T-cell lymphoma is a rare and heterogenous group of aggressive lymphomas, in which first-line treatment of choice is not settled. The authors describe the setting well and in a balanced way. Other than in ALCL, where BV-CHP is standard therapy, choice of treatment is under debate. CHOP or variants thereof are mostly used as is done in B-cell lymphomas, however, with greater success.
The authors have performed a systematic review and meta-analysis utilizing in total 28 studies with 1424 patients with PTCL of various subtypes. This is a significant number of patients in these rare diseases, but certainly a heterogenous cohort with regards to diagnostic subtype and trial designs.
Types of chemotherapy are grouped as CHOP, CHOP plus etoposide, gemcitabine-based or others.
Studies with a majority of ENKTL:s or studies focusing on CD30 positive or cutaneous lymphoma.
The authors find that OS is superior in the cohorts treated with CHOP plus etoposide compared to CHOP and gemcitabine-based seems at least numerically somewhat worse.

The overall design of this paper and the description of the field and the methods used seem sound, although I cannot evaluate details of the statistical methods. One can ask some questions:

The basis for classifying chemotherapy regimens seems a bit random. Often, the term “CHOP-like” is used encompassing among other combinations, also CHOEP and similar. Would it not be more reasonable to compare CHOP-like to gemcitabine-based chemo?

Often, etoposide, which has been reported to increase CR rates and toxicity, mostly affects EFS and not OS (German data) event though patients treated with more intensive chemo combinations tend to be younger. Here it seems to be the contrary, or maybe just a typo (lines 228-230)?

On lines 113 and 117 the abbreviation “CEOP” is used for two different combinations.
This trial does not does not answer the question whether autologous stem cell transplantation should be used for consolidation in first remission, but adds some data, not supporting ASCT. An eternal question.