- Email: [email protected]
Chemosensory and somatosensory regeneration after lingual nerve repair in humans
J Oral Maxillofac 552-l 3, 1997
Surg
Chemosensory and Somatosensory Regeneration After Lingual Nerve Repair in Humans JOHN
R. ZUNIGA, DMD, PHD,” AND CEIB L. PHILLIPS,
NING
CHEN,
MD,t
MPH, PHD*
Purpose: The objective of these studies was to measure the impact of Class IV and V lingual nerve injuries on taste sensitivity and taste receptor density of the anterior tongue before and after microneurosurgical repair. Materials and Methods: Citric acid detection threshold and suprathreshold magnitude response were measured on the anterior tongue in 12 adult volunteers with unilateral lingual nerve Class IV or V injuries. The right and left sides of the anterior tongue were tested at each session to assess the effect of nerve damage before and 1,3,6, and 12 months after repair. Whole-mouth threshold and suprathreshold scales of citric acid taste intensity were measured. The level of sensory impairment was scaled at each test session using a clinical neurosensory test algorithm. Finally, patients completed an 11 -item instrument survey that queried the patient’s perceived expectations of surgery on sensory, taste, and general health before surgery and the patient’s perceived outcome of surgery at each postrepair session. The patient’s perceived global satisfaction of surgery was also assessed. Results: All 12 patients failed to detect and scale citric acid and had complete or severe sensory impairment on the injured side of the anterior tongue. One year after repair, 50% of the patients demonstrated a substantial increase in the number of fungiform papillae, pores, and ratio of pores/papilla at the same time as they demonstrated the ability to detect and scale citric acid. After repair, patients perceived the greatest improvements in the categories of eating, chewing, feeling, and taste, and the least in speech. Conclusion: Lingual nerve repair may result in significant changes in somatosensory and chemosensory function and taste bud anatomy on the anterior tongue over time.
Complete or partial transection injuries of the lingual nerve occur with third molar odontectomy, reconstructive or ablative oral and maxillofacial surgery, mandibular orthognathic surgery, and trauma. When the lingual nerve has been sectioned, patients report varying degrees of sensory and taste dysfunction, with very little chance for spontaneous regeneration.i3’ In humans, the permanent loss of taste sensitivity on the ipsilateral side of the anterior tongue was universal on testing up to 5 years after transection injury.3’5 Case reports of atrophy of fungiform taste buds on the ipsilateral side of the anterior tongue after transection of the lingual nerve have been reported in the literature.3.6 We recently reported the absence of taste perception and reduction in the number of fungiform papillae and pores on the ipsilateral side of the anterior tongue after
* Associate Professor, Department of Oral and Maxillofacial Surgery, The University of North Carolina at Chapel Hill, Chapel Hill, NC. TFormerly, Visiting Research Fellow, Department of Oral and Maxillofacial Surgery, University of North Carolina at Chapel Hill, Chapel Hill, NC; Currently Vice-Chairman, Department of Oral and Maxillofacial Surgery, Nanjing Medical Center, Nanjing, Jiangsu, P.R.C. $ Research Professor, Department of Orthodontics, University of North Carolina at Chapel Hill, Chapel Hill, NC. Supported by National Institute of Dental Research grant DE10141 to JRZ. Address correspondence and reprint requests to Dr Zuniga: Department of Oral and Maxillofacial Surgery, University of North Carolina at Chapel Hill, School of Dentistry, Campus Box 7450, Chapel Hill, NC 27599-7450. 0 1997 American
Association
of Oral and Maxillofacial
Surgeons
0278-2391/97/5501-0001$3.00/O
2
Z&VGA, CHEN, AND PHILLIPS the partial or complete transection of the lingual nerve in 10 patients.7 Together, the published literature suggests that, in addition to sensory loss, patients will demonstrate loss or decrease in taste perception and taste bud density after partial or complete transection lingual nerve injuries attributable, in part, to the loss of the chorda-tympani contribution to the anterior tongue. If so, then the ability to measure the reduction in taste receptor density or taste perception on the anterior tongue may provide additional objective modalities to diagnose lingual nerve injuries during patient evaluation. Experimental studies in animals have shown that the suture anastomosis of the completely transected lingual nerve results in reformation of fungiform taste budsBs9 and the return of responses to taste stimulants applied to the anterior tongue.” In fact, regeneration of taste nerves in rodents has been characterized