Article Text


FRI0014 Factors predicting methotrexate toxicity
  1. M Hoekstra1,
  2. A Van Ede2,
  3. C Haagsma1,
  4. M Van de Laar1,
  5. R Laan2
  1. 1Department of Rheumatology, Medisch Spectrum Twente, Enschede, The Netherlands
  2. 2Department of Rheumatology, University Hospital St Radboud, Nijmegen, The Netherlands


Background Methotrexate is the most effective treatment in rheumatoid arthritis. The treatment is limited by adverse events. Gastro-intestinal toxicity and hepatotoxicity occur the most. Various different factors have been reported to contribute to methotrexate toxicity.

Objectives To determine predictive factors for developing methotrexate toxicity, and methotrexate withdrawal because of toxicity.

Methods Data from a double-blind placebo controlled multicenter study on patients with active rheumatoid arthritis were used. In this trial the effect of folic acid (1‑2 mg/day), folinic acid (2.5 mg/week) and placebo along methotrexate therapy (7.5-maximum 25 mg/week) was compared. The patients never had methotrexate before, had active disease and a creatinine clearance >50 ml/min. The three groups contained 133, 141 and 137 patients respectively. The follow-up was 48 weeks.

By means of multiple stepwise regression analysis we studied the relation between baseline variables and the following: hepatotoxicity (ALAT >3×ULN), the ten most frequent subjective adverse events, and methotrexate withdrawal because of toxicity.

With the different toxicities we studied specific independent factors, for example NSAID’s and alcohol use in hepatotoxicity.

Results Hepatotoxicity:

We found a significant association between folate suppletion and lack of hepatotoxicity (p < 0.0001). Smoking was negatively associated with hepatotoxicity, with less hepatotoxicity when the patient smoked (p = 0.003).

(P = 0.06 + 0.237 no folate – 0.023 smoking)

The ten most subjective adverse events:

  1. Upper abdominal pain was significantly related to corticosteroid use (p = 0.047) (P = 0.183 + 0.109 corticosteroid)

  2. Fatigue/malaise was significantly associated with smoking (p = 0.049) (P = 0.231–0.021 smoking)

  3. Diarrhoea is significantly associated with age and a positive gasro-intestinal history, e.g. peptic ulcers and abdominal surgery (p = 0.003 and p = 0.012) (P = 0.291–0.0037age + 0.087GI-history)

There were no significant associations between the other subjective adverse events (nausea, dizziness, headache, rash, stomatitis, cough and alopecia) and baseline characteristics.

Methotrexate withdrawal:

Significant factors predicting methotrexate withdrawal are folate suppletion, smoking and gastro-intestinal history. (p = 0.000, 0.009 and 0.019 respectively.

(P = 0.124 + 0.236 no folate – 0.027 smoking + 0.103GI-history)

Conclusion Folates strongly protect against hepatotoxicity and methotrexate withdrawal, but not against subjective adverse events. An interesting finding is the relation between smoking and hepatotoxicty and fatigue/malaise. In smokers the methotrexate withdrawal was significantly less. A positive gastro-intestinal history and age were predictive factors for the occurrence of diarrhoea. In this study a declining creatinine clearance was no risk factor.

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.