This prospective, assessor-blinded, 2-arm pilot cohort study was conducted at the rehabilitation unit at University Hospital San Paolo (Milan, Italy). On the basis of their ability to attend in-person sessions, each participant eligible for enrollment (N=16) engaged in either in-person PT treatment (ie, in-person emotional training; n=8) or in telerehabilitation treatment (ie, online emotional training; n=8). In both groups, treatments were led by trained physiotherapists. Motor outcomes (Sunnybrook Facial Grading System [SFGS] and SFGS with bilateral masseter contraction [SFGS with MassCon]), functional outcomes (Facial Disability Index; FDI), and psychological outcomes (Beck Anxiety Inventory; BAI) were assessed at baseline (T0) and at the completion of treatment (T1). Given the exploratory pilot nature of this study, which aimed primarily to assess feasibility and generate preliminary results to inform future confirmatory trials, and considering the rarity of the medical condition targeted by the inclusion criteria, no formal a priori sample size calculation was performed. However, the final sample (N=16) is consistent with those of previous pilot investigations conducted in comparable clinical populations [39,40]. A convenience sampling strategy was adopted, enrolling all consecutive eligible patients referred to the rehabilitation unit during the predefined recruitment period (November 2023 to November 2024). In addition, a prospective observational design was adopted due to several clinical (eg, emotional training is part of postoperative rehabilitation in our department) and logistical (eg, rarity of the target population and allocation to treatment according to patients’ real-world logistical circumstances) reasons.

The study protocol followed the Declaration of Helsinki. Considering that emotional training is common practice for patients undergoing triple innervation surgery at the rehabilitation unit of University Hospital San Paolo, and that patients with peripheral FNP would have been treated with the same protocol after triple innervation surgery despite not participating in the study, the research was considered as an observational study. Ethical approval was obtained from the Milan Area 1 Ethics Committee, protocol number 0048450/2022. All patients gave written informed consent before participation. All data collected in this study were handled in accordance with strict privacy and confidentiality standards. Records were anonymized through the use of coded identifiers, and no personally identifiable information was retained in either the research database or the investigational application. Hard-copy materials (eg, consent forms) were securely stored in a locked cabinet in a restricted-access laboratory, while electronic data were maintained on protected institutional servers with controlled access. Access to the coded datasets was limited to authorized members of the research team. Participants were not provided with financial compensation for their involvement in the study.

This study was conducted between January and October 2024. Participants were recruited from the hospital community based on the following inclusion criteria: (1) age between 18 and 75 years; (2) presence of unilateral facial palsy, confirmed through clinical history and anamnesis; (3) prior treatment with triple innervation surgery (as described in the work by Biglioli et al [14]); (4) a House-Brackmann scale grade of III or higher; (5) ability to understand and perform the tasks; and (6) capacity to provide informed consent. Exclusion criteria were (1) a diagnosis of major neurological, neuropsychological, or psychiatric disorders, confirmed through clinical history and anamnesis; (2) previous treatments for facial palsy other than triple innervation surgery [14]; and (3) changes in pharmacological therapy within 3 months prior to enrollment. Patients were instructed to maintain stable medication regimens throughout the study period and refrain from participating in additional rehabilitation programs.

SFGS, SFGS with MassCon, FDI, and BAI were administered both at baseline (T0) and the posttreatment time point (T1) by a single trained physiotherapist who was blinded to treatment group allocation but not to time points.

The SFGS is a widely used clinical scale designed to evaluate facial asymmetries and the presence of synkinesis comparing the paretic side of the face with the nonparetic side [51]. The scale consists of 3 components: the first (resting symmetry) involves assessing symmetry at rest, including the eye, cheek (nasolabial fold), and mouth; the second (symmetry of voluntary movement) rates symmetry during facial expressions (ie, during 5 standard facial expressions); and the third (synkinesis) grades the severity of synkinesis during these movements. Each component is assigned a score, and the weighted sum of these components generates a composite score ranging from 0 to 100, with a lower score indicating greater severity of facial palsy [51].

For this study, a modified version of the SFGS was used in which participants were instructed to perform the second component (symmetry of voluntary movement) while clenching their jaw (ie, activating the masseter muscles).

The FDI is a 10-item self-report questionnaire assessing 2 domains of facial disability: physical and psychosocial function [52]. It includes 2 subscales—one for physical function and the other for social and well-being function—each comprising 2 items. The scores for each subscale are converted to a 100-point scale, where a score of 100 denotes no impairment in function [52].

The BAI is a psychometrically validated, self-reported, 21-item scale commonly used to measure symptoms of anxiety [53]. The score for each item is summed, yielding a total score ranging from 0 to 63, categorized as minimal anxiety (0-7), mild anxiety (8-15), moderate anxiety (16-25), and severe anxiety (26-63) [53].

