Combining HR from same study sub-groups for meta-analysis

Hi all,

I’m working on a meta-analysis of randomized clinical trials in oncology and had a doubt I search the internet and could not find an adequate answer.

In one of the trials (out of five total), we have only the HR by a specific biomarker (positive and negative) but no overall HR is provided.

We would like to include the overall HR for all-comers (independent of HR) but I’m not sure how to derive this HR because they came from the same population (my instinct was to make a meta-analysis aggregation with fixed effects but searching the internet didn’t seemed right).

I have come also across this interesting paper by Hamilton (doi: 2010.1002/sim.3013) which provided a way to group different subgroups estimates but wanted to check with peers if this is the way to go and, if it is, if there’s any implementation in R or Stata.

Thank you all in advance,