Antimetabolites are structural analogs of natural nucleotide precursors or cofactors that compete with endogenous substrates for critical enzymes in nucleic acid synthesis. By inhibiting these enzymes with high affinity, they produce a state of nucleotide starvation that is selectively more toxic to rapidly dividing cells than to resting ones. The folate antagonists, the oldest class of antimetabolites, remain among the most widely used cytotoxic agents in clinical oncology and represent a paradigm for understanding how enzyme inhibition kinetics, cellular retention mechanisms, and renal pharmacokinetics interact to determine both efficacy and toxicity.1
Methotrexate (MTX) inhibits dihydrofolate reductase (DHFR), the enzyme that reduces dihydrofolate (DHF) to tetrahydrofolate (THF), the active reduced folate carrier required for one-carbon transfer reactions in de novo purine synthesis and thymidylate synthesis. By blocking DHFR, MTX depletes the intracellular THF pool, halting synthesis of thymidylate and de novo purines. The cytotoxic effect is predominantly expressed in S phase (deoxyribonucleic acid [DNA] synthesis phase), making MTX cycle-specific with activity dependent on sustained drug exposure during S phase. Two features amplify MTX retention and cytotoxicity beyond what simple DHFR inhibition would predict. First, cellular influx of MTX occurs via the reduced folate carrier (RFC, encoded by SLC19A1), and some tumors additionally accumulate MTX via proton-coupled folate transporter (PCFT). Second, once inside the cell, MTX undergoes polyglutamation by folylpolyglutamate synthetase (FPGS), an enzyme that adds multiple glutamate residues to the gamma-carboxyl group of MTX. Methotrexate polyglutamates (MTXPGs) are retained intracellularly for weeks because, unlike the parent drug, they cannot exit through the RFC. MTXPGs also directly inhibit thymidylate synthase (TS) and aminoimidazole carboxamide ribonucleotide transformylase (ATIC), a folate-dependent enzyme in de novo purine synthesis, compounding the antiproliferative effect beyond DHFR inhibition alone.1
The pharmacokinetics of MTX are tightly dependent on renal function. MTX is excreted approximately 80 to 90% unchanged in urine via glomerular filtration and active tubular secretion by the organic anion transporter OAT3 (organic anion transporter 3). Its plasma half-life at conventional doses is approximately 8 to 15 hours, but at high doses (above 1,000 mg/m²) the terminal elimination half-life extends substantially because tubular secretion becomes saturated. Any drug that reduces renal blood flow, GFR (glomerular filtration rate), or organic anion tubular secretion prolongs MTX plasma half-life and dramatically increases toxicity. The most clinically important interactions are with nonsteroidal anti-inflammatory drugs (NSAIDs), which reduce renal prostaglandin synthesis and decrease GFR, and with proton pump inhibitors (PPIs), which compete for OAT3-mediated tubular secretion. Both interactions can increase MTX plasma concentrations two- to tenfold in patients receiving high-dose MTX, causing life-threatening mucositis, nephrotoxicity, myelosuppression, and neurotoxicity. Third-space fluid accumulations (ascites, pleural effusions) sequester MTX and release it slowly back into plasma after the infusion ends, prolonging exposure. All NSAIDs and PPIs must be discontinued before high-dose MTX administration, and large effusions must be drained first.12
High-dose MTX (HD-MTX) with leucovorin rescue is used in osteosarcoma, ALL (acute lymphoblastic leukemia) consolidation and CNS (central nervous system) prophylaxis, diffuse large B-cell lymphoma (DLBCL), and primary CNS lymphoma. Leucovorin (5-formyltetrahydrofolate) is a reduced folate that directly enters the THF pool without requiring DHFR, bypassing the DHFR block. Given 24 to 42 hours after MTX administration, leucovorin rescues normal cells that have been depleted of THF while tumor cells, which typically accumulate more MTXPG (methotrexate polyglutamates) and have lower levels of efflux transporters, are more selectively killed. Rescue is timed by serial plasma MTX level monitoring: standard rescue is continued until plasma MTX falls below 0.05 to 0.1 micromolar. Delayed clearance (elevated MTX at 24 or 48 hours) mandates enhanced leucovorin rescue at higher doses and more frequent intervals. Glucarpidase (carboxypeptidase G2), an enzyme that cleaves the terminal glutamate from MTX, rapidly reducing plasma MTX concentrations by more than 98% within 15 minutes, is reserved for cases of severe MTX toxicity or failure of enhanced leucovorin rescue.2
Pemetrexed is a multitargeted antifolate that inhibits three folate-dependent enzymes: TS (thymidylate synthase), DHFR (dihydrofolate reductase), and GARFT (glycinamide ribonucleotide formyltransferase, an enzyme in de novo purine synthesis). Unlike MTX, pemetrexed is approved specifically for malignant pleural mesothelioma and non-squamous non-small cell lung cancer (NSCLC). Its polyglutamation is more extensive than MTX, making intracellular retention and TS inhibition particularly prominent. The dose-limiting toxicities of pemetrexed are myelosuppression and mucositis, both substantially ameliorated by supplementation with folic acid (400 micrograms orally daily, starting at least 5 days before the first pemetrexed dose) and vitamin B12 (cobalamin; 1,000 micrograms intramuscularly approximately 1 week before the first dose, then every 3 cycles). Without these supplements, severe and potentially life-threatening myelosuppression occurs in a disproportionate fraction of patients. Dexamethasone is given the day before, day of, and day after pemetrexed to reduce the incidence of skin rash, a common adverse effect. Pemetrexed is also renally cleared and requires dose modification when creatinine clearance falls below 45 mL/min; NSAIDs should be held for 2 days before and 2 days after pemetrexed administration in patients with normal or mild renal impairment, and for 5 days before and 2 days after in patients with moderate impairment.3
