xRead Articles - November 2022

46

B. Swendseid et al. / Oral Oncology 73 (2017) 43–47

Table 3 Relation of preoperative tissue diagnosis to management of facial nerve. A majority of patients who had a confirmed malignancy prior to surgery were able to avoid any removal of facial nerve branches. The frequency of partial and total nerve sacrifice was comparable between patients with and without preoperative tissue diagnosis of malignancy, with those with confirmed malignancies had slightly higher overall risk of some degree of nerve sacrifice. Relation of preoperative tissue diagnosis to management of facial nerve (N=75) Tissue diagnosis Malignant Suspicious for malignancy Inconclusive None performed Total patients 48 3 2 22 Facial nerve spared 27 (56.3%) 2 (66.7%) 1 (50%) 15 (68.2%) Partial nerve sacrifice 13 (27.1%) 1 (33.3%) 0 2 (9.1%) Total nerve sacrifice 8 (16.7%) 0 1 (50%) 5 (22.7%) Table 4 Preoperative predictive model for facial nerve sacrifice. Absence of facial nerve dysfunction and pain in the parotid region resulted in minimal risk of nerve sacrifice intraoperatively. Presence of both facial nerve dysfunction and parotid pain resulted in some degree of nerve sacrifice in all our patients, but nearly half were able to avoid total nerve resection. Presence of either facial nerve dysfunction or pain resulted in substantial risk of some degree of nerve sacrifice, but nearly 20% less than when both factors were simultaneously present. Number of patients requiring facial nerve sacrifice related to presence of preoperative predictors of sacrifice Nerve spared Partial sacrifice Total FN sacrifice Percent requiring sacrifice (%) Percent requiring total FN sacrifice (%) Total patients 45 (60%) 16 (21.3%) 14 (18.7%) 40 18.7 Normal nerve and no parotid pain 41 (89.1%) 4 (8.7%) 1 (2.2%) 10.9 2.2 Parotid pain 4 (23.5%) 7 (41.2%) 6 (35.3%) 76.5 With abnormal nerve 0 3 (42.9%) 4 (57.1%) 100 57.1 Without abnormal nerve 4 (40%) 4 (40%) 2 (20%) 60 20 Abnormal nerve only 0 5 (41.7%) 7 (48.3%) 100 48.3

benign frommalignant lesions of the parotid is beyond the scope of this manuscript but certainly can create a diagnostic dilemma for the head and neck surgeon. However, at high volume institutions the accuracy of FNA for malignant lesions is 93–97% [12,13], while a recent meta-analysis including 63 publications calculated the diagnostic accuracy of parotid FNA to be 96% [14].). In regard to our pathological determination of nerve invasion, another limita tion is that we assumed that a read of perineural invasion signified facial nerve invasion. It is possible that a portion of these patients had intact facial nerves with invasion of a different nerve. Our inclusion criteria also involved only patients in whom we could obtain a complete chart. This excluded approximately half of the parotidectomies for malignancy during the study timeframe. Addi tion of these additional cases could have affected the ultimate results. We would like to note that these predictive factors should not be used as a substitute for intraoperative decision making regarding facial nerve sacrifice. The facial nerve should be pre served whenever possible, with sacrifice reserved for gross tumor extension. Our model for evaluating the likelihood of nerve sacri fice is meant only for preoperative discussions with patients, and should not be a substitute for intraoperative clinical judgement. Even with these limitations the information gained from this data is helpful in discussing the likelihood of facial nerve sacrifice in those patients with malignant lesions of the parotid gland. While we were not surprised that facial nerve dysfunction, pre operative peri-parotid pain and total parotidectomy were associ ated with increased rates of nerve sacrifice, the secondary conclusions provide details of the risk in certain demographics that we did not expect. For example, a patient with normal facial nerve function and no facial pain will have minimal risk to their facial nerve if their process is benign, however our data illustrates that some form of nerve sacrifice will occur in up to 19% of these patients if the process turns out to be malignant. A clinician can choose to discuss this risk based on the likelihood that a patient’s process is benign. On the other hand, a patient worried about their confirmed tissue diagnosis of malignancy can be counseled that they can still have total nerve preservation 75% of the time so long as they have normal preoperative facial nerve function. That num ber rises to 89% if the patient also lacks facial pain. However, a

Pre-operative facial nerve dysfunction suggests invasion of the nerve or its blood supply by tumor, causing infarction and clinical dysfunction [2]. As expected, our study shows this is a strong pre dictor of eventual sacrifice. While 100% of the patients who pre sented with abnormal facial nerve function did require at least partial sacrifice, total sacrifice was only necessary half the time. The key point here is that only isolated nerve branches may be involved, and partial preservation is often possible. We found a statistically significant difference between patients with and without pre-operative pain regarding facial nerve sacri fice. While this is not always an indication of nerve invasion, it likely indicates a subset of cancers that are aggressive and thus can serve as a marker of possible nerve invasion in patients who do not manifest clinically identifiable facial nerve weakness. It is also no surprise that the need for total parotidectomy would increase the risk for facial nerve sacrifice, (58% vs 25% for superficial parotidectomy only). The need for deep lobe removal can be suggested by preoperative imaging showing tumor involve ment deep to the retromandibular vein, a radiographic marker sep arating the superficial from deep lobes. However, the ultimate decision to remove the deep gland is determined by the surgeon intraoperatively, and therefore this factor was not included in our predictive model. Still, when discussing preoperative planning with patients in whom imaging suggests need for total parotidec tomy, they should be counseled on the increasing likelihood of nerve sacrifice. We also found that even in the setting of a preoperative tissue diagnosis of malignancy, the entirety of the facial nerve can still be preserved more often than not (56.3% of the time) and in patients without pain or weakness it can be preserved 89.1% of the time. Patients ought to be counseled that while a nerve sacri fice is more common with malignant tumors than benign, patients still have a high likelihood of full facial nerve preservation if they lack abnormal facial nerve function and pre-operative pain. Some limitations in this study are its retrospective nature and variability in recording facial nerve function by the senior surgeons at our institution. Another limitation is inherent differences in sur gical technique and the individual surgeons’ threshold for nerve sacrifice. Inherent limitations in the ability of an FNA to determine