Pharmacokinetic adaptation of 5-fluorouracil: where are we and where are we going?

Fluorouracil plasma monitoring: systematic review and economic evaluation of the My5-FU assay for guiding dose adjustment in patients receiving fluorouracil chemotherapy by continuous infusion

BACKGROUND

5-Fluorouracil (5-FU) is a chemotherapy used in colorectal, head and neck (H&N) and other cancers. Dose adjustment is based on body surface area (BSA) but wide variations occur. Pharmacokinetic (PK) dosing is suggested to bring plasma levels into the therapeutic range to promote fewer side effects and better patient outcomes. We investigated the clinical effectiveness and cost-effectiveness of the My5-FU assay for PK dose adjustment to 5-FU therapy.

OBJECTIVES

To systematically review the evidence on the accuracy of the My5-FU assay compared with gold standard methods [high-performance liquid chromatography (HPLC) and liquid chromatography-mass spectrometry (LC-MS)]; the effectiveness of My5-FU PK dosing compared with BSA; the effectiveness of HPLC and/or LC-MS compared with BSA; the generalisability of published My5-FU and PK studies; costs of using My5-FU; to develop a cost-effectiveness model.

DATA SOURCES

We searched MEDLINE, EMBASE, Science Citation Index and other databases between January and April 2014.

METHODS

Two reviewers independently screened titles and abstracts with arbitration and consensus agreement. We undertook quality assessment. We reconstructed Kaplan-Meier plots for progression-free survival (PFS) and overall survival (OS) for comparison of BSA and PK dosing. We developed a Markov model to compare My5-FU with BSA dosing which modelled PFS, OS and adverse events, using a 2-week cycle over a 20 year time horizon with a 3.5% discount rate. Health impacts were evaluated from the patient perspective, while costs were evaluated from the NHS and Personal Social Services perspective.

RESULTS

A total of 8341 records were identified through electronic searches and 35 and 54 studies were included in the clinical effectiveness and cost-effectiveness reviews respectively. There was a high apparent correlation between My5-FU, HPLC and LC-MS/mass spectrometer but upper and lower limits of agreement were -18% to 30%. Median OS were estimated as 19.6 [95% confidence interval (CI) 17.0 to 21.0] months for PK versus 14.6 (95% CI 14.1 to 15.3) months for BSA for 5-FU+folinic acid (FA); and 27.4 (95% CI 23.2 to 38.8) months for PK versus 20.6 (95% CI 18.4 to 22.9) months for BSA for FOLFOX6 in metastatic colorectal cancer (mCRC). PK versus BSA studies were generalisable to the relevant populations. We developed cost-effectiveness models for mCRC and H&N cancer. The base case assumed a cost per My5-FU assay of £ 61.03. For mCRC for 12 cycles of a oxaliplatin in combination with 5-fluorouracil and FA (FOLFOX) regimen, there was a quality-adjusted life-year (QALY) gain of 0.599 with an incremental cost-effectiveness ratio of £ 4148 per QALY. Probabilistic and scenario analyses gave similar results. The cost-effectiveness acceptability curve showed My5-FU to be 100% cost-effective at a threshold of £ 20,000 per QALY. For H&N cancer, again, given caveats about the poor evidence base, we also estimated that My5-FU is likely to be cost-effective at a threshold of £ 20,000 per QALY.

LIMITATIONS

Quality and quantity of evidence were very weak for PK versus BSA dosing for all cancers with no randomised controlled trials (RCTs) using current regimens. For H&N cancer, two studies of regimens no longer in use were identified.

CONCLUSIONS

Using a linked evidence approach, My5-FU appears to be cost-effective at a willingness to pay of £ 20,000 per QALY for both mCRC and H&N cancer. Considerable uncertainties remain about evidence quality and practical implementation. RCTs are needed of PK versus BSA dosing in relevant cancers.