as being robust. Unfortunately, there is very little known about the recovery of taste after lingual nerve regeneration in humans. Despite proposals that restoration of taste function on the anterior tongue after lingual nerve repair was unlikely in humans,“.” recent reports have demonstrated the recovery of some taste function after lingual nerve repair.7”3X’4 Hillerup et all3 reported that 3.7 (1.1 to 4.6) years after repair, five of six patients detected sweet, sour, salty, and bitter stimulants applied to the repaired side of the anterior tongue. We recently demonstrated an increase in the number of regenerating fungiform taste receptors within 6 months of repair that was coincident with the recovery of sour taste detection and intensity scaling on the anterior tongue.7 Thus, recent studies are consistent with the prospect that, like animals, human taste receptors may regenerate and taste perception may recover after repair of sectioned lingual nerves. The primary purpose of this study was to measure the effect of injury and subsequent repair of the lingual nerve on taste in a clinical trial of patients electing surgery. The first objective of this study was to measure the effects of partial and complete transection of the lingual nerve on taste perception and fungiform taste receptor density on the anterior tongue. The value of distinguishing taste perception and receptor density on the injured and uninjured sides of the anterior tongue was to determine the diagnostic acumen of taste tests and provide baseline information about the regeneration of taste after repair. The second objective was to measure taste perception and fungiform taste receptor density on the repaired side of the anterior tongue for up to 1 year. Finally, different somatosensory modalities were measured to determine the relative contributions of the trigeminal nerve during regeneration of the lingual epithelium, and the relationship between the level of reinnervation of lingual nerve function and the maturation of taste buds and taste function.
Patients
and Methods
PATIENT SELECTION AND STUDY DESIGN Twelve paid volunteers were recruited to participate in this study. Each patient’s age, gender, and description of their lingual nerve injury and repair are provided in Table 1. Volunteers were nine women and three men (ages 23 to 45 years; mean age, 32) who satisfied the inclusion criteria of a healthy patient between the ages of 16 and 45 years, of any gender or race, with a unilateral Class IV or V lingual nerve injury of less than 1 year’s duration, with no historical or clinical evidence of neuropathic pain. Each patient was screened for metabolic, neurologic, and psychiatric disorders and were free of medications before testing. The average length of time from injury to repair was 5.25 ? 2.9 (standard deviation) months. The lingual nerve was injured during odontectomy in 10 cases and mandibular sagittal osteotomy in two cases. There were five right-sided and seven left-sided injuries. Each patient underwent a standardized surgical procedure, which included the transoral dissection of the lingual nerve in the paralingual sulcus of the posterior mouth, the transection of the nerve at the injured site to remove neuromatous tissue for biopsy, and a stumpto-stump coaptation using microsutures in eight cases or a single cable interpositional sural nerve graft in three cases. In one case, the lingual nerve was not repaired because of the extent of the injury, and this patient (no. 203) served as a positive control for repair. Using Seddon’s classification, a neurotmesis-type injury was confirmed by surgical findings and biopsy in every case. Using Sunderland’s classification, there were seven cases of Class IV (ie, partial transection) injury and five cases of Class V (ie, complete transection) injury. Each patient reviewed and signed an informed consent and received incentive payments for testing before and 1, 3, 6, and 12 months after repair. For each session, patients were requested not to eat or drink anything except water for at least 2 hours before testing. The studies were performed in a dental chair in a quiet room in the Clinical Taste Research Laboratory at the University of North Carolina at Chapel Hill, School of Dentistry. Each session required 2 hours of testing. TASTE SENSITIVITY AND TASTE RECEPTOR ANATOMY Detection of differences in sour taste sensitivity on the anterior tongue (ie, spatial testing) was performed with citric acid using a closed flow chamber delivery system previously described.” Briefly, citric acid solutions or distilled water were delivered to a closed spatially matched flow chamber attached to the injured and uninjured sides of the anterior tongue (1 cm from the anterior extent of the outstretched tongue and 0.5
4
REGENERATION
Table
1.
