The aminosalicylate drugs, collectively referred to as 5-ASA (5-aminosalicylic acid) agents, are the cornerstone of pharmacological management for mild to moderate ulcerative colitis (UC) and serve a more limited role in Crohn's disease (CD). Their therapeutic action is localized to the bowel mucosa, and the primary challenge in their formulation has been delivering the active moiety to the inflamed segment while minimizing systemic absorption.
Sulfasalazine, the first aminosalicylate developed, is a prodrug consisting of 5-ASA linked by an azo bond to sulfapyridine, a sulfonamide antibiotic carrier molecule. Colonic bacteria express azo-reductase enzymes that cleave this azo bond in the distal small intestine and colon, releasing mesalamine (the active 5-ASA moiety) and sulfapyridine locally at the site of IBD (inflammatory bowel disease) mucosal inflammation. The therapeutic activity of sulfasalazine resides entirely in the mesalamine component; sulfapyridine is the carrier that prevents premature jejunal absorption of the active moiety and is itself largely responsible for the adverse effects of sulfasalazine.1 Sulfapyridine is acetylated in the liver; slow acetylators accumulate higher sulfapyridine plasma concentrations and experience more adverse effects than rapid acetylators. The adverse effects attributable to sulfapyridine include nausea, headache, oligospermia (reversible on discontinuation), hemolytic anemia in patients with G6PD (glucose-6-phosphate dehydrogenase) deficiency, and rare but serious immune-mediated reactions including agranulocytosis and pulmonary toxicity. The requirement for folic acid supplementation with sulfasalazine arises because the sulfapyridine moiety competitively inhibits intestinal folate absorption; 1 mg of folic acid daily is recommended for all patients on sulfasalazine.
Mesalamine (also known as mesalazine in non-American nomenclature) formulations were developed to deliver 5-ASA to the target mucosa without the sulfapyridine carrier and its associated adverse effects. Multiple delivery strategies are employed. pH-dependent release formulations (Asacol, Delzicol) use enteric coatings that dissolve at specific luminal pH thresholds: Asacol HD (high-dose formulation) releases at pH 7.0, targeting the terminal ileum and proximal colon; Eudragit S-coated preparations release at pH 7.0; Eudragit L-coated preparations release at pH 6.0, targeting the distal small intestine and colon. MMX (multi-matrix) technology (Lialda, Mezavant) incorporates mesalamine in a hydrophilic lipophilic matrix within an outer enteric coating, achieving slow, extended release throughout the entire colon from a once-daily tablet, which substantially improves adherence.2 Controlled-release granule formulations (Pentasa) use ethylcellulose-coated microspheres that release mesalamine continuously from the duodenum throughout the colon, making them useful for small bowel CD. Topical formulations, including rectal suppositories (targeting the rectum and distal 10 to 15 cm), enemas (targeting the left colon to the splenic flexure), and foam preparations, deliver high mucosal concentrations to the distal colon with minimal systemic absorption and are preferred for proctitis and left-sided UC.
The mechanism of action of 5-ASA agents is multifactorial. Mesalamine inhibits arachidonic acid metabolism through COX (cyclo-oxygenase) enzyme inhibition, reducing prostaglandin and thromboxane synthesis at the mucosal level. It suppresses NF-kB (nuclear factor kappa-light-chain-enhancer of activated B cells) transcription factor activation in epithelial and immune cells, reducing the expression of pro-inflammatory cytokines including IL-1 (interleukin-1), IL-6 (interleukin-6), TNF-alpha (tumor necrosis factor-alpha), and IL-8 (interleukin-8). Mesalamine also reduces reactive oxygen species production by mucosal neutrophils and macrophages and stimulates mucosal production of protective prostaglandins and lipoxins.3 The net effect is attenuation of the mucosal inflammatory cascade without the systemic immunosuppression associated with corticosteroids or immunomodulators.
The clinical role of 5-ASA agents is most firmly established in UC, where they are effective for both induction of remission in mild to moderate disease and for long-term maintenance of remission. Combined oral plus rectal mesalamine produces higher mucosal drug concentrations and superior clinical outcomes than either route alone in left-sided and extensive UC, and this combination is now recommended as first-line therapy for mild to moderate active UC by major guidelines.2 The role of 5-ASA agents in CD is more limited: they are not as effective as in UC, with meta-analyses showing at best modest benefit for induction of remission in colonic CD and no demonstrated benefit for prevention of post-operative recurrence. Olsalazine is an azo-bonded dimer of two 5-ASA molecules cleaved by colonic bacteria; its principal limitation is dose-dependent secretory diarrhea. Balsalazide is a prodrug linking 5-ASA to an inert carrier (4-aminobenzoyl-beta-alanine) that is cleaved by colonic bacteria without producing a pharmacologically active carrier; it is generally better tolerated than sulfasalazine.