Before any high-dose MTX infusion, three conditions must be verified and, if necessary, corrected: (1) Normal or near-normal serum creatinine with a calculated creatinine clearance above the institutional threshold (typically above 50 to 60 mL/min) — do not use estimated GFR alone, obtain a 24-hour urine creatinine clearance in borderline cases; (2) all NSAIDs and PPIs discontinued for at least 48 to 72 hours, and penicillin-class antibiotics (which compete for OAT3) held or avoided; and (3) large pleural effusions or ascites drained to minimize the third-space reservoir effect. Proceeding without these three checks risks a prolonged MTX exposure for which even aggressive leucovorin rescue may be insufficient. A plasma MTX level at 24 and 48 hours post-infusion must be obtained to guide leucovorin dosing and rescue duration.
Pyrimidine analogs constitute the largest and most clinically diverse subgroup of antimetabolites, encompassing agents that target DNA (deoxyribonucleic acid) synthesis at the level of thymidylate synthesis, those that are incorporated into DNA or RNA (ribonucleic acid) and disrupt template function, and those that inhibit ribonucleotide reductase to deplete the deoxyribonucleoside triphosphate pool. Understanding the biochemical mechanism of each agent at the level of its specific enzymatic target is essential for predicting toxicity patterns, scheduling strategies, and resistance mechanisms.4
The central target shared by 5-fluorouracil (5-FU), its prodrug capecitabine, and pemetrexed is thymidylate synthase (TS), the enzyme that catalyzes the reductive methylation of deoxyuridine monophosphate (dUMP) to deoxythymidine monophosphate (dTMP) using 5,10-methylenetetrahydrofolate (5,10-CH₂-THF) as the methyl donor and reductant. This reaction is the sole de novo source of thymidylate; its inhibition blocks DNA synthesis by depleting the dTMP (and ultimately dTTP [deoxythymidine triphosphate]) pool, leading to thymine-less death. The mechanism of TS inhibition by fluorodeoxyuridine monophosphate (FdUMP), the active metabolite of 5-FU, is mechanistically distinctive: FdUMP forms a covalent, ternary complex with TS and 5,10-CH₂-THF in which the fluorine atom at the C5 (carbon-5) position of FdUMP, replacing the hydrogen that would normally be transferred to form dTMP, prevents catalytic completion. Because fluorine cannot be transferred (the C-F bond is approximately 100-fold stronger than the C-H bond), the ternary complex is essentially irreversible under physiological conditions, permanently inactivating the TS molecule it has bound. This mechanism is clinically exploited by administering leucovorin (folinic acid) concurrently with 5-FU: leucovorin increases intracellular 5,10-CH₂-THF concentrations, stabilizing and enhancing the ternary complex and increasing the depth and duration of TS inhibition.4
The two distinct modes of 5-FU cytotoxicity must be understood because they determine which toxicities predominate with different administration schedules. The first mode, TS inhibition via FdUMP, is responsible for predominantly S-phase-specific cytotoxicity through thymidylate starvation. The second mode involves incorporation of fluorouridine triphosphate (FUTP) into RNA (ribonucleic acid), where it disrupts RNA processing by inhibiting the pseudouridine synthase activity of rRNA (ribosomal RNA) processing enzymes, interfering with ribosome assembly and protein synthesis. FUTP can also be incorporated into DNA (after conversion to FdUTP), causing single-strand breaks and activating the base excision repair pathway in a futile attempt to remove incorporated fluorouridine. With bolus 5-FU administration, high peak concentrations favor FUTP generation and RNA disruption, producing predominantly myelosuppression (bone marrow cells are rapidly dividing and vulnerable to RNA dysfunction) and less mucositis. With continuous infusion 5-FU, sustained lower concentrations favor FdUMP generation and TS inhibition, producing predominantly mucositis and hand-foot syndrome (palmar-plantar erythrodysesthesia, a painful reddening and peeling of the palms and soles mediated by fluoropyrimidine accumulation in eccrine sweat gland cells). The clinical implication is that continuous infusion 5-FU is better suited to tumors in which TS inhibition is the primary mechanism (colorectal cancer with FOLFOX or FOLFIRI), while bolus 5-FU (MAYO clinic regimen) has a different toxicity profile but is less commonly used for colorectal cancer today.4
Leucovorin (5-formyl-THF) is converted intracellularly to 5,10-methylene-THF, the folate cofactor that participates in the ternary complex with FdUMP and TS. Adding leucovorin before or concurrent with 5-FU increases the availability of 5,10-CH₂-THF, driving more ternary complex formation and stabilizing those complexes already formed. This biochemical modulation increases the response rate of 5-FU by approximately twofold in metastatic colorectal cancer compared to 5-FU alone. The leucovorin dose used clinically (either low-dose leucovorin 20 mg/m² or high-dose leucovorin 200 to 500 mg/m²) is empirically derived; both produce intracellular 5,10-CH₂-THF levels well above the saturation threshold for ternary complex formation, and clinical trials have not demonstrated superiority of high-dose over low-dose leucovorin.