Patient
and Lingual
Nerve
Study
OF LINGUAL
NERVE IN HUMANS
Population Lingual Nerve Demographics
Patient No.
Gender
-4s W
Locatron
201
F F F M M F M F F F F F
27 26 29 23 44 26 45 27 29 43 43 26
Left Right Left Right Left Left Left Right Left Right Right Left
202 203 204 205 206 207 208 209 210 211 212
Classification”
V IV V IV IV V V IV V IV IV IV
Duration
(mo)
9 2 2
11 4 3 3 3 6 6 8 6
C!aLR
Repair
Odontectomy? Odontectomy Odontectomy Odontectomy Odontectomy Odontectomy Odontectomy Odontectomy BSSO BSSO Odontectomy Odontectomy
Graft Direct No repair$ Direct Direct Graft Direct Direct Graft Direct Direct Direct
* Sunderland partial (IV) or complete (V) transection. 7 Third molar extraction surgery. $ Positive control for repair.
cm from the midline). The internal dimensions of each chamber were identical (1,250 ,uL capacity: internal diameter = 14 mm, area = 1.54 cm2, perimeter = 44 mm) so that each tested area was spatially matched for each subject and for each session. Solutions were delivered to the chamber by an electric pump (Masterflex 7550-20, Cole Palmer Inst. Co, Chicago 11) at a flow rate of 1.5 mL/sec. Sour taste detection thresholds were measured with a temporal two-alternative forcedchoice procedure using the two-down one-up rule and an adaptation to the modified staircase.15 The citric acid stimuli were 16 concentrations of citric acid (J.T. Baker, #0122-01) that differed by 0.5 log unit serial dilutions (1 X 10e7 to 3 mol/L). Each citric acid stimulus was dissolved in room temperature distilled water. Room temperature distilled water served as the null stimulus and the rinsing solution. Citric acid and null stimuli were presented for 10 seconds; the intertrial interval was 30 seconds, during which the chamber was rinsed with distilled water before presentation of the next pair of stimuli. Scaling of citric acid taste intensity was accomplished with a modification of the magnitude matching procedure described by Stevens The magnitude responses to varying and MarksI brightnesses of a light source were used to normalize the magnitude response to taste stimuli. Delivery of visual stimuli was provided by a microcomputer-controlled tungsten lamp mounted inside a small windowed metallic box placed 30 cm in front and slightly to the left of the patient. Five levels of brightness were defined by approximately equal logarithmic increments of voltage over the range of 2.2 to 9.0 V. The citric acid stimuli were five concentrations differing by 0.5 log units of molar concentrations (0.003, 0.01, 0.03, 0.1,0.3 mom), presented in random order and in triplicate. After completion of each spatially matched citric
acid detection and scaling procedure, 5 mL 0.4% methylene blue was passed through the test chamber for 10 seconds. In humans, fungiform papillae are unstained by methylene blue, and are up to 1 cm in diameter, whereas pores are stained by methylene blue and are 10 to 50 pm in diameter.17 After staining, the chamber was removed, and images of the fungiform papillae and pores were recorded with a Zeiss OMPI dissecting microscope onto videotape (S-VHS-RT 120) with a solid-state color video camera (VK-C350, Hitachi, Lyndhurst, NJ). The patient and investigator viewed the images on a 20” color video monitor (CT-2010Y, Panasonic Comm. and Systems Co, Norcoss, GA). Videotapes of stained fungiform taste buds were reviewed using an S-VHS videoprinter (VP-1500, Codonics Inc, Middleburg Heights, OH). Fungiform papillae were identified and counted on images recorded at 40x, and pores were identified and counted from images recorded at 110X. After completing the citric acid spatial tests and videomicroscopy on both sides of the anterior tongue, the citric acid detection threshold and magnitude matching procedures were performed with a sip-and-spit method to examine taste sensitivity under whole-mouth conditions. SOMATOSENSORY
CAPACITY
AND IMPAIRMENT
SCALES
The level of somatosensory capacity on the anterior tongue was measured using a previously described clinical neurosensory test algorithm.‘* Three levels of testing were performed on the injured and uninjured sides of the anterior tongue at each session. Standard sensory procedures and psychophysical methodologies described by Essick” were used for testing. At level A testing, an evaluation of direction sensitivity using a No. 0 camel hair brush and two-point discrimination using a Boley’s gauge was performed. At level B test-
