2024 Volume 47 Issue 2 Pages 518-526
To investigate the risk of acute kidney injury (AKI) in patients with cancer following the initiation of proton pump inhibitors (PPIs) and potassium-competitive acid blocker (PCAB), considering sex and anti-cancer drug use. We conducted a self-controlled case-series study using the Japan Medical Data Center claims data from 12422 patients with cancer who were prescribed PPIs or PCAB between January 2017 and December 2019. Considering the timing of PPI or PCAB, control period (days −120 to −1), risk period 1 (days 0 to +30), and risk period 2 (days +31 to +365) were defined. To assess the incidence rate ratio (IRR) and 95% confidence interval (CI) as the risk ratio, we adjusted for anti-cancer drugs to assess the risk of AKI. Additionally, we also examined sex differences to identify the risk of AKI. AKI was observed in risk period 1 [2.05 (1.12–3.72), p = 0.0192], but a slight reduction was noted in risk period 2 [0.60 (0.36–1.00), p = 0.0481]. A sex-specific increase in the risk of AKI was observed only in males during risk period 1 [2.18 (1.10–4.32), p = 0.0260], with a reduction in risk period 2 [0.48 (0.26–0.89), p = 0.0200]. We identified an increased risk of AKI in patients with cancer starting PPIs or PCAB particularly in males within 30 d after PPI or PCAB initiation, emphasizing the need for vigilant monitoring and management of AKI in this patient population.
Proton pump inhibitors (PPIs) and potassium-competitive acid blocker (PCAB) are extensively prescribed for patients with gastric ulcers and gastroesophageal reflux disease (GERD), and for the prevention of gastric ulcers in patients taking low-dose aspirin. Although PPIs are commonly administered to patients, it is important to note that prolonged exposure to these medications can lead to various toxicities, such as dementia, pneumonia, enteric infection, type 2 diabetes, and osteoporosis.1–5) Furthermore, in certain advanced countries, such as the United States and Sweden, PPIs are available as over-the-counter drugs.6–8) The inappropriate use of PPIs and PCAB, including their long-term use, has become a global concern.9) This concern is particularly pronounced in the geriatric population, where the rate of inappropriate usage, including long-term usage, has been reported to be approximately 36%.10)
Several studies have highlighted the renal toxicity associated with PPIs, particularly concerning age-associated risk factors. Additionally, a higher incidence of renal toxicity was reported in individuals using PPIs compared to those using histamine-2 receptor antagonists (H2 blockers).11–13) One contributing factor to this discrepancy is that H2 blockers require dose adjustments in patients with chronic kidney disease (CKD), whereas such adjustments are not required for PPIs and PCAB. PPI exposure has been linked to an increased likelihood of postoperative dialysis and prolonged Intensive Care Unit (ICU) stay.14) Conversely, instances of acute kidney injury (AKI) have also been reported to be more prevalent in patients using PPIs, with a rate of 36.4 per 1000 patients compared to 3.54 per 1000 patients not using PPIs (p < 0.0001).15) The precise mechanisms underlying the onset of AKI in PPI users are not entirely clear; however, two potential mechanisms have been proposed. First, PPI use may decrease the glomerular filtration rate. Second, low magnesium levels induced by PPIs could also contribute to AKI onset.16,17) Notably, the use of PPIs and PCAB began in the late 1990s and 2015, respectively. Recent reports indicate that PCAB, as they belong to the same drug class, may potentially exhibit toxicities similar to those of PPIs.18) This is noteworthy because PCAB have not been clinically used for as long as PPIs.19)
The risk of deteriorating CKD or increasing mortality in patients with CKD associated with long-term PPI use is well known.14,20) In particular, 12% of patients with cancer have CKD stages G3–G5.21) In patients with cancer, the high prevalence of renal impairment can be attributed to a combination of factors, including the impact of the cancer itself, associated physical discomfort, and the effects of treatments such as medication therapies.22,23) On the other hand, a high prevalence of AKI was reported in patients with cancer, at approximately 12% as compared to 5–8% for patients without cancer.24–26) In addition, AKI in patients with cancer is associated with longer hospital stays, higher costs, and higher mortality rates.25) Thus, the onset of AKI in patients with cancer is a critical issue requiring appropriate care. In recent years, there have been many reports on the relationship between CKD and AKI. Patients with CKD are reported to be at a higher risk of developing AKI, with a 30-fold and 40-fold increased risk, respectively, in patients with estimated glomerular filtration rate (eGFR) levels of 15–29 (CKD Stage 4) and less than 15 mL/min/1.73 m2 (CKD Stage 5), compared to individuals with normal eGFR levels.27) Additionally, patients with AKI have been reported to have an 8.8-fold increased risk of developing CKD.28)
Previously, we reported that the combination of PPIs and nonsteroidal anti-inflammatory drugs (NSAIDs), cephalosporin, or fluoroquinolones is associated with a higher risk for AKI compared to PPI use alone.29) This suggests a synergistic effect contributing to AKI onset in PPI users. However, the onset of AKI in patients with cancer after initiating PPI or PCAB therapy has not been studied. Therefore, we investigated the onset of AKI after initiating PPI or PCAB treatment in patients with cancer using large claims data from Japan.