5-Fluorouracil and its oral prodrug capecitabine are among the most widely administered cytotoxic agents globally, forming the pharmacological backbone of treatment for colorectal, gastric, esophageal, breast, and pancreatic cancers. Their pharmacokinetics are governed by a single enzyme, dihydropyrimidine dehydrogenase (DPD), which determines both the rate of systemic catabolism and the risk of life-threatening toxicity in individuals with genetic DPD deficiency.5
5-FU is an intravenous fluorinated uracil analog. After entering cells via uracil transporters, it is converted to three active metabolites by competing anabolic pathways: FdUMP (via the thymidine phosphorylase/thymidine kinase route), FUTP (via the uridine phosphorylase/uridine kinase route), and FdUTP (via reduction of FUTP). The balance between these pathways is influenced by the intracellular concentrations of competing substrates, the activity levels of the activating enzymes, and the administration schedule (bolus versus continuous infusion, as discussed in Section 2). The primary catabolic pathway is hepatic oxidation by DPD (dihydropyrimidine dehydrogenase), which converts 5-FU to the inactive dihydrofluorouracil (DHFU) and accounts for more than 85% of administered 5-FU elimination. DPD activity in the liver is the dominant pharmacokinetic determinant of 5-FU plasma half-life (approximately 10 to 20 minutes after bolus infusion under normal DPD activity) and of steady-state plasma concentrations during continuous infusion. Because DPD activity varies substantially among individuals (approximately fivefold variation in the normal population) and is influenced by diurnal rhythms, the pharmacokinetics of 5-FU are highly variable between patients, contributing to the wide inter-patient variation in both toxicity and efficacy observed clinically.5
DPD deficiency is a pharmacogenomic variant affecting approximately 3 to 8% of the general population, caused by loss-of-function variants in the DPYD (dihydropyrimidine dehydrogenase) gene. The most clinically significant variant is DPYD*2A (IVS14 [intron 14]+1G>A, a splice-site mutation causing exon 14 skipping and complete loss of DPD activity in homozygotes). DPYD*13 (I560S) and c.2846A>T (D949V) are additional loss-of-function or reduced-activity variants. Patients with complete DPD deficiency (homozygous DPYD*2A) who receive standard 5-FU doses experience severe and potentially fatal toxicity: profound myelosuppression, life-threatening mucositis, neurotoxicity, and multi-organ failure occurring within days of the first dose. Patients with partial DPD deficiency (heterozygotes, affecting approximately 3 to 5% of the population) experience severe grade 3 or 4 toxicity at standard doses in approximately 25 to 30% of cases. DPYD genotyping before 5-FU or capecitabine administration is now recommended by the European Medicines Agency (EMA) and is increasingly adopted in North America. Carriers of DPYD*2A require a starting dose reduction of at least 50%; homozygotes should not receive fluoropyrimidines or should receive only with extreme caution and therapeutic drug monitoring. The uracil breath test (UBT) and plasma uracil level (a substrate that accumulates when DPD is deficient) are phenotypic alternatives to genotyping for pre-treatment DPD assessment.6
Capecitabine is an orally administered prodrug that undergoes a three-step enzymatic conversion to 5-FU. Step 1: intestinal carboxylesterases convert capecitabine to 5-deoxy-5-fluorocytidine (5-DFCR). Step 2: cytidine deaminase (CDA), expressed in liver and tumor tissue, converts 5-DFCR to 5-deoxy-5-fluorouridine (5-DFUR). Step 3: thymidine phosphorylase (TP), which is expressed at substantially higher levels in many tumor types (colorectal, breast, gastric) than in surrounding normal tissue, converts 5-DFUR to 5-FU at the tumor site. This tumor-preferential final activation step is the theoretical basis for capecitabine's improved tolerability profile compared to continuous infusion 5-FU. Because capecitabine generates 5-FU locally in tumor tissue (and to a lesser extent in liver and intestinal mucosa), it produces predominantly a continuous infusion-like toxicity profile: hand-foot syndrome is the most frequent and dose-limiting adverse effect, occurring in 50 to 60% of patients, while myelosuppression is generally milder than with intravenous 5-FU regimens. Diarrhea is also frequent and may be severe.6