ZUNIGA,
CHEN,
AND
PHILLIPS
ing, an evaluation of contact detection using Semmes Weinstein pressure anesthesiometers (Stoeling Co, Houston, TX) was performed. At level C testing, an evaluation of thermal pain sensitivity using a computer-controlled thermode, built at University of North Carolina from a prototype developed at the Neurobiology and Anesthesiology Branch, National Institute of Dental Research, was performed. The objective of this evaluation was to grade the level of impairment in sensory capacity as normal, mild, moderate, severe, or complete. An abnormal response indicated that the test site value was less than that of the control site or published “upper normative limits,” and a normal response indicated that the test and control sites exhibited comparable values within published normative limits. An abnormal response at levels A and B and no response at level C was graded as complete sensory impairment (viz, anesthetic). An abnormal response at levels A and B and an elevated response at level C was graded a severe sensory impairment. An abnormal response at levels A and B and normal response at level C was graded a moderate sensory impairment. An abnormal response at level A and a normal response at level B was graded a mild sensory impairment. Finally, a normal response at level A was graded as normal (note that “normal” refers to the patient’s sensory capacity and not his or her report of altered sensation) . PATIENT-REPORTED
EXPECTATIONS
AND
OUTCOMES
Patient-reported outcome variables were assessedat each sessionusing instrumented surveys described by Ostler and Kiyak” and used previously by uszl to measure expectations of surgical outcomes versus actual surgical experiences and problems with oral function. Figure 1 illustrates the 1l-item expectancies questionnaire distributed to the patient before surgery by a member of the surgical team, who explained the form. The 11-item survey measured two dimensions: oral function dimensions, including eating, speech, taste, feeling, pain, and chewing; and psychosocial dimensions, including appearance, health, self-esteem, work performance, and a category called “other.” These items were accompanied by a 7-point response scale indicating the expected changes in each item in the list. A parallel measure was distributed to the patients at each postrepair examination to assessthe perception of change on a scale of “much worse” to “much better.” The postrepair outcome scale was delivered personally to the patient before any examination, completed in private by the patient, and returned to a member of the team who filed the results in private. The postrepair outcome scale used the same 11-item 7point response scale so that patients could indicate how each item had been changed when compared with presurgical conditions. Finally, a global satisfaction
Expectancies Whenever we undergo any medical or dental treatment we usually expect the treatment to result in some change in our physical and psychological health. We may expect these changes to be positive or negative. You are probably anticipating some change after your nerve surgery. For each of the areas listed below, circle the number which best describes your expectations, f?om -3 @iI1 become much worse) to 0 (no change), to +3 (will become much better). Will be much worse Eating Speech Tasting Feeling Pain Chewing General appearance General health Feeling about self Performing at work Other (Specify _)
-3 -3 -3 -3 -3 -3 -3 -3 -3 -3 -3
Will be much better
No change -2 -2 -2 -2 -2 -2 -2 -2 -2 -2 -2
-1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1
0 0 0 0 0 0 0 0 0 0 0
+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1
+2 +2 +2 +2 +2 +2 Jr2 +2 +2 +2 +2
+3 +3 +3 +3 +3 +3 +3 +3 +3 +3 +3
FIGURE 1. Expectancy questionaire used to survey patient’s motivation for surgery and perceived deficit(s) of six functional and five psychosocial dimensions. A postsurgical scale questionaire was collected at each postrepair examination.
