Multivariate analysis of risk factors for cisplatin-induced nephrotoxicity in gynecological cancer

Transcription

1 doi: /jog J. Obstet. Gynaecol. Res. Vol. 43,. 12: J , Obstet. Gynaecol. December Res Multivariate analysis of risk factors for cisplatin-induced nephrotoxicity in gynecological cancer Yoshihiro Yamamoto 1, Kazushi Watanabe 2, Hiroshi Matsushita 2, Ikuto Tsukiyama 1, Katsuhiko Matsuura 1 and Akihiko Wakatsuki 2 Departments of 1 Pharmacy, 2 Obstetrics and Gynecology, Aichi Medical University School of Medicine, Aichi, Japan Abstract Aim: Risk factors for cisplatin-induced nephrotoxicity (CIN) vary by population. This study aimed to assess risk factors for CIN in patients with gynecological cancer. Methods: Patients who underwent cisplatin-based chemotherapy for gynecological cancer between January 2009 and December 2015 at Aichi Medical University School of Medicine were included in this study. CIN was defined according to the risk, injury, failure, loss, and end-stage kidney disease (RIFLE) criteria and classified as either risk (Class R) or injury (Class I). Analyses were performed using univariate and multivariate logistic regression models. Results: Among 112 patients enrolled, 30 had CIN. Multivariate analysis revealed that hydration with magnesium (odds ratio [OR], 0.223), history of cisplatin use (OR, 4.420), and hypoalbuminemia (OR, 4.170) were risk factors for Class R, and that frequency of cisplatin administration (OR, 5.620) and hydration with magnesium (OR, 0.216) were risk factors for Class I. Conclusion: This study confirmed that hydration without magnesium, history of cisplatin use, frequency of cisplatin administration, and hypoalbuminemia are significant risk factors for CIN. Key words: cisplatin, gynecological cancer, hypoalbuminemia, magnesium, nephrotoxicity. Introduction Cisplatin has been widely used for the clinical treatment of various solid tumors, although it causes nephrotoxicity as a major dose-limiting side-effect. The frequency of cisplatin-induced nephrotoxicity (CIN) is reported to be 28 42%, 1,2 with some patients exhibiting an irreversible decrease in glomerular filtration. 3 Cisplatin therapy cannot be continued if patients develop severe CIN, and subsequent chemotherapy with other anticancer drugs may be challenging given their limited availability. Hydration is the standard preventive approach against CIN, but some patients nevertheless present clinical symptoms of renal dysfunction. 3,4 Although a number of pharmacologic agents (e.g., amifostine, sodium thiosulfate, and magnesium) have been shown to decrease nephrotoxicity, 5 none play an established role. Therefore, the assessment of risks among patients with CIN is of clinical importance, since the development of CIN or a worsening of its symptoms can be prevented. In addition, careful consideration should be given to avoid coadministration of other potentially nephrotoxic agents and, if necessary, cisplatin should be switched with other platinum-containing drugs. Known risk factors for CIN include high cisplatin dose, concomitant use of other nephrotoxic drugs, female sex, older age, smoking, and hypoalbuminemia. 2,5,6 The dose of cisplatin used for gynecological cancer is usually 60 mg/m 2 or less, which is lower compared Received: February Accepted: June Correspondence: Dr Yoshihiro Yamamoto PhD, Department of Pharmacy, Aichi Medical University School of Medicine, Nagakute-city, Aichi , Japan. yamamo_0308@yahoo.co.jp Japan Society of Obstetrics and Gynecology 2017 Japan Society of Obstetrics and Gynecology1

