Magnetic resonance imaging of the TRAM flap donor site

High-field magnetic resonance imaging (Magnetom 42 SP, 1.0 Tesla) was performed a mean of 22 months (range, 8-42 months) postoperatively on 29 women who had undergone unilateral breast reconstruction with a transverse rectus abdominis musculocutaneous (TRAM) flap (19 free and 10 pedicled TRAM flaps). Fifteen T1-weighed, cross-sectional spin-echo images of the abdominal wall were obtained using a surface coil. The free TRAM flap was elevated sparing the lateral third of the rectus muscle. In the pedicled TRAM flap the whole rectus muscle was used. The patient groups were demographically similar, with no statistical differences in age or body mass index. In the free TRAM group the mean (+/-standard deviation) area of the rectus muscle in the upper third of the muscle (first five slices) was smaller on the operated side (376 +/- 135 mm2) than on the contralateral side (462 +/- 78 mm2), p = 0.02. The mean signal intensity (reflecting intramuscular fat content) of the upper third of the muscle was significantly higher on the operated side than on the nonoperated side, p = 0.04. Intramuscular fat content was also estimated and graded using an arbitrary scale from 1 to 4. The fat content of the upper third of the muscle was graded higher on the donor side (median, 3) than on the contralateral side (median, 2), p < 0.01. In pedicled TRAM flap patients, the remaining rectus was responsible for a mean of 47 +/- 5% of the distance between the lateral muscles, leaving a mean of 63 +/- 10 mm of the abdominal wall covered by fascia only. No hernias were detected in either group. This study shows that harvesting of a free TRAM flap seems to affect the quality of the donor rectus muscle over its whole length.