How Do I Treat…?

Practical perspectives on cancer treatment by thought leaders, explaining how they would approach the treatment of a patient in their area of expertise.

Sunday, March 9, 2014

ONLINE FIRST: How I Adjust Chemotherapy Doses for Patients with Renal Insufficiency


Professor of Medicine (Hematology/Oncology)

Professor of Pharmacology

Vanderbilt-Ingram Cancer Center

Nashville, Tenn.


On the surface, it may seem relatively easy to make appropriate chemotherapy dose modifications for patients with renal insufficiency. We generally have pharmacokinetic studies measuring renal clearance. We can estimate a patient’s renal function using a serum creatinine or glomerular filtration rate (GFR). If we have a patient who has half normal renal function and a chemotherapeutic drug where renal clearance accounts for 50 percent of overall clearance, we should make a dose reduction of 25 percent. If the drug is not cleared by the kidney, no dose reduction should be made. 


Unfortunately, things are not as simple as one might think. Several variables make dose adjustments for renal insufficiency more complicated. How do we accurately measure renal function? What is normal renal function?  Are drug metabolites formed? Are metabolites active? Are there multiple mechanisms for drug clearance? Can renal failure impact non-renal drug clearance?


I believe that most oncologists initially use a serum creatinine as a measure of renal function. However, a normal serum creatinine may not indicate “normal” renal function. Creatinine comes from skeletal muscle. If muscle mass remains steady and dietary creatinine intake remains steady (conditions that are not always met in cancer patients), the serum creatinine concentration is inversely proportional to creatinine clearance.


However 10 to 40 percent of urinary creatinine is secreted by the renal tubule. In early renal insufficiency, tubular secretion of creatinine will increase as GFR falls. Once a serum creatinine exceeds 1.5-1.8 mg/dl, the creatinine secretory process is saturated. In early renal failure, a small change in serum creatinine can mean a large drop in GFR. An increase in serum creatinine from 0.9 to 1.2 mg/dl corresponds to a decrease in creatinine clearance from 120 to 70 ml/min. Half of patients with a GFR < 60 ml/min have a “normal” creatinine so that oncologists should not assume that all patients with a serum creatinine under 1.5 mg/dl will have a GFR above 60 ml/min.


Measurement of Creatinine Clearance

Measurement of creatinine clearance (GFR) is a better indicator of renal function. GFR can be calculated by collecting a 24-hour urine sample and comparing the urine creatinine concentration with a serum creatinine (GFR = UCr x Volume/SCr). However, measuring creatinine clearance in the clinic is not practical due to time constraints and patients frequently not collecting all urine over 24 hours.


Estimates of creatinine clearance have been devised using serum creatinine and clinical variables. The Cockcroft-Gault and the MDMR (Modification of Diet in Renal Disease) estimates have been validated; they correlate well with each other and with some exception, make a reasonable estimate of GFR. Both can be used for dose modification in patients with renal failure.


However, the Cockcroft-Gault and MDRD estimates are less accurate in patients with unusual body mass (obesity, amputees) and in certain ethnic groups (specifically Asians). Patients with unusual diets or extremes of body weight may need to have a measured creatinine clearance.  


It is also important to remember that the MDMR is normalized to a 1.73 m2 body surface area (BSA). An MDMR-estimated GFR needs to be multiplied by a patient’s BSA to obtain a GFR in ml/min. A “normal” GFR varies considerably among normal individuals and depends on body size, sex, and age.


An average person’s GFR declines by 0.75 ml/min per year. A normal GFR for someone under 40 is roughly 100-120 ml/min, but by age 70, the GFR may average 60-70 ml/min. Although not scientifically sound, in practical terms most individuals seen in an oncology clinic with a GFR over 60 ml/min are considered to have “normal” renal function. Even with this loose definition of “normal,” approximately 12 to 20 percent of cancer patients presenting for chemotherapy will have an abnormal GFR


If we have an accurate measure of a patient’s renal function, can we simply measure the percentage of renal drug clearance for that drug and adjust the dose for our patients based on these measurements? This would make sense if the parent drug were the only active moiety and if non-renal clearance of drug was not affected by renal failure. However, active metabolites are often present, and non-renal clearance may be changed in patients with renal insufficiency.


