Risk factor analysis for cisplatin-induced nephrotoxicity with the short hydration method in diabetic patients

The occurrence of cisplatin (CDDP)-induced nephrotoxicity (CIN) has decreased with advancements in supportive care. In contrast, we reported that baseline diabetes mellitus (DM) complications significantly worsen CIN. This study aimed to determine further risk factors associated with CIN development in DM patients. Patients with thoracic cancer requiring DM pharmacotherapy, who received CDDP (≥ 60 mg/m2)-containing regimens using the short hydration method (n = 140), were enrolled in this retrospective multicenter observational study. The primary endpoint of the present study was the elucidation of risk factors (patient factors, DM medication influence, and treatment-related factors) associated with CIN development in patients with DM. Cisplatin-induced nephrotoxicity occurred in 22.1% of patients with DM. The median worst variation of serum creatinine levels and creatinine clearance (worst level − baseline level) was 0.16 mg/dL (range: − 0.12–1.41 mg/dL) and − 15.9 mL/min (− 85.5–24.3 mL/min), respectively. Multivariate logistic regression analyses identified female sex as the singular risk factor for CIN development in the DM population (adjusted odds ratio; 2.87, 95% confidence interval; 1.08–7.67, P = 0.04). Diabetes mellitus medication and treatment-related factors did not affect CIN development. In conclusion, our study revealed that female sex is significantly associated with CIN development in patients with DM and thoracic cancer.


Patient characteristics
In total, 140 patients with thoracic cancer who underwent DM pharmacotherapy and received CDDP-containing treatment (≥ 60 mg/m 2 ) were enrolled based on the eligibility criteria of this retrospective multicenter observational study (Fig. 1).Baseline patient characteristics are shown in Table 1.Approximately 84% of the participants were males, and the median age was 67 years (range 46-76 years).The percentage of patients with advanced cancer was 35.7% and 11.4% of those had undergone prior treatment.The proportion of patients who received chemoradiotherapy was 24.3%.Median baseline serum albumin level was 4.0 g/dL (2.2-5.0 g/dL), and that of HbA1c l was 7.0% (5.4-11.1%).Median baseline creatinine clearance was 93.7 mL/min (48.8-176.6 mL/min), and 2.9% of participants had less than 60 mL/min.The median CDDP dose in the first cycle was 80 mg/m 2 , and 2.9% of the participants received a dose reduction from treatment initiation.Patients who received co-administration of proton pump inhibitors (PPIs) or NSAIDs during treatment accounted for 36.4% and 20.7%, respectively.Dipeptidyl peptidase-4 (DPP-4) inhibitors or glucagon-like peptide-1 (GLP-1) analogs were administered to 75.7% of patients, followed by metformin (46.4%), sulfonylurea agents or glinides (32.9%), insulin (25.0%), sodium-glucose cotransporter 2 inhibitors (13.6%), and α-glucosidase inhibitors (10.0%).The total number of treatment cycles was 4 for 49.3%, one or two for 18.6%, 3 for 9.3%, and 5 or 6 for 2.1% each.In addition, 30% of the patients received a dose reduction during treatment because of gastrointestinal symptoms, neutropenia, febrile neutropenia, and fatigue; however, no patient received a reduction because of CIN.

Development of CIN and variation of serum creatinine levels and creatinine clearance
CIN was confirmed in 22.1% of patients.Median worst variation of serum creatinine (SCr) levels and creatinine clearance (worst level − baseline level) was 0.16 mg/dL (range: − 0.12-1.41mg/dL) and − 15.9 mL/min

Risk factor analysis for CIN development
Univariate and multivariate logistic regression analyses were performed to identify the independent risk factors for CIN development in patients with DM (Table 3).Multivariate analysis identified that female sex was the singular independent patient's risk factor (adjusted odds ratio; 2.87, 95% confidence interval; 1.08-7.67,P = 0.04).However, DM medication or CDDP-including treatment-related factors did not affect CIN development.Notably, the co-administration of NSAIDs, suggested factors in all patient populations in a meta-analysis 11 , was not suggested as a factor.The association between chemotherapy-induced gastrointestinal symptoms and CIN development was also assessed, as oral hydration plays an important role in the short hydration method.Patients exhibiting all-grade nausea, vomiting, and anorexia accounted for 58.6, 5.0, and 65.0% of the patients, respectively.Concerning problematic grade ≥ 2 symptoms that can affect oral hydration, the values were 20.7, 1.4, and 27.9%, respectively.However, its development was not associated with the development of CIN.

