All you need to read in the other general journalsBMJ 2010; 340 doi: https://doi.org/10.1136/bmj.c3004 (Published 08 June 2010) Cite this as: BMJ 2010;340:c3004
Support for standard treatment of node-positive early breast cancer
A trial compared three regimens of chemotherapy for early stage operable breast cancer with positive lymph nodes. Women were randomised to receive four cycles of doxorubicin and cyclophosphamide followed by four cycles of docetaxel (sequential ACT); four cycles of doxorubicin and docetaxel (doxorubicin-docetaxel); or four cycles of doxorubicin, cyclophosphamide, and docetaxel (concurrent ACT). The trial comprised 5351 women and lasted eight years⇑.
At the end of the study, survival was better with sequential ACT than with either doxorubicin-docetaxel or concurrent ACT (83%, 79%, and 79%, respectively), although, on the basis of a presumed synergistic effect of the triple combination of drugs, the researchers expected concurrent ACT to be better than sequential ACT. Disease-free survival for the three groups was 74%, 69%, and 69%, respectively. No change in practice is needed, because sequential ACT is standard treatment for these women.
Unexpectedly, survival was better in women with amenorrhoea for six months or more during the study, regardless of oestrogen receptor status. The editorialist (p 2122) hypothesises on possible explanations and discusses the implications for future research. Trials testing new chemotherapy drugs won’t be able to afford the size and duration of the studies that had assessed taxane based treatments. These trials will need to capture, in addition to clinical outcomes, biological samples before and during treatment, as well as include whole genome analyses.
Identifying city areas for delivery of community preventive interventions
A new type of analysis was used to identify neighbourhoods in one county in Georgia, US, where setting up community interventions might pay to improve survival after out of hospital cardiac arrest⇑. The researchers first estimated the frequency of cardiac arrests across 161 census tracts, which varied from 0.04 to 2.11 per 1000 persons, and then the frequency of cardiopulmonary resuscitations (CPRs) performed by bystanders, which varied from 0% to 100%. Empirical Bayes methods …