For left-sided and extensive UC (ulcerative colitis), guidelines recommend combining oral mesalamine with rectal mesalamine (suppository for proctitis, enema for left-sided disease). The combination produces higher mucosal drug concentrations throughout the affected colon, faster symptom response, and superior endoscopic remission rates compared with oral alone. Patients who do not respond adequately to oral 5-ASA (5-aminosalicylic acid) alone should have rectal therapy added before escalating to corticosteroids or immunomodulators. Adherence to maintenance therapy is a major predictor of long-term remission; once-daily MMX (multi-matrix) mesalamine formulations improve adherence compared with older three or four times daily regimens.
Sulfasalazine competitively inhibits intestinal folate absorption through its sulfapyridine carrier moiety. All patients on sulfasalazine must receive folic acid 1 mg daily to prevent folate deficiency, particularly women of childbearing potential where folate deficiency carries neural tube defect risk. Before starting sulfasalazine, screen for G6PD (glucose-6-phosphate dehydrogenase) deficiency, as sulfapyridine can precipitate hemolytic anemia in G6PD-deficient patients. In patients who are G6PD-deficient, sulfapyridine-free mesalamine formulations are preferred.
Corticosteroids are highly effective inducers of remission in both UC (ulcerative colitis) and CD (Crohn's disease), acting through genomic mechanisms to suppress the transcription of numerous pro-inflammatory genes including those encoding cytokines, adhesion molecules, and enzymes of arachidonic acid metabolism. The principal mechanism is binding of the glucocorticoid-receptor (GR) complex to glucocorticoid response elements (GREs) in gene promoter regions, inducing anti-inflammatory gene products and transrepressing NF-kB (nuclear factor kappa-light-chain-enhancer of activated B cells) and AP-1 (activator protein-1) transcription factor activity.4 Despite their efficacy for induction, corticosteroids have no role in IBD (inflammatory bowel disease) maintenance therapy; they do not alter the natural history of disease, do not prevent relapse, and are associated with cumulative toxicity that makes long-term use unacceptable. A patient requiring corticosteroids more than once per year, or who cannot taper off steroids without disease relapse, meets the definition of steroid-dependent disease and requires escalation to immunomodulators or biologics.
For mild to moderate active CD involving the ileum or right colon, budesonide controlled-release (CIR (controlled ileal release)) at 9 mg once daily is preferred over systemic prednisone for its superior adverse effect profile. Budesonide is a synthetic glucocorticoid with high topical anti-inflammatory potency that undergoes extensive first-pass hepatic metabolism, primarily by CYP3A4 (cytochrome P450 3A4), with oral systemic bioavailability of only 10 to 15% compared with approximately 80 to 100% for prednisone.5 This first-pass effect substantially reduces the HPA (hypothalamic-pituitary-adrenal) axis suppression, Cushingoid features, bone mineral density loss, and glucose intolerance associated with conventional systemic steroids, though these effects are not entirely absent, particularly with prolonged use beyond 8 weeks. Budesonide MMX (multi-matrix) formulations deliver drug throughout the colon and are approved for mild to moderate UC. The ileal-release budesonide formulation is not effective for colonic CD because the drug is released and absorbed in the terminal ileum before reaching colonic disease, so colon-specific budesonide formulations are required for colonic IBD.
Moderate to severe active UC that fails to respond to oral corticosteroids or 5-ASA (5-aminosalicylic acid) agents requires IV (intravenous) corticosteroids, with IV hydrocortisone 300 mg/day in divided doses or IV methylprednisolone 40 to 60 mg/day as the standard options. The ACTH (adrenocorticotropic hormone) alternative has largely been abandoned in favor of direct corticosteroid administration. Response to IV corticosteroids should be assessed formally at 3 days using validated clinical indices; patients who fail to demonstrate response by day 3 should be considered for rescue therapy with IV cyclosporine or infliximab rather than continuing IV steroids beyond 5 to 7 days without response.6 Approximately 30 to 40% of patients admitted with severe acute UC will require colectomy during that hospitalization or within 12 months if biologic rescue therapy is also unsuccessful.