questionnaire was delivered to the patient at each postrepair examination that required the patient to assign a numerical weight ranging from 0 to 4 (nil = 0, poor = 1, fair = 2, good = 3, excellent = 4) to indicate the overall satisfaction with the postrepair results. This survey was also self-administered in private and before examination. DATA
ANALYSES
The distribution of taste receptor anatomic variables (eg, fungifonn papilla density and ratio of pores, etc) was described by means, standard deviation (SD) and range values. The distribution of somatosensory tests results, the level of impairment, was numerically weighted (eg, normal = 1, mild = 2, moderate = 3, severe = 4, complete = 5) and described by means, SD, and range values. The distribution of scaled patient-reported expectancy and outcome results was numerically weighted (range, -3 to +3) and described by means, SD, and range values. A Wilcoxon signed rank test was used to determine the significant differences in anatomic variables between the injured and uninjured sides of the anterior tongue and to compare the presurgery expectancy score and sensory impairment level with the average postsurgery perception score and sensory level. For each anatomic measure, repeated measures analysis of variance with two within-subject factors (time and side) was used to assessthe pattern of recovery. All of the taste magnitude and detection threshold values were converted to logarithmic scales. At each point, for both the spatially matched and whole-mouth
6
REGENERATION
FIGURE 2. Clinical photographs of the anterior tongue in patients with unilateral lingual nerve transection injuries demonstrating the reduction in the number of grossly identifiable fungiform papillae on the injured side when compared with the uninjured side. The patient in A suffered from a Class V injury to the left lingual nerve. The patient in B suffered from a Class IV injury to the right lingual nerve.
procedures, least squares regression was used to describe the patient’s function between the log molar concentration and the geometric mean of the adjusted log response (each estimate of taste intensity was divided by the geometric mean of the patient’s estimate of brightness) for each side at each test session. The regression coefficient, calculated separately for time and side, was used as the outcome measures for evaluation of taste sensitivity. Repeated measures analysis of variance with two within-subject factors (time and side) was used to assess the pattern of recovery. Level of significance was set at .05. Results EFFECTS
OF INJURY
The findings indicated that Class IV and V injuries of the lingual nerve result in reduced numbers of fungiform papillae and pores on the anterior tongue in humans. Figure 2 shows clinical photographs of the ante-
OF LINGUAL
NERVE
IN HUMANS
rior tongue in two patients before repair, which demonstrate unilateral reductions in fungiform papillae without the aid of vital stains and videomicroscopy. The average number and standard deviation of stained fungiform papillae per chamber on the uninjured side of the anterior tongue was 16.6 t 8.6 (range, 7 to 37) and 10.9 t 6.8 (range, 4 to 23) on the injured side. The difference between the sides of the tongue was significant (P = .012). The average number and standard deviation of fungiform pores per chamber on the uninjured side of the anterior tongue was 50 2 29.6 (range, 21 to 117) and 9.8 + 8.4 (range, 0 to 25) on the injured side. The difference between sides of the tongue was statistically significant (P = ,005). The mean ratio of fungiform pores per papilla on the uninjured side of the anterior tongue was 3 + 0.8 (range, 2 to 4.8) and 0.8 +- 0.4 (range, 0 to 1.3) on the injured side. The difference between sides of the tongue was statistically significant (P = .OOS). Residual fungiform papillae were found on the injured side of the anterior tongue in every case. Figure 3 illustrates residual fungiform papillae in select patients. Residual fungiform papillae were generally located along the midline and anterior tongue tip and were frequently malformed; they were smaller, flatter, paler, polymorphic, and occasionally were filiform-like (ie, spinous processes extending from their centers). Normal-appearing residual