Background and importance Monitoring of methotrexate serum levels in osteosarcoma paediatric patients includes estimation of serum levels of methotrexate 24 hours after initiation of the infusion ([MTX_24h]), which allows folinic acid rescue to be started at adjusted doses. When pharmacokinetic estimation is not possible, the standard rescue (15 mg/m²/6 hours) is recommended and subsequently adjusted according to the real [MTX_24h].

Aim and objectives To evaluate the correlation and concordance of the estimated and real [MTX_24h], and the benefits of the estimation in comparison with the dosage by protocol.

Material and methods A retrospective study of paediatric patients treated with 12 g/m² methotrexate monitored by the pharmacy department from January 2014 to June 2020 was conducted. Estimated [MTX_24h] was determined with a Bayesian model with PKS software.

Variables collected were age, sex, number of cycles received, estimated and real [MTX_24h] and folinic rescue dose. Pearson and intraclass correlation coefficients between real and estimated [MTX_24h] were calculated. The agreement between the dosage of folinic acid by protocol and by estimating [MTX_24h] was assessed with the Cohen kappa coefficient.

Results 23 patients, 56.5% (13) men, median age 14 (4–17) years, received 152 cycles of methotrexate. The median number of cycles per patients was 8 (2–8). Median estimated [MTX_24h] was 7 (2–80) and real [MTX_24h] was 8 (1–85). The Pearson’s correlation coefficient and intraclass correlation coefficient for real and estimated [MTX_24h] were r=0.949 and CCI=0.974, respectively, indicating a high linear correlation and concordance between the two.

In 71.8% (94) of the cycles, the estimated folinic rescue matched with the dose which the patient should receive according to real [MTX_24h]. Assuming the dosing of folinic acid at 24 hours by protocol (15 mg/m²/6 hours) in all cases, only 35.1% (46) of patients would have received the correct dose. The Cohen kappa between the two methods was 0.189, indicating only slight agreement between both methods in favour of estimating [MTX_24h].

Conclusion and relevance Estimated and real [MTX_24h] showed high correlation and concordance, and in most cases the folinic acid rescue dose was correctly administered based on the estimated [MTX_24h]. These results seem to indicate that the estimation of [MTX_24h] and posterior estimation of folinic acid rescue are superior to systematic administration of 15 mg/m²/6 hours.

Conflict of interest No conflict of interest