The systemic adverse effect profile of corticosteroids in IBD is identical to that in other contexts and is dose- and duration-dependent. HPA axis suppression with adrenal insufficiency risk on abrupt withdrawal occurs after as little as 3 weeks of systemic steroid use. Bone mineral density (BMD) loss begins within the first few months of steroid therapy and is most rapid in the first year; calcium 1000 to 1200 mg/day plus vitamin D 800 to 1000 IU (international units)/day supplementation should accompany any corticosteroid course expected to last more than 3 months. Glucose intolerance and frank steroid-induced diabetes, hypertension, fluid retention, myopathy, cataracts, and avascular necrosis of the femoral head are all established risks of prolonged steroid use that reinforce the steroid-sparing imperative in IBD management. Osteoporosis screening with dual-energy X-ray absorptiometry (DEXA) scanning is recommended for patients requiring repeated or prolonged steroid courses.
Budesonide CIR (controlled ileal release) 9 mg/day: preferred for mild to moderate ileocaecal or right colonic CD (Crohn's disease). Superior adverse effect profile due to 85 to 90% first-pass hepatic metabolism by CYP3A4 (cytochrome P450 3A4). Not appropriate for severe or extensive colonic disease. Metabolized by CYP3A4 – significant interaction with CYP3A4 inhibitors (azole antifungals, macrolides, HIV protease inhibitors), which can markedly increase systemic budesonide exposure. Prednisone 40 to 60 mg/day: required for moderate to severe disease and for disease distribution beyond the ileum and right colon. Taper over 8 to 16 weeks. Never use for maintenance. If steroid dependence develops, escalate to immunomodulator or biologic therapy.
AZA (azathioprine) and 6-MP (6-mercaptopurine) are thiopurine immunomodulators used for maintenance of remission in both UC (ulcerative colitis) and CD (Crohn's disease) and as combination partners with anti-TNF biologics to reduce immunogenicity. AZA is a prodrug that is non-enzymatically cleaved in erythrocytes and intestinal cells to release 6-MP; 6-MP is then the substrate for three competing metabolic pathways that determine both efficacy and toxicity.7 The anabolic pathway catalyzed by HPRT (hypoxanthine-guanine phosphoribosyltransferase) converts 6-MP to 6-TGN (6-thioguanine nucleotides), the active immunosuppressive metabolites that are incorporated into dividing lymphocyte DNA (deoxyribonucleic acid), terminating cell proliferation and inducing apoptosis. The catabolic pathway catalyzed by XO (xanthine oxidase) converts 6-MP to 6-thiouric acid, an inactive metabolite. A third pathway catalyzed by TPMT (thiopurine S-methyltransferase) converts 6-MP to 6-MMP (6-methylmercaptopurine), an inactive metabolite that does not accumulate in bone marrow but is associated with hepatotoxicity at high concentrations.
TPMT is encoded by a highly polymorphic gene with clinically significant pharmacogenomic implications. Approximately 89 to 94% of the population has normal TPMT activity (two functional alleles); 6 to 11% are intermediate metabolizers (one functional and one non-functional allele); and 0.3% are poor metabolizers (two non-functional alleles). In TPMT poor metabolizers (PM), the catabolic 6-MMP pathway is non-functional, causing 6-MP to be channeled entirely into the HPRT pathway, generating extremely high 6-TGN concentrations that cause severe, potentially life-threatening myelosuppression at standard doses.7 Before initiating AZA or 6-MP, TPMT phenotyping (enzyme activity assay) or genotyping is mandatory to identify poor metabolizers who require dose reduction to approximately 10% of standard dosing or should avoid thiopurines entirely. TPMT intermediate metabolizers require dose reduction of approximately 30 to 50% and careful CBC (complete blood count) monitoring. Even in patients with normal TPMT activity, myelosuppression can occur, and WBC (white blood cell) monitoring every 3 months is standard practice during maintenance therapy.
