Unusual metastasis of hypernephroma in the maxillary sinus

OBJECTIVES: To determine the predictive value of intraoperative threshold stimulus for facial nerve outcome and the prevalence and prognostic value of persistent trains of activity and frequent spontaneous or mechanically induced contractions during acoustic neuroma surgery. STUDY DESIGN: Prospective recording and subsequent review of facial nerve activity. SETTING: Tertiary referral centre. PATIENTS AND METHODS: Consecutive patients undergoing acoustic neuroma surgery. Intraoperative facial nerve activity was digitised and stored on a personal computer for future analysis. Operative events were flagged. Recordings were available in 27 patients. MAIN OUTCOME MEASURES: Frequent mechanically induced contractions (< 20), prolonged trains of facial nerve activity (total time > 199 seconds), and facial nerve brainstem stimulus threshold were correlated with facial nerve outcome. RESULTS: A brainstem stimulus threshold > 0.1 mA was significantly associated with intermediate or poor facial nerve function (House-Brackmann grade > 2) on the sixth postoperative day, at 1 month and 6 months. Patients with normal or near-normal facial function on the first day and a threshold of > 0.1 mA were significantly more likely to develop a delayed facial nerve palsy. Frequent contractions were noted in 74% of patients and persistent train activity in 59%. Neither was predictive of facial nerve outcome. CONCLUSIONS: An elevated brainstem threshold is helpful in predicting delayed facial nerve palsy and suboptimal facial nerve outcome. Persistent train activity and frequent contractions, do not have major prognostic significance.     

Solid rod short segment anterior fusion in thoracolumbar scoliosis

BACKGROUND: Facial electroneurography (ENoG) is an established and reliable method for assessing neural degeneration in various conditions affecting the facial nerve. The facial nerve may be subclinically involved by parotid tumors, but estimating such involvement preoperatively may be difficult when facial function is normal. The hypothesis that preoperative ENoG: (1) can detect subclinical facial nerve degeneration as a measure of involvement by parotid tumors and (2) can predict facial nerve function following parotidectomy was prospectively evaluated in the present study. METHODS: Twenty-two patients undergoing parotidectomy for tumors were tested preoperatively with ENoG, and their facial nerve function was graded pre- and postoperatively (House-Brackmann system). Eight patients had malignant tumors and 14 benign tumors. RESULTS: In patients with malignant tumors, lower percentage of preoperative ENoG response indicated nerve involvement that was not evident on clinical examination and correlated significantly (p = .035) with postoperative facial nerve dysfunction. Preoperative ENoG reduction of greater than 80% was found in all patients whose facial nerve was infiltrated by tumor. In 14 patients with benign tumors, preoperative ENoG results had no correlation with postoperative facial function. CONCLUSIONS: In malignant tumors, even when facial function is clinically intact, a low preoperative ENoG response may predict facial nerve involvement by the tumor. The lower the preoperative ENoG response, the poorer is the expected postoperative facial nerve function. There was no such correlation in benign parotid tumors.     

Antiphospholipid syndrome associated with progressive systemic sclerosis

Hemifacial spasm is a neurological disorder due to abnormal hyperactivity of the facial nerve. The most common cause of hemifacial spasm is a neuro-vascular conflict in the cerebellopontine angle between a vascular loop and the root of the facial nerve (96% of cases). Tumors are the cause of hemifacial spasm in only 1% of cases). The authors present their results in 100 patients who underwent microvascular decompression for essential hemifacial spasm between 1990 and 1995. They used microsurgical and endoscopic procedures by a minimal retrosigmoid approach in all cases. The most common offending vessels were the posterior inferior cerebellar artery (70%), the vertebral artery (41%) and the anterior inferior cerebellar artery (28%). An aberrant vein was found in 2 cases. There were 38% of multiple artery-nerve conflicts. Physiopathology of hemifacial spasm is explained by two principal theories: in the ephaptic theory, hyperactivity and an abnormal nervous impulse pathway are due to a short demyelinated area on the nerve trunk caused by the offending vessel, inducing short circuiting between adjacent nerve fibers. In the nuclear theory, hyperactivity of the facial nerve is due to an abnormal and automatic activity of the facial nerve nucleus itself, induced by the vessel. The authors used pre and postoperative electromyographic tests and intraoperative electromyographic tests. Their results tend to prove the nuclear theory. Ninety per cent of the patients had a good result, with a mean follow-up time of 30 months in 60 cases. In 82% of the cases, there was a total recovery after a single procedure. There was no mortality and no facial palsy. Hearing loss occurred in less than 5%.     