Capecitabine is subject to the same DPD deficiency risk as intravenous 5-FU, and DPYD genotyping is equally applicable. The standard dose is 1,250 mg/m² twice daily for 14 days of a 21-day cycle (or 1,000 mg/m² twice daily when combined with oxaliplatin in XELOX or with docetaxel). Dose reduction is required in renal impairment (creatinine clearance 30 to 50 mL/min mandates a 25% dose reduction; below 30 mL/min capecitabine is contraindicated).6
A major pharmacokinetic drug interaction of clinical importance involves capecitabine and warfarin. Capecitabine inhibits CYP2C9 (cytochrome P450 2C9), the primary enzyme responsible for metabolizing the more potent S-warfarin enantiomer. Capecitabine administration in patients on warfarin anticoagulation can increase the international normalized ratio (INR) dramatically, sometimes to dangerous levels, within days to weeks of starting capecitabine. The interaction is unpredictable in magnitude and can occur even at low capecitabine doses. INR must be monitored weekly or more frequently in patients on warfarin receiving capecitabine, and the warfarin dose almost always requires substantial downward adjustment. When possible, switching from warfarin to a direct oral anticoagulant (DOAC) before capecitabine initiation avoids this interaction entirely, as DOACs are not metabolized by CYP2C9.4
DPD deficiency is not a rare curiosity: partial deficiency (heterozygous DPYD loss-of-function variants) affects approximately 3 to 5% of the population, meaning that in any oncology practice administering 5-FU or capecitabine to 100 patients per year, 3 to 5 patients are at substantially increased risk of severe toxicity at standard doses. The argument that genotyping is unnecessary because most patients tolerate standard doses conflates population-level safety with individual patient risk. Pre-treatment DPYD genotyping or phenotypic DPD assessment (plasma uracil, uracil breath test) is a patient safety measure that allows proactive dose individualization rather than reactive dose reduction after a severe toxic event. Any patient presenting with unexpectedly severe mucositis, myelosuppression, or neurological toxicity within the first week of 5-FU or capecitabine should have urgent DPYD testing ordered.
Cytarabine and gemcitabine are cytidine analogs that enter cells through nucleoside transporters, require intracellular phosphorylation to their active triphosphate forms, and exert cytotoxicity predominantly through incorporation into DNA (deoxyribonucleic acid), where they terminate chain elongation and trigger apoptosis. Both are S-phase-specific agents whose clinical efficacy is tightly linked to scheduling strategy and intracellular activation enzyme activity.7
Cytarabine (ara-C, cytosine arabinoside) is the defining drug of acute myeloid leukemia (AML) therapy and has been used in AML induction for more than five decades. It enters cells via nucleoside transporters (predominantly hENT1, human equilibrative nucleoside transporter 1) and is phosphorylated sequentially by deoxycytidine kinase (dCK), UMP-CMP (uridylate-cytidylate monophosphate) kinase, and nucleoside diphosphate kinase to the active triphosphate ara-CTP (cytosine arabinoside triphosphate). Ara-CTP is incorporated into DNA by DNA polymerase alpha and delta in competition with the natural substrate deoxycytidine triphosphate (dCTP). Once incorporated, ara-C at the 3-prime terminus of the growing chain produces a steric block to further elongation because the 2-prime-hydroxyl group of arabinose (which is in the 2-prime position rather than the 2-prime-deoxy position of normal deoxyribose) prevents the conformational change required for addition of the next nucleotide. This results in single-strand termination and, when opposite a template, triggers the DNA damage response and apoptosis. Ara-C is S-phase-specific because DNA polymerases are only active during S phase, accounting for the clinical scheduling of cytarabine as a 7-day continuous infusion in AML induction (the "7+3" regimen, combining 7 days of continuous infusion cytarabine with 3 days of an anthracycline).78