NUDT15 (nudix hydrolase 15) represents a second pharmacogenomic predictor of thiopurine myelotoxicity that is independent of TPMT and particularly important in East Asian populations. NUDT15 encodes a diphosphatase that inactivates thiopurine triphosphate metabolites; loss-of-function NUDT15 variants allow thioguanine triphosphates to accumulate in leukocytes, causing myelosuppression even in patients with normal TPMT activity. The NUDT15 R139C (arginine-to-cysteine substitution at position 139) variant (*2 allele) has an allele frequency of approximately 10% in East Asians but less than 0.2% in Europeans, making NUDT15 genotyping a recommended pre-treatment test for patients of East Asian ancestry.9 The Clinical Pharmacogenomics Implementation Consortium (CPIC) guidelines now recommend testing for both TPMT and NUDT15 before initiating thiopurine therapy, with dose reduction or alternative drug selection based on the results.
Therapeutic drug monitoring of thiopurine metabolites can guide dosing and distinguish inadequate response from non-adherence. 6-TGN (6-thioguanine nucleotide) levels in erythrocytes are measured as a surrogate for lymphocyte drug exposure; therapeutic 6-TGN concentrations for IBD (inflammatory bowel disease) have been defined as approximately 235 to 450 pmol per 8 × 108 erythrocytes. Patients with sub-therapeutic 6-TGN levels despite adequate doses may be rapid thiopurine metabolizers or non-adherent; those with supra-therapeutic levels are at increased myelosuppression risk. Elevated 6-MMP levels with low 6-TGN is a pattern called preferential methylation and is associated with hepatotoxicity (elevated ALT (alanine aminotransferase) and AST (aspartate aminotransferase)) and paradoxical therapeutic failure despite adequate dosing;88 this pattern can sometimes be corrected by adding low-dose allopurinol, a XO (xanthine oxidase) inhibitor, which blocks the catabolic pathway and shifts 6-MP metabolism toward the 6-TGN anabolic pathway, allowing dose reduction to avoid hepatotoxicity while maintaining therapeutic 6-TGN levels. This combination requires careful dose reduction of AZA to 25 to 33% of the original dose to avoid 6-TGN toxicity, and CBC monitoring must be intensified.10
The clinical role of thiopurines in IBD is as steroid-sparing maintenance agents rather than induction agents. Because the onset of therapeutic 6-TGN accumulation takes 3 to 6 months after starting AZA or 6-MP, thiopurines cannot induce remission rapidly enough to manage active disease flares; corticosteroids or biologics are required for induction while thiopurines are co-initiated for long-term maintenance. The combination of AZA with an anti-TNF (anti-tumor necrosis factor) biologic (infliximab or adalimumab) reduces the formation of anti-drug antibodies against the biologic and is more effective than either agent alone for induction and maintenance of remission in CD, as demonstrated in the SONIC (Study of Biologic and Immunomodulator Naive Patients in Crohn's Disease) trial.11 The risk of opportunistic infections, particularly Pneumocystis jirovecii pneumonia and herpes zoster, is elevated with combination immunosuppression; prophylaxis and vaccination decisions should be reviewed when starting combination therapy.
Allopurinol inhibits XO (xanthine oxidase), the enzyme that catabolizes 6-MP to inactive thiouric acid. When allopurinol is co-administered with AZA (azathioprine) or 6-MP (6-mercaptopurine) at full doses, the blocked catabolic pathway causes a massive increase in 6-TGN (6-thioguanine nucleotide) accumulation, which can cause fatal myelosuppression. The combination is intentionally used in IBD as a rescue strategy for preferential methylation (high 6-MMP, low 6-TGN), but ONLY with immediate reduction of the AZA dose to 25 to 33% of the original dose. Standard co-prescribing of allopurinol for gout in a patient on full-dose AZA without dose reduction is a potentially lethal drug interaction. Always check for thiopurine use before prescribing allopurinol.
MTX (methotrexate) is a folic acid antagonist that inhibits DHFR (dihydrofolate reductase), the enzyme responsible for reducing dihydrofolate to tetrahydrofolate (THF), the active folate cofactor required for one-carbon transfer reactions in purine and thymidylate synthesis. At the immunomodulatory doses used in IBD (inflammatory bowel disease; 15 to 25 mg once weekly), MTX's anti-inflammatory effects extend well beyond DHFR inhibition. The predominant anti-inflammatory mechanism at low doses is intracellular accumulation of MTX polyglutamate forms, which inhibit the 5-aminoimidazole-4-carboxamide ribonucleotide transformylase enzyme, leading to adenosine release from cells; adenosine acts on adenosine receptors on immune cells to suppress cytokine production, T cell proliferation, and neutrophil function.12 This adenosine-mediated anti-inflammatory mechanism is the primary basis for MTX's efficacy at doses that do not cause overt cytotoxicity.
