Recipient bed perfusion as a predictor for postoperative complications in irradiated patients with microvascular free tissue transfer of the head and neck area: a clinical analysis of 191 microvascular free flaps

Microvascular free tissue transfer is a reliable and well-established method for the reconstruction of large composite defects in the head and neck area. Various types of free flaps have been developed and further modified in order to restore soft tissue as well as bony defects in the head and neck area [1]. This can lead to both an adequate functional and aesthetic outcome. Meeting almost any requirement, more than 20 different types of microvascular free flaps for head and neck reconstruction are available today [2, 3]. Since its introduction in 1978, the radial forearm free flap (RFFF) has increased enormously in popularity due to its relative ease of harvest, reliable anatomy, pliable tissue, and the possibility of a multi-team approach leading to decreased operation time [4, 5]. Besides the most commonly used RFFF, the free latissimus dorsi flap (FLDF) and the scapular/parascapular free flap (SPFF) are two popular options for the reconstruction of head and neck defects. Both can be harvested alone or in combination based on a single vascular pedicle which allows the resurfacing of large composite defects [6]. In contrast to the fasciocutaneous RFFF and musculocutaneous FLDF, the osteomyocutaneous SPFF is also suitable for the reconstruction of defects involving both soft and bony tissues [7].

Even though the success rate of microvascular free tissue transfer exceeds 90% [8], postoperative complications such as impaired venous or arterial flow, wound infection, or hematoma may prove devastating for the patient [9]. To avoid potential flap failure, postoperative complications must be identified early and addressed sufficiently. Monitoring of free flap perfusion with a laser Doppler and tissue oxygen measurement are reliable methods for predicting ischemia in microvascular free flaps of the head and neck [10, 11]. Albeit the early success of microvascular free tissue transfer is based on the patency of the vascular pedicle rather than on random blood supply from the surrounding recipient bed, neovascularization into the graft is indispensable to maintain flap viability over time [12]. During this autonomization process, the free flap becomes independent from the vascular pedicle. However, there is no consent about when free flap autonomization is completed [13‐19]. As neovascularization is sprawling from the surrounding recipient bed into the flap, a good condition of the recipient bed is a prerequisite for successful free flap autonomization, and there is evidence that recipient bed vascularity may contribute to survival of ischemic microvascular free flaps in rats immediately after surgery [20]. However, the effects of recipient bed perfusion on postoperative complications remain unclear. In addition, the impact of previous radiotherapy on the recipient bed of the free flap is poorly understood in the context of flap autonomization, but it is well known that previous radiotherapy is associated with a higher risk of flap failure, wound-healing disorders, and complications due to endothelial alterations of the local vessels leading to an impaired blood supply [21‐28]. Therefore, the aim of the present study was to investigate differences between free flap and recipient bed perfusion with regard to postoperative complications within the first 2 weeks after microvascular free tissue transfer. The effects of a previous radiotherapy as well as differences between the various flap entities were also evaluated. We also investigated whether there was any sign of free flap autonomization during the first 2 weeks after surgery by evaluating free flap and recipient bed perfusion parameters, respectively.

This study included 191 patients who have undergone microvascular free tissue transfer in the Department of Oral and Maxillofacial Surgery at the University Hospital Erlangen, Germany, between October 2013 and July 2015. The patient population is described in Table 1. The study protocol was approved by the local ethical committee at Friedrich-Alexander-University of Erlangen-Nuremberg, Germany (No. 83_13B).

Four different points on the skin island of the free flap were measured. Additionally, in a subgroup of 101 free flaps, four standardized points in the recipient bed lying 2 cm distant to the outer margin of the free flap were included (Fig. 1). Standardized measurements were performed with oxygen-to-see (O2C, Lea Medizintechnik, Gießen, Germany) over an interval of 15 s on each of the selected points at postoperative days 1, 2, 3, 5, 7, and 14 after the patient was at rest and positioned horizontally for at least 10 min. Flap perfusion was quantified by white light spectroscopy (venous partial oxygen saturation of hemoglobin (SpO₂), relative amount of hemoglobin (Hb)) and laser Doppler (blood flow (AU = arbitrary units) and flow velocity (AU)) at 2 mm depth simultaneously. Measuring of perfusion parameters by white light spectroscopy is based on a Doppler shift caused by moving red blood cells which broaden the spectrum of the reflected light. SpO₂ and Hb are determined using tissue spectrometry. The light emitted into the tissue is absorbed or scattered depending on the Hb and SpO₂ [10].

Survival rates of microvascular free flaps range between 90 and 98% [29]. Besides relying on physical examination assessing color, turgor or capillary refill time, adjunctive techniques like laser Doppler of the vascular pedicle have proven to be a useful tool for evaluating free flap viability [30] and there is evidence that also perfusion of the surrounding recipient bed correlates with free flap survival not only in the long run, but also within the early postoperative days [20, 31]. Therefore, the present study is the first to investigate blood flow, flow velocity, hemoglobin concentration, and tissue oxygenation parameters of microvascular free flaps and the surrounding recipient bed within the first 2 weeks after surgery by using the O2C device, which delivers reliable real-time perfusion data via a non-invasive technique [10]. This includes parameters of irradiated and non-irradiated patients and both are considered with regard to the appearance of postoperative complications. We found that mean blood flow values of the recipient bed of irradiated patients were significantly lower compared to non-irradiated patients. Furthermore, it turned out that blood flow parameters of the irradiated recipient bed were significantly lower in patients who experienced early postoperative complications compared to patients who were not affected by a complication during the postoperative course. Remarkably, this difference was already present within the first postoperative day, which leads to the conclusion that sufficient recipient bed perfusion is essential for microvascular free flap viability immediately after surgery. In line with these observations, we conclude that monitoring of the recipient bed perfusion is particularly useful for the early prediction of postoperative complications in irradiated patients.