The rate-limiting step in ara-C activation is phosphorylation by dCK (deoxycytidine kinase). Loss of dCK expression is the primary mechanism of ara-C resistance. Competing catabolism by cytidine deaminase (CDA), which deaminates ara-C to the inactive arabinofuranosyluracil (ara-U), and by 5-nucleotidase, which dephosphorylates ara-CMP (cytarabine monophosphate), limits intracellular ara-CTP accumulation. The therapeutic window of cytarabine is further defined by hENT1 expression: high hENT1 expression predicts better ara-C uptake and sensitivity, and gemcitabine shares this transporter dependence. At high-dose cytarabine (HiDAC) regimens (2 to 3 g/m² every 12 hours for 3 to 6 days), the passive diffusion component of uptake becomes proportionally more important, potentially overcoming hENT1 limitation. HiDAC achieves cytotoxic ara-CTP concentrations in the cerebrospinal fluid (CSF) and is used in AML consolidation and CNS (central nervous system) treatment of aggressive lymphomas. The cerebellar toxicity that occurs with high-dose cytarabine (ataxia, dysarthria, and nystagmus from Purkinje cell loss, occurring in approximately 10 to 25% of patients receiving HiDAC) is dose-dependent and cumulative; age above 50 years and renal impairment are the primary risk factors. Neurological examination before each HiDAC dose cycle is required; cerebellar signs mandate immediate discontinuation.8
Gemcitabine is a difluorodeoxycytidine analog (dFdC) with a 2-prime,2-prime-difluoro substitution on the ribose ring rather than the 2-prime-hydroxy arabinosyl configuration of cytarabine. It shares hENT1-mediated uptake and dCK-dependent activation with cytarabine but has several additional pharmacological mechanisms that contribute to its broad spectrum of clinical activity and its superior potency compared to cytarabine in solid tumors. After phosphorylation to the triphosphate dFdCTP (gemcitabine triphosphate), incorporation into DNA produces a masked chain termination: unlike ara-C, which causes an obvious conformational block, incorporated dFdCTP allows addition of one further nucleotide before elongation terminates (a phenomenon called masked chain termination), rendering the incorporated gemcitabine relatively resistant to proofreading exonucleases. Gemcitabine diphosphate (dFdCDP) is a potent irreversible inhibitor of ribonucleotide reductase (RNR), the enzyme that converts ribonucleoside diphosphates to deoxyribonucleoside diphosphates. RNR inhibition by dFdCDP reduces intracellular dCTP concentrations, which in turn reduces competition with dFdCTP for DNA polymerase, increasing the efficiency of dFdCTP incorporation. This self-potentiating mechanism is unique to gemcitabine and explains why its intracellular pharmacology is more favorable than cytarabine in many solid tumor contexts.79
Gemcitabine is used as a single agent or in combination for pancreatic adenocarcinoma (gemcitabine plus nab-paclitaxel or gemcitabine plus erlotinib in advanced disease), NSCLC (non-small cell lung cancer, gemcitabine plus platinum as first-line doublet), bladder cancer (gemcitabine plus cisplatin as an alternative to MVAC [methotrexate, vinblastine, doxorubicin, cisplatin]), and relapsed ovarian cancer (gemcitabine plus carboplatin for platinum-sensitive relapse). Its toxicity profile differs from cytarabine: myelosuppression (predominantly thrombocytopenia and neutropenia) is dose-limiting, and a unique gemcitabine-specific toxicity is pulmonary toxicity, manifesting as dyspnea and interstitial infiltrates in approximately 10 to 15% of patients, usually within the first few cycles. Gemcitabine-induced hemolytic uremic syndrome (HUS), a thrombotic microangiopathy (TMA) characterized by microangiopathic hemolytic anemia, thrombocytopenia, and acute kidney injury, occurs rarely (less than 1% of patients) but is potentially fatal and requires immediate drug discontinuation. Flu-like symptoms and low-grade fever are common and usually manageable with acetaminophen (paracetamol).7
The S-phase specificity of cytarabine is the pharmacological rationale for the 7-day continuous infusion in AML induction. At any given moment, only approximately 20 to 40% of leukemic blast cells are in S phase. A bolus dose would kill this fraction but miss the cells that enter S phase in the hours following drug clearance. Continuous infusion at a steady cytotoxic plasma concentration ensures that every leukemic blast that enters S phase during the 7-day treatment window is exposed to ara-C at a cytotoxic concentration. This is the same kinetic rationale as for all S-phase-specific antimetabolites and illustrates why schedule is as important as dose for cycle-specific agents.