The Japan Medical Data Center (JMDC, Tokyo, Japan) claims database consists of anonymized data of over 10 million people (inpatient, outpatient, and pharmacy claims) aged ≤74 years from over 90% of the hospitals in Japan. The database includes working individuals with social insurance and their families. As of April 2021, the database represented 10% of Japan’s total population.30,31)
Case IdentificationThis retrospective study included approximately 370 thousand individuals who were prescribed PPIs and PCAB, excluding a combination drug for PPI or PCAB, from January 2017 to December 2019, as identified from the JMDC database. We applied two exclusion criteria: (1) individuals without a pre-existing cancer diagnosed prior to PPI or PCAB initiation and (2) those with PPI or PCAB exposure covering less than 50% of the proportion days cover32) until the last date of PPI or PCAB administration or up to 365 d after starting PPI or PCAB treatment (Supplementary Table 2). As a result, 12422 patients were analyzed (Fig. 1).
We adapted a self-controlled case series analysis33–35) of 12422 patients (Fig. 2). Briefly, we divided the study into three sequential periods: the control period (days −120 to −1), risk period 1 (days 0 to +30), and risk period 2 (days +31 to +365), with day 0 defined as the first exposure to PPI or PCAB in each patient (Supplementary Fig. 3). Exposure for the control period was set as the number of days from cancer diagnosis to PPI or PCAB initiation up to 120 d. We defined the control period as days −120 to −1 to include new users of PPI or PCAB and to maximize sample size for analysis within the allowable look-back period for identifying new users of PPI or PCAB from a clinical perspective. Additionally, we needed to observe and assess AKI just before the initiation of PPI or PCAB in each cancer patient in our study settings. Moreover, we set the risk period 1 as 30 d because PPIs or PCAB are typically prescribed for a limited duration of 4–8 weeks as the initial prescription, in accordance with the package contents. Furthermore, we set risk period 2 as up to 365 d to identify patients with long-term PPI or PCAB use from a clinical perspective. The risk period was set until the last PPI or PCAB administration date as >50% drug exposure. The date of PPI or PCAB exposure to the last date of PPI or PCAB exposure was defined as the exposure date for each patient (Supplementary Table 2). AKI (N17) and comorbidities were identified as per the ICD 10 code, and information on the use of anti-cancer drugs was considered in the analysis (Supplementary Table 1). In our study setting, we interpreted N17 as “severe AKI” according to Hwang et al.36) due to the higher sensitivity in severe AKI. Exposure to anti-cancer drugs is treated as binary data, categorized as either having been exposed or not in the control period, risk period 1, and risk period 2. Specifically, if patients were administered anti-cancer drugs during risk period 1, we categorized them as exposed to anti-cancer drugs during risk period 1 for the purpose of our calculations.
The primary endpoint was the incidence of AKI assessed through the incidence rate ratio (IRR). This ratio compared groups with and without anti-cancer drugs in risk period 1 or 2 to the control period. The secondary endpoint was sex differences in AKI incidence, which was evaluated by IRR after dividing the groups with and without anti-cancer drugs in risk period 1 or 2 compared to the control period in patients after starting PPI or PCAB.
Statistical AnalysisThe demographic and clinical characteristics of the patients are presented as numbers and percentages for categorical variables and mean and standard deviation (S.D.) for continuous variables at cohort entry. Incidence rate (IR) was calculated using AKI onset and person-days for exposure in each patient’s control, risk 1, and risk 2 periods, with a 95% confidence interval (95% CI) based on Poisson distribution. The IRR adjusted for anti-cancer drugs during risk periods 1 and 2 with respect to control period was calculated using a regression model based on a Generalized Estimating Equation with a Poisson link function. In the model, we classified patients as either having or not having anti-cancer drug use, irrespective of the study periods.
