Precision Management of a Patient With Dihydropyrimidine Dehydrogenase Deficiency and Liver-Predominant Metastatic Rectal Cancer Using Hepatic Arterial Floxuridine

Background

Dihydropyrimidine dehydrogenase (DPD), encoded by the DYPD gene, is the critical enzyme in the catabolism and detoxification of fluorouracil (FU) and its orally active prodrug, capecitabine.¹ However, patients with deficiency of DPD cannot safely receive these drugs. Precise strategies to manage metastatic colorectal cancer in patients with DPD deficiency have not been reported.

Many genetic polymorphisms in the DYPD gene exist, of which four specific variants have been consistently associated with National Cancer Institute Common Terminology Criteria for Adverse Events² grade ≥ 3 toxicities^1,3 and are recognized by the Clinical Pharmacogenetics Implementation Consortium.⁴ The most common no function variant (carrier frequency 1.6%) is c.1905+1G>A (previously IVS14+1G>A or DPYD*2A), which results in production of a nonfunctional enzyme,^4-8 although with significant phenotypic variation among patients.^9,10 The c.1129-5923C>G variant is the most common decreased function DPYD allele in Europeans (carrier frequency 4.7%).^4,8,11 The two other variants, which have been associated with absent and low enzyme activity respectively,⁸ include nonsynonymous c.1679T>G and c.2846A>T mutations.^12-15 These variants account for only a small fraction of reported FU toxicity.¹

Screening for patients receiving fluoropyrimidine-based therapies is not currently standard in the United States. We therefore identify patients with genetic polymorphisms in the DYPD gene only after they present with clinical toxicity. Practices vary around the world, however, and upfront DPD deficiency screening is recommended by the Dutch Pharmacogenetics Working Group¹⁶ and European Medicines Agency¹⁷ and is mandated in Germany.¹⁸ Among patients who have a known variant allele associated with toxicity or reduced DPD activity by functional testing, dose reduction and subsequent titration can allow for safer treatment with fluoropyrimidines.^19-21

The Clinical Pharmacogenetics Implementation Consortium has published a guideline to aid management of patients with DPD deficiency⁸; however, there is no published literature describing the unique use of the hepatic arterial infusion (HAI) of floxuridine (brand name FUDR) in these patients. We noted that when floxuridine is administered via HAI, it is converted predominantly to fluorodeoxyuridylate monophosphate (FdUMP), rather than to FU (Fig 1). We therefore hypothesized that floxuridine could inhibit tumor growth without increasing toxicity in patients with DPD deficiency. Here, to our knowledge, we report the first case of a patient with genotypic and phenotypic DPD deficiency who was treated with full-dose floxuridine via HAI without toxicity and with clinical response.

FIG 1.

Hepatic metabolism of FU, capecitabine, and floxuridine.²² If given by rapid intra-arterial injection, floxuridine can be converted to FU by deoxyuridine or TYMP and therefore potentially expected to cause toxicity in a patient with DPD deficiency. However, when floxuridine is given by hepatic arterial infusion (orange arrows), the metabolism shifts to the TK1/TYMS pathway. Because of rapid exposure to deoxyuridine or thymidine kinase within the liver, floxuridine can be converted to FdUMP and inhibit TYMS, which, in turn, can inhibit tumor growth in the liver with minimal conversion to FU (represented by red X). Adapted from 23. DHFU, dihydrofluorouracil; DPD, dihydropyrimidine dehydrogenase; DYPS, dihydropyrimidinase; FBAL, fluoro-beta-alanine; FdUMP, fluorodeoxyuridine monophosphate; FU, fluorouracil; FUPA, fluoro-beta-ureidopropionate; TK1, thymidylate kinase 1; TYMP, thymidylate phosphorylase; TYMS, thymidylate synthase; UPB1, beta-ureidopropionase.