Trans-petrous surgery in acoustic neuroma. Value of preoperative audiovestibular and facial investigation in the risk evaluation of facial nerve function

Between 1987 and february 1994, 162 consecutive patients with acoustic neuroma were operated on by an otoneurosurgery team, using transpetrous approaches (89% translabyrinthine, 8% middle fossa and 3% retrosigmoid). The relationship between the clinical, audiometric and vestibulographic characteristics and the post-operative facial nerve function were evaluated. In acoustic neuromas with cerebello-pontine component inferior to 3 cm without central neurologic signs (ic: central controlateral auditory and/or ipsilateral vestibular pathway alteration), good post-operative facial nerve function was achieved in 80% of cases. In acoustic neuromas superior to 3 cm with alteration of the central vestibular and auditory pathways, a good result was obtained in only 30% of cases which correlated negatively with preoperative facial dysfunction. These results underline the value of preoperative facial and audiovestibular examinations in predicting the postoperative facial nerve function following surgery for acoustic neuroma.     

Malignant epithelial parotid gland tumours: analysis and results in 65 previously untreated patients

BACKGROUND: Optimal management of malignant epithelial parotid tumours requires knowledge of the available therapeutic modalities and the different biological characteristics. The aim of the study was to review the characteristics of patients at presentation, histological classification, disease-free and overall survival rates, and the results of the applied treatment policy regarding the facial nerve and neck. METHODS: Between 1974 and 1995 a total of 65 patients was treated with curative intent for a previously untreated malignant epithelial parotid gland tumour. All patients underwent some type of parotidectomy, 20 of whom had an en bloc radical neck dissection. In selected cases the facial nerve or its branches were peeled off the tumour thus violating the objective of tumour-free margins and relying heavily on the efficacy of postoperative radiotherapy. In total 51 patients received postoperative radiotherapy. None of the patients was lost to follow-up. RESULTS: There were 12 locoregional failures (18 per cent). In only one of these 12 patients was salvage therapy successful; the remaining 11 patients died from the tumour. All but one of the eight patients with distant metastasis only died from the tumour. The estimated 5- and 10-year disease-free rates were 68 and 59 per cent respectively. The corresponding survival rates were 75 per cent and 67 per cent. A significant relationship could be observed between tumour stage and survival. The presence of lymph node metastases proved to be the strongest single prognostic factor. CONCLUSION: In selected cases a conservative approach towards the facial nerve is justified.     

Impending paradoxical embolus: clinical imaging and successful surgical treatment

AIM OF THE STUDY: Postoperative sensory component of the facial nerve after acoustic tumor surgery has received little attention in the literature. The object of the present investigation was to review this specific topic analyzing the postoperative frequency of taste and lacrimation (crocodile tears or dry eye) abnormalities. MATERIAL AND METHODS: 54 patients who underwent acoustic tumor removal were selected for this study. Each of these patients was recalled and pre and postoperative evolution of the sensory dysfunction were assessed. The latters were correlated with the facial function evaluated according to the House-Brackmann classification. RESULTS: Postoperative taste dysfunction (reduced or changed sensation) was complained by 38.5% of the patients. After surgery, 42.3% of the cases had crocodile tears, while in 59.6% altered tearing occurred. DISCUSSION: The present study, according to the Irving et al's experience confirmed a significant incidence of postoperative abnormal function of the sensory facial nerve. The influence of the motor component on these outcomes was variable. Lacrimation worsened when facial function was poor. On the other hand, grades V or VI did rarely manifest crocodile tears. Clinically, these findings implies the importance of a preoperative counseling of such particular aspect in the candidates to surgery of acoustic neuroma in order to adequately motivate them and, at the same time, to reduce their psychological discomfort.     

Intraoperative facial nerve monitoring (IFNM) predicts facial nerve outcome after resection of vestibular schwannoma

