br To further investigate the

To further investigate the possible correlation between various tumor factors and pathological nerve invasion, we conducted an addi-tional correlation analysis in the patients with facial nerve resection (n = 26). In summary, we did not find any correlations between pa-thological facial nerve invasion and tumor-grade, T and N status (Supplementary Table S2). Thus, it BMPO seems hard to predict pathological facial nerve invasion, just by tumor factors (tumor grade, T or N status). Rather, intraoperative findings may be important for the decision re-garding FN resection.
Discussion
Although parotid gland cancer with facial weakness is a serious clinical situation, clinical outcomes and treatment decisions regarding optimal surgical management are infrequently studied because of the rare incidence of parotid gland cancer with facial weakness. In our study, we tried to refine the surgical guidelines for facial nerve man-agement in patients with parotid gland cancer and pre-treatment facial weakness. Our results challenge the prevailing concept that pretreat-ment facial weakness with parotid gland cancer is a surgical indicator for facial nerve resection [5,13,19]. Although our study was limited by a small number of patients, it will provoke further discussion about this unsolved clinical problem.
The reported incidence of facial weakness in parotid gland cancers at presentation ranged from 13.6% to 34.7% of patients and was 13.5%
Table 2
Oncological outcomes at final follow-up according to subgroups (n = 43).
Groups TNM status at diagnosis1 Oncological outcomes Time to recurrence (months) Follow-up in cases without recurrence or
FN: facial nerve. 1 Tumor-Node-Metastasis staging according to 7th AJCC staging system: pTNM in cases with surgery, cTNM in cases with non-surgical management.
2 NED: no evidence of disease, DOD: death of disease, AWD: alive with disease.
3 See Supplementary Fig. S1.
Table 3
Survival analyses and risk factors for worse outcomes (n = 45).
Univariable analysis
Variables Categories Progression-free survival
Overall survival
Hazard ratio 95% Confidence interval P value Hazard ratio 95% Confidence interval P value
ACC vs. non-high-grade
Age Continuous
ACC: adenoid cystic carcinomas.
LN: Lymph node. 1 Please see Supplementary Table S1. Evaluation of multi-collinearity among variables.
Fig. 3. Pathological analysis of facial nerve invasion by tumors (n = 26, facial nerve-resected patients).
in our series [2,4,5,9,10,12]. Facial nerve paralysis at diagnosis is a significant prognostic factor in patients with parotid gland cancers [2,7,10]. In line with previous reports, our data demonstrated that most parotid gland cancers causing facial weakness have high-grade pa-thology, including adenoid cystic carcinoma. As expected, our patients showed around 40% survivals at 5 years.
In the literature, two papers indicated that facial nerve sacrifice to obtain clear surgical margins improved local control and long-term prognosis in adenoid cystic carcinomas [14,15]. However, there have been no reports in other types of parotid gland cancers with facial weakness. These other tumors are probably high-grade and have more aggressive clinical courses than adenoid cystic carcinomas. Our results raised questions about the clinical efficacy of facial nerve reconstruc-tion with facial nerve resection. Facial nerve graft requires recovery 
time of 6–12 months to allow nerve regeneration [20,21]. However, half of patients experienced disease recurrence, resulting in the salvage or other type of treatments. With lymph node metastasis, patients did not have enough time for facial nerve regeneration. Thus, we cautiously suggest that static techniques [22] for facial reanimation are better than nerve regeneration techniques for these patients.
Another thing to note was that facial weakness in parotid gland cancers did not always indicate local tumor invasion of the facial nerve, particularly with incomplete facial paralysis. In this study, no tumor invasion was identified in 30% of nerve-resected surgical specimens. Indirect causes, such as extrinsic compression, ischemia or local in-flammation, might cause dysfunction of the facial nerve [23,24]. Thus, decisions regarding facial nerve resection should be further based on intraoperative findings with/without radiological clues [25,26]. In our