Mandibular reconstruction involves the replacement of contractile tissues with immobile ones, and therefore, it will always be an approximation rather than a duplication of the pre-disease state . Fortunately, mandibular reconstruction with osseous and osteocutaneous fibula free flaps results in promising functional outcomes in the majority of patients. Although it is clear that mandibular reconstruction leads to positive outcomes, our current literature does not currently identify the clinical factors that may serve as predictors of postoperative function in these patients.
The importance of identifying predictors for both short-term and long-term swallowing function cannot be underestimated. Gaining a better understanding of possible outcome predictors are critical to recognize patients at a higher risk of dysphagia or aspiration and to develop individualized rehabilitation strategies for such patients in a timely fashion. Alas, the available evidence in the setting of microvascular mandibular reconstruction is very limited. In 1998, Wagner et al.  attempted to describe predictive factors for functional recovery following the microvascular reconstruction of segmental mandibular defects. In this study, most of the cohort was found to have an abnormal baseline swallowing function, which became consistently worse after the operation. The authors concluded that the skin paddle area was the single most important predictor of postoperative function , but their study was limited by its small sample size and an uneven distribution of oncologic vs. non-oncologic patients. In 2005, Seikaly et al. published a case series of ten patients who underwent modified barium swallow studies preoperatively, postoperatively, and after radiation therapy. In their series, they reported no instances of aspiration or laryngeal penetration in any of their swallow studies and found no significant differences in the rates of dysphagia at any of the assessed time periods .
Our retrospective study may help to address some of the weaknesses observed in previous studies. Firstly, our study is comprised of a larger and more contemporary patient cohort. Additionally, our patient cohort possesses an even distribution of relevant characteristics such as radiation exposure, malignant etiologies, and benign etiologies. The relative advantages associated with our patient cohort may provide a more accurate representation of the spectrum of clinical cases that reconstructive surgeons most often encounter in their practice.
Our swallowing outcomes compared favorably to the majority of recently published series [9, 17, 18]. Twenty-two percent of patients in the cohort were gastrostomy-dependent preoperatively, and 39% required gastrostomy at some point during the postoperative period. However, by the end of the follow-up period, 81% of them achieved an unrestricted oral diet. The literature in this topic is not recent, and the proportion of patients able to return to an unrestricted diet varies significantly between series. This is in part due to the differences in the definition of dietary goals, as well as the time frame considered for the assessment. However, most studies report a return to an oral diet that ranges between 70 and 90%, albeit with a wide distribution between what was is considered regular, soft, and liquid diet [9, 17, 18]. Long-term swallowing outcomes were recently presented in a series of 20 patients. The authors reported that 10 years after surgery, 70% of the patients were on a regular diet while the remaining 30% were on a soft diet .
Overall, 19% of patients remained gastrostomy-dependent throughout our follow-up period. This was associated with radiation exposure, presence of gastrostomy tube preoperatively, and older age. Radiation exposure is a well-known outlier for worse functional outcomes in head and neck reconstruction. Radiated patients present an increased oral transit time, greater pharyngeal residue, lower swallow efficiency, and shorter cricopharyngeal opening times when compared to radiation naïve patients . In addition to these direct effects, mucositis and xerostomia will further diminish the swallowing function in these patients. In the present series, the detrimental effects of radiation were apparent both in short- and long-term analysis; first with oropharyngeal dysphagia observed in postoperative VFSS, and subsequently with worse long-term swallowing outcomes.
The size of the soft tissue defect has been shown to be a predictor of swallowing outcomes in multiple series [12, 20, 21], and in this regard, our results are in partial agreement with the available evidence. Both variables aimed at documenting the extension of the soft tissue component (percentage of tongue resected and skin paddle area) were statistically correlated with the need for a gastrostomy tube postoperatively. We interpret this association as an indicator for short-term oropharyngeal dysphagia. However, in long-term analysis, neither of these variables achieved significance as a predictor for unrestricted diet. In light of these findings, we reviewed these cases and found that 50% of patients that needed a gastrostomy tube postoperatively ultimately achieved an unrestricted diet. All of these patients underwent aggressive swallowing therapy and were closely followed by a speech pathologist from the early phases of their recovery. We hypothesize that this intervention allowed for the compensatory mechanisms to overcome the functional deficits, which may also explain the lack of association between the extension of soft tissue defects and the long-term swallowing outcomes.
Perhaps the most interesting aspect of the study is the correlation of VFSS and clinical swallowing outcomes. To our knowledge, this is the first report to examine the role of early postoperative VFSS as an independent predictor for long-term functional outcomes in this setting. We found that the presence of either penetration or aspiration had a strong correlation with the need for the postoperative gastrostomy tube and failure to achieve an unrestricted diet. Encompassing with our practice pattern, none of our patients underwent speech or swallowing therapy between the time of surgery and the first postoperative VFSS. Therefore, we feel that VFSS truly provides the most accurate reflection of the functional deficits derived from surgery. Its results will not be skewed from patient compensation or from improvements after swallowing therapy. The ability to use the first postoperative VFSS as a long-term predictor for swallowing allows the physician and speech pathologist to identify patients at risk of persistent dysphagia. This evidence calls for the implementation of aggressive rehabilitation strategies early on for patients with an abnormal VFSS. While these patients would likely undergo rehabilitation anyways in most institutions, we feel that establishing this long-term correlation should perhaps change the time-frame and emphasis with which the therapy is implemented. This information also allows clinicians to have an earlier and more accurate estimation of the functional prognosis, and act as more effective patient counselors.
There are limitations of the study, which are inherent to its retrospective nature. Our VFSS penetration and aspiration data was collected from reports and not from the videos, which may introduce physician- or technician-dependent interpretation errors. The sample size of the study may also be considered as an additional limitation, although it compares favorably to most single-institution series. Despite these limitations, our study may offer clinically relevant information for reconstructive surgeons, as well as useful information to contribute toward future studies designed with the intent of improving the functional outcomes of patients undergoing microvascular mandibular reconstruction.