Long Survival and Severe Toxicity Under 5-Fluorouracil-Based Therapy in a Patient With Colorectal Cancer Who Harbors a Germline Codon-Stop Mutation in TYMS

We report the first clinical description of a patient with cancer with a heterozygous germline codon-stop mutation in the TYMS gene. The mutation g.657795_657826del, c.53_84del (NM_001071.2), p.Gln18Argfs*42 causes loss of function of one of the TYMS alleles, resulting in a truncated protein. This gene codifies for the target enzyme of 5-fluorouracil (5-FU), the basic treatment in colorectal cancer. The patient, diagnosed with metastatic colorectal cancer, had diarrhea and neutropenia grade 4 and mucositis and neurological toxicity grade 3 under 5-FU-based therapy and exceeded by more than 50% the average survival after metastasectomy. On the basis of the patient's characteristics and the key role of TYMS in 5-FU activity, we hypothesize that this mutation may contribute to the drug response and toxicities suffered by the patient.

Evaluation of 5-fluorouracil pharmacokinetic models and therapeutic drug monitoring in cancer patients

5-fluorouracil (5-FU) remains the cornerstone of all currently applied regimens for the treatment of patients with cancers of the gastrointestinal tract, breast, and head and neck. Unfortunately, a large variation in the clearance of 5-FU has been observed between patients, suggesting that some patients might receive nonoptimal 5-FU doses. However, therapeutic drug monitoring of 5-FU has been shown to result in reduced intra- and inter-individual variability in 5-FU plasma levels and pharmacokinetically guided dose adjustments of 5-FU-containing therapy results in a significantly improved efficacy and tolerability. To date, compartmental Michaelis-Menten elimination-based modeling has proven to be a sensitive and accurate tool for analyzing the pharmacokinetics of 5-FU and to identify patients with a dihydropyrimidine dehydrogenase deficiency. These Michaelis-Menten models also allow the use of a limited sampling strategy and offer the opportunity to predict a priori the 5-FU plasma concentrations in patients receiving adapted doses of 5-FU.

Personalized dosing via pharmacokinetic monitoring of 5-fluorouracil might reduce toxicity in early- or late-stage colorectal cancer patients treated with infusional 5-fluorouracil-based chemotherapy regimens

INTRODUCTION

Therapeutic plasma 5-fluorouracil (5-FU) levels are achieved in only 20% to 30% of patients with the current practice of administering 5-FU doses based on body surface area (BSA). Alternatively, 5-FU doses can be adjusted based on 5-FU pharmacokinetic (PK) monitoring. Although benefits of PK monitoring of 5-FU in metastatic colorectal cancer (CRC) have been reported, its utility among patients with early stage disease has not been reported.

PATIENTS AND METHODS

We retrospectively examined the effect of 5-FU PK monitoring in 84 CRC patients (49 stage IV and 35 stage II/III) receiving mFOLFOX6 (modifiedFOLFOX6; modified 5-fluorouracil, leucovorin, oxaliplatin protocol) or mFOLFIRI (modified 5-fluorouracil, leucovorin, irinotecan protocol). Forty-six of the 84 patients received 5-FU doses based on BSA and 38 received doses that were adjusted with PK monitoring. 5-FU plasma levels were measured using a nanoparticle immunoassay method.

RESULTS

5-fluorouracil PK monitoring significantly improved disease-free survival in stage II/III patients (P = .0429). There was also a trend towards improved progression-free survival among stage IV patients who had their 5-FU levels PK-monitored (P = .16). Moreover, 5-FU PK monitoring significantly reduced (P = .0437) and delayed (P = .0144) adverse effects in stage II/III patients. Toxicity occurred after the second 5-FU dose in the BSA group and after the sixth to seventh dose in the PK monitoring group. In stage IV patients, the onset of toxicities was also delayed with PK monitoring (P = .0605).

CONCLUSION

We provide evidence that PK monitoring of 5-FU is potentially beneficial for late stage and early stage CRC. These results contribute to the growing body of evidence regarding patient benefit when treatment decisions are based on the individual patient characteristics, in this case, a patients' 5-FU levels.