For example, cytosine arabinoside is primarily metabolized to uracil arabinoside, which has been felt to be a non-toxic metabolite. However, ara-U is cleared by the kidney and studies have suggested that high plasma concentrations of ara-U in renal failure patients receiving high dose ara-C may increase toxicity.


Irinotecan and imatinib are primarily cleared by hepatic metabolism. There is a suggestion that uremic toxins may decrease hepatic transport or metabolism of imatinib and irinotecan leading to increased toxicity. Thus, renal failure can impact clearance even if the parent drug does not undergo renal excretion.  Conversely, drug toxicity may not increase in patients with renal insufficiency even if the drug has significant renal excretion. 


Over 50 percent of oxaliplatin undergoes renal excretion. However, studies evaluating toxicity in patients with renal failure have found no measureable increase in toxicity down to a creatinine clearance of less than 20 ml/min. Thus, drug pharmacokinetics with estimates of renal clearance may not provide enough information to make appropriate dose adjustments in patients with renal insufficiency.


Studies Rare

Studies measuring drug toxicity and efficacy in a significant number of cancer patients with renal insufficiency, in addition to pharmacokinetics measurements, would provide better information to make informed dosing decisions. Unfortunately, such studies are rare. For some drugs, no toxicity or efficacy data is available. Toxicity evaluations in patients with renal insufficiency are sometimes available but often in very small series. Information that dose modifications for renal impairment result in similar antitumor responses (response rates, PFS, and overall survival) is almost never available.



Given the lack of perfect information to make recommendations for dose adjustments for renal failure patients, what do I recommend when presented with a patient needing chemotherapy that has a creatinine clearance below 60 ml/min?


I would first review the package insert and/or ask my pharmacist if there is any information regarding studies of this drug in patients with renal insufficiency. Pharmaceutical companies may be a good source of information. I generally recommend using a drug where we have fairly good information (carboplatin is an example) or using a drug with minimally renal excretion (less than 20% of total clearance).


It is important to know what the goals of treatment are. For palliation, minimizing risk by avoiding a drug with potential for increased toxicity may be a better plan. If the goal is potential cure, then using a drug with the potential for high risk of toxicity may be justified. Chemotherapy is sometimes appropriate for patients on dialysis.  In addition to dose adjustments for lack of any renal function, it is important to remember that drugs may be cleared by dialysis. Chemotherapeutic agents should generally be administered just following a dialysis session.


There are many review articles and book chapters providing suggestions as to dose modifications for renal insufficiency. These suggestions are at times based only on “expert opinion.” I categorize the information regarding the need for dose adjustments of chemotherapy drugs into three groups:

·    Grade 1 evidence is that a drug (or active metabolite) is renally excreted, that there is evidence of increased toxicity using standard doses in patients with renal failure and that there is some data as to dose adjustments for renal failure. Chemotherapy drugs falling into that category include: capecitibine, carboplatin, cisplatin, lenalidomide, methotrexate, and oxaliplatin. For carboplatin, oxaliplatin, and lenalidomide, good data for dose adjustment exist, and I would be comfortable using these agents.  Equations for adjusting carboplatin doses based on GFR are well-known. Oxaliplatin can safely be used down to a GFR of 20 ml/min. Dosing guidelines for lenalidomide are available. In myeloma patients with renal insufficiency, lenalidomide therapy may improve renal function. I tend to avoid cisplatin and methotrexate in patients with renal insufficiency as the risks are just too great without good information for dose adjustments. The risk of standard-dose cisplatin is worsening renal insufficiency to the point of dialysis. If there is no concern about worsening renal function, then cisplatin is a reasonable choice.

·    My Grade 2 evidence includes drugs that are excreted via the kidney and for which there are case reports or small series suggesting increased drug toxicity in renal failure. Suggested dose modifications are only guesses without much data or where toxicity studies are conflicting. Category 2 drugs include: arsenic trioxide, bleomycin, high-dose cytarabine, etoposide, fludarabine, imatinib, melphalan, pentostatin, sorafanib, sunitinib, topotecan, and vandetinib. 

·    Category 3 drugs either have significant renal excretion or a case report noting increased toxicity in a patient with renal toxicity but minimal or conflicting information. Category 3 drugs include chorambucil, cyclophosphamide, daunorubicin, epirubicin, erbulin, ifosfamide, irinotecan, lomustine, nitrosoureas, and pemetrexed.