Purine analogs exploit the dependence of lymphoid and myeloid malignancies on de novo and salvage purine synthesis pathways. Their clinical efficacy in diseases where other cytotoxics fail reflects the high activity of purine salvage enzymes in lymphocytes relative to other cell types. The most clinically critical pharmacogenomic consideration in this class is thiopurine methyltransferase (TPMT) activity, which determines 6-mercaptopurine dosing safety and is the paradigm for pharmacogenomics-guided cancer therapy.10
6-Mercaptopurine (6-MP) is a thiopurine prodrug used primarily in ALL (acute lymphoblastic leukemia) maintenance therapy and, as its prodrug azathioprine (which is cleaved to 6-MP in vivo), in inflammatory bowel disease and organ transplant immunosuppression. 6-MP requires intracellular activation by HGPRT (hypoxanthine-guanine phosphoribosyltransferase) to thioinosine monophosphate (TIMP), which is then converted through a series of enzymatic steps to the active TGNs (thioguanine nucleotides), principally 6-thioguanine triphosphate (6-TdGTP). TGNs are incorporated into DNA (deoxyribonucleic acid) in place of deoxyguanosine, triggering mismatch repair-mediated apoptosis and directly inhibiting de novo purine synthesis via feedback on the committed steps of the purine synthesis pathway. The primary catabolic pathway for 6-MP is methylation by TPMT (thiopurine S-methyltransferase), which converts 6-MP to the inactive metabolite 6-methylmercaptopurine (6-MMP) and its nucleotides, and oxidation by xanthine oxidase (XO) to the inactive 6-thiouric acid.10
TPMT activity is trimodally distributed in the population: approximately 90% of patients have high (wild-type) TPMT activity, approximately 10% have intermediate activity (heterozygous for one low-activity allele), and approximately 0.3% have low or absent activity (homozygous for low-activity alleles). Patients with low TPMT activity cannot methylate 6-MP efficiently and therefore shunt virtually all 6-MP through the HGPRT pathway, accumulating very high TGN (thioguanine nucleotide) concentrations and experiencing life-threatening myelosuppression at standard doses. ALL maintenance protocols now routinely include TPMT genotyping or phenotyping before starting 6-MP; patients with intermediate TPMT activity receive a 30 to 50% dose reduction and those with low activity receive 10 to 20% of the standard dose.10
The interaction between 6-MP and allopurinol is among the most dangerous drug interactions in clinical oncology. Allopurinol is a xanthine oxidase inhibitor used in gout and for hyperuricemia prophylaxis. Because XO is a primary catabolic enzyme for 6-MP, allopurinol co-administration blocks 6-MP catabolism, increasing 6-MP plasma concentrations approximately fourfold and dramatically increasing TGN accumulation and myelotoxicity. This interaction has caused fatal aplasia in patients receiving standard 6-MP doses with allopurinol. The standard management is to reduce the 6-MP dose to 25% of the usual dose when allopurinol cannot be avoided. The same interaction applies to azathioprine, which is metabolized to 6-MP in vivo; azathioprine doses must be reduced to 25 to 33% when combined with allopurinol. For tumor lysis syndrome prevention in patients who will receive 6-MP or azathioprine, rasburicase rather than allopurinol should be used as the uric acid-lowering agent whenever possible.10
Fludarabine is a fluorinated purine nucleoside analog that, after dephosphorylation to 2-fluoroadenosine (2-F-ara-A) in plasma, enters cells via nucleoside transporters and is phosphorylated by dCK (deoxycytidine kinase) to the active triphosphate 2-F-ara-ATP (fludarabine triphosphate). 2-F-ara-ATP inhibits DNA (deoxyribonucleic acid) polymerase alpha, DNA primase, and DNA ligase, blocking DNA synthesis; it is also incorporated into DNA, producing chain termination. Additionally, fludarabine inhibits ribonucleotide reductase and RNA (ribonucleic acid) polymerase II. Fludarabine is used in CLL (chronic lymphocytic leukemia) and low-grade lymphoma, typically as part of the FCR (fludarabine, cyclophosphamide, rituximab) regimen in fit CLL patients with mutated IGHV (immunoglobulin heavy chain variable region gene).11
Its most clinically significant adverse effect is profound and sustained immunosuppression: fludarabine is selectively toxic to CD4⁺ T lymphocytes (CD4 [cluster of differentiation 4] is the surface marker identifying T-helper lymphocytes), depleting them to very low levels for months to years after therapy. This immunosuppression creates risks for opportunistic infections including PCP (Pneumocystis jirovecii pneumonia), viral reactivations (CMV [cytomegalovirus], herpesvirus, Epstein-Barr virus), and fungal infections. PCP prophylaxis with trimethoprim-sulfamethoxazole (TMP-SMX) is mandatory during and for at least 6 to 12 months after fludarabine-containing regimens. Antiviral prophylaxis against herpesvirus reactivation (acyclovir or valacyclovir) is also standard. Fludarabine can induce or trigger autoimmune hemolytic anemia (AIHA) in CLL patients who have a pre-existing positive direct antiglobulin test (DAT); AIHA developing during fludarabine therapy is an indication to discontinue the drug immediately, as continuation worsens hemolysis.11
Cladribine (2-CdA, 2-chlorodeoxyadenosine) is a purine analog with high selectivity for lymphocytes because these cells have high dCK (activating enzyme) activity and low 5-nucleotidase (deactivating enzyme) activity, leading to selective TGN accumulation. It is the drug of choice for hairy cell leukemia, producing durable complete remissions in more than 90% of patients with a single 7-day continuous infusion course. It is also active in CLL, low-grade lymphomas, and Waldenstrom macroglobulinemia. Like fludarabine, cladribine causes profound immunosuppression requiring PCP and herpesvirus prophylaxis. Clofarabine is a second-generation purine analog combining structural features of both cladribine and fludarabine, designed to resist deamination by adenosine deaminase (ADA) and to achieve higher intracellular triphosphate concentrations. It is approved for relapsed or refractory ALL in pediatric patients and has activity in AML (acute myeloid leukemia). Its dose-limiting toxicities are myelosuppression and hepatotoxicity; systemic inflammatory response syndrome (SIRS) is a recognized adverse effect during clofarabine infusion, managed with supportive care and steroid premedication in some protocols.7
Every patient starting 6-MP or azathioprine should have TPMT status established (genotype or phenotype) before the first dose. Low TPMT activity at standard doses causes life-threatening aplasia. The dosing strategy must be documented in the chart and communicated to all prescribers. Separately, every prescriber seeing a patient on 6-MP or azathioprine must know that adding allopurinol at any dose, without reducing the thiopurine dose to 25%, risks fatal myelosuppression. This interaction is not theoretical: multiple fatalities have occurred from this drug combination. Both checks should be part of every oncology and gastroenterology new-patient protocol when thiopurines are prescribed.