Data management and analysis were performed using JMP 16 (SAS Institute Inc., Cary, NC, U.S.A.) and R version 4.2.2 (R Foundation for Statistical Computing, Vienna, Austria).
Ethics DeclarationsThe data utilized in this study were obtained from commercially available JMDC databases. The Inc. used in this study anonymized processed information based on Japan’s Personal Information Protection Law, and individual informed, and individual informed consent was not required for its provision and use. In addition, according to the ethical guidelines for clinical research in Japan, research using anonymized processed information is not subject to review by ethical review committees.
This study included 12422 patients with cancer (male/female: 6141/6281, 54.9 ± 10.4 years old) who were new PPI or PCAB users. The PPIs or PCAB administered for patients were esomeprazole, lansoprazole, vonoprazan, omeprazole, and rabeprazole (Table 1). The baseline comorbidities were hypertension (28.3%), dyslipidemia (25.4%), diabetes mellitus (19.0%), and osteoporosis (7.0%). The most frequent cancer types were digestive organs (32.5%), breast (18.2%), neoplasms stated or presumed to be primary, lymphoid, hematopoietic, and related tissue (9.3%), and female genital organs (7.9%). Anti-cancer drugs were administered in the control (12.2%, n = 1519), risk 1 (16.0%, n = 1987), and risk 2 periods (31.9%, n = 2049). The most frequent anti-cancer drugs were antimetabolites (4.9%), platinum compounds (4.3%), plant alkaloids and other natural products (4.1%), monoclonal antibodies, and antibody-drug conjugates (2.9%) (Supplementary Table 2).
| Characteristic | Total | With anti-cancer drugs | Without anti-cancer drugs |
|---|---|---|---|
| n = 12422 | n = 2939 | n = 9483 | |
| Age, year | |||
| Average [S.D.] | 54.9 ± 10.4 | 54.0 ± 10.4 | 54.9 ± 10.6 |
| The number of patients, % | |||
| Female | 6281 (50.6%) | 1384 (47.1%) | 6281 (50.8%) |
| Male | 6141 (49.4%) | 1555 (52.9%) | 6141 (49.6%) |
| Comorbidity [ICD code], % | |||
| Hypertension [I10–I15] | 3511 (28.3%) | 802 (27.2%) | 3511 (28.5%) |
| Dyslipidemia [E78] | 3155 (25.4%) | 631 (21.5%) | 3155 (25.6%) |
| Diabetes mellitus [E10–E14] | 2360 (19.0%) | 570 (19.4%) | 2360 (19.2%) |
| Osteoporosis [M80–M82] | 869 (7.0%) | 137 (4.7%) | 871 (7.0%) |
| Chronic kidney disease [N18] | 242 (1.9%) | 49 (1.7%) | 244 (1.9%) |
| PPI and PCAB [ATC code], number of patients, % | |||
| Esomeprazole [A02BC05] | 3390 (27.3%) | 853 (29.0%) | 2537 (26.8%) |
| Lansoprazole [A02BC03] | 3194 (25.7%) | 974 (33.1%) | 2220 (23.4%) |
| Vonoprazan [A02BC08] | 2868 (23.1%) | 485 (16.5%) | 2383 (25.1%) |
| Omeprazole [A02BC01] | 1541 (12.4%) | 335 (11.4%) | 1206 (12.7%) |
| Rabeprazole [A02BC04] | 1429 (11.5%) | 292 (9.9%) | 1137 (12.0%) |
| Cancer types [ICD code], % | |||
| Digestive organs [C15–C26] | 4033 (32.5%) | 818 (27.8%) | 3215 (33.9%) |
| Breast [C50] | 2260 (18.2%) | 429 (14.6%) | 1831 (19.3%) |
| Neoplasms, stated or presumed to be primary, of lymphoid, hematopoietic, and related tissue [C81–C96] | 1154 (9.3%) | 563 (19.2%) | 591 (6.2%) |
| Female genital organs [C51–C58] | 978 (7.9%) | 200 (6.8%) | 778 (8.2%) |
| Respiratory and intrathoracic organs [C30–C39] | 961 (7.7%) | 317 (10.8%) | 644 (6.8%) |
| Male genital organs [C60–C63] | 671 (5.4%) | 64 (2.2%) | 607 (6.4%) |
| Urinary tract [C64–C68] | 628 (5.1%) | 118 (4.0%) | 510 (5.4%) |
