Facial Synkinesis Resting Phenotypes and Surgical Implications

Importance: Facial synkinesis manifestations are well established, but no trait-based phenotypic stratification exists with implications for static and dynamic treatment planning. Objective: To present clinically observable clusters of resting features in patients with stabilized facial synkinesis with implications for guiding potential static and dynamic treatment considerations. Design, Setting, and Participants: This single-center retrospective cross-sectional study used videographic and photographic data from January 2009 to March 2025 from a tertiary academic referral center. Inclusion criteria required standardized images and videos captured at least 16 months after facial paralysis onset, prior to synkinesis operations, and at least 3 months following chemodenervation. Patients without all standardized photos and videos or bilateral synkinesis were excluded. Determinations were made by 3 independent evaluators. Exposures: This clustering study of resting phenotypes did not involve any patient exposures. Main Outcomes and Measures: The primary outcome was classification of patients into clinically observable phenotypic clusters. Secondary outcomes included comparisons of etiological and medical characteristics between the identified phenotypes, with clinician-graded eFACE facial nerve scale scores quantifying facial synkinesis severity. Results: Of 549 consecutive patients with facial synkinesis evaluated, 160 were included in the study. Overall, 128 patients (80%) were female, and the mean (SD) age at paralysis onset was 41 (16) years. Median (IQR) follow-up time was 1.1 (0.2-3.1) years. Mean (SD) eFACE synkinesis score was 53 (16); median time from paralysis onset to evaluation was 5.4 (range, 1.4-36.0) years. Three phenotypes were identified. Type 1, identified in 93 patients (58%), represented notable resting hypertonicity and facial asymmetry, with a mean (SD) eFACE score of 49 (16). Type 2, identified in 47 patients (29%), represented symmetry throughout the face with a mean (SD) eFACE score of 58 (16), which was significantly less severe than type 1 (mean difference, 9 points [95% CI, 3.68-14.48 points]). Type 3, identified in 20 patients (13%), represented localized hypotonicity on the affected side, with a mean (SD) eFACE score of 55 (20). Clustering analysis of dynamic appearance did not yield clinically usable results. Conclusions and Relevance: In this cross-sectional study, clinically meaningful clusters of resting facial synkinesis phenotypes represent regional hypertonicity in most cases, symmetry in others, and regional hypotonicity in atypical cases. This phenotypic framework may be useful in systematically evaluating patients and devising static and dynamic surgical treatment strategies.