Intraoperative facial nerve monitoring (IFNM) is a suitable technique for intraoperative facial nerve identification and dissection, especially in large vestibular schwannomas (VS) (acoustic neuroma). To evaluate its feasibility for estimating functional nerve outcome after VS resection 60 patients underwent surgery using IFNM. Out of this group the last 40 patients were included in a prospective study evaluating the prognostic value of various IFNM parameters (proximal and distal absolute EMG amplitude, stimulation threshold, and proximal-to-distal amplitude ratio) for prediction of initial postoperative facial nerve function and recovery of function. Stimulation threshold and absolute EMG amplitude proximally at the brain stem were both predictive for postoperative nerve function. Good initial facial nerve outcome (modified House Brackmann grading, mHB degree I and degree II) was found in 15/16 patients with a proximal EMG amplitude greater than 800 microV and in 19/22 patients with proximal stimulation threshold less than 0.3 mA. Sixteen of 16 patients with proximal stimulation threshold equal to or greater than 0.3 mA had moderate-to-severe facial palsy (mHB degree III or worse). Six of six patients without evokable proximal amplitude initially had insufficient nerve function (mHB degree IV). Intraoperative decrease of the proximal amplitude was associated with an unfavourable outcome, whereas distal amplitudes usually stayed unchanged. Mean distal EMG amplitudes were also found to be decreased with poor nerve function, which may mean that the tumour had already affected the nerve. A proximal amplitude of 300 microV or less and a proximal-to-distal amplitude ratio below 1:3 were found in the absence of functional recovery in 6/8 (75%) and 5/6 (83%) patients with initial mHB degree IV, respectively. Two patients with initial mHB degree IV improved to mHB degree III despite intraoperative evidence of missing functional nerve integrity. Therefore, functional recovery cannot be predicted by IFNM in all cases of anatomical nerve preservation. We conclude that a minimum follow-up period of 1 year may still be advisable even in certain patients without evidence of intraoperative functional nerve integrity.     

Submandibulectomy in the treatment of pathology of the submandibular gland: our experience 1970-1995

Submandibular gland excision is proposed in the treatment of neoplastic and non neoplastic diseases; this surgical procedure can be performed by transoral or transcervical approach. The aim of the study is to demonstrate that cervical approach must be preferred because it is safer and allows a wider exposition of the surgical field. From 1970 to June 1995, 54 patients (47 with chronic sialadenitis, 7 with benign tumors and 7 with malignant tumors) were submitted to excision of the submaxillary gland. Of the 54 resections performed, 2 were completed with "functional" cervical lymphadenectomy and 1 with Radical Neck Dissection in pts. with malignant neoplasms. There were no postoperative deaths; complications occurred in 1 patient (1/54 = 1.8%) as a iatrogenic permanent lesion of the maxillary branch of the facial nerve (in detail 0/47 patients with benign disease and 1/7 (14.7%) patients with malignant disease). The cervical approach for the resection of the submaxillary gland is preferred to the transoral approach for the lower risk of iatrogenic lesions of the lingual and hypoglossal nerves and the possibility of curative resections in case of malignant neoplasms. A regulated and experimented technique through the cervical approach also lowers the risk of a lesion of the maxillary branch of the facial nerve.     

Biological effects and cellular uptake of c-myc antisense oligonucleotides and their cationic liposome complexes

OBJECTIVE: To review patient selection and factors affecting the decision for surgical or non-surgical management of patients with trauma-induced facial nerve palsy. STUDY DESIGN: Retrospective review of 85 consecutive cases occurring over a 6-year period. METHODS: Between 1984 and 1990, 85 cases of facial nerve palsy following trauma were treated. Patient assessment included clinical, audiological, radiological and EMG examinations. Depending on test results, patients were either given medical treatment or underwent total facial nerve decompression surgery. RESULTS: Recovery was achieved in all 33 medically treated patients. Among the 52 surgically-treated patients (61.2%), 47 had immediate facial nerve palsy at admission (90.4%). Onset of facial palsy was delayed in 5 others, including 3 for whom the delay to onset was unknown. Surgical accesses used were the middle fossa and transmastoidal (60%), translabyrinthine (11%), and pure transmastoidal (29%) routes depending on hearing loss, location of fracture line and general patient condition. Lesions were predominantly found in the geniculate ganglion area. A nerve gap was found in only 13.5% of the cases. Two years post-operatively, 92.3% of the patients had grade I-III recovery. No grade V or VI cases were observed. DISCUSSION AND CONCLUSION: The rarity of severe nerve lesions encountered in surgically-treated patients raises the question of better candidate selection for surgery. At present, imaging techniques, particularly radiology, can help to choose the best indication. Immediate total facial palsy associated with a clear-cut fracture going through the Fallopian canal is perhaps the only case requiring surgery. When the delay of onset is unknown, an EMG indicating a total denervation process is the decisive argument.     

