|Home | About | Journals | Submit | Contact Us | Français|
Cisplatin is administered in combination with massive hydration to avoid renal toxicity, making its administration difficult in an outpatient setting. Although a short hydration protocol for cisplatin has been recently developed, its safety is not fully understood.
Consecutive patients with lung or other cancer and an Eastern Cooperative Oncology Group performance status of 0–2 who were receiving chemotherapy containing cisplatin at a dose of ≥60 mg/m2 in a single administration were evaluated. Seventy-four patients were treated with a short hydration protocol consisting of 1750–2250 ml of hydration with mannitol and magnesium supplementation over a period of 3.75–4.75 h on Day 1. Sixty-nine patients were treated with a conventional hydration protocol consisting of 2100–2600 ml of hydration over 6.5–7.5 h on Day 1 with pre- and post-hydration on Days 0, 2 and 3. Toxicity was then compared between the two groups.
An elevated serum creatinine level ≥grade 1 was significantly less frequent in the group receiving the short hydration protocol than in the group receiving conventional hydration. Other toxicities were similar between the two groups. Consequently, the completion rate for the planned treatment in the short hydration group (73.0%, 54/74) was significantly higher than that in the conventional hydration group (53.6%, 37/69).
Short hydration is safe, making cisplatin-containing chemotherapy easier to perform.
Cisplatin is one of the most widely used drugs and is a major component of standard combination chemotherapy for various malignancies (1). With the dose-limiting toxicity of nephrotoxicity (2) and the nuisance of strong emetogenicity (3), however, its administration traditionally has required massive hydration and the extensive use of antiemetic agents resulting in a preferred avoidance of its use, especially in an outpatient setting. Minimizing hydration without increasing the risk of renal toxicity would make the use of cisplatin more convenient. Despite a massive and conventional hydration protocol consisting of pre-hydration with 1000–2000 ml over 4 h and subsequent hydration with 1500–3000 ml over 6 h per day during and after its administration, as recommended by the package leaflets in Japan and the USA, many studies have reported renal toxicity with cisplatin (4–10). To improve the situation, a further massive hydration protocol including as much as 2.5 l on the first day of treatment (11) and the fractionated administration of cisplatin on three to five consecutive days (12) have been reported. These attempts, however, inevitably make the use of cisplatin more inconvenient, especially in an outpatient setting.
On the other hand, several attempts to minimize the hydration volume and duration have been made. Besides hydration, other strategies to minimize nephrotoxicity as a result of cisplatin have involved the inclusion of magnesium and mannitol in the hydration protocol. Based on the fact that magnesium depletion increases the severity of platinum-induced nephrotoxicity (13), Evans et al. (14) reported the clinical relevance of routinely providing magnesium supplementation during hydration for cisplatin administration (15). As to the use of diuretics, mannitol has been suggested to reduce the urine concentration of cisplatin decreasing its nephrotoxicity (16,17), although some controversy regarding this point presently exists (18).
As the plasma concentration of protein-unbound platinum decreases rapidly and becomes undetectable at 2 h after intravenous administration, a long period hydration may not be required (10). In 2007, Tiseo et al. (19) reported the feasibility of a hydration protocol consisting of 2 l of saline supplemented with magnesium and diuretics (10,20). In addition, the National Comprehensive Cancer Network has made a short hydration protocol for cisplatin publically available. These trends have encouraged clinical researchers to perform retrospective (21) and prospective (22–26) studies to re-evaluate short hydration protocols for the administration of cisplatin. These single arm studies have demonstrated the safety of short hydration protocols, but only one retrospective study, with a limited patient number, has compared both short and conventional hydration protocols to examine safety (27). In this study, the toxicities (including an elevated serum creatinine level) with conventional and short hydration were retrospectively compared to further validate the feasibility of the short hydration for cisplatin.
