Since the RFF is drained by a deep and a superficial system, there is still a debate as to which system is more reliable for anastomoses. Anatomically, the superficial system of the RFF is largely represented by the cephalic vein, which collects the blood from the dorsal venous arch and courses along the dorsolateral border of the radius toward the antecubital fossa, where it communicates with the median cubital vein, the basilic vein, and also with the deep venous system of the RFF via a perforating vein. The vein then further ascends to the deltopectoral fossa, traveling within the groove between the biceps and brachialis muscles, where it perforates the clavipectoral fascia to drain into the axillary vein. According to an anatomic dissection by Reid and Taylor [430], the cephalic vein was absent in 2 out of their 50 specimens. The average diameter of the cephalic vein is about 5 mm (1–12 mm) [275].

The communicating vessel between the cephalic vein and the deep system was first noted by Soutar et al. in 1983 [511], followed by a number of more detailed anatomical and clinical descriptions. Tahara et al. [523] found this vein to be present in all but one of their 188 patients, and despite the complex procedure to dissect the pedicle up to the brachial vessels, they emphasized to include this vein because it allows to drain both venous systems with only one anastomosis. Whereas a similarly high reliability of this communicating vein was also described by Valentino et al. [566], Thoma et al., who classified the branching pattern of the deep and superficial system into five different groups, could find the communicating vein in only 62% of their cases. Of these cases, the median cubital vein either split into the cephalic or median basilic vein (group 1) or did not bifurcate (group 2). In 18%, both deep veins merged and formed a sizeable common trunk, but did not anastomose with the superficial system, and in the other two groups, both deep veins did not merge, but had a suitable caliber for anastomoses (5%) or were different in size with one vein significantly larger than the other (15%).

According to a review by Medard de Chardon et al. [363], there are supporters of performing anastomoses to either the deep [118, 242, 243, 510, 570] or the superficial veins [246, 339, 400], because they believe those veins represent the dominant drainage system. Other authors consider both systems to have an equal capacity to drain the flap [136, 268, 557], but they prefer the superficial system because the veins have a larger diameter, making anastomoses easier. Other authors state that the deep veins are more adequate to drain the distal part of the skin paddle [243, 370] or believe it to be more reliable than the superficial system in elderly patients [118]. Moreover, the superficial veins were proposed to serve as a backup in case of the need for revision surgery [370].

The question of whether to perform a single or double venous anastomosis was discussed extensively in the literature. Again, a very useful summary was provided by Minoun et al. who could identify authors suggesting only one venous anastomosis, whether it is superficial or deep [154, 268, 510], because they considered a single venous outflow to be sufficient. Moreover, when performing a second anastomosis, this would lead to an increased operating time and reduction of venous blood flow, which could increase the risk for thrombosis. A significant decrease in flow velocity was also confirmed by a study by Hansono et al., who therefore argued against two venous anastomoses. To evaluate the efficacy of the cephalic vein, it was proposed to temporarily clip the deep veins and to anastomose the flap to the superficial system only if there were no signs for venous congestion [429]. On the other hand, many other authors recommend performing two venous anastomoses routinely [61, 198, 247, 306, 543, 566] preferably one from the deep and another from the superficial system [244]. Using the double anastomoses technique, failure rates between 0% and 1% were reported in primary reconstructive procedures [5, 443]. This was confirmed by a systematic review and metaanalysis including almost 7000 free flaps by Riot et al. [439], who found a significantly lower failure rate for double venous anastomosis. This finding was also evident for most of the studies focusing on the RFF [244, 446, 490, 530].

Beyond the relevance of the cephalic vein to drain RFF in free-tissue transfers, it can be left attached to the donor site after extended dissection up to the clavicle, if a long venous pedicle is required in cases of vessel-depleted necks or during salvage maneuvers [188, 275, 634]. With this technique, the cephalic vein is exposed along its course at the arm directly or via endoscopic guidance and together with the flap tunneled subcutaneously over the clavicle into the defect, so that only an arterial anastomosis is necessary, and the dissection of the cephalic vein is possible in less than 30 min [422]. Moreover, the vein is a useful source for vein grafts because of its large caliber, superficial location, and ease of handling [188] (◘ Figs. 2.1 and 2.2).