Background: The elevation of tissue in any form is accompanied by some degree of devascularisation. Accurate determination of skin flap viability has long been a goal of plastic surgeons. The use of fluorescent indocyanine green tissue angiography (FITA) provides a useful clinical procedure for skin and soft tissue viability assessment of both mastectomy flaps and autologous tissue transfers.
Methods: 64 breast reconstructions performed in 43 patients undergoing various reconstructive procedures were evaluated intra-operatively with fluorescent indocyanine green perfusion using the Novadaq SPY laser fluorescence camera. Areas of dubious perfusion were resected even when clinical assessment suggested some degree of flow. Images were recorded and correlated with post-operative flap survival.
Results: 64 mastectomy skin flaps, 23 unipedicle TRAM flaps, 2 bipedicle TRAM flaps, 2 free TRAM flaps , 1 DIEP flap, 3 latissimus dorsi flaps and 26 expander insertions.  Poorly perfused  skin was resected based on the perfusion mapping intra-operatively, although in 5 patients skin was preserved as being helpful for closure with seemingly adequate perfusion based on clinical assessment even though the SPY images suggested hypoperfusion.  1 TRAM flap demonstrated marginal perfusion along the edge of zone lll but the tissue was preserved for shaping of the breast as it appeared perfused clinically. Of 5 patients with poor mastectomy flap perfusion demonstrated on SPY imaging, 4  developed post-operative necrosis (3 smokers and 1 patient with systemic lupus) and  1 non-smoker developed blistering, all in a pattern predicted by FITA. All reconstructions were salvaged using the original procedure. The TRAM flap patient developed predicted necrosis requiring debridement and closure and the residual flap remained soft and final shape was excellent .  This represents a skin sparing mastectomy necrosis rate of only 6.3% (including both smokers and non-smokers) compared with quoted rates in the literature of 15% and 25% in non-smokers and smokers respectively. In all of our patients, SPY images demonstrating hypoperfusion correlated accurately with areas that subsequently necrosed. If these had been resected primarily, the problem would have been reduced to almost 0%. All patients went on to complete successful reconstruction using the original technique. Similarly, skin and fat necrosis occurred in only 1 TRAM flap (3.7%)  which could also have been avoided had the area under suspicion been resected primarily in favor of a smaller final breast volume. All of the remaining autologous flaps healed without clinically significant fat necrosis. This represents a dramatic reduction when compared with reported palpable fat necrosis rates of 5-50% depending on the series.
Conclusion:  Novadaq SPY perfusion scans were almost 100% accurate in predicting skin flap viability post-operatively. Familiarity with the technique has resulted in the authors relying increasingly on the initial perfusion scan as an accurate assessment of likely skin necrosis post-operatively. The technique is simple, safe and reproducible, taking only minutes to perform during surgery. Reduction in delayed wound healing secondary to skin flap necrosis after breast reconstruction,  reduces patient morbidity and the cost of subsequent wound care and diminishes the risk of developing palpable fat necrosis in TRAM flap reconstruction.