Normal distributions of the dependent variables were assessed using the Shapiro-Wilk test for normality [54]. As all variables passed the test (P>.05), parametric statistical analysis was applied. To evaluate the effects of time, treatment, and the interaction between time and treatment, changes in the outcomes at each time point were examined using a 2-way repeated-measure ANOVA with treatment as the between-subject factor. Post hoc analysis was performed using the Holm-Bonferroni correction. Finally, the Pearson correlation coefficient was calculated to explore relationships among age, time from onset of facial palsy to surgery, time from surgery to rehabilitation, time from onset of facial palsy to rehabilitation, and the change in outcome scores (∆; calculated as ∆ score = T1 score − T0 score). A P value of less than .05 was considered statistically significant for all analyses. Data were analyzed using JASP (version 0.16.3) for Windows (JASP Team).

Despite the pivotal importance of PT, the most effective PT approach after surgically treated peripheral FNP is still unclear, with a telerehabilitation-based emotional training protocol showing promising foundations given the specific characteristics of the facial impairments. In this exploratory pilot study, we confirmed the feasibility and potential clinical value of a telerehabilitation-based emotional training protocol in improving motor symptoms, facial disability, and anxiety in patients with peripheral FNP treated with triple innervation surgery also when compared with an in-person approach. However, these findings should be considered preliminary.

Emotional training is a novel therapeutic approach based on multiple phases that has shown preliminary promising results in the motor rehabilitation of FNP [30]. The first phase comprises a sensory-based intervention in which patients are asked to recognize the sensations induced by external stimuli on the lesioned hemiface (eg, intrinsic characteristics of objects or surfaces and passive movements performed by the physiotherapist) and perform facial movements—or imagine them, given that both motor imagery and actual movements activate the same neural circuits [48,49]. The rationale behind this first phase relies on the theories of motor control [55] that conceptualize the normal control of movement as a cognitive task [32] in which individuals integrate sensory inputs (feedback control), prior knowledge, and volitional intention (feedforward control) to select appropriate motor programs and regulate motor output [56,57]. As reported by numerous studies, sensory stimulation reshapes cortical motor representations [58] given the interplay between motor (primary motor cortex, supplementary motor area, and premotor cortex) and somatosensory (primary and secondary somatosensory cortices and the sensory thalamus) regions [59,60]. In neurological motor rehabilitation, sensory-based interventions have been shown to enhance adaptive motor cortical plasticity and promote functional recovery [61-63]. Nevertheless, cognitive processes (eg, information perception and the elaboration, selection, and execution of motor plans) [31,64] play a critical role in mediating these mechanisms, supporting functional interactions between the body and the environment [65,66]. In addition, combining different types of sensory information stimulates the activity of associative brain areas that are important for learning, neuronal plasticity, and recovery [67]. Although evidence remains limited [64,68-71], promising findings suggest that sensory-based interventions can significantly improve the ability to perform activities of daily living and enhance quality of life [71] even for people with peripheral FNP [72,73].

The second phase aims to restore authentic facial expressions, emotional congruence, emotional symmetry, and spontaneity [74], which is crucial for optimal motor recovery [75]. At the neurophysiological level, facial movements in response to emotions are mediated by a widespread neural network involving various cortical regions that project to the facial nucleus, integrating motor planning, somatosensory feedback, and limbic system activity [76,77]. While corticofacial projections from the lateral frontal lobe contribute to the voluntary control of facial movements [24], the cingulate cortex plays a pivotal role in emotional expression through facial mimics, with projections from the amygdala providing the necessary emotional context [24]. For example, among the 2 different neural pathways controlling laughter [25], one, which involves primarily the frontal operculum [78,79] and the primary motor cortex, represents the neural basis of voluntary, nonemotional, conversational laughter; the other, which includes the pregenual anterior cingulate [80,81], ventral striatum and nucleus accumbens, amygdala, and hypothalamus [82], is involved in the production of emotional laughter, with the supplementary motor area likely modulating the production of both [27]. Notably, training voluntary movements to express different emotions helps reconnect the limbic-motor pathways [24], supporting the synergy between voluntary and emotional facial movements [24]. This is particularly important for facial nerve lesions, where voluntary movement and emotional expression are often dissociated [83]. The final phase of rehabilitation focuses on functional exercises designed to restore the ability to perform essential daily activities. These functional tasks, such as eating, speaking clearly, or demonstrating facial expressiveness during conversations, involve complex and coordinated facial muscle activity [84]—which is why these exercises are included as the final part of the treatment. By training patients in real-life contexts, they are engaged in meaningful activities that mirror their daily life. This approach has been shown to enhance neuroplasticity and functional recovery through task-specific repetition [85]—a principle known as “use-dependent cortical reorganization” [86]. In addition, focusing on real-world tasks aligns rehabilitation with the patient’s meaningful goals, improving motivation and adherence to the treatment protocol [87-89].

