13.2 Prevention of decreased renal function during cancer drug therapy
(1) Overview
CQ3: Is reduction of anticancer drug doses recommended for mitigating toxicity in patients with decreased renal function?
Recommendation grade: Weakly recommended (suggestion)
Recommendation
When using drugs that lead to an increased risk of adverse drug events in patients with decreased renal function, dose reduction is recommended. However, when the goal is to cure cancer, doses must ultimately be determined with consideration of the balance between risks and benefits.
Summary
When using agents that lead to an increased risk of adverse drug events in patients with decreased renal function, dose reduction is recommended. However, when the objective is to cure cancer, doses must ultimately be determined with consideration of the balance between risks and benefits.
Background and Objectives
The kidneys are an elimination pathway for many anticancer drugs and their metabolites; therefore, renal impairment can delay the excretion and metabolism of anticancer drugs, potentially resulting in increased toxicity and thus necessitating consideration of dose reduction [
81]. For patients with decreased renal function, dose reduction is also sometimes considered for anticancer drugs that are metabolized in the liver. For example, dose reduction is considered necessary when irinotecan is administered to dialysis patients [
82‐
85]. For sorafenib, as well, a drug that is primarily metabolized in the liver, some believe that dose reduction should be considered [
86]. The present draft summarizes evidence related to dose reduction and presents principles for dose reduction for major anticancer drugs.
Commentary
Answering CQ3 requires studies comparing frequencies of adverse drug events between normal doses and reduced doses in patients with decreased renal function; however, the search formula used in the present guidelines yielded no relevant literature. Such studies present ethical issues and are considered difficult to conduct. Much of the available evidence [
87‐
90] comes from studies that compared the frequencies of adverse drug events in patients with normal renal function and patients with decreased renal function (reduced doses) [
87‐
90]. However, there are very few such studies; thus, the quality of the evidence is judged to be extremely low (D: Almost no confidence in effect estimates).
Consideration of the balance between benefits and risks is particularly important in determining recommendation levels, but due to the paucity of evidence regarding the efficacy of treatment with reduced doses, our recommendation is weak.
However, in real-world clinical settings, attempts have been made to reduce doses in accordance with renal function and to control plasma drug concentrations; these attempts have yielded a small number of studies and guidelines that serve as references. One such attempt with carboplatin dosing is the Calvert formula, which calculates doses using target AUC and Ccr as estimated with the Cockcroft-Gault equation based on the results of a phase I clinical trial (see CQ10 for details) [
91]. Another study has reported a revised Calvert formula based on data from Japanese patients [
92].
Although there are no comprehensive guidelines regarding dose reduction methods in Japan, the Japanese Society of Nephrology and Pharmacotherapy [
93] has presented opinions on dose reduction methods for several anticancer drugs (Table
5); in addition, there are various books with information on dose reduction for anticancer drugs [
94]. Outside of Japan, the United States FDA [
95] and the European Medicines Agency [
96] have published guidelines calling for the inclusion of methods of administration for patients with decreased renal function in package inserts for all types of drugs; these publicly available package inserts may also serve as a reference for dose reduction.
Table 5
Dose reduction methods for major anticancer drugs in patients with decreased renal function
For anticancer drugs, the therapeutic range and the toxic range are extremely close to each other. Therapeutic drug monitoring is considered useful for preventing toxicity in such cases; in fact, therapeutic drug monitoring has been proven effective in randomized clinical trials for some anticancer drugs [
97,
98]. However, at present, attempting to measure blood concentrations of anticancer drugs is not standard practice.
A realistic desirable approach for patients with decreased renal function is to begin anticancer drug administration by referring to the above-mentioned dose adjustment guidelines, monitor adverse events more closely than normal, and consider adjusting doses in future treatment. In patients for whom the objective is to cure cancer, doses must ultimately be determined in consideration of the balance between risks and benefits.
(2) Platinum-based drugs
CQ4: Is risk factor assessment recommended for predicting cisplatin-induced AKI?
Recommendation grade: Weakly recommended (suggestion)
Recommendation
Reported predictors of cisplatin-induced AKI include hypoalbuminemia; smoking; female sex; age (1.03-fold increase in risk per year of age); concomitant use of other anticancer drugs; comorbid cardiovascular disease or diabetes; advanced cancer; and total cisplatin dose. In order to prevent cisplatin-induced AKI, risk factors should be assessed prior to drug administration.
Summary
Reported predictors of cisplatin-induced AKI include hypoalbuminemia; smoking; female sex; age (1.03-fold increase in risk per year of age); concomitant use of other anticancer drugs; comorbid cardiovascular disease or diabetes; advanced cancer; and total cisplatin dose. However, among existing studies, there is no consistent definition of AKI, there are no clear threshold values for risk factors, and there are no established measures for cases with risk factors. Thus, many issues remain for further investigation.
Background and Objectives
Cisplatin, a key drug in treatment for many types of cancer, is one of the most commonly used anticancer drugs. However, cisplatin is known to produce side effects such as myelosuppression, intestinal toxicity, and neurotoxicity; another crucial side effect, nephrotoxicity, is a potential subsequent cisplatin dose-limiting factor. One-third of patients receiving cisplatin are presumed to have comorbid AKI [
104], which often results in the limitation of subsequent doses of cisplatin. Furthermore, AKI sometimes develops into chronic tubulointerstitial fibrosis and irreversible chronic tubulopathy, which may further progress to CKD [
105,
106]. The present draft examines risk factors that may serve as predictors of cisplatin-induced AKI.
Commentary
Cisplatin-induced renal injury is considered to manifest primarily as proximal tubular injury, particularly in the S3 segment [
107]. Cisplatin is absorbed from the basolateral surface into cells and injures mitochondrial DNA, thereby activating apoptosis. Intracellular accumulation of cisplatin results in inflammation, oxidative stress, and ischemic injury [
105]. Hypomagnesemia is also considered to cause renal injury. Magnesium is thought to be involved in active transport mechanisms in the renal tubules. Sobrero et al. have supposed that hypomagnesemia leads to an increased concentration of cisplatin in renal tubular cells, thereby causing proximal tubular injury [
108].
In an investigation by de Jongh et al. [
109] of weekly-dose cisplatin for 400 patients with locally advanced or metastatic cancer, 36% of patients received cisplatin alone, 49% received cisplatin + etoposide, and 15% received cisplatin + paclitaxel. A total of 116 patients (29%) demonstrated a reduction in Ccr of ≥ 25%, while 29 patients (7%) were unable to continue cisplatin due to nephrotoxicity. Independent predictors of post-cisplatin nephrotoxicity as determined by multivariate analysis were paclitaxel coadministration (odds ratio [OR] 4.0, p = 0.001), hypoalbuminemia (OR 3.5, p = 0.006), smoking (OR 2.5, p = 0.002), female sex, and old age. According to age group, the risk of nephrotoxicity was 26% among patients aged < 48 years, and increased with age to 35% for patients aged 48-62 years and 41% for patients aged > 62 years; the risk of nephrotoxicity increased 1.03-fold per year (OR 1.03, p = 0.007). Regarding gender, the risk of nephrotoxicity was twice as high for women as for men (OR 2.0, p = 0.025). Another study reported that cisplatin excretion capacity is lower in women than in men [
110]; however, the cause of this difference is unknown. The involvement of smoking in nephrotoxicity has been surmised to be the effect of oxidant stress [
111]; however, one possibility that cannot be ruled out is that smoking causes cardiovascular disease, which secondarily leads to post-cisplatin nephrotoxicity. Also, in hypoalbuminemia, an increased concentration of unbound cisplatin is considered to enhance nephrotoxicity [
109]. The cited study, which defines nephrotoxicity as a reduction in Ccr of ≥ 25%, is not strictly an assessment of predictors of AKI.
In an investigation of 425 patients treated with cisplatin (total dose 220 mg/m
2 [median]), Stewart et al. [
112] reported that in multivariate analysis, the factors that predicted maximum increases in serum Cr up to 4 weeks after cisplatin treatment were serum albumin, serum potassium, body surface area, and number of administrations. However, this study contains flaws: renal function was assessed with serum Cr alone, and the authors’ method of assessing maximum increases in serum Cr up to 4 weeks after cisplatin treatment is neither a well-established method nor period for assessment. Furthermore, anticancer drugs were used in combination with many other drugs; thus, the degree to which cisplatin contributes to changes in renal function is unknown.
