Oncological outcomes of dose reductions in cisplatin due to renal dysfunction for patients with metastatic urothelial carcinoma

Abstract Objective To investigate whether dose reductions in cisplatin due to renal dysfunction were associated with worse clinical outcomes in metastatic urothelial carcinoma (UC) patients. Patients and methods One hundred and fifty one metastatic UC patients who received first‐line gemcitabine plus cisplatin (GC) salvage chemotherapy without a previous history of peri‐surgical chemotherapy were included in this retrospective study. Patients with endogenous creatinine clearance of 60 mL/min or more were treated with a full dose of cisplatin, while those with 45‐59 and 30‐44 mL/min were treated with 75% and 50% doses, respectively. Patients were divided into three groups based on the average administered dose of cisplatin of 100% (Group A, N = 43), 99%‐75% (Group B, N = 59), and less than 75% (Group C, N = 49), and therapeutic responses and the toxicity of GC were compared. Results Complete response rates were 9.3%, 13.6%, and 14.3% in groups A, B, and C, respectively. One‐year progression‐free survival rates were 22.9%, 31.1%, and 36.7% in groups A, B, and C with no significant differences. One‐year cancer‐specific survival rates were 56.1%, 71.1%, and 68.3% in groups A, B, and C with no significant differences. A multivariate Cox's regression analysis showed that the dose of cisplatin was not an independent prognostic factor for disease progression and cancer death. Furthermore, there were no significant differences in the incidence of severe adverse events. Conclusions Dose reductions in cisplatin due to renal dysfunction did not worsen clinical outcomes for metastatic UC.


| INTRODUC TI ON
The gemcitabine plus cisplatin (GC) regimen and high-dose intensity methotrexate, vinblastine, doxorubicin, and cisplatin (MVAC) regimen with growth factor support are used as first-line systemic therapies for locally advanced or metastatic urothelial carcinoma (UC). Unfortunately, UC patients with renal impairment (glomerular filtration rate: GFR < 60 mL/min) are classified as cisplatinunfit in the current guidelines and account for approximately half of UC patients. [1][2][3][4][5] Specifically, a total of 37% of patients treated with radical nephroureterectomy (RNU) for upper tract urothelial carcinoma (UTUC) had a preoperative estimated GFR of 60 mL/ min/1.73 m 2 or more, which decreased to 16% after RNU. 5 Although the gemcitabine plus carboplatin (GCarbo) regimen and immune-checkpoint inhibitors have been used as first-line systemic therapies for cisplatin-unfit UC patients, 6-8 a meta-analysis of randomized phase II and III trials revealed that cisplatin-based chemotherapy achieved a significantly higher complete response (CR) and overall response rates than carboplatin-based therapy in metastatic UC patients with normal renal function. 9 When the metastatic UC patients with renal impairment become resistant to Anti-PD-1/PD-L1 antibody, those patients receive less effective regimen such as GCarbo, gemcitabine plus paclitaxel (GP) or gemcitabine monotherapy.
Currently, societies of nephrology recommend dose reductions in cisplatin for patients with renal insufficiency of creatinine clearance (CrCl) between 30 and 60 mL/min. [10][11][12] Among their recommendations, unified protocol based on Kintzel's indication is especially popular. 10,11,13 Kintzel et al reported an indicator of cisplatin dose reductions in patients with impaired renal function based on empiric adjustment due to nephrotoxicity. This indicator shows that patients with CrCl of more than 60 mL/min received a full dose of cisplatin, while those with 46-60 and 30-45 mL/min received 75% and 50% doses, respectively. 13 However, UC patients with renal impairment (GFR < 60 mL/min) are classified as cisplatinunfit in the current urology guidelines. 1

| Patient samples
We retrospectively identified 238 metastatic UC patients ( TanyN1-3M0 or TanyNanyM1) who underwent first-line GC salvage chemotherapy between 2008 and 2020 at our seven institutions, consisting of Keio University Hospital and six affiliated institutions. All cases were histopathologically diagnosed with UC by excision or biopsy of the primary lesion or biopsy of a metastatic lesion. Eighty-one patients with a previous history of peri-surgical chemotherapy, two patients on dialysis and four patients with evidence of a squamous cell carcinoma or adenocarcinoma histology were excluded from the present study ( Figure 1). We divided 151 patients into three groups based on the average administered dose of cisplatin of 100% (Group A, N = 43), 99%-75% (Group B, N = 59), and less than 75% (Group C, N = 49). We examined clinical backgrounds and therapeutic responses to GC chemotherapy in the three groups.