papillae were found, but they were generally devoid of identifiable pores. The mean sour detection threshold on the uninjured side of the anterior tongue before repair was 0.97 f 1.2 mol/L (range, 0.008 to 3), which was not different from previously published normative values.15 The mean threshold on the injured side of the anterior tongue was >3 t 0 mol/L. These findings demonstrated that none of the patients could detect the difference between a stimulus of 3 mol/L citric acid and water on the injured side. Citric acid solutions more concentrated than 3 mol/L were not used to avoid burning the tongue mucosa. The mean whole-mouth threshold was 0.0017 -+ 0.002 mol/L (range, 0.00006 to O.Ol), which was not different from previously published normative values.22 Figure 4 illustrates the intensity rating power function curves of increasing suprathreshold concentrations of citric acid under the three conditions. The averaged patient’s responses and citric acid concentrations are plotted on logarithmic axes. The data for responses under the three conditions were extended for visual comparison. The averaged regression coefficient values for whole-mouth conditions was 2 2 0 for y-intercept and 0.4 + 0.2 for slope, which was not different from previously published normative values7 The mean intercept was 1.2 f 0.5 and mean slope was 0.32 I 0.24 on the uninjured side of the anterior tongue before repair, which was not different from previously published normative values. The mean intercept was 0.18 f 034. and the mean slope was
ZUNIGA,
CHEN,
AND
7
PHILLIPS
Table 2. Patient-Perceived Microsurgical Treatment
FIGURE! 3. Videophotographs of the methylene blue-stained anterior tongue in select patients. A, Many identifiable fungiform papillae are seen on the uninjured surface side of the anterior tongue after removal of a flow chamber. B. For comparison, the injured surface side of an anterior tongue contains very few identifiable fungiform papillae. C, Normal fungiform papillae with pores (arrows) on the uninjured side of the anterior tongue are demonstrated at a higher power. D, Residual fungiform papillae on the surface of the injured side of the anterior tongue are shown at a higher power to be smaller, paler, dysmorphic, and devoid of pores when compared with normal fungiform papillae in C. E, Two residual fungiform papillae on the surface of the injured side of the anterior tongue at higher power appear to be filiform-like (spinous).
Dimension
Mean
Speech Feeling Taste Eating Chewing Pain Appearance Health Self Work Other
1.2 2.0 2.0 2.1 2.2 1.9 0.9 0.9 0.4 0.2 0
Motives
for
(SD)
Range
(1.2)
o-+3 +1-+3 -k-t3 +1-f3 o-+3 0-+3 o--t3 o-+3 o-+3 o-+1 0
co.71
(0.5)
(0.6) (0.9)
(1.0)
(2.0) (2.0) (0.9) (0.4) 0
0.01 ? 0.05 on the injured side of the anterior tongue, which was significantly different from the uninjured side before repair (P = .0006 and .0002, respectively). Patients could not detect or scale the increasing concentrations of citric acid solutions on the injured side of the anterior tongue before repair. Before repair, the sensory impairment rating on the uninjured side of the anterior tongue was graded as normal (mean = 1 _f 0) in every case, and there were no subjective reports of altered sensation. The mean level of sensory impairment on the injured side before repair was 4.75 I- 0.4, and the difference between sides was significant (P = .002); nine patients demonstrated complete and three patients demonstrated severe sensory impairment. A positive trigger was elicited from the ipsilateral retromolar trigone in every caseon clinical examination before repair. Table 2 illustrates the rank order of patient-perceived expectancies for surgery. As found in previous studies, oral function dimensions, including improved taste and feeling, were most important at the presurgical assessment,” and
FIGURE 4. Mean intensity ratings of citric acid plotted against molar concentration before repair. The standard error of the normalized logarithmic means are marked. The figure graphs the magnitude responses under whole mouth conditions (m); spatially matched regions of the uninjured side of the anterior tongue (+); and spatially matched regions of the injured side of the anterior tongue (*).
MOLAR CONCENTRATION
(Log)
8
REGENERATION
Table
3.