In an examination of 1,721 patients treated with cisplatin, Mizuno et al. [
113] found, in multivariate analysis, that cancer stage 4 diagnosis (OR 1.8, p = 0.011) and total cisplatin dose were risk factors for moderate AKI (1.5-1.9-fold increase in serum Cr within 7 days of cisplatin treatment), while comorbid cardiovascular disease, comorbid diabetes mellitus, and cancer stage 4 diagnosis were risk factors for severe AKI (≥ 2.0-fold increase in serum Cr within 7 days of cisplatin treatment).
Several studies have thus reported predictors of AKI. However, these studies do not present a consistent definition of AKI, and no studies have utilized the RIFLE or AKIN classifications. Furthermore, there are no clear threshold values for risk factors, and there are no established measures for cases with risk factors. Thus, many issues remain for further investigation.
CQ5: Are divided doses of cisplatin recommended for preventing nephrotoxicity?
Recommendation grade: Strongly advised against
Recommendation
Divided doses of cisplatin are not recommended for preventing nephrotoxicity, as the significance of this practice has not been established.
Summary
Divided doses of cisplatin are not recommended for preventing nephrotoxicity, as the significance of this practice has not been established.
Background and Objectives
The fact that the kidneys are the primary organs that excrete platinum-based agents, particularly cisplatin, is related to the nephrotoxicity induced by these agents; this nephrotoxicity is considered to be caused by tubular necrosis. Nephrotoxicity is often prevented or alleviated by hydration via large-volume fluid replacement, or by administration of magnesium. Although some physicians prefer to use divided doses of platinum-based agents to prevent or alleviate nephrotoxicity, some studies with pediatric cancer patients have reported that nephrotoxicity is less frequent with continuous drug administration than with divided doses. The present draft examines the recommendation level for the current practice of administering divided doses of cisplatin with the intention of alleviating nephrotoxicity.
Commentary
At present, there are no articles detailing prospective randomized clinical trials on divided doses of platinum-based agents with alleviation of nephrotoxicity as the primary endpoint. No studies have directly examined the nephrotoxicity prevention effect of divided doses in adult subjects, while there are only three observation studies that have compared divided doses of cisplatin to other administration methods. The content and results of these studies are summarized below.
Forasteiere et al. compared 5 divided, intermittent doses of cisplatin (each administered over 20 minutes) to the same total dose administered via continuous infusion (24 hours) [
114]. The subjects were patients with head and neck cancer; 6 patients received cisplatin 30 mg/m
2 via continuous infusion (24 hours) for 5 days, while another 5 patients received cisplatin 30 mg/m
2 via intermittent intravenous bolus (20 minutes) for 5 days; the two groups were compared in terms of total platinum concentration, free platinum concentration, and adverse events. Although the continuous infusion group demonstrated an extremely low maximum unbound platinum concentration compared to the intermittent bolus group, the exposure to unbound platinum (AUC) was 1.5-2 times higher in the continuous infusion group. An assessment of subclinical nephrotoxicity based on measurement of urinary excretion of the renal enzymes NAG and alanine found no differences between the two groups in nephrotoxicity, or in hearing loss or nausea/vomiting. In contrast, myelosuppression and hypomagnesemia were observed frequently in the continuous infusion group, suggesting that total platinum exposure contributes to nephrotoxicity more than does peak concentration. Because adverse events in continuous administration were clinically acceptable, the authors recommended larger-scale trials with continuous infusion of cisplatin. However, because this study compared two forms of divided doses against each other, the merits of divided doses remain unknown.
Ikeda et al. investigated the optimal administration method for combined 5-FU + cisplatin therapy in patients with gastric cancer and esophageal cancer [
115]. The study compared pharmacokinetic differences (AUC and C
max) in 12 courses of therapy for 9 subjects. Comparisons were made among three groups: 4 courses of cisplatin 80 mg/m
2 (2 hours), 4 courses of 20 mg/m
2 (2 hours) for 5 days, and 4 courses of 100 mg/m
2 (120 hours). In all three groups, 5-FU was continuously infused at a dose of 800 mg/m
2 (24 hours) for 5 days. The authors concluded that continuous infusion is the pharmacokinetically optimal administration method; however, this method has not been recognized to be superior in terms of adverse events.
Takahashi et al. also compared pharmacokinetics and nephrotoxicity according to different cisplatin administration methods (5 divided doses, 24 hours continuous infusion, 12 hours continuous infusion, 6 hours continuous infusion); they found no differences in clinical adverse events [
116].
The above-cited three studies found no difference in nephrotoxicity based on the cisplatin administration method; thus, there is no basis for the active recommendation of divided doses. Therefore, due to the current absence of appropriately designed studies, there is no basis for actively recommending divided doses of cisplatin for the prevention of nephrotoxicity. On the other hand, the National Comprehensive Cancer Network’s 2014 bladder cancer guidelines state that divided doses (35 mg/m
2 on days 1 and 2 or days 1 and 8) may be considered for patients with borderline renal function or minimal dysfunction [
117]. However, no references are cited for these proposed divided doses. In addition, the therapeutic effect of divided doses is unclear.
Nonetheless, some believe that continuous administration of cisplatin is safe for preventing and alleviating nephrotoxicity. Erdlenbruch et al. compared pharmacokinetics and nephrotoxicity between two groups: a group of 4 pediatric osteosarcoma patients who received continuous infusion of 120 mg/m
2 cisplatin over 72 hours, and a group of 6 pediatric medulloblastoma patients who received 1-hour bolus infusions of 40 mg/m
2 cisplatin per day for 3 consecutive days [
118]. The divided dose group demonstrated a peak concentration of free platinum 19 times that of the continuous infusion group, as well as a lower minimum GFR and a higher rate of persistent nephrotoxicity within 1 year after the completion of cisplatin therapy. The authors concluded that continuous administration of cisplatin is less nephrotoxic than divided doses.
CQ6: Is hydration (≥3 L/day) during cisplatin administration recommended for mitigating nephrotoxicity?
Recommendation grade: Strongly recommended
Recommendation
Hydration (≥3 L/day) during cisplatin administration is recommended for mitigating nephrotoxicity.
Summary
The nephrotoxicity of cisplatin was established at the preclinical level (animal trials); therefore, cisplatin dosage regimens were formulated from the outset using hydration and other forms of supportive therapy. Consequently, despite the absence of high-quality evidence from sources such as randomized clinical trials, hydration is strongly recommended during cisplatin administration.
Background and Objectives
Platinum-based agents are renally excreted anticancer agents used to treat various forms of cancer; these agents are well known to be nephrotoxic. Cisplatin is particularly nephrotoxic and has thus been examined frequently. The primary measures for preventing cisplatin nephrotoxicity are hydration and administration of diuretics. The issues of fluid replacement volume and the use versus non-use of diuretics are covered in another CQ and are thus not discussed here.
Commentary
Answering the present CQ fundamentally requires a randomized clinical trial examining the use versus non-use of hydration during cisplatin therapy in human subjects; however, our search formula did not retrieve any such trials. Most of the existing relevant literature consists of reviews that discuss nephrotoxicity. As a basis for recommending fluid replacement, one typical review [
119] cites an animal trial [
120] that divided dogs into a control group, a prehydration group, and a mannitol infusion group; nephrotoxicity was alleviated in the latter two groups. Many other reviews have also recommended forced diuresis with hydration and diuretics. Cisplatin was developed in the 1970s; as a result of the different development methodologies of today, as well as the fact that cisplatin was known early in its development to be nephrotoxic, the absence of validation studies in human subjects is considered inevitable. Consequently, the quality of the evidence is assessed as extremely poor (D: Almost no confidence in effect estimates).
In the evidence for various existing cancer drug therapies, implementation plans prescribe normal hydration when using cisplatin. For other platinum-based agents (carboplatin, etc.), hydration is not typically prescribed. The Japanese package insert for cisplatin states in the dosage section that hydration is to be performed before, during, and after administration; however, the package insert for carboplatin does not contain such instructions. In the United States as well, the package insert for cisplatin calls for hydration, whereas the package insert for carboplatin does not (rather, it specifies that, unlike with cisplatin, massive hydration and forced diuresis are normally not to be performed).