| Thyroid and other endocrine glands [C73–C75] | 584 (4.7%) | 26 (0.9%) | 558 (5.9%) |
| Ill-defined, secondary, and unspecified sites [C76–C80] | 482 (3.9%) | 228 (7.8%) | 254 (2.7%) |
| Lip, oral cavity, and pharynx [C00–C14] | 247 (2.0%) | 70 (2.4%) | 177 (1.9%) |
| Eye, brain, and other parts of central nervous system [C69–C72] | 149 (1.2%) | 63 (2.1%) | 86 (0.9%) |
| Melanoma and other malignant neoplasms of skin [C43–C44] | 116 (0.9%) | 9 (0.3%) | 107 (1.1%) |
| Mesothelial and soft tissue [C45–C49] | 113 (0.9%) | 28 (1.0%) | 85 (0.9%) |
| Bone and articular cartilage [C40–C41] | 46 (0.4%) | 6 (0.2%) | 40 (0.4%) |
| Days from cancer diagnosis to PPI or PCAB administration, % | |||
| >365 | 7088 (57.1%) | 918 (31.2%) | 6,170 (65.1%) |
| 120–365 | 1471 (11.8%) | 473 (16.1%) | 998 (10.5%) |
| <120 | 3863 (31.1%) | 1,548 (52.7%) | 2,315 (24.4%) |
| Anti-cancer drugs [ACT code], number of patients, % | |||
| Antimetabolites [L01B] | 605 (4.9%) | 605 (20.6%) | — |
| Platinum compounds [L01XA] | 531 (4.3%) | 531 (18.1%) | — |
| Plant alkaloids and other natural products [L01C] | 504 (4.1%) | 504 (17.1%) | — |
| Monoclonal antibodies and antibody drug conjugates [L01F] | 364 (2.9%) | 364 (12.4%) | — |
| Cytotoxic antibiotics and related substances [L01D] | 299 (2.4%) | 299 (10.2%) | — |
| Alkylating agents [L01A] | 286 (2.3%) | 286 (9.7%) | — |
| Protein kinase inhibitors [L01E] | 197 (1.6%) | 197 (6.7%) | — |
| Other antineoplastic agents [L01XX] | 97 (0.8%) | 97 (3.3%) | — |
| Proteasome inhibitors [L01XG] | 9 (0.1%) | 9 (0.3%) | — |
PPI: proton pump inhibitor, PCAB: potassium-competitive acid blocker.
We analyzed each anti-cancer drug in our study population. Platinum agents, known to be a high-risk class for renal failure, were one of the highest IR (in 28.44 per 100000 person-days) in risk period 1. Furthermore, no incidence of AKI including other renal impairments was observed during risk period 1 among patients receiving monoclonal antibodies and antibody drug conjugates, known to have a low risk for AKI (Table 2, Supplementary Table 2).
| Control period | Risk period 1 | Risk period 2 | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Observation periods (person-days) | No. of events | Incidence rate per 100000 person-day | Observation periods (person-days) | No. of events | per 100000 person-day | Observation periods (person-days) | No. of events | Incidence rate per 100000 person-day | |
| Alkylating agents [L01A] | 28044 | 1 | 3.57 | 16240 | 0 | 0.00 | 124726 | 3 | 2.41 |
| Antimetabolites [L01B] | 61348 | 3 | 4.89 | 19261 | 3 | 15.58 | 234449 | 13 | 5.54 |
| Plant alkaloids and other natural products [L01C] | 50545 | 3 | 5.94 | 20179 | 1 | 4.96 | 217330 | 11 | 5.06 |
| Cytotoxic antibiotics and related substances [L01D] | 28519 | 1 | 3.51 | 13371 | 1 | 7.48 | 93209 | 1 | 1.07 |
| Protein kinase inhibitors [L01E] | 22649 | 0 | 0.00 | 4344 | 1 | 23.02 | 50284 | 4 | 7.95 |
| Monoclonal antibodies and antibody drug conjugates [L01F] | 39436 | 0 | 0.00 | 15696 | 0 | 0.00 | 171592 | 6 | 3.50 |
| Platinum compounds [L01XA] | 49973 | 4 | 8.00 | 14065 | 4 | 28.44 | 155967 | 9 | 5.77 |
| Proteasome inhibitors [L01XG] | 778 | 0 | 0.00 | 1352 | 1 | 73.96 | 14087 | 0 | 0.00 |
| Other antineoplastic agents [L01XX] | 9886 | 1 | 10.12 | 3218 | 1 | 31.08 | 40729 | 1 | 2.46 |
AKI, acute kidney injury; PCAB, potassium-competitive acid blocker; PPI, proton pump inhibitor.