2 to other cancer regimens. Moreover, patients with gynecological cancer tend to be younger compared to those with other cancers, and present with characteristics different from other cancer patients (e.g., cardiac disease and smoking). Risk factors and odds ratios for CIN vary due to differences in populations, 2,7,8 and no study to date has assessed risk factors for CIN associated with gynecological cancer. Thus, a detailed assessment of CIN among patients with gynecological cancer may help to reduce the incidence of CIN in this population. To this end, we retrospectively investigated risk factors for CIN in patients with gynecological cancer at Aichi Medical University School of Medicine. Methods Eligibility criteria We screened inpatients who underwent cisplatin-based chemotherapy for gynecological cancer between January 2009 and December 2015 at Aichi Medical University School of Medicine. Patients were eligible if they had an Eastern Cooperative Oncology Group performance status (PS) of 0 to 2. Patients with severe renal dysfunction (estimated glomerular filtration rate [egfr] < 30 ml/min) were excluded from the study. For patients whose egfr was <50 ml/min, we used a cisplatin dose of 50 75% of the usual dosage. The Institutional Review Board decided to exempt this study from the usual review process. Study approval was obtained from the president of the Aichi Medical University Hospital, Aichi, Japan (approval number: 2016-H091). Hydration protocol Patients receiving cisplatin adhered to the prescribed hydration protocol of our hospital. The hydration protocol was drawn up as follows: 2500 ml of fluids (lactate Ringer s solution) to be infused prior to the administration of cisplatin, and 500 ml, 1000 ml, and 1000 ml of fluid (lactate Ringer s solution) to be infused for 3 days after the administration of cisplatin. We established the hydration protocol that mannitol and diuretic were not used from the beginning; these drugs were used if the patient s urineflow was below 100 ml/h for 2 days following cisplatin administration despite additional infusion solution. Furthermore, magnesium sulfate was added from 2012 in this protocol (day 1, 15 meq; day 2 3, 5 meq). Collection of data All data were obtained from electronic charts. The following data were collected for each patient: age, body surface area, body mass index, gynecological carcinoma tumor type, disease stage, PS, chemotherapy regimen, prior treatment, cisplatin dose, cumulative dose of cisplatin per square meter, frequency of cisplatin administration, history of cisplatin use in previous chemotherapy regimens, baseline renal function (serum creatinine, egfr), and hydration with or without magnesium. Information on regular administration of non-steroidal anti-inflammatory drugs (NSAID), hypertension, history of cisplatin use, diabetes mellitus, and hypoalbuminemia was collected in terms of their presence or absence. CIN was evaluated according to the risk, injury, failure, loss, and end-stage kidney disease (RIFLE) criteria, 9 which are widely used to assess acute nephrotoxicity. The RIFLE criteria define risk (Class R) as a proportional increase in serum creatinine of more than 50% or a decrease in egfr of more than 25% relative to baseline after cisplatin administration, and injury (Class I) as a proportional increase in serum creatinine of more than 100% or a decrease in egfr of more than 50% relative to baseline after cisplatin administration. Class I includes Class R. The calculation for egfr was as follows: egfr = 194 Scr age We collected the worst level for each renal function parameter throughout the course of treatment in all patients, and classified CIN into Class R or Class I according the RIFLE criteria, in reference to previous reports. 10,11 Statistical analyses Categorical variables were compared using Fisher s exact test. Univariate logistic regression was used to identify factors associated with the incidence of CIN (Class R or I). In addition, multivariate analysis was performed to determine independent risk factors; covariates that were significant in univariate analysis were incorporated into multivariate analysis. Stepwise selection, including all significant parameters, was performed to select more reliable predictive independent parameters. All statistical analyses were performed with EZR (Saitama Medical Center, Jichi Medical University), a graphical user interface for R (The R Foundation for Statistical Computing). Specifically, it is a modified version of R commander designed to include statistical functions frequently used in biostatistics. All reported P-values were two-sided, and P < 0.05 was considered statistically significant Japan Society of Obstetrics and Gynecology 2017 Japan Society of Obstetrics and Gynecology 1881