Fungiform
Taste
Receptor
Response
to Microsurgical
OF
* t z 5
l-Year
N
FP*
Fpi-
Fp/FP:
5 5 1
13.4 (8.1) 8.4 (6.2) 15
11.8 (9.1) 7.6 (10) 9
0.8 (0.3) 0.7 (0.5) 0.6
Mean and standard deviation of identifiable fungiform Mean and standard deviation of identifiable fungiform Mean and standard deviation of the ratio of fungiform Difference within the same group and among groups
papillae. pores. pores/papilla. was statistically
psychosocial dimensions were least important. Based on mean scoresper dimension, the most important oral function expectancy among presurgical patients was chewing (2.2 + 0.9) and the least was speech (1.2 k 1.2). EFFECTS
OF REPAIR
Of the 12 patient volunteers, 1 did not complete all the postsurgical testing and was eliminated from postrepair analyses, and 1 did not have a repair performed because of the extent of injury and served as a positive control for repair. Repeated measuresanalysis of variance for the remaining 10 patients indicated that the pattern of difference between the injured and uninjured sides of the tongue was not consistent over time for the number of fungiform taste pores (P = .05), and the spatially matched detection threshold (P = .017), and was marginally significant for the number of fungiform papillae (P = .07). The pattern was consistent for the whole mouth detection threshold (P = .55). Five of the 10 patients demonstrated some recovery of taste function on the repaired side by 1 year and were considered responders. The remaining five patients demonstrated no recovery of taste function on the repaired side by 1 year and were considered nonresponders. The mean number of identifiable fungiform papillae (FP), pores (Fp) and the ratio of pores per papilla (Fp/FP) on the repaired side of the anterior tongue before and 1 year after repair in the five responders, five nonresponders, and positive control patient are tabulated in Table 3. There were measurable differences between the pre- and l-year postrepair mean numbers of FP (p = 5.4), Fp (,u = 23.8), and Fp/PP (II = 1.24) in the responders. By comparison, the pre- and l-year postrepair differences in mean numbers of FP (p = 2.2), Fp (p = 2.2), and Fp/FP (p = 0.36) in the nonrespondersand positive control patient was less. The data point out that neither the nonrespondersnor the positive control patient demonstrated a significant change in fungiform receptor density over time; however, the responders demonstrated an in-
significant
NERVE
IN HUMANS
Repair
Presurgical
Responders Nonresponders Positive control
LINGUAL
FP
18.8 (9.5)s 10.6 (5.3) 22
Postsurglcal FP
38.8 (19)§ 9.8 (6.9) 19
Fp/FP
2.1 (0.1)s 0.9 (0.4) 0.8
(P < .OOOl)
crease in fungiform receptor density over time. Figure 5 illustrates fungiform papillae and pores stained before and 1 year after repair in a responder patient. The increase in the density of fungiform papillae and pores are demonstrated, as well asthe detectable differences in receptor morphology over time. The mean citric acid detection threshold decreasedover time on the repaired side in the responders (1.3 t 1.5 mol/L) by 1 year.
FIGURE 5. Videophotographs of the injured sides of the anterior tongue before and after repair in a patient who recovered somatosensory and chemosensory function. A, The injured side of the anterior tongue before repair at low magnification, showing the reduced number of identifiable fungiform papillae. B, For comparison, the repaired side of the anterior tongue 1 year after repair, showing me increase in identifiable fungiform papillae. C, The injured side of the anterior tongue before repair in A at a higher magnification. D, For comparison with C, the repaired side of the anterior tongue 1 year after repair. E, The injured side of the anterior tongue before repair at high magnification, showing residual fungiform papillae. F, For comparison with E, the repaired side 1 year later, showing regenerate fungiform papillae with identifiable pores (arrows).
ZUNIGA,
CHEN,
AND
PHILLIPS
wx.4R~ENlR4noN,~~
IOURWC~WITIoN(t4,
FIGURE 6. Mean intensity ratings of citric acid plotted against molar concentration after repair over time. In each figure, magnitude scaling under whole-mouth conditions are indicated by 0, scaling of the uninjured side of the anterior tongue are indicated by q , and scaling on the repaired side are indicated by n . The test sessions are indicated as A, presurgical, B, 1 month postrepair, C, 3 months postrepair, D, 6 months postrepair, and /Z, 12 months postrepair.
