The distal branch of the RLN is preserved to the maximal extent possible when a tumor invading the RLN is resected. When the nerve is transected at the entrance thereof to the larynx, the adductor branch of the RLN is identified by removing the inferior-posterior portion of the thyroid cartilage after disconnection of the cricothyroid joint. Because the adductor branch is thin, it is sometimes difficult to link the nerve stumps with three stitches. If both ends are appropriately fitted, one or two stitches may yield good postoperative vocal results. Direct suture of both stumps is technically and functionally possible only when the gap between the ends is less than 5 mm, because tension at the suture site is unacceptable. A free nerve graft is placed between the severed ends if direct anastomosis is not possible. A segment of the great auricular nerve (GAN) usually serves as such a graft (Fig. 5.4) because the graft surgical field is close to the RLN, and the GAN and RLN are of similar diameter. The third method is nerve transfer, in which the ACN is sutured to the peripheral end of the RLN when the other end of the RLN is unavailable for reconstruction. Nerve interposition is preferred to nerve transfer, because the motor supply to the intrinsic laryngeal muscles originates from the RLN and not the ACN. We believe that using the RLN as the regenerating nerve is better for restoring normal vocal function than other methods (including anastomosis with the ACN). In a few patients in whom the peripheral end of the RLN was unavailable for reconstruction, we performed nerve–muscle pedicle (NMP) flap implantation combined with AA. Figure 2.​6 illustrates the methods used for laryngeal reinnervation.

The superior root of the ACN is located dorsal to the anterior belly of the omohyoid (OH) muscle and anterior to the internal jugular vein. The ACN is followed caudally until that nerve branch enters the sternohyoid (SH) muscle after looping around the inferior root (Fig. 5.5). A main branch is utilized for anastomosis with the peripheral end of the RLN. This step is performed under microscopic guidance to ensure an exact fit between the cut ends of the nerve fibers; between two and four ties of 9-0 nylon thread are placed.

Thirty-two patients underwent primary nerve reconstruction at the Department of Otolaryngology-Head and Neck Surgery, Kumamoto University Hospital, between March 2000 and March 2014. Table 5.1 shows the clinical profiles of these patients. In 19 patients, the duration of UVFP prior to surgery ranged from 1 to 62 months (median: 3.5 months); thus, most patients underwent nerve reconstruction within a few months after onset of UVFP. The ACN was utilized for nerve transfer in 17 patients, and in 1 patient with uterine cancer metastasis, the ACN of the unaffected side was moved to anastomose with the contralateral peripheral RLN stump.

The vocal function of each patient was assessed in terms of aerodynamics, acoustic analysis, and auditory impression, both before surgery and 1, 6, and 12 months postoperatively. The parameters measured were maximum phonation time (MPT), mean airflow rate (MFR), jitter, shimmer, harmonics-to-noise ratio (HNR), grade overall, greatness (using the GRBAS scale, which is an acronym for grade overall, rough, breathy, asthenic, and strained; 0 = normal, 1 = slight, 2 = moderate, 3 = extreme; see Sect. 4.​2), and range of pitch. Auditory perceptual assessment using the GRBAS scale was conducted by three different listeners in random order. The mean values of G and B were recorded for each patient. Because the patients varied in terms of the presence or duration of UVFP prior to surgery, we examined vocal function after nerve reconstruction. The Mann–Whitney test was used for statistical analysis, and P < 0.05 was considered statistically significant. Table 5.2 shows the mean ranges of vocal function (except for jitter, shimmer, and HNR) of healthy Japanese men and women aged 50–59 years [5]. The normal ranges of acoustic parameters were taken from the manufacturer’s practical guide to the Multi-Dimensional Voice Program (Kay-Pentax).

Temporal changes in the vocal functions of 22 patients, who underwent primary RLN reconstruction and who were followed-up over 12 months, are shown in Figs. 5.6, 5.7, 5.8, and 5.9. All vocal parameters exhibited continuous and significant improvement during the 12-month follow-up period. Although jitter and shimmer also significantly improved, the normal ranges of these parameters were not attained. However, all other parameters recovered to values within the normal ranges by 12 months after surgery.