3 Table 1 Patient characteristics Characteristics n-cin (n =82) ClassR(n = 30) Class I (n =10) Age (years) 57 (22 77) 60 (39 75) 58 (40 70) BMI, kg/m ( ) 21.2 ( ) 22.4 ( ) BSA, m ( ) 1.49 ( ) 1.44 ( ) Type of gynecological cancer Ovarian cancer Endometrial cancer Cervical cancer Others Stage I II III IV PS Chemotherapy regimen CPT-11 + CDDP CDDP + RT Monthly TP ADM + CDDP Monthly DP Weekly TP Weekly DP CDDP + VP IFM + CDDP gitecan + CDDP Prior line of treatment, n Dose of cisplatin, mg ( ) (40 95) ( ) Cumulative dose of cisplatin, mg/m ( ) ( ) ( ) Frequency of administration of cisplatin 6 (2 12) 6 (2 12) 6 (2 10) Baseline renal function Serum creatinine, mg/dl 0.60 ( ) 0.55 ( ) 0.56 ( ) egfr, ml/min 79 (38 137) 89 (62 146) 84.5 (62 107) Hydration with magnesium Yes Regular use of NSAID Yes Hypertension Yes History of cisplatin use Yes Diabetes mellitus Yes (Continues) Japan Society of Obstetrics and Gynecology 2017 Japan Society of Obstetrics and Gynecology3

4 Table 1 (continued) Characteristics n-cin (n = 82) Class R (n =30) ClassI(n =10) Hypoalbuminemia Yes Data are n or median (range). ADM, doxorubicin; BMI, body mass index; BSA, body surface area; CDDP, cisplatin; CPT-11, irinotecan; DP, docetaxel plus cisplatin; egfr, estimated glomerular filtration rate; IFM, ifosfamide; n-cin, no cisplatin-induced nephrotoxicity; NSAID, non-steroidal anti-inflammatory drugs; PS, performance status; RT, radiation; TP, paclitaxel plus cisplatin; VP-16, etoposide. Results Patient characteristics We analyzed 112 patients with gynecological cancer who had received cisplatin during the study period. Among these, 30 patients developed CIN (Class R, n = 30; Class I, n = 10). Patient characteristics are summarized in Table 1. The study population included younger women at a median age of 58 years, of whom approximately 95% were PS 0 1; the number of patients with diabetes mellitus and hypertension was low and there were no patients with cardiac disease. The default dose of cisplatin was 40 mg/m 2 in 47 patients, 50 mg/m 2 in 40, and 60 mg/m 2 in 35, with a median dose-intensity of 97.7% (interquartile range: %), 95.2% ( %), and 88.6% ( %), respectively. patients were routinely given mannitol or diuretic in this study. Twenty-six patients required readministration of Table 2 Univariate analysis in Class R Factors CIN event Odds 95%CI P-value ratio (+) / ( ) Age (years) <65 20 / ( ) / 18 BMI <25 26 / ( ) / 10 Stage 3 19 / ( ) / 24 PS / ( ) / 3 Prior line of treatment <3 23 / ( ) / 15 Dose of cisplatin <40 mg/m 2 15 / ( ) 1 40 mg/m 2 15 / 40 Cumulative dose of cisplatin <500 mg/m 2 29 / ( ) mg/m 2 1/4 Frequency of administration of cisplatin <7 25 / ( ) / 7 Baseline egfr 60 ml/min 30 / 73 0 ( ) 0.11 <60 ml/min 0 / 9 Hydration with magnesium Yes 10 / ( ) < / 25 Regular use of NSAID Yes 6 / ( ) / 74 Hypertension Yes 2 / ( ) 1 28 / 76 History of cisplatin use Yes 9 / ( ) / 73 Diabetes mellitus Yes 2 / ( ) 1 28 / 76 Hypoalbuminemia 3.5 mg/dl 17 / ( ) <0.01 <3.5 mg/dl 13 / 11 BMI, body mass index; CI, confidence interval; CIN, cisplatin-induced nephrotoxicity; egfr, estimated glomerular filtration rate; NSAID, non-steroidal anti-inflammatory drugs; PS, performance status Japan Society of Obstetrics and Gynecology 2017 Japan Society of Obstetrics and Gynecology 1883