Patients eligible for inclusion in this study had been treated with chemotherapy containing cisplatin at a dose of 60 mg/m2 or higher in a single administration and in a frontline setting with or without irradiation for any purpose and with an Eastern Cooperative Oncology Group performance status (PS) of 0–2. For patients treated in the Department of Medical Oncology, Chiba University Hospital, the hydration protocol for cisplatin was completely switched as of September 2010 from a conventional protocol to a short hydration protocol. Patients treated in the department between September 2010 and April 2013 with the short hydration protocol constituted the short hydration group. Patients treated during the period from February 2009 to September 2010 with the conventional hydration method constituted the conventional hydration group. Prior to September 2010, the Department of Medical Oncology and the Department of Respirology of Chiba University Hospital collaborated in treating patients with lung cancer; thus, the conventional hydration group consisted of patients treated in both departments. At the time of transition from conventional hydration to the short hydration protocol, some patients were initially treated using conventional hydration and were then treated using the short hydration protocol for subsequent chemotherapeutic cycles: these patients were excluded from this study. The required number of patients was not predetermined for the statistical evaluation, and all eligible patients treated during the study period, as described above, were enrolled in the study. To roughly match the patient number, the enrollment period for the conventional hydration group was defined as described above.
Details of the hydration methods for the two groups are shown in Table 1. Briefly, the conventional hydration consisted of pre-hydration with 1000–1500 ml over 4–6 h on the day before Day 1 and hydration with 2100 ml plus a varied amount of hydration (0–500 ml) dependent on the combined chemotherapeutic agent over 6.5–7.5 h on Day 1, followed by post-hydration with 1100 ml over 4–4.5 h on Days 2 and 3, resulting in a total volume of 5.1–6.3 l over 4 days. On the other hand, the short hydration protocol consisted of 1750 ml plus a varied amount of hydration (0–500 ml) dependent on the combined chemotherapeutic agent over 3.75–4.75 h on Day 1 only. Oral hydration on Days 2 and 3 was not routinely recommended and was not recorded for the short hydration. For emesis prophylaxis, intravenous treatment with dexamethasone at a dose of 13.2 mg (or 9.9 mg when combined with aprepitant) was administered on Days 1–3 for the conventional hydration protocol, whereas it was delivered intravenously at a dose of 9.9 mg on Day 1 followed by oral administration at a dose of 8 mg on Days 2 and 3 for the short hydration protocol; an 5HT3 antagonist of the physicians' discretion was administered on Day 1 and optionally on Days 2 and 3 for the conventional hydration protocol, whereas palonosetron at a dose of 0.75 mg was administered on Day 1 for the short hydration protocol. Oral aprepitant at a dose of 125 mg on Day 1 and 80 mg on Days 2 and 3 was administered for the conventional hydration protocol after its approval in Japan (i.e. after December 2009) and was administered for the short hydration protocol throughout the study period.
All data consisting of the hydration method, the dose of chemotherapeutic agents including cisplatin and the toxicities including an elevated serum creatinine level were collected retrospectively from individual clinical records. All the toxicities were defined according to the Common Terminology Criteria for Adverse Events (CTCAE, version 4.0). The serum creatinine level was measured using an enzymatic method. In the latest CTCAE, two methods are provided for determining whether an elevated creatinine level exists: one is based on the upper limit of the normal range (ULN) for serum creatinine at each institute (ULN-based), and the other is based on the pretreatment baseline creatinine score for each patient (baseline-based). The ULN values at our institute are 1.1 and 0.8 mg/dl for men and women, respectively. This study used these two criteria, and a worse grade was provided. Laboratory examinations were performed on Day 1 before the administration of cisplatin for each treatment course and were optionally repeated during each course at the physicians' discretion. As the examinations were repeated more frequently for inpatients than for outpatients, only the data obtained on Day 1 of each treatment cycle were collected and analyzed. Categorical variables were compared using the Fisher exact test. Continuous variables were compared using the Wilcoxon rank sum test. All descriptive statistics were expressed as medians with their ranges. All the analyses were performed using JMP, version 10 (SAS Institute Japan, Tokyo, Japan). All P values were two-tailed, and P < 0.05 was considered statistically significant. All the patients provided written informed consent to undergo chemotherapy; no additional consent was needed for this study because of its retrospective nature. This retrospective study was approved by the ethical committee of the Graduate School of Medicine, Chiba University (accession #309).