In consideration of the above information, and of the balance between benefits and risks, hydration during cisplatin administration is strongly recommended. Hydration is not recommended during administration of other platinum-based agents such as carboplatin. In the past, hydration during cisplatin administration commonly consisted of approximately 2 L saline solution or half-normal saline solution prior to cisplatin and ≥ 1 L saline solution or half-normal saline solution after cisplatin. In regard to “short hydration”, which reduces this hydration volume and uses oral rehydration, please see CQ7.
CQ7: Is short hydration recommended during cisplatin administration?
Recommendation grade: Weakly recommended (suggestion)
Recommendation
When administering cisplatin on an outpatient basis, short hydration is recommended with consideration of renal function, performance status (PS), and age. However, performing short hydration safely requires sufficient oral rehydration and establishment of sufficient urinary output. Short hydration is intended for patients who, from day 0 to day 3 of chemotherapy, can consume a normal amount of food and undergo an additional ~1,000 mL of hydration per day. When oral rehydration is insufficient, it is necessary to modify the environment to enable rapid hydration via intravenous infusion.
Summary
When administering cisplatin, short hydration is recommended with consideration of renal function, PS, and age. Performing short hydration safely requires sufficient oral rehydration; short hydration is intended for patients who, from day 0 to day 3 of chemotherapy, can consume a normal amount of food and undergo an additional ~1,000 mL of hydration per day. When oral rehydration is insufficient, the environment must be modified to enable rapid hydration via intravenous infusion. In addition, short hydration requires establishment of sufficient urinary output via diuretics (mannitol or furosemide), supplementation of magnesium and potassium, and confirmation of serum electrolyte levels.
Background and Objectives
Before and after administration of cisplatin, hydration must be performed in order to prevent nephrotoxicity. In Japan, standard practice is to replace 1,000-2,000 mL fluid over the course of ≥ 4 hours before and after cisplatin and to administer cisplatin diluted with ≥ 500-1,000 mL infusion solution over the course of ≥ 2 hours. However, this hydration is performed over a long period of time and requires hospitalization. A number of studies have examined methods of hydration for preventing cisplatin-induced nephrotoxicity. Here, we have examined the safety of short hydration via 2,000-2,500 mL fluid replacement over the course of approximately 4 hours.
Commentary
In 2007, Tiseo et al. [
121] reported the results of a retrospective two-center observational study regarding the safety of high-dose cisplatin (≥ 75 mg/m
2) administered with short hydration. Following administration of approximately 2,000 mL saline solution and furosemide over the course of 4 hours on the day of cisplatin administration, nephrotoxicity resulted in withdrawal of chemotherapy in 5 of 107 subjects (4.6%); among these 5 subjects, 2 demonstrated Grade 2 nephrotoxicity according to National Cancer Institute Common Toxicity Criteria. In Japan, Horinouchi et al. [
122] and Hotta et al. [
123] have conducted small-scale prospective trials with patients who received 75 mg/m
2 and 60 mg/m
2 cisplatin, respectively. With short hydration incorporating potassium, magnesium, and mannitol, elevations in serum Cr of Grade 2 or higher (based on the reference range upper limit in the Common Terminology Criteria for Adverse Events ver. 4.0) occurred in 2.2% (1/44) and 0% (0/46) of subjects, respectively. In all other literature we assessed [
124‐
130], compared to conventional hydration, short hydration was found not to increase the incidence of nephrotoxicity and was concluded to be safe; the results of these studies were judged to be consistent. The short hydration method assessed in the present CQ is as follows: a total of approximately 1,600-2,500 mL fluid is replaced over the course of approximately 4 hours; potassium and magnesium are supplemented; and urinary output is established via diuretics (furosemide and mannitol). In contrast, the United States National Comprehensive Cancer Network presents a chemotherapy order template of a total of 1,000-3,000 mL hydration before and after cisplatin administration at a rate of 250-500 mL/h for many carcinomas [
131]. In Japan, the Japan Lung Cancer Society and the Japanese Society of Medical Oncology have created a guide that mentions short hydration, stating that short hydration can be performed safely if the target patients are limited to those who meet certain criteria [
132]. These target patients fulfill conditions such as the following: age < 75 years, serum Cr value below the center’s reference value, Ccr ≥ 60 mL/min, PS of 0-1 on the Eastern Cooperative Oncology Group scale, no pleural effusion or ascites, cardiac function capable of withstanding approximately 500 mL/h hydration (left ventricular ejection fraction ≥ 60% on echocardiography, etc.), completion of appropriate antiemetic therapy, and ability to receive approximately 1,000 mL/day oral hydration from day 0 to day 3 of cisplatin administration. Therefore, consciousness of disease and assurance of adherence are important when selecting patients. In addition, in the event of serious side effects or insufficient water intake, fluid replacement should be performed at a center that can adapt rapidly to such circumstances. All of the target studies were observational studies; therefore, the quality of the overall evidence was initially graded as C (weak). There were judged no serious problems with risk of bias, indirectness, inconsistency, imprecision, or publication bias, all of which downgrade the quality of evidence. In addition, intervention effects, dose-response gradient, and confounders, all of which improve the quality of evidence, were judged not to apply; therefore, the overall quality of evidence was ultimately graded as C (weak).
CQ8: Are diuretics recommended for preventing cisplatin-induced nephrotoxicity?
Recommendation grade: Weakly recommended (suggestion)
Recommendation
We cannot definitively recommend diuretics for preventing cisplatin-induced nephrotoxicity. Such a preventive effect has not been proven in small-scale randomized clinical trials; therefore, there is no basis for recommending diuretics to achieve this effect. Nevertheless, diuretics are used to prevent nephrotoxicity during cisplatin treatment, which has been widely performed since the 1970s. The efficacy and safety of diuretics for preventing nephrotoxicity have been confirmed in large-scale clinical trials of cisplatin and other therapies. Therefore, there is also no basis for rejecting the use of diuretics to prevent cisplatin-induced nephrotoxicity.
Summary
We cannot definitively recommend diuretics for preventing cisplatin-induced nephrotoxicity. Such a preventive effect has not been proven in small-scale randomized clinical trials; therefore, there is no basis for recommending diuretics to achieve this effect. Nevertheless, diuretics are used to prevent nephrotoxicity during cisplatin treatment, which has been widely performed since the 1970s. The efficacy and safety of diuretics for preventing nephrotoxicity have been confirmed in large-scale clinical trials of cisplatin and other therapies. Therefore, there is also no basis for rejecting the use of diuretics to prevent cisplatin-induced nephrotoxicity.
Background and Objectives
High-dose cisplatin administration was first reported to be possible with the combined use of hydration and diuretics in the 1970s. Since then, the osmotic diuretic mannitol and the loop diuretic furosemide have been used to prevent nephrotoxicity in the administration of cisplatin. Here, we examine whether these diuretics are effective for preventing cisplatin-induced nephrotoxicity.
Commentary
A phase I clinical trial for cisplatin has demonstrated that nephrotoxicity is a dose-limiting factor [
133]. Hydration and diuretics have been used in attempts to rapidly excrete toxic platinum metabolites and reduce their duration of contact with the renal tubule in order to alleviate nephrotoxicity. However, pharmacokinetic analysis has shown that diuretics do not affect the half-life of free platinum following cisplatin administration and are considered to reduce the urinary excretion rate and increase serum platinum concentration [
134,
135]. Even if diuretics are effective in preventing nephrotoxicity, the mechanism of this effect is not sufficiently understood.
Hayes et al. were the first to report that fluid replacement and mannitol alleviate cisplatin-induced nephrotoxicity. High-dose cisplatin (120 mg/m
2) was administered to 60 patients with the combined use of hydration and mannitol. The 52 patients who were analyzed demonstrated only temporary increases in serum Cr; no serious nephrotoxicity was observed. Serum Cr increased by < 2 mg/dL in almost all patients; although 10 patients demonstrated greater increases, 9 of those patients demonstrated reduced renal function at baseline and were at high risk for nephrotoxicity [
136]. Following this report, diuretics have been used in the majority of subsequent clinical trials for therapies that include cisplatin.