| GC regimen
The GC regimen used in the present study was as follows. UC patients with normal renal function were administered gemcitabine 1,000 mg/m 2 on days 1, 8, and 15 and cisplatin 70 mg/m 2 on day 2 every 28 days. Concurrent radiation and split doses of cisplatin were not performed. To select the dose of cisplatin based on Kintzel's indication, 13 we measured endogenous CrCl using 24-hour urine specimens obtained immediately prior to each cycle of GC chemotherapy.
According to Kintzel's indication with modifications, patients with endogenous CrCl of equal to or more than 60 mL/min were treated with a full dose of cisplatin, whereas those with endogenous CrCl of 45-59 and 30-44 mL/min were treated with 75% and 50% doses, respectively. In contrast, the dose of gemcitabine was not affected by the level of endogenous CrCl. The gemcitabine dose was reduced to approximately 80% according to the physician's discretion as appropriate when repeated grade 3 or 4 adverse events were observed and general status was expected to deteriorate due to adverse effects. Granulocyte colony-stimulating factor (G-CSF) was administered when the neutrophil count was less than 500/mm 3 and was not used prophylactically. The GC dose in each case was calculated based on the average value of all cycles in individuals.

| Response evaluation
Tumor measurements were assessed radiologically using computed tomography scans before the start of the GC regimen. All patients

| Statistical analysis
Radical surgery included total cystectomy for bladder cancer and RNU for UTUC. Local recurrence was defined as relapse within the retroperitoneal field of exenteration. Visceral metastasis was categorized as the liver, lung, bone (including bone marrow), and others (including the adrenal gland, peritoneum, skin, and other organs not classified elsewhere). Lymph node metastasis was defined as regional or distant lymphadenopathy.
CSS and PFS curves were estimated by the Kaplan-Meier method and compared using the Log-rank test. Clinical and pathological parameters were assessed in multivariate models using Cox's proportional hazard regression models with a stepwise forward selection method. Comparisons of the distribution of binary and non-ordered categorical variables were performed using the χ 2 test. Differences in the mean endogenous CrCl values were analyzed using the Mann-Whitney U test. A P-value of less than .05 was considered to indicate significance. Analyses were performed using the SPSS version 26.0 statistical software package.

| Clinical features according to the cisplatin (CDDP) dose administered
The clinical features of patients are shown in Table 1 The means ± standard deviations of endogenous CrCl and GC doses among groups of A, B, and C are shown in Table 2.

| Adverse events according to the CDDP dose administered
No significant differences were observed in overall toxicity between the three groups. Grade 3/4 leukopenia occurred in 30.2%, 37.3%, and 36.7% of patients in groups A, B, and C, respectively (Table 5).  Figure S1. In group A, the mean endogenous CrCl value obtained just before the first cycle of chemotherapy was 95.3 mL/ min, which was not significantly higher than that immediately prior to the third (84.1 mL/min, P = .129) and fifth cycles (80.7 mL/min, P = .099). In group B, the mean endogenous CrCl value obtained just before the first cycle of chemotherapy was 83.4 mL/min, which was significantly higher than that immediately prior to the third (64.3 mL/min, P = .024) but not different from that immediately prior to the fifth cycles (65.3 mL/min, P = .252). In group C, the mean endogenous CrCl value obtained just before the first cycle of chemotherapy was 54.9 mL/min, which was not significantly higher than that immediately prior to the third (51.8 mL/min, P = .517) and fifth cycles (48.1 mL/min, P = .209).   Cisplatin binds to blood plasma proteins, binding exceeds 90% within a few hours of its administration, and only free (non-binding) cisplatin is cytotoxic. [19][20][21] The plasma level of free cisplatin is important for anti-tumor activity and nephrotoxicity. [22][23][24] Since the area under the curve (AUC) of free cisplatin is higher and total body clearance is lower in patients with chronic renal failure than in those with normal renal function, dose reductions in cisplatin are considered for these patients.

TA B L E 4 Uni-and multivariate Cox's regression analyses of disease progression and cancer-specific death
Gemcitabine is rapidly metabolized to difluorodeoxyuridine by cytidine deaminase and disappears from plasma. The peak plasma concentration, elimination half-life, clearance, and AUC of gemcitabine are not changed regardless of renal function. 25,26 Gemcitabine is cytotoxic, whereas difluorodeoxyuridine is considered to be a non-cytotoxic metabolite.
The present study had several limitations. It was performed with a retrospective and non-randomized design and a small sam-

| CON CLUS IONS
We demonstrated that gemcitabine plus a reduced dose of cisplatin did not have a negative impact on clinical outcomes and has potential as a therapeutic option for metastatic UC with endogenous CrCl of 30-60 mL/min.

CO N FLI C T O F I NTE R E S T
There are no conflicts of interest.