A total of 143 patients were enrolled in this retrospective study, with 69 patients in the conventional hydration group and 74 patients in the short hydration group. The patient characteristics are shown in Table 2. The median age and sex distribution were not significantly different between the two groups. Although most of the patients in both groups had lung cancer (either small cell or non-small cell), the conventional hydration group included atypical carcinoid and pleomorphic carcinoma in one patient each, while the short hydration group included head and neck cancer in eight patients, cancer of unknown primary site in two patients, thymic carcinoma in two patients and Merkel cell carcinoma, bladder carcinoma and neuroendocrine carcinoma of the kidney in one patient each. Concurrent radiotherapy was delivered for 19 and 29 patients in the conventional hydration and short hydration groups, respectively. Significantly more patients were treated with a higher dose of cisplatin in the conventional hydration group than in the short hydration group, whereas the number of cycles delivered was significantly higher in the short hydration group than in the conventional hydration group. Patients treated with a reduced cisplatin dose were less frequent in the short hydration group than in the conventional hydration group. Although the administration of cisplatin in the conventional hydration group was performed exclusively in an inpatient setting, 65.9% of the administrations in the short hydration group were performed in an outpatient setting.
The changes in the serum creatinine levels after cisplatin administration using either the conventional or short hydration protocols are illustrated in Figs 1 and and2.2. Figure 1 shows a box plot depicting the overall range and the 25th–75th percentile values of the serum creatinine level before treatment and after each cycle of cisplatin treatment for the conventional hydration and short hydration groups. While the serum creatinine value in the short hydration group remained stable throughout the treatment, the values for the conventional hydration group increased significantly as the treatment progressed. An elevated serum creatinine level of ≥grade 1 was significantly more infrequent in the short hydration group than in the conventional hydration group, that is, 5.8% with conventional hydration versus 0% with short hydration in Cycle 1 (P = 0.052), 15.8 versus 2.9% in Cycle 2 (P = 0.023), 18.6 versus 5.1% in Cycle 3 (P = 0.049) and 21.0 versus 3.8% in Cycle 4 (P = 0.015, Fisher exact test. See Supplementary data, Table S1 for detailed data according to grade). When adjusted according to the cisplatin dose, a similar trend in the occurrence of ≥grade 1 serum creatinine elevation was observed for a dose level of 80 mg/m2, whereas the comparisons at dose levels of 60 and 70–75 mg/m2 were inadequate because of the limited sample numbers (Fig. 2 and Supplementary data, Table S1). The comparison of the worst grade of serum creatinine level in all available data, not limited to the data at Day 1, at each cycle between the two groups disclosed more prominent results favor for the short hydration group as expected, presumably because of bias due to the data sampling frequency (Supplementary data, Table S2).
A comparison of hematological toxicities between the two groups demonstrated a significantly higher incidence of leukocytopenia in the conventional hydration group; however, no differences were observed between the two groups when the analyses were limited to patients treated with cisplatin at a dose of 80 mg/m2 (Table 3). No significant differences in non-hematological toxicities, including cardiac events, were observed between the two groups (Table 4).
As shown in Table 5, 53.6% (37/69) of the patients in the conventional hydration group and 73.0% (54/74) of the patients in the short hydration group completed the planned cycles of cisplatin treatment, and the completion rate was significantly higher in the latter group than in the former group. Regarding the causes of treatment discontinuation, the incidence of an elevated serum creatinine level was significantly lower in the short hydration group than in the conventional hydration group.
Despite recent advocacy for a short hydration protocol when administrating cisplatin, this protocol does not seem to be widely accepted. In fact, a questionnaire conducted in Japan in 2009 revealed that 84% of the responding institutions administered cisplatin using a conventional hydration protocol in an inpatient setting, and only 4% of the responding institutions treated patients with a platinum-containing regimen in an outpatient setting (28). A similar situation has been reportedly in UK, with varied hydration methods being used depending on the institution (29). The area of greatest trepidation in adopting the new method seems to be the possibility of an increase in the incidence of nephrotoxicity. This study seems to eliminate this concern. Nephrotoxicity, as represented by an increased serum creatinine value, was not severer in the short hydration group than in the conventional hydration group. In contrast, the incidence of an elevated creatinine value was lower in the short hydration group than in the conventional hydration group. When comparing laboratory data, the frequency of examination is an important consideration. In this study, conventional hydration was performed exclusively in an inpatient setting, while 65.9% of the short hydration protocols were performed in an outpatient setting. This difference might have influenced the frequency of laboratory examinations in the two groups. Therefore, only data obtained immediately before the administration of cisplatin in each cycle, that is, on Day 1 of each cycle, was extracted for analysis so as to enable an impartial comparison. Consequently, the serum creatinine level after each cycle of cisplatin administration (on Day 1 of the next cycle) was found to have increased in value in the conventional hydration group, whereas the values in the short hydration group remained stable for six cycles, with statistically significant differences observed between the groups for each cycle. In addition, a serum creatinine elevation ≥grade 1 was significantly more common in the conventional hydration group than in the short hydration group. These facts might have resulted in the more favorable outcome, including a higher number of delivered cycles and a higher rate of patients in whom the planned chemotherapy was completed, in the short hydration group than in the conventional hydration group (Table 2). A stratified analysis according to the cisplatin dose also produced similar results: a ≥grade 1 elevation in the serum creatinine level showed a trend favor for the short hydration group over the conventional hydration group when 80 mg/m2 of cisplatin was administered, although comparisons of the two groups at dose levels of 60 and 70–75 mg/m2 of cisplatin were difficult because of the limited number of patients in these categories. No statistical differences in hematological and non-hematological toxicities, except for leukocytopenia, were observed between the two groups. Although the underlying mechanism responsible for the lower incidence of leukocytopenia in the short hydration group than in the conventional hydration group is unclear, the imbalance in chemotherapeutic regimens between the two groups or the multiplicity of the statistical analyses may explain this phenomenon.