Discussion
We previously reported that patients with DM develop CIN at a significantly higher rate in a short hydration method 4 .Autophagy reportedly protects against CIN 8,13,14 , and its induction at the proximal tubule, where CIN mostly appears, is significantly suppressed in type 2 DM, which all participants met in previous and present studies 4,15 .Thus, we consider that the main mechanism of CIN degradation in type 2 DM is the reduction in renal autophagy.Owing to advances in supportive care, the development rate of CIN has significantly decreased; however, approximately 30% of patients with DM developed symptoms in our previous study 4 .Consequently, we aimed to identify further risk factors for the early detection and treatment of CIN in patients with DM.The development rate of CIN was 22.1%, and female sex was identified as the single most significant risk factor for CIN development in patients with DM.Previous studies have reported that women are at a higher risk of developing CIN 9,[16][17][18] .One possible hypothesis is sex differences in autophagy 19,20 .Tao et al. reported that lower visceral adiposity in female mice arises from more active estradiol-estrogen receptor α signaling, which regulates autophagy and adipogenesis 20 .However, sex was not associated with CIN development in the entire patient population in our previous studies 3,4 .In previous reports reporting the results, a short hydration method was not used, and the evaluated CDDP dosage was different from ours 3,4,9,[16][17][18] , which might have induced the incongruence in the results.In addition, as autophagic activity at the proximal tubule is considered to be suppressed in patients with type 2 DM, further suppression in females may have affected the results.However, as the detailed mechanism for this finding remains elusive, further studies are required.
In contrast, the co-administration of NSAIDs was not associated with CIN development in patients with DM.In our previous studies evaluating the risk factors or meta-analyses of CIN in an all-patient population, NSAIDs co-administration was significantly associated with its development 3,11 .The main mechanism of NSAID-induced nephrotoxicity is a decrease in prostaglandin levels owing to cyclooxygenase-1 inhibition 21 .In contrast, prostaglandin E 2 (PGE 2 ) signaling may contribute to the progression and development of type 2 DM with insulin resistance, elevated fasting glucose, and glucose intolerance 22 .Fenske et al. reported that plasma PGE 2 metabolite levels are twice as high in patients with type 2 DM than in controls 22 .In addition, we have reported that most NSAIDs do not affect renal autophagy in vitro 14 .Thus, we believe that the higher baseline PGE 2 levels in patients with DM compensate for the renal influence of NSAIDs.However, it is necessary to compare the effects of concomitant NSAIDs use on CIN development between patients with and without DM complications, including PGE 2 levels, to understand this phenomenon fully.
Autophagy is nephroprotective against CIN 8,13,14 , and autophagic activity is induced in type 1 DM and suppressed in type 2 DM 15 .Furthermore, renal damage caused by CDDP was reportedly reduced in streptozotocin-administered mice, which are recognized as models of type 1 DM 23,24 .The effect of autophagy can differ between DM types; therefore, it is necessary to evaluate the impact of the DM type on CIN clinically, focusing on autophagy.
Managing chemotherapy-induced nausea and vomiting (CINV) is important for sufficient oral hydration.Thus, adequate prevention and appropriate management of breakthrough gastrointestinal symptoms in accordance with the current guidelines are essential.In this study, adequate CINV management, considering CIN, was performed, resulting in no association between adverse gastrointestinal effects and CIN.Consequently, we should try to manage CINV with adequate understanding from the medical team.
This study had some limitations.First, this study was retrospectively conducted in a relatively small population, as patients with DM were limited to 14% of the total population.Second, as all patients in the present study had type 2 DM and renal pathology is different between type 1 and 2 DM 15 , the present results may differ from those of patients with type 1 DM.Third, we excluded patients receiving immune-checkpoint inhibitors (ICIs) as we tried to evaluate the direct association between the evaluated factors and CIN development.However, because combination treatment with ICIs is one of the main treatment strategies for several types of cancer, an assessment of this latter population is needed.Finally, we did not evaluate the genetic backgrounds or autophagy levels of the patients.Organic cation transporter 2 (OCT2), which transports CDDP to the proximal tubule, has single-nucleotide polymorphisms (SNPs) 25 .In addition, the 808G>T SNP in OCT2 was shown to ameliorate CIN without altering its disposition 26 .Therefore, the patients' genetic backgrounds and/or other unknown factors may have affected the results.Furthermore, as we consider that autophagy plays an important role in the development of CIN in patients with DM, its assessment is required.Considering these limitations, our preliminary results should be validated by future studies.
In conclusion, our study revealed that female sex is significantly associated with CIN development in a short hydration method in patients with DM and thoracic cancer.Management of high-risk populations is crucial; therefore, further evaluation, particularly regarding prophylaxis, is necessary.
All patients met the following baseline criteria: (1) age ≥ 20 years; (2) CDDP administration by a short hydration method; (3) 0 to 2 Eastern Cooperative Oncology Group performance status (ECOG-PS); (4) detailed patient information available from medical records; and (5) sufficient renal or liver function for treatment induction.Patients who received ICI co-administration, were transferred to the hospital during chemotherapy, discontinued the treatment during the first cycle, and whose dosage of magnesium premedication was outside 8 mEq were excluded.The primary endpoint of the present study was the elucidation of risk factors (patient factors, DM medication influence, and treatment-related factors) associated with CIN development in patients with DM.Based on our previous report that found CIN