We compared incidence rate ratio adjusted for anti-cancer drugs in risk periods 1 and 2 with those in the control period. A total of 31 events of AKI occurred during the control period. The observation periods were 1185488 person-days, and the IR (95%CI) per 100000 person-day was 2.61 (1.78–3.71; Table 3). The number of AKI events during risk period 1 and risk period 2 were 16 and 27, respectively, resulting in an IR of 5.95 (3.40–9.66) and 2.29 (1.51–3.34) per 100000 person-day, respectively. Compared to the control period, the IRR in risk periods 1 and 2 were 2.05 (1.12–3.72, p = 0.0192), and 0.60 (0.36–1.00, p = 0.0481), respectively.
| Group and variable | No. of events | Observation periods (person-days) | Incidence rate per 100000 person-day | Incidence rate ratio adjusted for anti-cancer drugs (95%CI) | p-Values |
|---|---|---|---|---|---|
| All Patients (n = 12422) | |||||
| Control period | 31 | 1185488 | 2.61 (1.78–3.71) | 1.00 | ref |
| Risk period 1 | 16 | 268998 | 5.95 (3.40–9.66) | 2.05 (1.12–3.72) | 0.0192 |
| Risk period 2 | 27 | 1176710 | 2.29 (1.51–3.34) | 0.60 (0.36–1.00) | 0.0481 |
| Female patients (n = 6281) | |||||
| Control period | 10 | 638904 | 1.57 (0.75–2.88) | 1.00 | ref |
| Risk period 1 | 3 | 130119 | 2.31 (0.48–6.74) | 1.35 (0.37–4.93) | 0.6461 |
| Risk period 2 | 8 | 506047 | 1.58 (0.68–3.11) | 0.76 (0.31–1.87) | 0.5440 |
| Male patients (n = 6141) | |||||
| Control period | 21 | 546584 | 3.84 (2.38–5.87) | 1.00 | ref |
| Risk period 1 | 13 | 138879 | 9.36 (4.98–16.01) | 2.18 (1.10–4.32) | 0.0260 |
| Risk period 2 | 19 | 670663 | 2.83 (1.71–4.42) | 0.48 (0.26–0.89) | 0.0200 |
Incidence rate was calculated using AKI onset and person-days for exposure in each patient’s control, risk 1, and risk 2 periods, with a 95% CI based on Poisson distribution. The incidence rate ratios adjusted for anti-cancer drugs were calculated using a regression model based on a Generalized Estimating Equation with a Poisson link function. AKI, acute kidney injury; CI, confidence interval; PCAB, potassium-competitive acid blocker; PPI, proton pump inhibitor.
Following the subgroup analysis, in male patients adjusting anti-cancer drugs, 53 events of AKI were observed from the control period to risk period 2, whereas in female patients, 21 events were observed (Table 3). The IR per 100000 person-days was consistently higher in males than in females throughout the study period. Among males, IRR for anti-cancer drugs with 95%CI were higher in risk period 1 (2.18 [1.10–4.32], p = 0.0260), with a risk reduction in risk period 2 (0.48 [0.26–0.89], p = 0.0200). In females, no significant risk enhancement was observed during risk period 1 (1.35 [0.37–4.93], p = 0.6461) or risk period 2 (0.76 [0.31–1.87], p = 0.5440; Table 3, Fig. 3).
AKI, acute kidney injury; PPI, proton pump inhibitor; PCAB, potassium-competitive acid inhibitors.
This study evaluated AKI in patients with cancer who started PPI or PCAB. We found that the risk was observed in male patients within 30 d after PPI or PCAB initiation.