Ostrow et al. conducted the first trial comparing mannitol and furosemide. Twenty-two patients with advanced cancer resistant to existing therapies received 100 mg/m
2 cisplatin. The subjects were assigned to one of two groups; one group received 37.5 g mannitol by 6-hour infusion, while the other group received 40 mg furosemide by intravenous injection 60 minutes prior to treatment. All subjects underwent hydration with 1 L normal saline following cisplatin administration. Nephrotoxicity (defined as Ccr ≤ 50 mL/min or serum Cr > 2 mg/dL) occurred in 28% of the 22 courses in the mannitol group and 19% of the 25 courses in the furosemide group. Mean Ccr values in the mannitol group and the furosemide group were 34 mL/min and 26 mL/min, respectively. Although the mannitol group demonstrated a tendency toward more severe nephrotoxicity, the difference was not statistically significant. Therefore, the interpretation is that neither diuretic was demonstrated to be superior [
137].
In a prospective randomized phase II trial, Al-Sarraf et al. compared the incidence of post-cisplatin nephrotoxicity between hydration alone and hydration + mannitol. Nephrotoxicity occurred following initial cisplatin administration in 30% of patients who received hydration only and in 15% of patients who received hydration + mannitol. The overall incidence of nephrotoxicity in the hydration only group and the hydration + mannitol group was 39% and 32%, respectively. Thus, mannitol demonstrated a prevention effect against cisplatin-induced nephrotoxicity in the initial administration of cisplatin; however, this effect was not evident in subsequent administrations [
138].
Santoso et al. conducted a randomized comparative trial to compare the cisplatin-induced nephrotoxicity prevention effects of hydration (500 mL normal saline), hydration + furosemide (40 mg), and hydration + mannitol (50 g). Forty-nine patients with gynecologic cancers underwent therapy with 75 mg/m
2 cisplatin + paclitaxel or 5-FU and were randomly assigned to one of the three above-described combination therapies. A total of 15 women were assigned to the hydration only group, while 17 were assigned to the hydration + furosemide group and 17 were assigned to the hydration + mannitol group. The three groups demonstrated nearly equal Ccr at baseline. However, following cisplatin therapy, Ccr (± standard deviation) in the hydration only group, the hydration + furosemide group, and the hydration + mannitol group was 80.4 (±33.5), 81.4 (±23.3), and 60.6 (±26.8) mL/min, respectively; thus, the hydration + mannitol group demonstrated a result significantly inferior to that of the other two groups [
139]. Several issues have been highlighted in relation to this trial: the trial was discontinued due to poor outcomes in the hydration + mannitol group, the sample size was small, the dose of mannitol was larger than in previous trials, and urine collection for Ccr lacked rigor, among other issues. Thus, despite being a randomized comparative trial, its quality is lacking.
Therefore, although diuretics have been widely used since the 1970s to prevent nephrotoxicity induced by platinum-based agents, no randomized comparative studies have clearly demonstrated that diuretics are effective to that end, and there is no sufficient basis to recommend diuretics for that purpose. The European Society of Clinical Pharmacy Special Interest Group on Cancer Care states in their recommendations on the prevention of cisplatin nephrotoxicity that there is no reason to recommend the use of diuretics [
140]. However, diuretics have been widely used in the administration of cisplatin for many years, and this approach has been used to create evidence for various therapies; thus, the safety of diuretics in cisplatin administration is well established. In short hydration as well, which has been attempted in recent years, the use of diuretics is presumed, and they have been reported to be safe. Therefore, without proof that diuretics pose evident risks, there is little basis to advise against the use of these agents.
CQ9: Is magnesium recommended for preventing cisplatin-induced nephrotoxicity?
Recommendation grade: Weakly recommended (suggestion)
Recommendation
Administration of magnesium, which can be expected to prevent hypomagnesemia, has been indicated to affect renal function favorably; therefore, administration of magnesium is recommended for preventing nephrotoxicity.
Summary
Prophylactic administration of magnesium can be expected to prevent hypomagnesemia and alleviate nephrotoxicity following the administration of cisplatin.
Background and Objectives
Due to enhanced excretion primarily from the kidneys as well as intestinal toxicity, cisplatin administration frequently results in hypomagnesemia, which has been reported to potentially cause nephrotoxicity. Therefore, prophylactic administration of magnesium is anticipated to alleviate nephrotoxicity.
Commentary
In our searches for studies comparing nephrotoxicity in the use versus non-use of magnesium in patients receiving high-dose cisplatin, we found two randomized comparative trials and one retrospective analysis [
141‐
143].
Willox et al. randomly assigned 17 cancer patients (16 patients with testicular cancer and 1 patient with an ovarian dysgerminoma) scheduled to receive cisplatin into a magnesium treatment group and a non-magnesium treatment group; the non-magnesium group subsequently demonstrated significant tubular damage (high NAG value) [
141]. Bodnar et al. conducted a double-blind, randomized comparison of magnesium administration and non-administration (placebo) in ovarian cancer patients scheduled to receive cisplatin; the magnesium group subsequently demonstrated significantly favorable GFR compared to the placebo group [
142]. Although both of the above studies indicate that prophylactic administration of magnesium affects renal function favorably, the sample sizes were small, and the endpoints and statistical hypotheses were unclear. Therefore, although magnesium can be anticipated to prevent nephrotoxicity, this preventive effect has not been definitively verified.
However, prophylactic administration of magnesium is inferred to prevent hypomagnesemia and consequently alleviate adverse reactions such as nephrotoxicity, while adverse reactions induced by prophylactic administration of magnesium are minor; considering these points, prophylactic administration of magnesium is currently recommended.
CQ10: Is carboplatin dose setting based on renal function recommended?
Recommendation grade: Strongly recommended
Recommendation
In adult cancer patients receiving carboplatin, there is insufficient evidence to prove that the method of setting doses based on renal function following the establishment of a target AUC increases therapeutic effects and reduces side effects compared to the general method of determining doses based on body surface area. However, the setting of doses based on renal function is both reasonable and widespread in daily clinical practice.
Summary
In adult cancer patients receiving carboplatin, there is insufficient evidence to prove that the method of setting doses based on renal function following the establishment of a target AUC increases therapeutic effects and reduces side effects compared to the general method of determining doses based on body surface area. However, the setting of doses based on renal function is both reasonable and widespread in daily clinical practice. Therefore, we graded our recommendation as “strong”.
Background and Objectives
The platinum-based agent carboplatin is almost completely excreted from the kidneys following administration; therefore, its pharmacokinetics can be predicted based on GFR. Furthermore, AUC, an indicator of drug exposure volume in the body, is closely correlated with hematotoxicity and antitumor effect. Consequently, it is now widespread practice to set carboplatin doses based on GFR after establishing a target AUC. In many cases, GFR is substituted by Ccr. The present draft examines the validity of the routine clinical practice of setting carboplatin doses based on renal function.
Commentary
Carboplatin is a platinum-based agent with a broad antitumor spectrum that primarily includes gynecologic cancer and lung cancer. Because carboplatin is almost completely excreted from the kidneys following administration, its pharmacokinetics can be predicted based on GFR [
144]. Furthermore, AUC, an indicator of drug exposure volume in the body, is strongly correlated with thrombocytopenia and other forms of hematotoxicity, which limit carboplatin doses; AUC is also strongly correlated with antitumor effect. Therefore, individual differences in the side effects and therapeutic effects of carboplatin can be explained by individual differences in AUC arising from pretreatment GFR [
145]. Setting carboplatin doses based on GFR after establishing a target AUC minimizes individual differences in AUC, which can consequently be expected to reduce the risks of serious hematotoxicity and undertreatment. Based on this idea, Calvert et al. created a formula for setting carboplatin doses based on GFR; this formula, called the Calvert formula, remains widely used in daily clinical practice today [
146].
$$ {\text{Calvert formula}}:{\text{ dose }}\left[ {\text{mg}} \right] \, = {\text{ target AUC }}\left[ {{\text{mg}}/{\text{mL }} \times { \hbox{min} }} \right] \, \times \, \left( {{\text{GFR }}\left[ {{\text{mL}}/{ \hbox{min} }} \right] \, + { 25}} \right) $$
To calculate doses, the previously established target AUC and the patient’s GFR are entered into the Calvert formula. Based on clinical trials, AUC is set at 5-7; however, analysis of a model for ovarian cancer patients demonstrates that while the antitumor effect of carboplatin nearly plateaus at an AUC of 5-7, thrombocytopenia and other forms of hematotoxicity are enhanced as AUC increases [
147]. Similar formulas for calculating carboplatin doses based on renal function have been created by Egorin et al. [
148,
149] and Chatelut et al. [
150]; however, the Calvert formula remains the most widely used due to its simplicity. In any case, setting dosages based on renal function is reasonable. However, no clinical trial has prospectively examined the setting of doses based on renal function from the perspective of increasing therapeutic effects and reducing side effects in comparison to typical methods based on body surface area; thus, the evidence for setting doses base on renal function is insufficient.