When comparing the overall blood flow and velocity parameters of the free flap and the recipient bed, we could observe an approximation of both curves, but the differences were still statistically significant at the end of the investigated period. This leads to the conclusion that there is a process of autonomization of the free flap, but it is not fully completed within the first 14 postoperative days.

Overall, a total of 18.3% (35/191) of patients required surgical revision but only one single flap failure could be observed in our study group, indicating the importance of early detection and operative intervention for flap salvage. In accordance with the current literature [32], we observed a higher number of postoperative complications in patients who had received radiotherapy before the surgery (36.4% in irradiated compared to 23.5% in non-irradiated patients).

Our data show that perfusion parameters, especially flow and velocity, significantly vary in different types of free microvascular flaps, which has to be taken into consideration during postoperative monitoring. In particular, RFFF shows higher blood flow and velocity parameters compared to SPFF and FLDF. The O2C device detects flow parameters in the capillaries at 2-mm depth [10, 33]. Therefore, varying flow parameters might be explained by substantial differences in flap volume and weight, with RFFF being smaller and lighter compared to SPFF and FLDF. Regarding the postoperative course of flow and velocity in the present study, a significant increase in blood flow during the investigating period was observed. But reaching postoperative day 5, a reduced increase or even a slight decrease with a consecutive flattening of the flow- and velocity- curves of RFFF could be observed. We hypothesize that during the first postoperative days, the reconnection of the existing capillaries known as inosculation has not yet occurred so free flaps are exclusively supplied via the flap pedicle. During this process endothelial cells of the recipient bed, mediated through tip-stalk cell selection and shuffling [34], invade in preexisting vascular channels of the flap. As a result, this process leads to further blood supply independent from the vascular pedicle (Fig. 6) and different studies suggest that accelerating the connection of those life-sustaining pipelines may help to enhance the viability of implanted tissue constructs [35, 36]. Conclusively, the flattening of the flow and velocity curve might be associated with a successfully ongoing inosculation and therefore an autonomization of free flaps with regard to the blood supply. However, we did not particularly evaluate neovascularization, so the approximation of both curves may also be a result of autoregulatory processes within the flap and the recipient bed. We could not observe similar flattening of blood flow and velocity parameters in SPFF and FLDF, indicating that inosculation and flap autonomization may be delayed compared to RFFF. In accordance with other studies [10, 14], our results show neither an increase in relative hemoglobin nor in venous partial oxygen saturation. On the contrary, all flaps even show a slight decrease of SpO₂ from the seventh postoperative day, which we hypothesized could be due to postoperative healing processes associated with local inflammatory responses.

After reviewing the current literature, most studies show only minor differences with a slight superiority of RFFF in terms of recipient site survival or postoperative complications when comparing various free flaps for head and neck reconstruction [37‐40], which is consistent with our observations. There is broad consensus that free flap viability mainly relies on the axial blood supply through the vascular pedicle on the early postoperative days [41]. In this critical time frame, obstruction of the vascular pedicle can lead to flap failure and may prove disastrous for the patient. Once the ongoing inosculation reaches a critical threshold, the flap becomes independent from the vascular pedicle. As mentioned above, we hypothesized that inosculation and autonomization of RFFF proceeds earlier in the postoperative period compared to FLDF and SPFF, which could explain the lower number of postoperative complications we have seen in RFFF compared to other free flap entities. RFFF is the most frequently used free flap entity in our department which, in fact, could also lead to a lower complication rate due to a higher expertise.

Moreover, higher mean blood flow values of the non-irradiated recipient bed compared to the irradiated recipient bed were not present until the fifth postoperative day, indicating that this could be the first sign of blood flow through the capillaries from the surrounding recipient bed into the flap and therefore the beginning of a proceeding inosculation, which is consistent with other studies, suggesting free flap autonomization between the postoperative days 5–15 [13‐18]. This correlates with our previous observation that blood flow and velocity parameters decrease from postoperative day five on, which we attribute to successfully ongoing inosculation. However, there are no differences between the various free flap entities, which is contrary to our hypothesis that autonomization of RFFF may start earlier compared to FLDF and SPFF.

Previously, we concluded that impaired blood flow parameters are probably a result of endothelial alterations due to prior radiation. If our hypothesis is true, this difference should not show up in non-irradiated recipient beds and this is what we have seen in our study. However, we could only show a correlation between reduced recipient bed perfusion and the occurrence of complications in the early postoperative period. Therefore, further studies must clarify the particular function of recipient bed perfusion for free flap viability in the early days after surgery. Also the link between approximating flow curves and histological alterations should be further investigated.