The renal impairment associated with PPIs has been reported to result from their inhibition of the proton pump expressed in the kidneys.20) PCAB also inhibit proton pump via competing with potassium ions.37) Therefore, in this study, we investigated the role of PPIs and PCAB as exposure factors in development of AKI. Following subgroup analysis, risk enhancement of AKI was only observed in male patients in the anti-cancer drugs group due to sex differences (Supplementary Table 3, Supplementary Fig. 1a, 1b). This pattern was also observed in the analysis adjusted for anti-cancer drugs among all patients, including patients without CKD (Fig. 3, Supplementary Fig. 2, Table 3, Supplementary Table 4). It is well documented that AKI is prevalent in males.38–41) In particular, George et al. reported that risk factors for postoperative AKI in lung cancer were age and male sex.42) Previous studies reported that PPI-induced renal toxicity, such as CKD exacerbation, was not associated with male sex.43) However, in this study, we revealed that PPIs or PCAB toxicity in patients with cancer undergoing anti-cancer drug treatment was particularly pronounced within the initial 30 d after initiating PPI or PCAB therapy, leading to the onset of AKI. Gender differences in renal dysfunction have been reported, with males experiencing a more rapid decline in average GFR compared to females.44) Factors that may contribute to this gender difference include the protective effect of nitric oxide, which is less likely to decrease in females than in males, as well as the renal-protective properties of estrogen.45,46) For these reasons, we included gender as a secondary outcome in our study.
The risk of AKI adjusted for anti-cancer drugs in patients with cancer was high within 30 d of starting PPIs or PCAB (Table 3, Fig. 1). The timing was also observed after subgroup analysis with anti-cancer drugs (3.02 [1.36–6.72]), especially in male patients taking anti-cancer drugs (3.05 [1.24–7.50]). A previous report confirmed that acute interstitial nephritis occurs 4 or 5 d after starting PPIs and develops into irreversible interstitial nephritis.47–49) Myers et al. reported that omeprazole-induced interstitial nephritis occurs within approximately 1.8 months.50) These results suggest that PPI- or PCAB-induced renal toxicity in patients with cancer is likely to have a similar risk timing as PPI- or PCAB-renal toxicity, at least within 1 month. In Japan, the PPI or PCAB administration period is generally limited to 4–8 weeks for reflux esophagitis.51,52) This could be desirable for monitoring patients in a clinical setting for at least 4 weeks, especially patients with the risk factors identified in our study.
We observed the differences in risk of AKI among different classes of anti-cancer drugs (Table 2). Cytotoxic anti-cancer drugs such as platinum compounds, alkylating drugs, and antimetabolites are well known for their potential to cause renal toxicity.53–55) In contrast, vascular endothelial growth factor inhibitors, such as monoclonal antibodies or tyrosine kinase inhibitors, have rarely been reported; however, they exhibit renal toxicity, as previously documented in case reports.56,57) AKI is generally uncommon, but severe and clinically significant adverse events can occur. In this respect, we observed a trend towards nephrotoxic effects depending on the anti-cancer drugs; unfortunately, significant differences could not be observed due to the limited sample size. Therefore, we included all classes of anti-cancer drugs in our study by adjusting for them.
Patients who developed AKI with platinum agents were 0.8% (4/531) during the control period, 0.7% (4/570) during risk period 1, and 1.3% (9/685) during risk period 2 in our study (Table 2, Supplementary Table 2). While the incidence of AKI induced by cisplatin is reported to range from 6.5 to 30%.58,59) In addition, severe AKI accounting for only 1.7%.59) Possible reasons for the discrepancy from previous reports are partially explained by (1) the use of N17, which may selectively extract cases of high severity AKI, and (2) the inclusion of group of platinum agents, such as oxaliplatin, nedaplatin, and carboplatin, which have a lower risk of acute kidney injury compared to cisplatin. In fact, Hwang et al. reported that N17 is associated with high sensitivity for severe AKI as compared to that of low sensitivity for non-severe AKI.36) Since we use N17, there is a possibility that we identified only severe AKI cases in our study setting. Therefore, we need carefully interpreted our study results. According to this, we defined N17 as interpreting “severe AKI” in our study settings. Considering the possibility that only severe AKI cases were extracted in our study, the observed incidence of 0.7 to 1.3% may be reasonable as compared to reported percentages for 1.7% in patients with severe AKI treated with cisplatin.59)
A risk reduction during risk period 2 for male patients were observed (0.48 [0.26–0.89], p = 0.0200; Table 3). Previous research has shown a high incidence of AKI in patients with cancer around the time of cancer diagnosis. Briefly, the cumulative onset of AKI after a cancer diagnosis reaches 27% during the first 5 years, and patients experiencing AKI including other renal impairments account for 17.5% within the first year after cancer diagnosis.60) This can partially explain our results, which indicated a high risk during the control period, likely attributed to the timing being closer to the cancer diagnosis. This suggests that PPIs or PCAB can potentially act synergistically to induce AKI in patients with cancer with a history or plan to receive anti-cancer drugs, especially in males within 30 d of PPI or PCAB initiation.