In the process of creating the Calvert formula, the investigators employed actual GFR based on clearance of EDTA labeled with
51Cr, a radioisotope of chromium. In Japan, the gold standard for GFR is inulin clearance; although measurement of inulin clearance is covered by health insurance, the procedure is cumbersome, and thus, Ccr is often used instead in daily clinical practice. However, not only does serum Cr undergo glomerular filtration, approximately 20-30% of it is secreted from the renal tubules; consequently, Ccr yields higher values than does GFR, a point which requires caution. The serum Cr values used to calculate Ccr are measured with the Jaffé method and an enzymatic method. The Jaffé method is affected by non-specific substances in serum; therefore, the measured value of serum Cr is approximately 0.2 mg/dL higher than the true serum Cr value. However, in calculations of Ccr, this measurement error is cancelled out by the difference with GFR resulting from tubular secretion; therefore, in effect, Ccr calculated with serum Cr as determined by the Jaffé method approximates GFR. With an enzymatic method, on the other hand, serum Cr measurements are precise, and Ccr values are higher than GFR values. Consequently, the Calvert formula, which uses Ccr as a substitute for GFR, engenders a risk of excessive carboplatin dosing. Since the mid-1990s, most medical centers in Japan have used the enzymatic method, whereas the United States and Europe have used the Jaffé method until recently. Caution is necessary when interpreting clinical trials of carboplatin conducted outside Japan. One proposed measure is to add 0.2 to enzymatic method-based serum Cr values when calculating Ccr [
151,
152]. When GFR is used, it is calculated with the Japanese Society of Nephrology’s GFR estimation formula (eGFR) without correcting for body surface area (see CQ1). However, many Japanese clinical trials currently use Ccr calculated from enzymatic method-based serum Cr values in place of GFR. When evidence from these trials is used in clinical practice, regardless of the presence of evident bias between the actual AUC and the target AUC, what is effectively being used is GFR estimated with the same methods as in the relevant trials. However, considering the objective of individualized patient doses based on renal function, renal function should be accurately assessed from the clinical trial stage in order to prevent bias between actual AUC and target AUC.
In the United States, since 2010, serum Cr has been measured by isotope dilution mass spectrometry, which are as accurate as the enzymatic method. Accordingly, establishing an upper limit for the GFR used in the Calvert formula (125 mL/min) is recommended to avoid the excessive carboplatin dosing that would result from overestimation of renal function. In gynecology, for extremely low serum Cr, a lower limit (0.7 mg/dL) is sometimes established. With these methods, it must be noted that actual AUC is larger than target AUC in the majority of patients, while actual AUC is smaller than target AUC in some patients with favorable renal function.
The “GFR+25” component of the Calvert formula corresponds to total carboplatin clearance; “GFR” corresponds to renal clearance, while the constant “25” corresponds to non-renal clearance. Non-renal clearance depends primarily on an individual’s physical size. The Calvert formula was created in the United Kingdom; when using the Calvert formula for Japanese individuals, who are physically smaller on average than Caucasians, the proportion of non-renal clearance increases relative to GFR, particularly in patients with severely decreased renal function, thereby potentially leading to excessive carboplatin dosing [
153].
Measurement of Ccr requires urine collection (typically for 24 hours); thus, estimates calculated based on serum Cr values are sometimes used as a substitute for GFR. Formulas for estimating Ccr include the Cockcroft-Gault equation and the Jelliffe equation. Formulas for estimating GFR in the United States and Europe include the MDRD equation, the CKD-EPI equation, and the Wright formula; for Japanese patients, the previously described GFR estimation formula (eGFR) is used (see CQ1). When using these formulas, the patient’s background (e.g., racial differences and pathologies) and the serum Cr measurement method must be noted. Furthermore, these formulas presume that serum Cr is stable; when renal function fluctuates markedly (such as in the acute phase of renal failure) or when muscle mass is greatly reduced (such as in sarcopenia or undernutrition), renal function will be overestimated.
(3) Other agents
CQ11: Is urine alkalinization recommended for preventing nephrotoxicity in high-dose methotrexate therapy with leucovorin rescue?
Recommendation grade: Strongly recommended
Recommendation
Urine alkalinization is recommended for preventing nephrotoxicity in methotrexate with leucovorin rescue.
Summary
In methotrexate with leucovorin rescue, in addition to urine alkalinization and diuresis via sufficient hydration, monitoring of blood methotrexate levels is recommended. In addition, increased doses and prolonged administration of leucovorin in accordance with serum methotrexate levels are recommended.
Background and Objectives
Methotrexate with leucovorin rescue was developed in the 1970s; supportive therapies such as urine alkalinization and diuresis via sufficient hydration were generally established by the 1990s. The present draft re-examines this method based on recent findings.
Commentary
More than 90% of methotrexate is excreted from the kidneys. In animal experiments, methotrexate nephrotoxicity has been shown to arise from the accumulation of methotrexate or its metabolite 7-OH-MTX in the renal tubules. The solubility of methotrexate and its metabolites depends on urinary pH; this solubility is considered to increase five- to eight-fold with an increase in pH from 6.0 to 7.0 [
154]. Methotrexate with leucovorin rescue was developed in the 1970s with the following theoretical basis: when high-dose (generally ≥ 500-1000 mg/m
2) methotrexate is administered, the methotrexate is passively incorporated into cancer cells; after a certain amount of time has elapsed, leucovorin is administered as a methotrexate antidote and is passively incorporated by healthy cells capable of doing so, thereby rescuing those cells. Methotrexate with leucovorin rescue has been demonstrated as effective against osteosarcoma, acute leukemia, and malignant lymphoma; however, in the 1970s, the frequency of drug-related deaths was at a high rate of approximately 6% [
155]. The pathological explanation for these drug-related deaths emphasized the following: methotrexate nephrotoxicity results in delayed excretion of methotrexate itself, thereby aggravating myelosuppression and other serious adverse events [
155]. Subsequently, in addition to diuresis with urine alkalinization [
156] and sufficient hydration [
157], monitoring of blood methotrexate levels has become widespread, as have increased doses and prolonged administration of leucovorin based on serum methotrexate levels [
158]. In accordance with these technical improvements, deaths related to methotrexate with leucovorin rescue have decreased; data for 3,887 cases of osteosarcoma aggregated in 2004 showed a rate of deaths related to methotrexate with leucovorin rescue of 0.08% [
159]. Based on the above, although there is no evidence from randomized comparative trials, the establishment of urinary output through urine alkalinization and sufficient hydration are recommended to prevent nephrotoxicity in methotrexate with leucovorin rescue. However, in the above-cited 2004 data, nephrotoxicity Grade ≥ 2 (WHO criteria, serum Cr levels 1.5-3.0 × upper limit of normal) was observed in 68 patients (1.8%); the methotrexate with leucovorin rescue-related mortality rate among these patients was 4.4%, thus remaining high [
159]. Increased doses of leucovorin are reported to be effective for delayed excretion of methotrexate resulting from nephrotoxicity [
160].
The efficacy of recombinant enzymes for directly degrading methotrexate in plasma has recently been reported in a prospective trial [
161] and a retrospective analysis [
162]; these recombinant enzymes have been approved in the United States, but not in Japan. Methotrexate is a small molecule with a molecular weight of only 454.44 and thus can be removed by hemodialysis. On the other hand, approximately 50% of methotrexate binds to proteins, and its volume of distribution is tens of liters; in 4 hours, hemodialysis removes only 10.8% of methotrexate (according to drug information). However, the use of high-flux membranes in hemodialysis has been reported in case studies to remove methotrexate more efficiently [
163,
164], thus making this technique worthy of consideration as a therapeutic approach.