Case Presentation

A 40-year-old man was diagnosed with metastatic moderately differentiated rectal adenocarcinoma, microsatellite stable and RAS/RAF wild type. Positron emission tomography/computed tomography demonstrated the primary rectosigmoid neoplasm, regional lymphadenopathy, and bilobar hepatic metastases. The carcinoembryonic antigen level was 3,884 ng/mL. The patient received systemic chemotherapy with the fluorouracil + oxaliplatin regimen, consisting of FU, leucovorin, and oxaliplatin. He presented 12 days later in distress with fever to 101.6°F, oropharyngeal mucositis, and profound (grade 4) neutropenia (absolute neutrophil count [ANC] of 0.0 K/μL). He experienced complete alopecia. His clinical presentation was determined to be the result of FU toxicity related to DPD deficiency.

The patient described has provided consent for publication of this material related to the patient.

DPYD gene mutation analysis (Quest Diagnostics, San Juan Capistrano, CA²⁴) identified one copy of the IVS14+1G>A mutation (c.1905+1G>A, DPYD*2A), which is associated with a dysfunctional DPD protein and increased risk for adverse reactions to fluoropyrimidines.²⁵

The patient required a prolonged 4-week recovery and then received irinotecan monotherapy for 5 months before disease progression. He had a response over the next 9 months to continuous treatment with irinotecan-containing doublet therapy, first with the anti-EGFR monoclonal antibody panitumumab and subsequently with rechallenge of oxaliplatin (irinotecan + oxaliplatin), but then experienced painful progression of the primary rectal tumor and clinical deterioration with significant hepatomegaly and new ascites.

He received short-course radiation to the primary rectal tumor (2,500 cGy over five fractions).²⁶ During the radiation course, the patient was admitted with a pain crisis. Physical examination and repeat imaging showed increased hepatosplenomegaly, moderate volume ascites, and a new right pleural effusion (Fig 2). Laboratory workup was notable for pancytopenia with an ANC of 800/μL, a hemoglobin level of 8.6 g/dL, and a platelet level of 97,000/μL.

FIG 2.

CT of abdomen before and after initiation of floxuridine via HAI pump. (A) Axial and (B) coronal baseline (before floxuridine) CT images of the abdomen after progression on IROX and before initiation of floxuridine via HAI pump. This scan shows hepatomegaly, splenomegaly, and ascites with numerous bilobar hepatic metastases. (C) Axial and (D) coronal (after floxuridine) CT images of the abdomen after four cycles of floxuridine at full dose. This scan shows the decrease in size and attenuation of multiple hepatic metastases with no evidence of ascites and reduced hepatomegaly. CT, computed tomography; HAI, hepatic arterial infusion; IROX, irinotecan + oxaliplatin.

For management of the painful liver metastases, we considered the remaining standard options of regorafenib or trifluridine/tipiracil, neither of which offered a promising chance of major tumor response. Although there were no prior reports of the use of HAI of floxuridine in patients with DPD deficiency, floxuridine metabolism, when administered by HAI, shifts to the thymidine kinase (TK) 1/thymidylate synthase (TYMS) pathway and thus does not require DPD for metabolism (Fig 1). Therefore, after completing radiation, the patient underwent HAI pump placement. He received the first treatment by HAI of floxuridine, which is usually administered in the outpatient clinic, during his hospitalization. The standard dose of floxuridine via HAI at our institution is [(0.12 mg/kg/day) × wt(kg) × (30mL)) / pump flow rate] and dexamethasone [(1 mg/day*30) / pump flow rate] on day 1 of each cycle over 14 days of a 28-day cycle.^27,28 Out of caution, this patient received a first dose of [(0.04 mg/kg/day) × wt(kg) × (30mL)) / pump flow rate] and dexamethasone [(1mg/day*30) / pump flow rate] on day 1 of each cycle over 14 days of a 28-day cycle, or 33% of the standard dose. We planned to titrate the dose using the patient's clinical tolerance of the drug as a precision strategy. He was monitored with daily complete blood counts and comprehensive metabolic panels and was discharged from the hospital on postoperative day 14, at which time the ANC had improved to 1,700/μL, with a hemoglobin level of 8.2 g/dL and a platelet count 104,000/μL.