Single arm retrospective studies performed by Brock and Alberts (30), Tiseo et al. (19) and Lavole et al. (21) with patient numbers of 147, 107 and 357 showed the feasibility of short hydration for cisplatin administration. Four single arm prospective studies, one involving 44 patients and performed by Horinouchi et al. (22), one involving 46 patients and performed by Hotta et al. (23), one involving 45 patients and performed by Ninomiya et al. (24) and one involving 46 patients and performed by Horinouchi et al. (26) also demonstrated the safety of short hydration for cisplatin. In addition, Ouchi et al. (27) retrospectively compared a short hydration protocol performed in 13 patients with a conventional hydration protocol performed in 17 patients and showed a similar safety between the two groups. The present retrospective study comparing short hydration with conventional hydration and involving a relatively larger patient population provides further confirmation of the safety of a short hydration protocol.
Interestingly, this study suggested that the short hydration protocol had a safer toxicity profile than the conventional hydration protocol, with a significantly lower incidence of nephrotoxicity as assessed using the serum creatinine level and equivalent incidences of other hematological and non-hematological toxicities. Although the study performed by Horinouchi et al. was a single arm investigation, this research group had retrospectively evaluated the renal toxicity of cisplatin administered using a conventional hydration protocol in a previous study (31). When the two reports by these authors were directly compared, elevated serum creatinine levels of grades 1 and 2 were observed in 16% (7/44) and 2% (1/44), respectively, of the patients who received the short hydration protocol (22), whereas grade 1 and 2/3 elevations were observed in 37% (231/617) and 17% (104/617), respectively, of the patients who received conventional hydration (31). Special caution should be paid, however, in interpreting these data because the observation methods, such as the measurement schedule for serum creatinine, may have differed between the two studies.
The limitations of this study include its retrospective nature. Also, in addition to the hydration volume and duration, other factors could have influenced the outcome. First, the majority of patients in the short hydration group were treated in an outpatient setting, whereas the patients in the conventional hydration group were exclusively treated in an inpatient setting. Second, the doses of cisplatin and the agents that the cisplatin was combined with differed between the two groups. To minimize any potential bias, all patients during the study period were included in the study, only the laboratory data obtained on Day 1 of each cycle in each group were analyzed and stratified comparisons according to cisplatin dose were also performed. Despite these efforts, some other biases might not have been eradicated. Finally, the use of mannitol as a diuretic and magnesium supplementation was only performed in the short hydration group. In this connection, magnesium supplementation was used in all six studies mentioned above. These studies, however, varied in the use of mannitol; three studies used it (22,23,30), while the studies by Tiseo et al. (19) and Ouchi et al. (27) used furosemide and that by Lavole et al. (21) did not use any diuretic. Therefore, an optimal short hydration protocol remains to be established.
In conclusion, a short hydration protocol for cisplatin administration is safe and easy to perform, resulting in a lower burden to patients and caregivers. Further efforts should be made to optimize the short hydration protocol.
This work was conducted and funded by the Cancer Professional Training Plan of the Ministry of Education, Culture, Sports, Science and Technology of Japan.