On the other hand, in combination chemotherapy that includes standard-dose methotrexate, i.e., cyclophosphamide, methotrexate, and 5-FU (CMF) for breast cancer and methotrexate, vinblastine, doxorubicin, and cisplatin (M-VAC) for urothelial carcinoma, there is no definitive evidence showing that leucovorin and urine alkalinization are useful for preventing nephrotoxicity. In addition, the combined use of non-steroidal anti-inflammatory drugs in combination chemotherapy that includes standard-dose methotrexate is reported to exacerbate adverse events; therefore, methotrexate should not be used in combination with non-steroidal anti-inflammatory drugs [
154].
CQ12: Is withdrawal or reduction of angiogenesis inhibitors recommended when proteinuria is observed?
Recommendation grade: Strongly recommended
Recommendation
When proteinuria is observed during administration of angiogenesis inhibitors, withdrawal or reduction of these drugs is recommended upon consideration of the grade of proteinuria and of the risks/benefits of continued drug therapy.
Summary
Administration of angiogenesis inhibitors requires regular measurement of blood pressure, urinalysis for early detection of hypertension and proteinuria, and proactive administration of antihypertensive agents for sufficient control of blood pressure. If proteinuria manifests, temporary withdrawal of angiogenesis inhibitors or continued treatment with reduced doses are reasonable options; however, in the case of grade 1 proteinuria, for patients with advanced cancer, another option is to continue treatment upon consideration of the risks and benefits. When proteinuria is grade 2 or higher, angiogenesis inhibitors are temporarily withdrawn or reduced, and the patient is treated by a nephrologist as necessary.
Background and Objectives
Angiogenesis inhibitors, which are clinically applied in the treatment of various carcinomas, inhibit tumor angiogenesis primarily by suppressing the VEGF pathway. The actions and adverse events of angiogenesis inhibitors differ from those of cytotoxic anticancer drugs. Proteinuria, like hypertension, is an adverse event that occurs during treatment with angiogenesis inhibitors [
165]. Proteinuria and microalbuminuria have been demonstrated to be independent risk factors for renal disease and cardiovascular disease [
166]; thus, when proteinuria manifests during administration of angiogenesis inhibitors, appropriate management is necessary. There are many different types of angiogenesis inhibitors, each of which is indicated for a different carcinoma and has a different treatment regimen. Angiogenesis inhibitors are administered upon initiation of drug therapy for carcinomas that in most cases occur in a solitary kidney, such as advanced renal cell carcinoma. Furthermore, angiogenesis inhibitors are sometimes administered alone and also sometimes used as part of multidrug therapy. With this diverse background, the incidence of proteinuria during administration of angiogenesis inhibitors has been determined to differ for each individual agent [
165]. According to Japanese special drug use surveillance, during the administration of bevacizumab in 2,696 cases of advanced colorectal cancer, proteinuria occurred in 4.60% of cases; proteinuria was serious in 0.11% of these cases [
167]. The incidence of proteinuria during the administration of sunitinib in 2,141 cases of advanced renal cell carcinoma and gastrointestinal stromal tumors was 1.59%, versus an incidence of 1.20% in advanced renal cell carcinoma and 2.98% in gastrointestinal stromal tumors [
168]. During the administration of sorafenib in 3,335 cases of advanced renal cell carcinoma, the incidence of proteinuria was 0.71%, with no cases of serious proteinuria reported [
169]. In a phase II clinical study of 64 Japanese patients with cytokine-refractory metastatic renal cell carcinoma, proteinuria occurred in 58% of patients, 9% of whom developed serious proteinuria of grade 3 or higher [
170].
Commentary
Angiogenesis inhibitors, i.e., VEGF pathway inhibitors, result in proteinuria during treatment; although the precise mechanism of onset of this adverse effect has not been determined, the presumed mechanism is a breakdown of glomerular structure and filtration function originating from the inhibition of VEGF production by podocytes [
171]. Angiotensin-converting enzyme inhibitors and ARBs dilate efferent arterioles, reduce intraglomerular pressure, and reduce proteinuria; therefore, administration of angiogenesis inhibitors must involve regular measurement of blood pressure, urinalysis for early detection of proteinuria, and proactive administration of antihypertensive agents for sufficient control of blood pressure [
165].
While the incidence of proteinuria is different for each individual angiogenesis inhibitor, the risk of proteinuria is considered to be dose-dependent [
172,
173]. Thus, when proteinuria manifests, reduction or temporary withdrawal of angiogenesis inhibitors are practical options. In fact, in a clinical trial that investigated the therapeutic effects of various molecularly targeted agents, many patients who demonstrated grade 2 or higher proteinuria during treatment were able to resume treatment following dose reduction or withdrawal [
174]. When advanced cancer patients with limited outcomes develop grade 1 proteinuria during treatment with angiogenesis inhibitors, withdrawal or reduction is not always necessary; rather, the decision must be based on an examination of the benefits/risks of continued drug therapy and on the individual patient’s wishes. However, cases of nephrotic syndrome have been confirmed during treatment with various types of angiogenesis inhibitors [
175‐
177]. In cases in which proteinuria worsens despite temporary withdrawal or reduction of angiogenesis inhibitors, referral to a nephrologist should be considered [
165].
CQ13: Is reduction of bisphosphonates and anti-RANKL antibodies recommended for patients with decreased renal function?
Recommendation grade: Strongly recommended
Recommendation
Reduction of bisphosphonates is recommended for patients with decreased renal function. However, reduction of anti-RANKL antibodies is not recommended for patients with decreased renal function.
Summary
Reduction of bisphosphonates is recommended for patients with decreased renal function. However, reduction of anti-RANKL antibodies is not recommended for patients with decreased renal function.
Background and Objectives
Injectable bisphosphonates have been established as useful for improving malignancy-induced hypercalcemia and inhibiting skeletal-related events associated with bone lesions resulting from multiple myeloma or bone metastases from solid tumors (defined as pathologic fracture, radiotherapy for bone lesions, surgery for bone lesions, spinal cord compression, and hypercalcemia). The primary bisphosphonates used in treatment for malignancies in Japan are zoledronic acid and pamidronate. Pamidronate is approved for malignancy-induced hypercalcemia and osteolytic bone metastases from breast cancer, whereas zoledronic acid is approved for malignancy-induced hypercalcemia and bone lesions resulting from multiple myeloma or bone metastases from solid tumors. In Europe, intravenous ibandronate is approved for the inhibition of bone metastasis-related events. One known adverse event associated with bisphosphonates is nephrotoxicity; the present draft examines the necessity of dose reduction in accordance with renal function.
Commentary
High-dose (90-360 mg/month) pamidronate has been reported to induce glomerulosclerosis and acute tubular necrosis, as well as to accelerate acute renal failure and nephrotic syndrome [
178]. A subsequent examination of dosage and administration time found that nephrotoxicity was mild when pamidronate 90 mg was administered over the course of ≥ 3 hours, while a phase III trial found no significant pamidronate-induced nephrotoxicity compared to placebo. Based on these results, the American Society of Clinical Oncology (ASCO) guidelines were revised to specify that when pamidronate 90 mg is administered over the course of ≥ 2 hours, reduction of the pamidronate dose is unnecessary even when Ccr is 30-60 mL/min; the guidelines also specify that when Ccr is < 30 mL/min, further prolongation of pamidronate administration time (4-6 hours) or pamidronate dose reduction is recommended [
179,
180].