Hypomethylating agents represent a mechanistically distinct subclass of pyrimidine antimetabolites whose primary cytotoxic effect is not nucleotide starvation but reactivation of epigenetically silenced tumor suppressor genes through inhibition of DNA (deoxyribonucleic acid) methyltransferase. Understanding their unique mechanism and the broad resistance mechanisms that affect the entire antimetabolite class is essential for predicting cross-resistance patterns and guiding combination strategies.12
Azacitidine and decitabine are cytidine analogs in which a nitrogen atom replaces the carbon at position 5 of the cytosine ring (azacytosine), preventing the normal methylation of cytosine residues in newly synthesized DNA. After phosphorylation to their respective triphosphates and incorporation into DNA during S phase, they form a covalent, irreversible bond with the catalytic cysteine residue of DNMT1 (DNA methyltransferase 1), the maintenance methyltransferase responsible for replicating the parental methylation pattern on newly synthesized daughter strands. This trapping of DNMT1 on DNA triggers its proteasomal degradation, depleting cellular DNMT1 levels. As cells divide in the continued presence of drug, the methylation marks on CpG (cytosine-phosphate-guanine dinucleotide) islands in the promoters of epigenetically silenced genes are progressively lost from daughter strands, a process called passive demethylation. Tumor suppressor genes silenced by promoter hypermethylation in myelodysplastic syndrome (MDS) and AML (acute myeloid leukemia), including p15/CDKN2B (cyclin-dependent kinase inhibitor 2B), p16/CDKN2A (cyclin-dependent kinase inhibitor 2A), and CDH1 (E-cadherin), can be reactivated when their promoter CpG islands are demethylated, restoring differentiation, apoptosis responsiveness, and cell cycle checkpoint function. At high doses, azacitidine and decitabine also exert direct cytotoxic effects through DNA damage and apoptosis, but the clinical doses used in MDS and AML are chosen to maximize the epigenetic effect over direct cytotoxicity.12
The structural difference between azacitidine and decitabine determines their subcellular distribution and clinical use. Azacitidine is a ribonucleoside that is incorporated into both RNA (ribonucleic acid) and DNA (after reduction by ribonucleotide reductase to the deoxyribonucleoside form). The RNA incorporation of azacitidine contributes additional mechanisms of action including disruption of RNA processing and protein synthesis. Decitabine is a deoxyribonucleoside that is incorporated exclusively into DNA, making its DNMT1 trapping and demethylating effect more focused. Azacitidine is administered subcutaneously or intravenously for 7 days every 28 days (75 mg/m²/day) in MDS and is also approved in AML unfit for intensive chemotherapy. Decitabine is administered intravenously (20 mg/m²/day for 5 days every 28 days) in MDS and AML. Both agents are also available in oral formulations (oral azacitidine, CC-486 [the trade development code for oral azacitidine], approved for AML maintenance after intensive chemotherapy). The clinical benefit of hypomethylating agents in MDS often requires three to six treatment cycles before a response is apparent, reflecting the time required for passive demethylation to accumulate across multiple cell divisions. Patients who discontinue prematurely before an adequate trial period may not derive the full benefit.13
The principal resistance mechanisms across the antimetabolite class share several themes. Reduced drug uptake is relevant to all agents that rely on transporter-mediated entry: loss of RFC (reduced folate carrier, encoded by SLC19A1) expression or mutations in SLC19A1 reduce MTX (methotrexate) transport; loss of hENT1 expression (encoded by SLC29A1) reduces cytarabine and gemcitabine uptake. Reduced activating enzyme expression is the most drug-specific resistance mechanism: loss of dCK expression causes resistance to cytarabine, gemcitabine, cladribine, and clofarabine; loss of HGPRT (hypoxanthine-guanine phosphoribosyltransferase) expression