Influence of the recipient vessel on fat necrosis after breast reconstruction with a free transverse rectus abdominis myocutaneous flap.

The effect of the selection of recipient vessels on the rate of fat necrosis after microsurgical reconstruction of the breast remains largely unknown. Our aim was to evaluate the incidence of fat necrosis after unilateral breast reconstruction with a free transverse rectus abdominis myocutaneous (TRAM) flap after anastomosis with either the internal mammary vessels or the thoracodorsal artery and vein. Consecutive patients who had unilateral reconstruction with a free, muscle-sparing TRAM flap at two tertiary care centres over a 6-year period were identified. The incidence of fat necrosis, defined as postoperative firmness of 1 cm or more persisting for 3 months or more after anastomosis was calculated. To control for the effect of potentially confounding variables (body mass index (BMI), history of preoperative or postoperative radiation, previous abdominal operation, smoking, and hospital) we did a matched-cohort study. A total of 840 unilateral muscle-sparing TRAM flaps were done using either the internal mammary (n = 109) or the thoracodorsal (n = 731) vessels. Evaluation of the entire cohort showed that the incidence of fat necrosis after the two anastomoses was 13 (12%) compared with 130 (18%), respectively (p = 0.17). To control for the effect of confounding variables, 98 patients who had internal mammary anastomoses were matched 1:1 with 98 patients who had thoracodorsal anastomoses. Pair-wise comparisons showed that the incidence of fat necrosis was significantly higher when the thoracodorsal vessels were used (29; 30%) compared to when the internal mammary vessels were used (12; 12%; p = 0.002). Our results showed that a higher rate of fat necrosis may be seen after muscle-sparing TRAM flap reconstruction after anastomosis to the thoracodorsal vessels than with the internal mammary vessels. The exact mechanisms of this association are unknown and warrant additional investigation.