Several phase III trials of zoledronic acid for breast cancer with osteolytic lesions, multiple myeloma, lung cancer, and other solid tumors began with a protocol of 4 mg or 8 mg administered over the course of 5 minutes; however, the group that received 8 mg over 5 minutes demonstrated a high incidence of nephrotoxicity, thereby necessitating a two-stage protocol amendment. First, the administration duration was extended from 5 minutes to 15 minutes; second, the 8 mg dose was reduced to 4 mg. These amendments reduced the incidence of nephrotoxicity induced by zoledronic acid to a rate equal to that induced by placebo or by pamidronate (the control group) [
181‐
185]. In 2005, Novartis Pharmaceuticals filed a package insert revision with the FDA stating that for patients with decreased renal function (Ccr 30-60 mL/min), the dosage of zoledronic acid was to be reduced to achieve the same AUC as that for patients with a Ccr of 75 mL/min (specifically, doses of 3.5 mg, 3.3 mg, and 3.0 mg for patients with a Ccr of 50-60 mL/min, 40-49 mL/min, and 30-39 mL/min, respectively). The revised ASCO guidelines specify the following: the recommended dose and duration for zoledronic acid is 4 mg over ≥ 15 minutes; when Ccr is 30-60 mL/min, the dose should be reduced in accordance with the recommendation in the package insert; and when Ccr is < 30 mL/min, zoledronic acid should not be administered. In a retrospective analysis of 220 patients, Shah et al. reported that when zoledronic acid doses were adjusted as recommended in the package insert, patients with decreased renal function demonstrated the same incidence of acute renal failure (as an adverse event associated with zoledronic acid) as patients with normal renal function [
186].
Ibandronate, for which an intravenous formulation is approved in Europe, is used to inhibit associated skeletal-related events associated with bone metastasis (in Japan, ibandronate is approved only in oral form for osteoporosis). Ibandronate is considered to be associated with the lowest incidence of nephrotoxicity of all intravenous bisphosphonates [
187] However, the package insert recommends that when Ccr is > 50 mL/min, infusion time should be extended from 15 minutes to 1 hour; and when Ccr is < 30 mL/min, the dose should be reduced from 6 mg to 2 mg.
Anti-RANKL antibodies were developed to treat bone metastasis; in a phase III trial, anti-RANKL antibodies were significantly superior to zoledronic acid in inhibiting skeletal-related events. Renal impairment did not occur; thus, dose adjustments in accordance with renal function are considered unnecessary [
188]. However, patients with Ccr < 30 mL/min and ESRD patients requiring dialysis were excluded from the trial; therefore, it is necessary to consider the potential onset of serious hypocalcemia and to assess the suitability of anti-RANKL antibodies carefully for patients with severe renal impairment.
(4) Maintenance dialysis patients
CQ14: Is dialysis therapy recommended for drug removal following cisplatin administration in maintenance dialysis patients?
Recommendation grade: Weakly advised against (suggestion)
Recommendation
The majority of cisplatin binds to tissue and proteins and remains in the body even if dialysis is performed, with the resultant potential for a post-dialysis rebound. Therefore, dialysis therapy for drug removal is not recommended following cisplatin administration for maintenance dialysis patients, regardless of timing.
Summary
The majority of cisplatin binds to tissue and proteins and remains in the body even if dialysis is performed, with the resultant potential for a post-dialysis rebound. Therefore, dialysis therapy for drug removal is not recommended following cisplatin administration for maintenance dialysis patients, regardless of timing. However, this is only an expert opinion based on case reports; further clinical study is necessary to close the evidence-to-practice gap.
Background and Objectives
Due to concerns regarding accumulated toxicity following cisplatin administration, ESRD patients sometimes undergo dialysis for drug removal. In the present draft, we assess the efficacy of dialysis therapy for drug removal following cisplatin administration.
Commentary
Cisplatin rapidly binds to plasma proteins upon entering the blood, thereby undergoing conversion from unbound cisplatin (free Pt) to protein-bound cisplatin (≈total Pt). One side effect of cisplatin is nephrotoxicity; dialysis patients, whose renal function has already been eliminated, may instead face problems such as myelotoxicity and peripheral neuropathy.
Aside from case reports, there are very few systematic studies of the pharmacokinetics of cisplatin in dialysis patients. In their investigation of the pharmacokinetics of cisplatin in five patients who developed gastric cancer during maintenance dialysis, Miyakawa et al. reported the following results. When cisplatin was administered concurrently with the initiation of dialysis, the concentration of free Pt in blood rapidly decreased and was below measurable levels following dialyzer use; the concentration of total Pt fluctuated relatively sharply in the early stage and subsequently decreased gradually. When dialysis was initiated 1 hour after cisplatin administration, changes in concentrations of free Pt and total Pt in blood were considered to be the same as when cisplatin and dialysis were initiated simultaneously. However, the study does not specify which patients began cisplatin and dialysis immediately and which patients began dialysis 1 hour after administration of cisplatin [
189]. In the same year as the above study, Miyakawa et al. reported on the pharmacokinetics of cisplatin in 2 gastric cancer patients undergoing maintenance dialysis; however, it is unclear whether these 2 patients were also included in the other study [
190].
In all case reports, aside from a report by Inozume et al. stating that, “In order to maximize the effect of the key drug cisplatin, we elected to perform dialysis the day after cisplatin” [
191], dialysis was initiated 30 minutes to 1 hour following the administration of cisplatin. Without allowing a certain interval following cisplatin administration, free Pt will be eliminated from the blood by dialysis before it can bind to plasma proteins, thus greatly reducing the antitumor effect. Patients with normal renal function who receive cisplatin demonstrate a biphasic pattern in which the blood concentration of cisplatin increases sharply in the early stage (α-phase), then gradually decreases (β-phase). This pattern is also observed in patients with chronic renal failure; the α-phase is considered to be result of the entry of cisplatin into tissue [
192]. The β-phase is the result of excretion of cisplatin from the kidneys; in patients with renal failure, this reduction in the blood concentration of cisplatin is either diminished or completely absent. This biphasic pattern is also observed in reports in which dialysis is initiated 30 minutes to 1 hour following cisplatin administration.
Cisplatin administered
in vivo binds to proteins in plasma and tissue in a short time, after which it is not dialyzed; therefore, in 3.5 to 4 hours of dialysis, only approximately 10% of cisplatin is removed [
193,
194]. Most cisplatin removed from the body is free Pt; the majority of cisplatin binds to tissue and proteins, thus remaining in the body even when dialysis is performed. A post-dialysis rebound results in a renewed increase in free Pt in blood [
191,
195‐
200]. In addition, as the volume of accumulated cisplatin increases, the rate of cisplatin removed by dialysis further decreases [
198,
201].
The above-cited studies demonstrate that although most free Pt can be removed by dialysis, the majority of cisplatin binds to tissue and proteins and thus remains in the body even when dialysis is performed, thereby potentially resulting in a post-dialysis rebound and a consequent renewed increase in cisplatin concentration. Therefore, the answer to the present CQ can be considered to be, “Even when dialysis is performed following cisplatin administration, not only is the cisplatin removal rate roughly a mere 10%, a rebound phenomenon occurs; therefore, regardless of timing (whether immediately after cisplatin or 30 minutes to 1 hour after cisplatin), dialysis is not recommended for the removal of cisplatin”. However, this is an only an expert opinion based on case reports; further clinical study is necessary to close the evidence-to-practice gap. When cisplatin is administered to dialysis patients, a 50-75% dose reduction is recommended [
202,
203]. When dialysis is performed following cisplatin administration, caution is necessary regarding cisplatin accumulation.
(5) Particular comorbidities
CQ15: Is rasburicase recommended for preventing tumor lysis syndrome?
Recommendation grade: Strongly recommended
Recommendation
Rasburicase is recommended for preventing tumor lysis syndrome.
Summary
The suitability of rasburicase for preventing tumor lysis syndrome (TLS) has been described according to risk in the Japanese Society of Medical Oncology’s Tumor Lysis Syndrome Practice Guidance [
204]; rasburicase has also been reported to reduce the need for hemodialysis. Rasburicase reduces uric acid levels, prevents nephropathy, and is effective in preventing TLS.
Background and Objectives
Rasburicase is a recombinant version of urate oxidase that rapidly metabolizes uric acid to allantoin. Compared to uric acid, allantoin is much more soluble in urine; this metabolism thus rapidly reduces uric acid levels in blood. Administration of rasburicase requires caution regarding the following three points: 1) rasburicase is an enzyme preparation and may thus trigger hypersensitivity reactions; 2) antibody formation has been reported, thereby rendering repeated administration is not recommended; 3) rasburicase is contraindicated in patients with glucose-6-phosphate dehydrogenase deficiency. The present draft examines whether rasburicase is recommended for the prevention of TLS.