causes resistance to 6-MP and 6-thioguanine; loss of FPGS (folylpolyglutamate synthetase) expression reduces MTX polyglutamation and intracellular retention, reducing efficacy at a given extracellular MTX concentration. Target amplification is exemplified by DHFR (dihydrofolate reductase) gene amplification causing MTX resistance, and TS (thymidylate synthase) amplification or TS point mutations causing 5-FU and pemetrexed resistance. Increased expression of the catabolic enzyme CDA (cytidine deaminase) causes resistance to cytarabine and gemcitabine by increasing deamination to inactive metabolites; DPD (dihydropyrimidine dehydrogenase) overexpression in tumors causes 5-FU resistance. Finally, loss of mismatch repair (MMR) confers resistance to the thiopurines (6-MP, 6-TG) and to agents whose primary mechanism of lethality is MMR-mediated apoptosis triggered by the DNA mismatch.7
Clinicians and patients should be counseled that hypomethylating agents do not produce rapid cytoreduction like conventional intensive chemotherapy. The mechanism, progressive passive demethylation across multiple cell divisions, requires sequential cycles of drug exposure and cell division before sufficient promoter CpG methylation loss accumulates to reactivate tumor suppressors. In MDS, the median time to first response with azacitidine is three to four cycles, and patients who achieve a response often do so between cycles 4 and 6. Discontinuing HMA therapy after one or two cycles due to absence of visible response deprives patients of the time required for the epigenetic mechanism to operate. A minimum of 4 to 6 cycles should be administered before declaring treatment failure, provided hematological tolerance allows continuation.
Jolivet J, Cowan KH, Curt GA, Clendeninn NJ, Chabner BA. The pharmacology and clinical use of methotrexate. N Engl J Med. 1983;309(18):1094–1104.
doi:10.1056/NEJM198311033091805Howard SC, McCormick J, Pui CH, Buddery RL, Henderson ES. Preventing and managing toxicities of high-dose methotrexate. Oncologist. 2016;21(12):1471–1482.
doi:10.1634/theoncologist.2015-0164Adjei AA. Pemetrexed (Alimta): a novel multitargeted antineoplastic agent. Clin Cancer Res. 2004;10(12 Pt 2):4276s–4280s.
doi:10.1158/1078-0432.CCR-040010Longley DB, Harkin DP, Johnston PG. 5-Fluorouracil: mechanisms of action and clinical strategies. Nat Rev Cancer. 2003;3(5):330–338.
doi:10.1038/nrc1074Miura K, Kinouchi M, Ishida K, et al. 5-FU metabolism in cancer and orally-administrable 5-FU drugs. Cancers (Basel). 2010;2(3):1717–1730.
doi:10.3390/cancers2031717Deenen MJ, Meulendijks D, Cats A, et al. Upfront genotyping of DPYD*2A to individualize fluoropyrimidine therapy: a safety and cost analysis. J Clin Oncol. 2016;34(3):227–234.
doi:10.1200/JCO.2015.63.1325Galmarini CM, Mackey JR, Dumontet C. Nucleoside analogues: mechanisms of drug resistance and reversal strategies. Leukemia. 2001;15(6):875–890.
doi:10.1038/sj.leu.2402114Estey EH. Acute myeloid leukemia: 2021 update on risk-stratification and management. Am J Hematol. 2020;95(11):1368–1398.
doi:10.1002/ajh.25975Heinemann V, Xu YZ, Chou T, et al. Inhibition of ribonucleotide reduction in CCRF-CEM cells by 2',2'-difluorodeoxycytidine. Mol Pharmacol. 1990;38(4):567–572.
Lennard L. TPMT in the treatment of Crohn's disease with azathioprine. Gut. 2002;51(2):143–146.
doi:10.1136/gut.51.2.143Keating MJ, O'Brien S, Albitar M, et al. Early results of a chemoimmunotherapy regimen of fludarabine, cyclophosphamide, and rituximab as initial therapy for chronic lymphocytic leukemia. J Clin Oncol. 2005;23(18):4079–4088.
doi:10.1200/JCO.2005.12.051Stresemann C, Lyko F. Modes of action of the DNA methyltransferase inhibitors azacytidine and decitabine. Int J Cancer. 2008;123(1):8–13.
doi:10.1002/ijc.23607Fenaux P, Mufti GJ, Hellstrom-Lindberg E, et al. Efficacy of azacitidine compared with that of conventional care regimens in the treatment of higher-risk myelodysplastic syndromes: a randomised, open-label, phase III study. Lancet Oncol. 2009;10(3):223–232.
doi:10.1016/S1470-2045(09)70003-8