5 cisplatin due to cancer recurrence or metastasis during the survey period and we defined these patients as having a history of cisplatin use. Univariate analysis and multivariate analysis Tables 2 and 3 show the relation between each factor determined by univariate analysis. In Class R, univariate analysis revealed hydration with magnesium, history of cisplatin use, and hypoalbuminemia to be significant risk factors for CIN. Multivariate analysis using these factors showed that hydration with magnesium (odds ratio [OR], 0.223; P < 0.01), history of cisplatin use (OR, 4.420; P =0.02),andhypoalbuminemia (OR, 4.170; P < 0.01) were independent risk factors for Class R CIN (Table 4). In Class I, univariate analysis revealed that frequency of cisplatin administration, hydration with magnesium, and hypoalbuminemia were significant risk factors for CIN. There were significantly high correlations between Class I and frequency of cisplatin administration (OR, 5.620; P = 0.04) and hydration with magnesium (OR, 0.216; P =0.04). Discussion We found that hypoalbuminemia, history of cisplatin use, and hydration without magnesium were risk factors for Class R CIN, whereas hydration without magnesium and frequency of cisplatin administration were risk factors for Class I CIN. In this study, hydration with magnesium was associated with a decreased risk of CIN in both Class R and Class I. We previously reported that hydration with magnesium provided protective effects against CIN. 10,11 In the present study, multivariate analysis revealed that hydration with magnesium also has a nephroprotective effect. Table 3 Univariate analysis in Class I Factors CIN event Odds 95%CI P-value ratio (+) / ( ) Age (years) <65 6 / ( ) / 24 BMI <25 8 / ( ) / 12 Stage 3 7 / ( ) / 32 PS / 6 0 ( ) / 96 Prior line of treatment <3 7 / ( ) / 19 Dose of cisplatin <40 mg/m 2 6 / ( ) mg/m 2 4/51 Cumulative dose of cisplatin <500 mg/m 2 10 / 97 0 ( ) mg/m 2 0/5 Frequency of administration of cisplatin <7 7 / ( ) / 9 Baseline egfr 60 ml/min 10 / 93 0 ( ) 0 <60 ml/min 0 / 9 Hydration with magnesium Yes 3 / ( ) / 38 Regular use of NSAID Yes 3 / ( ) / 91 Hypertension Yes 1 / ( ) / 95 History of cisplatin use Yes 3 / ( ) / 87 Diabetes mellitus Yes 0 / 8 0 ( ) 1 10 / 94 Hypoalbuminemia 3.5 mg/dl 5 / ( ) 0.04 <3.5 mg/dl 5 / 19 BMI, body mass index; CI, confidence interval; CIN, cisplatin-induced nephrotoxicity; egfr, estimated glomerular filtration rate; NSAID, non-steroidal anti-inflammatory drugs; PS, performance status Japan Society of Obstetrics and Gynecology 2017 Japan Society of Obstetrics and Gynecology5

6 Table 4 Multivariate analysis in Class R and Class I Factors Odds ratio 95%CI P-value Class R Baseline egfr 60 ml/min <60 ml/min Hydration with magnesium Yes ( ) <0.01 History of cisplatin use Yes ( ) 0.02 Hypoalbuminemia 3.5 mg/dl ( ) <0.01 <3.5 mg/dl Class I Frequency of administration < ( ) Hydration with magnesium Yes ( ) 0.04 Regular use of NSAID Yes ( ) 0.19 Hypoalbuminemia 3.5 mg/dl ( ) 0.12 <3.5 mg/dl CI, confidence interval; egfr, estimated glomerular filtration rate; NSAID, non-steroidal anti-inflammatory drugs. Hypoalbuminemia was associated with CIN (Class R) in this study, which is consistent with previous reports, 2,5 although the mechanism is unclear. Cisplatin is a highly protein-bound drug, and therefore, the plasma concentration of unbound cisplatin may increase in patients with hypoalbuminemia, causing enhanced renal toxicity. CIN is related to the peak plasma concentration and/or the area under the plasma concentration time curve of non-protein-bound cisplatin. 12 In addition, patients with hypoalbuminemia might be at risk for malnourishment and poor PS. Based on these results, we believe that careful attention should be paid to CIN in patients with hypoalbuminemia. Our results showed that frequency of cisplatin administration is a risk factor for Class I CIN, and that a history of cisplatin use is a risk factor for Class R CIN. These results are in line with previous studies reporting frequency of administration and cumulative dose of cisplatin as risk factors for CIN. 5 Patients who respond to initial platinum-based therapy and have a long interval of time to recurrence are recommended to undergo retreatment with platinum-containing drugs, including cisplatin, in gynecological cancer. Therefore, there are likely more patients who had been readministered cisplatin in this particular population compared to those with other cancers. We speculate that a history of cisplatin use is a risk factor that contributes to the