Google Scholar]
- 35. Howard SC, Jones DP, Pui CH. The tumor lysis syndrome. N Engl J Med. 2011;364(19):1844–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Hoy WE, Hughson MD, Singh GR, Douglas-Denton R, Bertram JF. Reduced nephron number and glomerulomegaly in Australian Aborigines: a group at high risk for renal disease and hypertension. Kidney Int. 2006;70(1):104–10. [DOI] [PubMed] [Google Scholar]
- 37. Stewart JH, McCredie MRE, Williams SM. Geographic, ethnic, age-related and temporal variation in the incidence of end-stage renal disease in Europe, Canada and the Asia-Pacific region, 1998-2002. Nephrol Dial Transpl. 2006;21(8):2178–83. [DOI] [PubMed] [Google Scholar]
- 38. Luyckx VA, Shukha K, Brenner BM. Low nephron number and its clinical consequences. Rambam Maimonides Med J. 2011;2(4):e0061. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Hocher B, Adamski J. Metabolomics for clinical use and research in chronic kidney disease. Nat Rev Nephrol. 2017;13(5):269–84. [DOI] [PubMed] [Google Scholar]
- 40. Bellomo R, Ronco C, Kellum JA, Mehta RL, Palevsky P. Acute renal failure - definition, outcome measures, animal models, fluid therapy and information technology needs: the Second International Consensus Conference of the Acute Dialysis Quality Initiative (ADQI) Group. Crit Care. 2004;8(4):R204–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study were provided by the Cancer Control Center, Osaka International Cancer Institute. These data are not publicly available due to institutional and privacy restrictions but may be available from the corresponding author upon reasonable request and with permission from the data provider.