Prediction of Excretion and Toxicity of High Dose Methotrexate in Children and Adolescents With ALL

Overview

Each year approximately 2,900 children and adolescents less than 20 years old are diagnosed with acute lymphoblastic leukemia or acute lymphoblastic lymphoma in the United States. (For the purposes of this protocol, ALL will be used to refer to patients with either acute lymphoblastic leukemia or acute lymphoblastic lymphoma as patients are treated in the same manner.) High-dose methotrexate (HDMTX; 5 g/m2) remains an important component of standard treatment for most ALL patients. However, high plasma and intracellular MTX concentrations (defined as a MTX level of \>1 µmol/L at 42 hours and \> 0.40 µmol/L at 48 hours) can quickly lead to acute kidney, bone marrow, liver, skin, central nervous system, and gastrointestinal toxicities requiring extended hospitalization and delays in subsequent chemotherapy treatments. This study seeks to identify more sensitive markers of kidney injury that could serve as better predictors of delayed excretion and/or toxicity of HDMTX. This study is a pilot repeated-measures feasibility study. Hypothesis 1: Directly measured GFR (mGFR, a type of test to measure the filtering rate of kidneys) by iohexol clearance obtained prior to HDMTX will demonstrate greater sensitivity and specificity for prediction of delayed MTX excretion and/or toxicity in children and adolescents with ALL than serum creatinine (sCr) alone or sCr used for eGFR calculation. If this study proves that mGFR is a better predictor of delayed MTX excretion and/or toxicity, then another study will be developed in the future to determine if modifying the HDMTX dose or adjusting supportive care based on mGFR will prevent delayed clearance and toxicity without impacting patient survival. Hypothesis 2: Those participants prospectively demonstrating delayed MTX excretion or toxicity will exhibit elevation of kidney injury biomarkers less than 24 hours following initiation of HDMTX infusion compared to pre-chemotherapy measurements. These biomarkers will increase prior to a measurable sCr elevation.

Using this approach, retrospective clinical data from our patient population has identified a significant number of HDMTX treatments associated with delayed MTX excretion (defined as a MTX level of \>1 µmol/L at 42 hours and \> 0.40 µmol/L at 48 hours) and secondary toxicity. From January 2012-May 2013, 16 ALL patients received 64 total HDMTX doses (4 cycles each). At 48 hours post-HDMTX infusion, 27 doses (42%) exhibited delayed MTX clearance \[median plasma MTX level 0.41 µmol/L (range, 0.01-1.6 µmol/L) above the 48-hour clearance level of 0.40 µmol/L\]. Mucositis, cytopenias, skin rashes, and subsequent chemotherapy delays were noted in 31%, 31%, 6%, and 34% of doses respectively. Toxicities in combination with delayed MTX clearance occurred in 17 doses (27%) and toxicity without delayed clearance occurred in 18 doses (28%). Also, a mean increase in sCr of 17% (range 0-57.6%) above pre-HDMTX baseline was observed. Although measurements of sCr are used per the standard of care as an indirect marker of kidney injury, the known delay in sCr rise following nephrotoxic (drug-induced) kidney injury (KI) eliminates its use as a point of care surrogate marker for kidney toxicity. Therefore, this study seeks to identify more sensitive markers of kidney injury that could serve as better predictors of delayed excretion and/or toxicity of HDMTX. This study is a pilot repeated-measures feasibility study. Substantial evidence demonstrates sCr, alone or used to estimate glomerular filtration rate (eGFR), overestimates kidney filtering function when compared with direct measurement of GFR \[3\]. A new, commonly used equation for estimating the GFR uses values of sCr, blood urea nitrogen (BUN), and cystatin C. Conversely, a direct measurement of kidney filtering function (mGFR) can be accomplished by determining the plasma (part of the blood) iohexol (Omnipaque300, GE Healthcare) clearance rate (rate at which the iohexol is removed from the body by the kidneys) over a period of 300 minutes, and has been documented as a sensitive and accurate measure of renal filtering function \[4\]. Plasma clearance of iohexol is an optimal method of determining mGFR, particularly in children, because it is not radioactive, is non-ionic, has a low osmolality, has been proven to be a safe and non-toxic X-ray contrast medium, and is excreted from the plasma exclusively by the kidneys \[5\]. Iohexol is a well-known contrast agent that is already FDA-approved for use in urographic procedures, angiographic procedures, and for the measurement of GFR (See Appendix I for more information on the history and safety of iohexol use for GFR measurement). Iohexol clearance from plasma has been shown in several studies to be a suitable method of determining GFR in adults \[6\] and children \[7, 8, 9\]. Given that standard oncology care relies on sCr values and eGFR, the kidney filtering function is likely often overestimated, which can result in an increased incidence of delayed MTX excretion and subsequent toxicity. Therefore, mGFR may provide an optimal method to more accurately determine kidney filtering function and thereby allow prediction of which ALL patients will develop delayed MTX excretion and/or toxicity. The utility of mGFR in predicting MTX excretion delay and/or toxicity has not been previously studied. In addition, earlier recognition of kidney injury following HDMTX administration could also be evaluated using biomarkers in blood and urine. If these biomarkers are found to be useful in diagnosing early kidney injury, these tests may become a new clinical point of care in the future that could allow earlier and thus more effective clinical intervention in ALL patients. These interventions include increased intravenous hydration, which decreases the associated end-organ toxicities. Urinary kidney injury molecule-1 (KIM-1) and clusterin, serum cystatin C, and plasma fibroblast growth factor 23 (pFGF23) are biomarkers that appear earlier and are more sensitive than sCr in the setting of kidney injury. In fact, a detectable increase in sCr does not occur until 24-48 hours after a primary renal injury \[2\]. KIM-1, a type 1 transmembrane glycoprotein, is not expressed in normal kidneys, but is exhibited at high levels in proximal tubule epithelial cells and is excreted into the urine within 6-12 hours after toxic or ischemic injury \[10, 11, 12\]. Similarly, clusterin expression is induced by glomerular, tubular, or papillary kidney injury, with increased urine concentrations at 8, 24, and 48 hours post-toxic exposure \[12, 13\]. Cystatin C, a cysteine protease inhibitor, is not affected by age, gender, race, or overall muscle mass, making it a useful glomerular filtration marker in early KI. In a prospective study, cystatin C rise preceded an increase in sCr by 1-2 days \[14\]. FGF23 increases as early as within one hour of acute KI induction in mice and remains elevated in chronically compromised GFR, thereby serving as a useful biomarker in both acute and chronic KI \[15\]. In addition, plasma FGF23 has recently been show to be a strong predictor for risk of adverse outcomes in adults with acute kidney injury (AKI), with the median FGF 23 level obtained within 24 hours of admission being 5.5 times higher in patients with AKI compared to "control" ICU patients without kidney injury and almost 20 times higher than the upper limit of normal \[16\]. Therefore, KIM-1, clusterin, cystatin C, and FGF23 could provide earlier detection of kidney injury related to HDMTX when compared with sCr measurements. The utility of these biomarkers in identifying early kidney injury following treatment with HDMTX has not been previously studied. This study will compare these biomarkers with sCr pre-, intra-, and post-HDMTX treatment to determine their utility in detecting early kidney injury.