Taken together, our findings reinforce the conclusions of several systematic reviews on the feasibility of telerehabilitation in neurological conditions [33,90]. Recently, a randomized controlled trial investigated the effectiveness of video exercise–based telerehabilitation on motor and psychological symptoms in patients with peripheral FNP [35]. Patients in the telerehabilitation group not only found the treatment satisfactory and accessible but also demonstrated a significant improvement in FDI scores comparable to those receiving routine in-person care. Furthermore, both groups exhibited similar improvements in anxiety after treatment [35]. Despite the limitations of our study design, our findings are consistent with this evidence, suggesting the need to expand upon the value of telerehabilitation services for peripheral FNP, also considering that both patients and therapists generally have a positive attitude toward the integration of digital technologies [91,92]. Following the telerehabilitation-based emotional training, our patients reported improvements comparable to those of the face-to-face treatment in facial nerve function and facial symmetries (both at rest and during voluntary movement) even when considering the modified version of the SFGS (SFGS with MassCon). This is consistent with the only study assessing the effectiveness of emotional training in peripheral FNP, although the latter was performed in an in-person fashion [30]. Indeed, the authors reported that 50 patients who underwent face-to-face emotional training had improved House-Brackmann scale scores and symmetry of the face both at rest and during movements. In addition, in contrast to our study, only patients with iatrogenic peripheral FNP but not those who underwent surgery had gains in social function and subjectively perceived well-being. In our study, all patients experienced significant improvements in social and well-being function at the end of the treatment both in the in-person and online groups. This result was further confirmed by improved anxiety assessments. This is of particular importance given the substantial psychosocial burden that patients with facial impairments experience [93]. Nevertheless, the nonrandomized allocation and limited sample size warrant a cautious interpretation of these positive findings. Moreover, comparisons should be interpreted cautiously as the aforementioned study [30] was a retrospective analysis with no control group and the authors performed an in-person treatment protocol. However, in this study, self-reported physical function was greater at the end of the in-person treatment when compared to telerehabilitation treatment. At least for physical functioning, the therapeutic alliance and the therapist-patient relationship, which are strengthened by the direct, in-person interaction, and the emotional bond inherent to in-person rehabilitation may have had a positive effect, as previously suggested [33,94].

This study has some limitations. First, the small sample size may have underpowered the study and increased the risk of inflated effect estimates (eg, from correlations), whereas the use of subjective and operator-dependent assessments might have affected the results and interpretation. The nonrandomized allocation could have affected the internal validity of the study, whereas the absence of a follow-up assessment prevents conclusions regarding the durability of the observed improvements. In addition, our neurophysiological hypotheses should be confirmed through neuroimaging studies, which were not feasible at this time. Second, although our patients were asked to maintain a stable pharmacological treatment throughout the study, the pharmacological therapies of the patients were not documented. The different biochemical effects induced by the medications could have influenced motor recovery, as previously suggested [95,96]. Similarly, we included surgically treated patients with peripheral FNP with multiple palsy etiologies (eg, neurinoma and trauma). The different underlying mechanisms of facial dysfunction might have affected the results and reduced the external validity of our findings.

The findings of this pilot study indicate that emotional training could be integrated into routine postoperative care for patients with peripheral FNP after triple innervation surgery, offering measurable benefits in facial symmetry, voluntary motor performance, social functioning, and anxiety reduction. The intervention’s emphasis on both motor and psychosocial domains aligns well with the complex functional demands of patients undergoing facial reanimation, who often experience challenges that extend beyond motor impairment alone. Interestingly, the comparable improvements observed with the in-person protocol and telerehabilitation delivery suggest that emotional training can be implemented flexibly without compromising clinical outcomes, providing a viable option for individuals with mobility limitations, work constraints, or psychosocial discomfort associated with facial impairment. The structured nature of emotional training, centered on restoring selective motor control, improving emotional expression, and reducing maladaptive compensatory patterns, offers clinicians a reproducible framework that can be tailored to individual patient needs. Clinicians may consider telerehabilitation as a practical strategy to maintain continuity of care and reduce barriers to treatment access.

In conclusion, this exploratory pilot study provides preliminary evidence that telerehabilitation-based emotional training may have beneficial effects on motor, physical, and psychosocial functioning, as well as on anxiety, in surgically treated patients with peripheral FNP. Given the study’s observational design and limited sample size, these findings should be interpreted cautiously and considered hypothesis generating rather than confirmatory. Further studies with optimized designs (eg, double-blind, parallel-group trials), larger sample sizes, and more objective and informative assessments (such as neurophysiological or neuroimaging evaluations) are warranted to confirm and extend these findings.