Treatment methods
The short hydration method has been described in detail in previous reports 3,27,28 .All regimens basically included antiemetic therapy according to the current national guidelines 29 ; palonosetron 0.75 mg on day 1, oral aprepitant 125 mg on day 1 and 80 mg on days 2 and 3 or intravenous fosaprepitant 150 mg on day 1, and dexamethasone 9.9 mg infusion on day 1 and 8 mg orally on days 2-4.These prophylactic medications and other additional antiemetic drugs were administered at the physicians' discretion.

Evaluation of CIN and other adverse effects
Toxicities in all subsequent treatment cycles were graded according to the Common Terminology Criteria for Adverse Events (CTCAE), version 5.0.Renal function was evaluated based on SCr elevation measured using an enzymatic method.In this study, CIN was defined as a SCr elevation of ≥ 1.5 times or > 0.3 mg/dL from the baseline levels as referenced in previous reports 4,9,10,30 .
All analyses were performed using the JMP statistical software version 16.1 (SAS Institute Japan, Tokyo, Japan).P-values less than 0.05 were considered statistically significant.

Table 3 .
Univariate and multivariate analyses of the risk factors associated with the incidence of CIN in DM patients.
*P < 0.05.ECOG Eastern Cooperative Oncology Group, BSA Body surface area, PPIs proton pump inhibitors, NSAIDs nonsteroidal anti-inflammatory drugs, DM diabetes mellitus, DPP-4 dipeptidyl peptidase-4, GLP-1 glucagon-like peptide-1.Considering the cutoff at our facility, serum albumin levels less than 4.1 g/dL were defined as hypoalbuminemia.Liver dysfunction: grade 1 or higher aspartate aminotransferase, alanine aminotransferase, and total bilirubin levels.Renal dysfunction: creatinine clearance of less than 60 mL/min.Vol:.(1234567890)Scientific Reports | (2023) 13:17126 | https://doi.org/10.1038/s41598-023-44477-wwww.nature.com/scientificreports/ 4n 27% patients with DM4, we decided to include approximately 3-4 covariates in the multivariate analysis, resulting in the necessity of enrolling approximately 150 participants.This study was approved by the Ethical Review Board for Life Science and Medical Research of Hokkaido University Hospital (Approval Number: 020-0515) and the Institutional Review Boards of KKR Sapporo Medical Center, Sapporo Medical University, Teine Keijinkai Hospital, and NTT Medical Center Sapporo, and was conducted in accordance with the Declaration of Helsinki and the STROBE statement.The requirement for formal consent for this study was waived by the Ethical Review Board for Life Science and Medical Research of Hokkaido University Hospital and the Institutional Review Boards of KKR Sapporo Medical Center, Sapporo Medical University, Teine Keijinkai Hospital, and NTT Medical Center Sapporo.