Generally, the sooner a functional connection is made between a regenerating axon and a muscle fiber, the more likely it is that the extent of functional recovery of the muscle will be close to normal. Currently, however, no definitive clinical data are available on how long it is acceptable to wait from onset of UVFP to nerve reconstruction ensuring functional recovery of the muscle. The longest such interval in the present series was 14 months in a patient who had thyroid papillary cancer (thus delaying reinnervation); vocal function recovered to normal over the 12-month follow-up period. Of the nine patients in the report by Maronian et al. [6] who underwent laryngeal reinnervation procedures, eight developed a normal or improved voice. The interval between UVFP onset and surgery exceeded 12 months in all but one case. The longest interval was 9 years, after which the voice was noted to have improved. Denervation usually causes a drastic reduction in muscle fiber numbers and, ultimately, the total loss thereof. However, the duration of atrophy of human intrinsic laryngeal muscles remains unclear, because this varies among muscles [7]. Further work is required to determine the longest duration from UVFP onset to reinnervation that allows successful RLN reconstruction.

Another concern is the unfavorable effect of long-lasting nerve compression. Two such cases were encountered in our present series. In both patients, the RLNs were flat at the site of tumor adhesion because of chronic compression. One patient underwent nerve transfer using the intact ACN and experienced excellent vocal function postoperatively. The other patient underwent direct anastomosis of the severed ends because the gap was less than 5 mm; limited improvement in vocal function was noted (MPT: 7.1 s, G: 1.3, B: 1.0). Endoneurial tubes within a macroscopically distorted nerve may also be distorted, resulting in suppression of nerve regeneration per se or growth of sprouting fibers into the endoneurial tubes. Although we only have little evidence, we believe that any portion of a nerve that has been flattened by tumor compression should be removed and nerve reconstruction should be performed to optimize functional recovery.

Unless the ACN has been resected during prior surgery, reinnervation of the TA muscle is attempted together with AA. If the ACN cannot be bilaterally located because of the presence of cicatricial tissue, or if electrical stimulation of the ACN does not elicit SH muscle contraction, AA alone is performed, and type I thyroplasty or intracordal injection is scheduled for a later date.

Forty-three UVFP patients with severe breathy dysphonia underwent NMP flap implantation combined with AA between July 2002 and April 2011. In total, 10 patients were excluded from analysis: 2 required perioperative tracheostomy, 1 experienced wound opening caused by postoperative bleeding, 2 developed local infections after surgery, and 5 were lost to follow-up within 1 year. Consequently, 33 patients were finally enrolled (16 males, 17 females, age range 28–82 years, median age 59.5 ± 13.2 years). The average duration of UVFP prior to surgery was 24.9 months (range 1–366 months, median 9.0 months). The follow-up period ranged from 12 to 84 months (mean 29.0 months, median 26.0 months). The causes of UVFP are listed in Table 5.3.

Eleven UVFP patients with severe breathy dysphonia underwent nerve transfer combined with AA between October 2001 and November 2008. The primary etiologies of UVFP were associated with previous surgeries to treat a basal meningioma (1), esophageal cancer (1), a right vertebral artery aneurysm (1), a neck tumor (1), a vagal schwannoma (1), an aortic aneurysm (3), and a mediastinal tumor (1). The remaining two patients had left bulbar palsy (Wallenberg syndrome) (one) and an idiopathic etiology (one). Three patients were excluded from analysis; two patients experienced wound opening caused by postoperative bleeding, and one patient developed a recurrence of esophageal cancer. Thus, the vocal functions of eight patients (four males and four females) were followed-up over 2 years. The time from onset of paralysis to surgery ranged from 6 to 52 months (mean 23.0 ± 16.5 months).

The Mann–Whitney test was used in statistical analysis and P < 0.05 was considered statistically significant.