Background In May 2008, the European Medicines Agency (EMA) granted authorisation to Pemetrexed as a first-line treatment for locally advanced or metastatic non-small cell lung cancer (NSCLC), other than predominantly squamous cell (non-SC) histology patients. A phase III trial compared Pemetrexed with Gemcitabine, both in association with cisplatin, and found a similar overall survival between both groups (10.3 months). The EMA authorisation was only based in a subgroup analysis of this phase III trial. In late 2008, the first Gemcitabine generic was commercialised.

Purpose Our aim was to highlight the limited evidence of the quality of Pemetrexed’s efficacy considered on its approval and its impact on its use.

Material and methods The literature was reviewed and a retrospective analysis of the first-line treatment options in non-SC NSCLC in our hospital was made between July 2013 and June/2017.

Results An opinion article was published in January 2018 in JAMA Oncology.¹ It discussed if an approval based in a subgroup analysis of a clinical trial, predefined but never tested in a phase III trial design for its validation, was strong enough to influence clinical practice. It is well known that any data retrieved from a clinical trial subgroup analysis is indicative and non-conclusive. It is uncertain when a subgroup analysis should influence clinical practice. The non-SC NSCLC treatment guidelines replace Gemcitabine for Pemetrexed as a first option, with evidence level II, using efficacy and not safety as a reason, which could be an argument.

In the 4 years’ analysed, 71 patients were treated with Pemetrexed and 22 patients with Gemcitabine, both associated with platin. The cost difference per patient (six cycles considered) was €10 554 (€7 49 334 for the 71 patients).

Conclusion Pemetrexed was preferred to Gemcitabine as a first-line treatment of non-SC NSCLC, beside its limited evidence quality. A change in clinical practice should require better evidence levels. In our hospital, this change in clinical practice had a relevant economic impact.

References and/or acknowledgements 1. JAMA Oncol 2018;4:17–8. doi:10.1001/jamaoncol.2017.1944

2. J Clin Oncol 26:3543–51. doi:10.1200/JCO.2007.15.0375

No conflict of interest.