Dual rearrangement of immunoglobulin and T-cell receptor genes in a case of Philadelphia chromosome-positive acute leukemia

In the medical treatment of facial nerve paralysis a large number of different techniques have been developed to restore the function of the facial nerve. These include (a) the ipsilateral nerve grafting (e.g., partial hypoglossal-facial, spinal accessory-facial, partial glossopharyngeal-facial), (b) crossfacial nerve grafting and (c) temporal muscle flaps or even free muscle transfers. None of these techniques uses the masseteric nerve as a graft for reconstruction of the facial nerve. This preliminary report deals with the anatomical basis, which could lead to a new technique. The masseteric nerve leaves the infratemporal fossa through the mandibular notch, accompanied by the masseteric artery. At this level the nerve consists in nine of 36 cases studied of only one branch (25.0%), in 17 cases of two branches (47.0%), in nine cases of three (25.0%), and in the remaining case of four branches (2.8%). There are three main reasons for considering the masseteric nerve as a possible donor for at least the orbicular branch of the facial nerve: (1) The approach to the mandibular notch is quite simple; (2) since the nerve consists of two or more branches in 75.0% of the cases, severe dysfunction of the masseter muscle should not occur; (3) if there is complete denervation of the masseter muscle, its function may be taken over by the temporalis muscle.     

Intraoperative monitoring of facial nerve antidromic potentials during acoustic neuroma surgery

The present paper presents monopolar recording of facial nerve antidromic potentials as an alternative technique to facial electromyography for the continuous monitoring of the facial nerve during acoustic neuroma surgery. The investigation involved 22 patients undergoing acoustic neuroma surgery via a retrosigmoid approach (tumour sizes ranging from 5 to 28 mm). Bipolar electrical stimulation of the marginalis mandibulae was performed to elicit facial nerve antidromic potentials. Stimulus intensity ranged from 2 to 6 mA with a delivery rate of 7/sec. A silver wire monopolar electrode positioned intracranially on the proximal portion of the acoustic facial bundle was used to record antidromic potentials. To define the specific origin of the action potentials and acquire normative data, monopolar and bipolar recordings of facial nerve antidromic potentials were performed in 15 subjects undergoing retrosigmoid vestibular neurectomy for Meniere's disease. The average facial nerve antidromic potential latency was 4.2 (+/- 0.6) msec in subjects with acoustic neuroma and 3.3 (+/- 0.2) msec in subjects with Meniere's disease. Facial nerve antidromic potentials furnished near real-time information about intraoperative facial nerve damage and postoperative facial nerve function during acoustic neuroma surgery. Facial nerve antidromic potentials may provide additional information to conventional EMG. They allow the use of endplate blockers, yield quantitative estimation of facial nerve conduction properties in terms of amplitude and latency, and allow actual continuous monitoring of the facial nerve.     

Surgical pathology of parotid gland tumors

The records of 27 patients operated for parotid tumors were reviewed retrospectively. Pleomorphic adenoma was the most frequent tumor (37.1%) and required subtotal parotidectomy in all cases. Twenty percent presented permanent facial paralysis of the marginal mandibular branch. No recurrence has been observed in five years of follow-up. Warthin's tumor, found in 11.1% of patients, was removed by either superficial or subtotal parotidectomy. Parotidean cysts were observed in 7.4% and were excised by superficial parotidectomy. The malignant tumors included squamous cell carcinoma (22.2%), adenoid cystic carcinoma (14.8%), melanoma (3.7%), and renal-cell metastasis (3.7%). All were treated by total parotidectomy with conservation of the facial nerve in 67%. Twenty-five percent had postoperative facial paralysis and 33% developed Frey's syndrome. Thirty-three percent died in the next 5 years from locoregional metastases.     

A comparison of electroneuronography with facial nerve latency testing for prognostic accuracy in patients with Bell's palsy

The purpose of this study was to evaluate the ability of electroneuronography (ENOG), also called evoked electromyography (EEMG), and facial nerve latency testing (FNLT) to assess the prognosis of facial nerve palsy, using the House-Brackmann facial nerve grading system as criterion. From 1988 to 1994 these tests were employed at the ORL Clinic of the University of Ioannina in 250 patients with idiopathic facial nerve palsy. The ENOG test results indicated that when the amplitude of the compound muscle action potentials ranged from 51% to 95% of the normal value, 97% of the patients achieved complete functional recovery (grade I) within at least 2 months. When the muscle action potential decreased to a value below 51% of normal values, prognosis for recovery was considerably worse. FNLT test results indicated that as the latency time extended, the recovery grade of the facial nerve worsened. When latency time was within the normal range (group A patients), about 92% of patients had complete functional recovery. In contrast all patients having either a very extended latency time or unable to be monitored (groups C and D) demonstrated incomplete functional recoveries that ranged from grade II to grade VI. Comparing each test with the House-Brackmann facial nerve grading system, we ascertained that the percent accuracy for ENOG was 97.6%, and that for FNLT was 94.4%.     