The preferred route of MTX administration in CD (Crohn's disease) is IM (intramuscular) or SC (subcutaneous) injection at 15 to 25 mg once weekly, rather than oral administration. This preference has a clear pharmacokinetic basis: oral MTX bioavailability is highly variable (25 to 100%) due to saturable intestinal absorption at the folate carrier transporter (RFC1 (reduced folate carrier 1)), and bioavailability decreases substantially at doses above 15 mg. Parenteral administration bypasses intestinal absorption, providing reliable and consistently high systemic exposure at doses that would be unpredictably absorbed orally.12 All patients on MTX must receive folic acid supplementation at 1 mg daily (or 5 mg once weekly on a non-MTX day) to reduce the incidence of MTX-related mucositis, nausea, alopecia, and cytopenias without reducing the drug's anti-inflammatory efficacy, because these adverse effects arise from DHFR inhibition in rapidly dividing non-immune tissues while the anti-inflammatory effect occurs through the adenosine mechanism.
Hepatotoxicity is the most important chronic adverse effect of MTX in IBD. At the cumulative doses relevant to long-term IBD management, MTX causes hepatic fibrosis in a dose-dependent manner; risk factors include pre-existing liver disease, obesity, type 2 diabetes, excessive alcohol consumption, and total cumulative dose exceeding 1.5 g. LFT (liver function test) monitoring with ALT (alanine aminotransferase) and AST (aspartate aminotransferase) is performed at baseline and every 4 to 8 weeks during ongoing therapy. Persistently elevated transaminases (greater than twice the upper limit of normal on two consecutive measurements) warrant dose reduction; persistent elevation despite dose reduction requires discontinuation and hepatology evaluation. Liver biopsy to assess fibrosis stage is no longer routinely recommended for all patients but should be considered when cumulative doses exceed 1.5 g in high-risk individuals. Non-invasive fibrosis markers, including the Fibroscan (transient elastography) and the FIB-4 (fibrosis-4) index, a serum-based fibrosis score using age, ALT, AST, and platelet count), are increasingly used as surrogate monitoring tools.
MTX is absolutely contraindicated in pregnancy due to its teratogenicity and abortifacient properties; it is a potent inhibitor of trophoblast proliferation and fetal folate metabolism and has caused spontaneous abortion, fetal death, and major congenital malformations at doses used therapeutically. MTX must be discontinued at least 3 months before planned conception in women and ideally 3 months before in men, though the male fertility evidence for 3-month washout is less robust. During MTX therapy, reliable contraception is mandatory. The CBC (complete blood count) must be monitored at baseline and every 4 to 8 weeks during therapy to detect bone marrow suppression; cytopenias are more likely in patients with renal impairment, as MTX is predominantly renally excreted and accumulates with reduced GFR (glomerular filtration rate). MTX is contraindicated when estimated GFR falls below 30 mL/min/1.73m2.13
The comparative position of MTX versus thiopurines in IBD management is influenced by disease phenotype, sex, reproductive considerations, and prior treatment history. MTX is indicated for steroid-sparing maintenance in CD when thiopurines have failed or are contraindicated; evidence for MTX in UC (ulcerative colitis) is substantially weaker, with randomized controlled trials showing no benefit over placebo for UC maintenance, making thiopurines the preferred immunomodulator for UC. In CD, MTX and thiopurines have comparable efficacy as monotherapy steroid-sparing agents, and the choice between them often depends on TPMT (thiopurine S-methyltransferase) status, NUDT15 (nudix hydrolase 15) status, sex (MTX is preferred in men with CD because it avoids thiopurine-associated non-Hodgkin lymphoma risk, which is higher in men; MTX teratogenicity makes it less preferred in women of reproductive age), and tolerability. The methylenetetrahydrofolate reductase C677T (MTHFR C677T) polymorphism, which reduces folate metabolism enzyme activity, increases MTX toxicity risk modestly and may warrant higher folic acid supplementation doses, though routine MTHFR genotyping before MTX is not currently standard practice.13
MTX (methotrexate) is teratogenic and abortifacient. It must be discontinued at least 3 months before planned conception in women. Reliable contraception is mandatory during therapy and for 3 months after discontinuation. In men, MTX is also discontinued at least 3 months before a partner attempts conception, due to effects on spermatogenesis, though data supporting the male washout period are less definitive. Never prescribe MTX to a patient who may be pregnant or is actively attempting conception without explicit discussion and documented contraception plan. Counsel patients at initiation and at every prescription renewal.