Four weeks after his first dose of floxuridine, the patient's liver function tests and blood counts were stable and he received the full therapeutic dose of floxuridine via HAI of [(0.12 mg/kg/day) × wt(kg) × (30mL)) / pump flow rate] and dexamethasone [(1 mg/day * 30) / pump flow rate] on day 1 of each cycle over 14 days of a 28-day cycle. He tolerated this well and received floxuridine via HAI for 11 months. His initial restaging scans demonstrated continuous response in his hepatic metastases without progression at other disease sites, and his blood counts remained stable (Fig 2). The carcinoembryonic antigen level continued to trend down (Fig 3).

FIG 3.

CEA trend throughout the course of treatment. Time zero represents the start of FOLFOX and the first CEA value documented. The patient was admitted with severe neutropenia, fever, and diarrhea 12 days after receiving his first dose of FOLFOX (black square). CEA began to trend down after only one cycle of FOLFOX although he required 4 weeks to recover. He started treatment with irinotecan at 1 month and received treatment for nearly 5 months before progression (red bar and shaded area). He then transitioned to irinotecan + panitumumab (teal bar and shaded area), but experienced disease progression with concurrent rise in CEA. Although CEA did come down from its peak during treatment with IROX (orange bar and shaded area), this peak was thought to be in part related to inflammation. He overall had clinical progression on IROX, and the CEA remained markedly elevated compared with the pretreatment baseline. Progression was at both the primary rectal tumor and liver, and his CEA declined after receiving palliative radiotherapy to the rectum (black diamond and gray bar). The CEA again rose because of progressive liver disease and possibly inflammation from HAIP placement but plateaued after only two doses of floxuridine (purple bar and shaded area) and has been subsequently downtrending. CEA, carcinoembryonic antigen; CT, computed tomography; FOLFOX, fluorouracil + oxaliplatin; HAIP, hepatic arterial infusion pump; IROX, irinotecan + oxaliplatin.

Discussion

To our knowledge, this is the first report of a patient with genomic and phenotypic DPD deficiency treated with a fluoropyrimidine by HAI and the first precision approach to managing metastatic rectal cancer in a patient with DPD deficiency. If given by rapid intra-arterial injection, floxuridine would be converted to FU^28,29 by deoxyuridine or thymidine phosphorylase and would be expected to cause toxicity in a patient with DPD deficiency. However, when floxuridine is given by protracted HAI, the metabolism shifts to the TK1/TYMS pathway.^30,31 Because of exposure to tissue deoxyuridine or TK within the liver,³⁰ floxuridine can be converted to FdUMP and inhibit TYMS, which, in turn, can inhibit tumor growth in the liver with minimal conversion to FU (Fig 1). Since this metabolic pathway obviates the need for drug catabolism by DPD, floxuridine administered by HAI may be a novel therapeutic option for patients with DPD deficiency with liver metastases.

Prior publications support the clinical relevance of this metabolic pathway. None of the genes associated with floxuridine/FU metabolism are significantly correlated with survival or recurrences in patients treated with adjuvant floxuridine via HAI, possibly since floxuridine activity is not achieved via conversion by TP to FU, but instead via direct conversion of floxuridine to FdUMP.³² Similarly, there is no correlation between TP or DPD levels and floxuridine sensitivity.³³ Finally, hepatic removal of floxuridine with hepatic arterial administration is efficient, with 94%-99% of the drug extracted; by contrast, hepatic removal of FU is less efficient, with only 19%-51% of the drug extracted.³⁴

This case demonstrates the feasibility and efficacy of administering floxuridine via HAI in a patient with clinically relevant DPD deficiency. Our precision treatment strategy leverages the underlying pharmacology of protracted infusional floxuridine and suggests a basis for future investigation into novel treatment strategies for patients with DPD deficiency and tumors that are routinely treated with fluoropyrimidine-based therapy.

AUTHOR CONTRIBUTIONS

Conception and design: Melissa Lumish, Jennifer Thackray, Andrea Cercek

Collection and assembly of data: Melissa Lumish, William R. Jarnagin, Andrea Cercek

Data analysis and interpretation: All authors

Manuscript writing: All authors

Final approval of manuscript: All authors

Accountable for all aspects of the work: All authors

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated unless otherwise noted. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or ascopubs.org/po/author-center.

Open Payments is a public database containing information reported by companies about payments made to US-licensed physicians (Open Payments).