This study has several limitations. First, we identified the AKI cases using the ICD 10 code N17, which was found to have approximately 30% sensitivity but over 90% specificity in renal transplant patients.61,62) On the other hand, Hwang et al.36) reported that while the sensitivity to detect patients with particularly low severity of AKI is limited, it is possible to identify patients exhibiting an increase in serum creatinine levels (indicative of renal impairment). They concluded that their research findings could serve as guidance for interpreting the N17 code. Consequently, aligning with these interpretations, we opt to interpret the N17 code as “severe AKI” in our study. In this study, there is a potential evaluation bias toward only severe AKI, as Hwang’s report by using N17. Therefore, caution is warranted in interpreting the results, acknowledging the possibility that non-severe AKI cases may not have been adequately assessed. This could have partially affected our results. Second, lansoprazole has been reported to enhance renal toxicity induced by cisplatin, a platinum-based anti-cancer drug, both in vivo and in vitro.63) However, due to the limited sample size of patients with AKI in our study, we were unable to conclusively determine the relationship between the onset of AKI after starting PPIs or PCAB and the use of specific classes of anti-cancer drugs. The transition from AKI to CKD is already known,28,64) and there have been reports of an increased risk of CKD due to AKI caused by anti-cancer drugs.51) However, the risk of AKI was robustly observed after adjusting for anti-cancer drugs in the group without CKD (Supplementary Table 4, Supplementary Fig. 2).
In conclusion, this study provided insights about the risk of AKI, particularly within 30 d after starting PPI or PCAB treatment in male patients with cancer. Our study emphasizes the importance of risk monitoring for the onset of AKI with cancer to avoid the risk of anti-cancer treatment termination due to life-threatening outcomes. Further research is warranted to elucidate the precise mechanisms underlying this interaction and to guide clinical practice for safer medication use, particularly in this specific patient population.
This work was supported by JSPS KAKENHI (Grant Number: 20K07186) and a Grant from Showa University Translational Research in 2023.
All the authors satisfied the ICMJE-recommended criteria. KM and KS conducted the study. KM, AW, and YK extracted raw data from the JMDC Inc. database. KM conducted the statistical analysis, and EI confirmed the analysis. KM and KS drafted the manuscript. KM completed the study. All authors contributed to the discussion in the manuscript. All the authors have agreed to publish this manuscript.
Showa University (KM) and JMDC Inc. collaborated on this project in accordance with a collaborative research agreement. JMDC Inc. provides a large amount of claims data. According to the analysed results, JMDC Inc. did not intervene in data implementation. KM received honorarium fees for presentations from JMDC.inc, Nippon Kayaku, Abbvie, and Eisai. KS (Showa University) received honorarium fees for presentations from Nippon Boehringer Ingelheim and Taiho. EI received lecture and consulting fees from Bristol Myers Squibb, Eisai, Nippontect Systems, Cyberdyne Inc., and Chugai.
The Department of Hospital Pharmaceutics, School of Pharmacy, Showa University, received funds from Ono for a contract research project in accordance with a collaborative research agreement.
As a potential conflict of interest, the department of Hospital Pharmaceutics received research grants from Nippon-Kayaku, Ono, Shionogi, Bayer, Daiichi Sankyo, Eisai, Mochida, and Taiho. The other authors declare no conflicts of interest associated with this manuscript.
JMDC (Japan Medical Data Centre, Tokyo, Japan) claims database is commercially available from JMDC, Inc. The data sharing policy is available from the JMDC website (https://www.jmdc.co.jp/en/bigdata), and the database created by JMDC was detailed by Nagai et al.31,32) Our research data are not available because of the restrictions imposed by the JMDC policy on provide 3rd party to researchers with other objectives.
This article contains supplementary materials.