development of CIN not only because it is associated with an increased frequency of administration and the cumulative dose of cisplatin, but also due to the irreversible nature of CIN. This report is the first to suggest that focus should be placed on the relationship between history of cisplatin use and CIN. Cumulative dose of cisplatin was not shown to affect CIN, perhaps due to the small number of patients in this study. Limitations of this study include the retrospective design and small sample size. Further prospective studies with a larger sample size are needed. The study population included patients who were younger and showed good performance, and are thus unlikely to be affected by hypertension, diabetes mellitus, or cardiovascular disease. Consequently, we could not examine their associations. Further clinical studies will be needed to address this issue. In this study, our findings revealed a significant association of CIN with hypoalbuminemia, history of cisplatin use, and hydration without magnesium in Class R CIN; and hydration without magnesium and frequency of cisplatin administration in Class I CIN. Thus, the substitution of the other platinum drugs for cisplatin might be considered for patients who require readministration of cisplatin and those with hypoalbuminemia. In addition, hydration with magnesium might reduce the risk of CIN from our results as well as previous studies 8,10,11 ; therefore, we recommend administering hydration supplemented with magnesium for patients receiving cisplatin-based chemotherapy. Acknowledgments This study received no foundational support Japan Society of Obstetrics and Gynecology 2017 Japan Society of Obstetrics and Gynecology 1885

7 Disclosure The authors declare no conflicts of interest. References 1. Lebwohl D, Canetta R. Clinical development of platinum complexes in cancer therapy: An historical perspective and an update. Eur J Cancer 1998; 34: de Jongh FE, van Veen RN, Veltman SJ et al. Weekly high-dose cisplatin is a feasible treatment option: Analysis on prognostic factors for toxicity in 400 patients. Br J Cancer 2003; 88: Cvitkovie E, Spauldling J, Bethune V, Martin J, Whitmore WF. Improvement of cis-dichlorodiammineplatinum (NSC ): Therapeutic index in an animal model. Cancer 1977; 39: Wittes RE, Brescia F, Young CW, Magill GB, Golbey RB, Krakoff IH. Combination chemotherapy with cisdiamminedichloroplatinum (II) and bleomycin in tumors of the head and neck. Oncology 1975; 32: Miller RP, Tadagavadi RK, Ramesh G, Reeves WB. Mechanisms of cisplatin nephrotoxicity. Toxins 2010; 2: Sánchez-González PD, López-Hernández FJ, López-voa JM, Morales AI. An integrative view of the pathophysiological events leading to cisplatin nephrotoxicity. Crit Rev Toxicol 2011; 41: Mizuno T, Ishikawa K, Sato W et al. The risk factors of severe acute kidney injury induced by cisplatin. Oncology 2013; 85: Kidera Y, Kawakami H, Sakiyama T et al. Risk factors for cisplatin-induced nephrotoxicity and potential of magnesium supplementation for renal protection. PLoS One 2014; 9: e Bellomo R, Ronco C, Kellum JA, Mehta RL, Palevsky P, Acute Dialysis Quality Initiative Workgroup. Acute renal failuredefinition, 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: Yamamoto Y, Watanabe K, Tsukiyama I et al. Nephroprotective effects of hydration with magnesium in patients with cervical cancer receiving cisplatin. Anticancer Res 2015; 35: Yamamoto Y, Watanabe K, Tsukiyama I, Yabushita H, Matsuura K, Wakatsuki A. Hydration with 15 meq magnesium is effective at reducing the risk for cisplatin-induced nephrotoxicity in patients receiving cisplatin ( 50 mg/m2) combination chemotherapy. Anticancer Res 2016; 36: Reece PA, Stafford I, Russell J, Khan M, Gill PG. Creatinine clearance as a predictor of ultrafilterable platinum disposition in cancer patients treated with cisplatin: Relationship between peak ultrafilterable platinum plasma levels and nephrotoxicity. JClinOncol1987; 5: Japan Society of Obstetrics and Gynecology 2017 Japan Society of Obstetrics and Gynecology7