A biological basis for instantaneous centres of rotation of the vertebral column

BACKGROUND: The morbidity of treating benign parotid tumors is an important issue because the majority of patients are young. We therefore undertook this study to evaluate the outcome for these patients treated at the Princess Margaret Hospital. METHODS: A retrospective review was conducted on all patients registered with benign parotid tumors between 1970 and 1987. RESULTS: Seventy-six patients were included in this review; postoperative radiotherapy (XRT) was administered to 55 patients. Median follow-up time was 12.5 years. Postoperative XRT was particularly important for patients with recurrent disease. The risk of both temporary and permanent facial nerve paralyses increased with each operation. No patients in this study developed a malignancy. CONCLUSIONS: We recommend that all patients with benign parotid tumors be treated with a complete parotidectomy with preservation of the facial nerve. Consideration to postoperative XRT should be given to patients in whom there was tumor spillage, residual disease, or recurrences.     

Sindbis virus infection of neonatal mice results in a severe stress response

A free vascularized rectus abdominis muscle graft with a long motor nerve was used for reconstruction of unilateral established facial paralysis in one stage. Regarding the procedure, the pedicle vessels were anastomosed to the recipient vessels in the ipsilateral face, and the motor nerve of the muscle was sutured to the contralateral facial nerve. The advantages of using the rectus abdominis muscle are as follows: (1) the muscle is very thin, not bulky, (2) the muscle can be split easily to reduce the volume, (3) the intercostal nerve is long enough (more than 20 cm) to reach the contralateral facial nerve for suturing, (4) the pedicle vessels are large and long, (5) it is possible to carry out simultaneous operations with two teams, and (6) the donor-site morbidity is minimal. The disadvantages of this method are that complicated surgical dissection is required to obtain a motor nerve and that a postoperative abdominal hernia may occur.     

Temporalis muscle for facial reanimation. A 13-year experience with 224 procedures

A procedure for temporalis muscle transposition was used to reanimate the paralyzed face in 219 patients. In most cases, facial paralysis had followed an operation to remove an acoustic tumor. Analysis of the results showed this procedure to be highly successful and the method of choice, alone in cases of long-standing facial paralysis or to augment the effects of facial nerve grafting or hypoglossal-facial nerve anastomosis, in reanimating the mouth. It was successful in restoring a smile to 80% of the 219 patients and provided overall improvement in mouth function in 96%. Complications occurred in 21% of patients, with the most common being infection (12% of patients). Since one of us began to use the procedure to reanimate the eye and mouth, results of temporalis muscle transposition have been improved by the following: (1) using the procedure to reanimate the mouth only; (2) performing revision surgery, most often tightening the corner of the mouth (25% of patients), as indicated; (3) transposing only the midsection of the muscle; (4) implanting a prefabricated Silastic prosthesis to fill the muscle defect; (5) when indicated, lengthening the muscle with polytef (Gore-Tex+); and (6) placing the muscle in a tunnel lateral to the superficial musculoaponeurotic system to avoid injuring the underlying facial nerve should some spontaneous recovery of facial nerve function be possible.     

Intraoperative facial nerve monitoring in acoustic neuroma surgery

Intraoperative facial nerve monitoring simultaneously using electromyography and mechanical pressure sensors is being used in retrosigmoid and translabyrinthine approaches for acoustic neuroma resection. Insulated electrified microsurgical instruments and air drills are used to stimulate the facial nerve with a pulsed, constant current through bone and tumor, before the facial nerve is visually encountered. Electrical stimulation is used to help locate the facial nerve, map the course of the facial nerve within tumor, warn the surgeon of unexpected facial nerve locations, and help predict facial nerve function postoperatively. In 57 unmonitored cases a House-Brackmann (H-B) grade I or II result was obtained in 77 percent of small, 81 percent of medium, and 60 percent of large tumors. In 64 monitored cases H-B grade I or II was obtained in 88 percent of small, 79 percent of medium, and 90 percent of large tumors. Overall, facial nerve outcomes were better after monitored procedures (p < 0.02). A modified H-B classification for acute facial nerve injury is introduced to grade facial weakness immediately postoperatively and until function is stable at 1 year. In the unmonitored group there were five (9%) cases with a complete facial paralysis, facial nerve intact (i.e., acute H-B grade VIA) and seven (13%) cases with the facial nerve transected (i.e., acute H-B grade VIB). In the monitored group there were five (8%) acute H-B grade VIA and two (3%) acute H-B grade VIB results. In the unmonitored group of large tumors, there were statistically more patients with an acute H-B grade VIB result (p < 0.05).(ABSTRACT TRUNCATED AT 250 WORDS)     