Commentary
The suitability of rasburicase for preventing TLS has been described according to risk in the Japanese Society of Medical Oncology’s Tumor Lysis Syndrome Practice Guidance [
204]; for high-risk and moderate-risk patients, in cases in which uric acid levels continually increase despite the use of allopurinol and febuxostat, or in cases in which hyperuricemia is observed at diagnosis, rasburicase should be administered or at least considered [
204]. The TLS preventive effect of rasburicase has been demonstrated in a phase III study that randomly assigned subjects at high risk for TLS to rasburicase only (0.20 mg/kg/day, days 1-5), rasburicase plus allopurinol (rasburicase 0.20 mg/kg/day, days 1-3; allopurinol 300 mg/day days 3-5), or allopurinol only (300 mg/day, days 1-5). In comparison to the allopurinol only group, the rasburicase only group demonstrated a significantly lower incidence of laboratory TLS
1 [
205]. In a number of other studies conducted in children, rasburicase significantly reduced uric acid levels compared to allopurinol [
206,
207]. Regarding the nephropathy preventive effect of rasburicase, a systematic review of multiple clinical trials conducted in leukemia and lymphoma patients found that hemodialysis was performed for 0-2.8% of patients who used rasburicase, versus 15.9-25.0% of patients who did not use rasburicase; thus, the use of rasburicase tended to reduce the need for hemodialysis [
208]. Rasburicase has also been shown to reduce uric acid levels in patients at high risk for TLS in a number of randomized comparative trials [
209,
210]. The above-cited studies demonstrate that rasburicase reduces uric acid levels, prevents nephropathy, and is effective for preventing TLS.
CQ16: Is plasmapheresis recommended for anticancer drug-induced thrombotic microangiopathy?
Recommendation grade: Weakly advised against (suggestion)
Recommendation
Due to the absence of definitive evidence, plasmapheresis is currently not recommended for anticancer drug-induced thrombotic microangiopathy (TMA). Although plasmapheresis has been observed to inhibit the progression of TMA-induced renal impairment in a handful of isolated cases, the efficacy of plasmapheresis for this purpose has not been properly assessed and therefore is currently not recommended.
Summary
Plasmapheresis is currently not recommended for anticancer drug-induced TMA due to the dearth of reliable evidence on its efficacy for this purpose. Although there are several case series and cross-sectional studies regarding mitomycin C, these studies have been no assessments of therapy with plasmapheresis alone. Because, plasmapheresis is often performed following hemodialysis or is combined with drug therapy based on anti-platelet drugs and steroids. On the other hand, regarding TMA-induced renal impairment, several reports have only stated that plasmapheresis inhibited further worsening of renal function. In addition, plasmapheresis is often combined with hemodialysis. Thus, the usefulness of plasmapheresis against drug-induced TMA has not truly been assessed.
Background and Objectives
Thrombotic microangiopathy is a disorder that presents with thrombocytopenia, microangiopathic hemolytic anemia, and organ dysfunction. Classical TMAs include thrombotic thrombocytopenic purpura (TTP) and Shiga toxin-induced hemolytic-uremic syndrome (HUS), in both of which activity of a disintegrin-like and metalloproteinase with thrombospondin type 1 motifs 13 (ADAMTS13) is reduced. However, there is a great deal that remains unknown regarding the pathology of TMA, while the pathology of TMA is diverse; therefore, in 2013, all TMAs other than TTP and HUS were defined as atypical hemolytic-uremic syndrome (aHUS), for which diagnostic criteria were created [
212]. While TTP can present with both congenital and acquired ADAMTS13 deficiency, most cases involve acquired deficiency, in which anti-ADAMTS13 autoantibodies are involved. Therefore, plasmapheresis is the first-line treatment for acquired TTP. The objectives of plasmapheresis are ADAMTS13 replenishment; removal of anti-ADAMTS13 antibodies; and removal of unusually large von Willebrand factor multimers (UL-vWFM), which are multimers composed of a hemostatic factor called von Willebrand factor. For HUS, plasmapheresis has not been established as effective and is used primarily as supportive therapy. Plasmapheresis is also used for aHUS induced by complement system abnormalities; however, due to the diverse etiology of aHUS, the efficacy of plasmapheresis for this purpose has not been established. Drug-induced TMA includes TTP caused by production of immunological autoantibodies against ADAMTS13 associated with ticlopidine and other antiplatelet drugs; for this form of TTP, plasmapheresis is effective. On the other hand, calcineurin inhibitors such as cyclosporine and tacrolimus are not associated with ADAMTS13 deficiency; these drugs considered to cause aHUS, which primarily induces vascular endothelial injury without decreasing ADAMTS13 activity, for which plasmapheresis is often ineffective. Many cases of drug-induced TMA are considered to present with a pathology resembling that of aHUS; however, the detail mechanism of drug-induced TMA remains poorly understood. Anticancer drugs that induce TMA include mitomycin C, cisplatin, bleomycin, gemcitabine, pentostatin, and sunitinib [
213].
Although plasmapheresis is combined with antiplatelet drugs and steroids to treat TMA, there is no established treatment. The present draft examines the efficacy of plasmapheresis for anticancer drug-induced TMA.
Commentary
In regard to the efficacy of plasmapheresis for mitomycin C-induced TMA, a case series of 4 patients [
214] reported the results of antiplatelet drugs + plasmapheresis (3-4 L) administered 5-7 times over 1-2 weeks. Two patients demonstrated rapid improvement in platelet count, red blood cell count, and other hematologic parameters, as well as a tendency toward recovery of renal function within 6 weeks. Another patient continued to demonstrate decreased renal function following plasmapheresis; however, over the following ≥ 4 months, renal function gradually improved. The last patient, despite demonstrating an increase in platelet count following plasmapheresis, did not show improvement in renal function and subsequently died. No relationship was demonstrated in these patients between overall mitomycin C dose and TMA onset; thus, no definitive conclusion was reached about the usefulness of plasmapheresis. In some cases, plasmapheresis is combined with antiplatelet drugs (e.g., dipyridamole and sulfinpyrazone) or hemodialysis (conditions unknown); therefore, the effects of plasmapheresis monotherapy are difficult to assess.
In regard to cancer-associated HUS, a cross-sectional study of patients in a national registry with hematocrit ≤ 25%, platelet count < 10 × 10
4 μL, and serum Cr ≥ 1.6 mg/dL (accounting for 99% of patients receiving mitomycin C and 68% of patients receiving 5-FU) [
215] found that of the 37 patients who underwent plasmapheresis, 11 patients (30%) responded to treatment, while 26 patients (70%) either did not respond to treatment or worsened. In a case series of 12 patients who developed TMA following mitomycin C-containing chemotherapy regimens [
216], all patients demonstrated renal failure at the time of diagnosis; although 2 of these patients demonstrated low serum Cr values of 1.8 mg/dL and 2.7 mg/dL, the remaining 10 patients demonstrated serum Cr values of 3.4-9.6 mg/dL. Six of these patients underwent 2 L plasmapheresis 3 times over the course of 1-2 weeks while also receiving antiplatelet drugs or steroids. However, only 1 of these patients responded to plasmapheresis; this patient was also receiving steroids, azathioprine, and dipyridamole.
In a cross-sectional study of breast cancer patients, plasmapheresis was performed 2-49 times (median 46 times) to treat TMA which had developed following high-dose chemotherapy with cyclophosphamide, cisplatin, and carmustine + autologous bone marrow stem cell transplantation [
217]. High-dose chemotherapy was performed for 581 patients, 15 of whom (2.6%) developed TMA; 4 of these patients survived. Of the 15 patients who developed TMA, 12 underwent steroid therapy + plasmapheresis. Survival following TMA diagnosis was 2-76 days (median 41 days); 3 of the 4 survivors had undergone plasmapheresis a mean 50 times.
In a case series of 9 patients who developed TMA among a total of 2,586 patients receiving gemcitabine [
218], the median time to development of TMA in the 9 patients was 8 months (3-18 months) following a total gemcitabine dose of 19.2 g/m
2 (9-56 g/m
2). Six of the patients survived, while 3 died. Of these 9 patients, 5 underwent plasmapheresis. Among these 5 patients, 2 died, while the other 3 developed chronic renal failure; of these latter 3 patients, 2 required dialysis. Of the 3 patients whose renal function recovered, none underwent plasmapheresis; however, the report does not explain the details of this recovery.