Los Angeles—Many pharmacists are unclear about how they should be dosing carboplatin and would like an organization to come forward and provide standards, according to a panel discussion during the annual meeting of the Hematology/Oncology Pharmacy Association (HOPA).
“Currently, there is no uniform standard, so unfortunately if you come to Yale in New Haven, Temple in Texas or MD Anderson, your carboplatin dose will be different,” said Scott Soefje, PharmD, MBA, BCOP, the associate director of oncology pharmacy services and the PGY2 oncology residency director at the Smilow Cancer Hospital at Yale-New Haven, in Connecticut.
Carboplatin dosing is calculated using estimates of a patient’s renal function. A variety of formulas have been used, but none are perfect, according to the panel (Table).
For example, the Egorin formula is troublesome because clinicians do not agree on the proper platelet nadir. “You can ask 10 people and get 10 different answers,” said Jon Herrington, PharmD, BCPS, BCOP, the PGY1 residency assistant program director at the Scott & White Memorial Hospital and Vasicek Cancer Center, in Temple, Texas.
The Calvert formula, which was designed to simplify the measurement of carboplatin dosing, relies on glomerular filtration rate (GFR) measured using the radioisotope Cr-ethylenediaminetetraacetic acid (Cr-EDTA). However, according to Dr. Herrington, Cr-EDTA is expensive and not readily available at all clinical practice sites or hospitals. Substitution of an estimate of creatinine clearance (CrCl) calculated using formulas such as Cockcroft-Gault in place of the GFR in this formula is routine in clinical practice. Although CrCl is always slightly higher than GFR, the two estimates of renal function are used interchangeably in the Calvert formula, said Dr. Herrington, but he noted that determining an accurate CrCl for the Calvert formula is challenging.
Another method, the Chatelut equation, provides an assessment that doesn’t use radioisotopes and doesn’t require a calculation of CrCl, but is complicated, according to Dr. Herrington. “It has a lot of different numbers.” He also noted that the research that supported its use in the first place (J Natl Cancer Inst 1995;87:573-580) has not been corroborated in further studies. The Chatelut equation also provides a wide variance compared with the Calvert method.
A Look at the Literature
Dr. Herrington highlighted results from a study in the British Journal of Cancer that assessed the accuracy of three equations to estimate GFR (eGFR) for use in the Calvert equation (2012;107:1310-1316). In the study, the investigators selected 288 patients referred for a radionuclide GFR during a three-year period; eGFR was calculated using the Modification of Diet in Renal Disease (MDRD), Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) and Cockcroft-Gault equations. Compared with the reference standard (rGFR), MDRD, CKD-EPI and Cockcroft-Gault equations overestimated carboplatin dose by 12%, 14% and 8%, respectively. The researchers recommended that the Cockcroft-Gault should be used if rGFR is unavailable (Figure).
This recommendation supports a National Cancer Institute (NCI) recommendation made five years ago in favor of the Cockcroft-Gault equation for CrCl estimations in adults, but questions remain about this approach, according to Judith Smith, PharmD, BCOP, an associate professor of gynecologic oncology and reproductive medicine at The University of Texas MD Anderson Cancer Center, in Houston. There is uncertainty regarding which weight should be used for calculating CrCl and how the equation should be adjusted for obesity, Dr. Smith said. According to the literature, she noted, actual body weight should be used in most nonobese patients—the Cockcroft-Gault formula was designed and validated using actual body weight. Adjusted body weight, however, is appropriate for patients with a body mass index greater than 30 kg/m.
Also, it is not clear whether it is appropriate to round serum creatinine levels to prespecified values in elderly or cachectic patients. “Serum creatinine levels can run quite low in patients with low muscle mass or the elderly,” said Dr. Smith. “We see a lot of variation … in practice in recommendations of what you should round off to in that scenario. Rounding off to 1 will generally underestimate creatinine clearance.” Clinicians often round off to 0.7 or 0.8, she said, but “there is no magic answer. Until we have a standard, use the lower limit of serum creatinine assay parameters at your institution.”
She also pointed out that wide interlaboratory variation exists in reported serum creatinine. This led the National Kidney Disease Education Program to published recommendations in 2006 to recalibrate serum creatinine assays to an isotope dilution mass spectrometry (IDMS) traceable reference method. All laboratories were expected to comply by Dec. 31, 2010, but this never happened, said Dr. Smith. In some patients with normal renal function, the new standardized IDMS method produced creatinine values that were, on average, 10% to 20% lower than older, non-IDMS values. In patients with low serum creatinine, the IDMS method generated abnormally low values, leading to an overestimation of CrCl and higher calculated carboplatin doses.
HOPA Survey Shows Lack of Consensus
A multidisciplinary HOPA survey conducted in 2010 sheds light on the widely varying practice patterns. Of the 525 respondents, 80.5% of whom were pharmacists, 96.5% said they used the Calvert equation and 94.1% said they use estimated CrCl when calculating this equation. When asked which methods they used to make this estimation, many used Cockcroft-Gault (90%); some used Jelliffe (37.9%); and a few individuals employed MDRD (2.5%) or another method (3.5%). When asked which body weight they used to determine calculations for obese patients, responses fluctuated widely. Survey participants were asked, ‘do you use an adjusted/assigned value for serum creatinine when below (less than) your laboratory normal limit?’ fifty-six percent answered affirmatively. When asked whether or not they had a cap for CrCl when dosing carboplatin, 48% responded “yes.”
The NCI recommends capping GFR used in the Calvert formula at 125 mL/min for the majority of patients, and the survey showed that many clinicians are complying with this. Although capping at 125 mL/min may be suitable for many patients, it might not be the most appropriate method for everyone, said Dr. Soefje. This being the case, Dr. Smith recommended that clinicians consider goals of therapy and use their clinical judgment.
“What we have concluded from this whole [survey] process is there is no standard,” said Dr. Soefje, adding, “We believe there should be a standard method.” Such a standard, he said, would state that Cockcroft-Gault is the preferred equation to calculate CrCl in the Calvert equation, that adjusted body weight is recommended for estimating renal function in obese patients, and that clinicians should round up low CrCl values to their institutional lower limit of normal. Additionally, he said, capping CrCl at 125 mL/min is appropriate for the majority of patients.
Dr. Soefie is not alone in seeking a standard approach to carboplatin dosing. At the end of the session, Robert Ignoffo, PharmD, a professor of pharmacy at Touro University California and a clinical professor emeritus at the University of California, San Francisco, also recommended that dosing standards be developed, and a large majority of audience members agreed.
Dr. Soefje reported being on the speaker’s bureau for Amgen, Eisai and Millennium. Drs. Herrington, Smith and Ignoffo reported no relevant financial conflicts of interest.