Methylcobalamin treatment of Bell's palsy

Sixty patients with Bell's palsy were included in an open randomized trial. Patients were assigned into three treatment groups: steroid (group 1), methylcobalamin (group 2) and methylcobalamin + steroid (group 3). Comparison between the three groups was based on the number of days needed to attain full recovery, facial nerve scores, and improvement of concomitant symptoms. The time required for complete recovery of facial nerve function was significantly shorter ( p < 0.001) in the methylcobalamin (mean of 1.95 +/- 0.51 weeks) and methylcobalamin plus steroid groups (mean of 2.05 +/- 1.23 weeks) than in the steroid group (mean of 9.60 +/- 7.79 weeks). The facial nerve score after 1-3 weeks of treatment was significantly more severe (p < 0.001) in the steroid group compared to the methylcobalamin and methylcobalamin plus steroid groups. The improvement of concomitant symptoms was better in the methylcobalamin treated groups than the group treated with steroid alone.     

The Bor?s experience of endoscopic thoracic sympathicotomy for palmar, axillary, facial hyperhidrosis and facial blushing

OBJECTIVE: To study the outcome of endoscopic thoracic sympaticotomy (ETS) for palmar, axillary, facial hyperhidrosis and facial blushing. SUBJECTS: 1152 patients, 59% women and 41% men. INTERVENTION: ETS was performed by transection of the sympathetic chain where it overlies the second and third rib. The nerve was divided also over the fourth rib in patients with axillary hyperhidrosis. Questionnaires were sent to all patients. MAIN OUTCOME MEASURES: The effect of surgery was assessed by a 10 grad visual analogue scale (VAS) by the patients. The results were divided into effect rate (the effect on the symptom) and overall satisfaction rate, taking into account any side effects and complications apart from the effect. RESULTS: The response rate was 90%. The mean follow up time, effect rate and overall satisfaction rate were: 38 months for palmar hyperhidrosis, 99.4% and 87%; 26 months for axillary hyperhidrosis, 94.5% and 68%; 31 months for facial hyperhidrosis, 97% and 76%; 8 months for facial blushing, 96% and 85%. CONCLUSION: ETS is a very effective procedure in palmar, axillary and facial hyperhidrosis and facial blushing. The overall satisfaction rate is very good for palmar hyperhidrosis and facial blushing, not equally good but acceptable for facial hyperhidrosis. The lower satisfaction rate in patients with axillary hyperhidrosis makes this a questionable indication for ETS.     

Prediction of prognosis in peripheral facial nerve paralysis using submandibular gland scintigraphy

In this study, we evaluated the ability of submandibular gland scintigraphy to predict the prognosis of peripheral facial nerve paralysis. METHODS: Submandibular gland scintigraphy was performed in 78 patients with acute peripheral facial nerve paralysis. After injection of 180-370 MBq [99mTc]pertechnetate, serial 1-min images were acquired for 25 min. At 15 min after injection of radionuclide, ascorbic acid was administered intraorally to stimulate salivary secretion. Regions of interest were set manually on both submandibular glands, and time-activity curves were generated. The ratios of peak count density (PCR) and washout (WR) of the affected side to the normal side were calculated. Parameters of > or = 0.8 suggested normal affected submandibular function and indicated a good prognosis. RESULTS: Complete recovery of facial nerve paralysis was observed in 52 of 78 patients. The sensitivity, specificity and accuracy of PCR for a good prognosis were 79%, 50% and 69%, and those of WR were 85%, 77% and 82%, respectively. Positive and negative predictive values for a good prognosis were 76% and 54% in PCR and 88% and 71% in WR, respectively. When WR obtained within 14 days of the onset was used, positive and negative predictive values for a good prognosis were 94% and 73%, respectively. None of the eight patients who had values of <0.8 for both parameters within 14 days of the onset recovered completely. CONCLUSION: Submandibular gland scintigraphy can serve as a reliable indicator to predict the prognosis of acute peripheral facial nerve paralysis in its early symptomatic period.