Test TPMT and NUDT15 before starting any thiopurine. TPMT poor metabolizers and NUDT15 homozygous loss-of-function patients require major dose reductions or should use MTX instead.
AZA/6-MP are preferred for UC maintenance over MTX, which lacks proven efficacy in UC. For CD, both are acceptable; MTX is preferred in men (lower lymphoma risk from thiopurines) and when thiopurine intolerance or failure has occurred.
Neither AZA/6-MP nor MTX induces remission rapidly. Always bridge with corticosteroids or a biologic during the 3 to 6-month thiopurine onset period. Starting a thiopurine without a bridge drug in active disease is inadequate therapy.
MTX teratogenicity requires mandatory contraception. In women of reproductive age with CD, discuss family planning at initiation and factor it into drug choice.
Monitor CBCs and LFTs regularly on all immunomodulators. For AZA/6-MP: CBC every 3 months. For MTX: CBC and LFTs every 4 to 8 weeks.
Klotz U, Maier K, Fischer C, Heinkel K. Therapeutic efficacy of sulfasalazine and its metabolites in patients with ulcerative colitis and Crohn's disease. N Engl J Med. 1980;303(26):1499–1502.
doi:10.1056/NEJM198012253032602Rubin DT, Ananthakrishnan AN, Siegel CA, Sauer BG, Long MD. ACG clinical guideline: ulcerative colitis in adults. Am J Gastroenterol. 2019;114(3):384–413.
doi:10.14309/ajg.0000000000000152Desreumaux P, Colombel JF. Mechanisms of action of aminosalicylates. Aliment Pharmacol Ther. 1996;10(Suppl 2):41–49.
doi:10.1046/j.1365-2036.1996.22164009.xCoutinho AE, Chapman KE. The anti-inflammatory and immunosuppressive effects of glucocorticoids, recent developments and mechanistic insights. Mol Cell Endocrinol. 2011;335(1):2–13.
doi:10.1016/j.mce.2010.04.005Greenberg GR, Feagan BG, Martin F, et al. Oral budesonide for active Crohn's disease. N Engl J Med. 1994;331(13):836–841.
doi:10.1056/NEJM199409293311303Turner D, Walsh CM, Steinhart AH, Griffiths AM. Response to corticosteroids in severe ulcerative colitis: a systematic review of the literature and a meta-analysis. Clin Gastroenterol Hepatol. 2007;5(1):103–110.
doi:10.1016/j.cgh.2006.09.033Lennard L. The clinical pharmacology of 6-mercaptopurine. Eur J Clin Pharmacol. 1992;43(4):329–339.
doi:10.1007/BF02220605Dubinsky MC, Lamothe S, Yang HY, et al. Pharmacogenomics and metabolite measurement for 6-mercaptopurine therapy in inflammatory bowel disease. Gastroenterology. 2000;118(4):705–713.
doi:10.1016/S0016-5085(00)70140-5Yang SK, Hong M, Baek J, et al. A common missense variant in NUDT15 confers susceptibility to thiopurine-induced leukopenia. Nat Genet. 2014;46(9):1017–1020.
doi:10.1038/ng.3060Sparrow MP, Hande SA, Friedman S, et al. Allopurinol safely and effectively optimizes tioguanine metabolites in inflammatory bowel disease patients not responding to azathioprine and mercaptopurine. Aliment Pharmacol Ther. 2005;22(5):441–446.
doi:10.1111/j.1365-2036.2005.02583.xColombel JF, Sandborn WJ, Reinisch W, et al. Infliximab, azathioprine, or combination therapy for Crohn's disease. N Engl J Med. 2010;362(15):1383–1395.
doi:10.1056/NEJMoa0904492Feagan BG, Rochon J, Fedorak RN, et al. Methotrexate for the treatment of Crohn's disease. N Engl J Med. 1995;332(5):292–297.
doi:10.1056/NEJM199502023320503van Assche G, Dignass A, Bokemeyer B, et al. Second European evidence-based consensus on the diagnosis and management of ulcerative colitis, part 3: special situations. J Crohns Colitis. 2013;7(1):1–33.
doi:10.1016/j.crohns.2012.09.005