In three of the five responders, the mean threshold difference between the injured/repaired (0.15 t 0.04 mol/L) and the uninjured sides (0.16 5 0.2 mol/L) was nearly identical at 1 postrepair year. Neither the five nonresponders nor positive control patient could detect 3 mol/L of citric acid applied to the repaired side of the tongue after 1 year. The pattern of difference between the regression coefficients of the injured and uninjured sides was not consistent over time (P = .04). There was a statistically significant difference in uninjured and injured regression coefficients at all times but the 6-month interval (p = 0.07). At all other times, the mean difference was greater than 0.23 (P 5 .Ol). Figure 6 demonstrates that responders were able to detect and scale citric acid on the uninjured side of the anterior tongue and whole mouth both before and at each session after repair. Figure 6 also demonstrates that responders were unable to scale citric acid before repair, but demonstrated the progressive increase in the ability to detect and scale citric acid on the repaired side over time. Two of the five responders were first able to scale citric acid stimuli 3 months after repair; two after 6 months, and one at 1 year. By 1 year, three of the five responders were able to scale the whole range of stimuli concentrations, whereas two scaled the weaker concentrations of citric acid as less intense than the same citric acid concentrations applied to the uninjured side. Neither the nonresponders nor the positive control subject could scale citric acid on the repaired side over the same time. Nine of 10 patients (90%) demonstrated an increase in somatosensory capacity on the repaired side of the
anterior tongue by 1 year. The mean score for level of sensory impairment of the nine patients was 2.1 2 0.9 at the l-year postrepair visit. The difference between the presurgical score (4.75 2 0.45) and l-year postrepair score was statistically significant (P = .OOOl). The distribution of sensory impairment levels ranged from normal to moderate impairment; the mean level of sensory impairment was first significantly different from presurgical levels at 3 months. One of 10 patients (10%) was unable to detect either somatosensory or chemosensory stimuli on the anterior tongue at the lyear postrepair examination. The positive control subject was unable to detect either somatosensory or chemosensory stimuli on the ipsilateral anterior tongue over the same period. Figure 7 illustrates the 12-month postrepair mean outcome score versus the mean presurgical expectation score for each functional and psychosocial dimension queried. Analysis of patient’s expectations and actual experiences with microsurgery showed that the functional and psychosocial expectations were achieved through surgery; in all areas, patients reported minimal to moderate fulfillment. The ranking of postrepair outcome scores was very similar to presurgical expectation scores, for which functional scores were more important and achieved greater scores than psychosocial scores. Among the functional outcomes scores, the greatest improvement was noted for chewing (1.1 + 0.8) and the least for speech (0.2 f 0.7). Among the patients who demonstrated taste recovery, the category of taste was scored higher (1.0 + 1) than by nonresponders (0.2 I 0.6) or the positive control patient (0). Among patients who demonstrated
10
REGENERATION
MUCH BElTER
NO
3
OF
LINGUAL
IN HUMANS
5 1% ;
FIGURE 7. Mean ratings of presurgical expectations score (0) vs 12-month postrepair outcomes score q for each functional (speech, feeling, taste, eating, chewing, and pain) and psychosocial (appearance, health, self, work, and other) dimension queried. The standard deviations of the means are marked.
0
CHANGE
0
EXPECTATIONS 12 MONTH
MUCH WORSE
NERVE
SCORE
OUTCOME
SCORE
m3
sensory recovery, the category of feeling was scored higher (1.3 t 1.2) than for the positive control patient (-1) or the patient without sensory recovery (0). Figure 8 shows graphic plots of the global satisfaction scores of the surgical outcomes by the patients at the different postrepair examinations. The highest scores were recorded at the l-year postrepair examination and the least at the l-month postrepair examination, suggesting there was a relationship between the progressive improvement in surgical outcome and global satisfaction. Discussion The reduction in the number of identifiable fungiform papillae, pores, and the ratio of pores per papilla, and the absenceof citric acid detection and magnitude
EXCELLENT
GOOD
FIGURE 8. Mean rating of patient’s satisfaction score with surgery at each postrepair examination. The standard deviations of the means are marked.
4
1
3
FAIR
2
POOR
1
NIL
scaling on the injured side of the anterior tongue demonstrated that Class IV and V lingual nerve injuries in humans result in a sensorineuraltaste disorder characteristic of dystrophic ageusia.The progressive increase in the number of identifiable fungiform papillae, pores, and the ratio of pores per papilla along with increased response sensitivity to citric acid stimuli on the repaired side of the anterior tongue in 50% of cases demonstrated that taste may regenerate after repair of sectioned lingual nerves in humans. Our finding that fungiform taste buds atrophy after lingual nerve injury points out that epithelial staining to count the number of taste pores in papillae may serve as a useful quantitative test for the diagnosis of Class IV and V lingual nerve injuries. The basis for staining was that fungiform taste pores are thought to be a classic example of neurotrophically dependent
0
1
6
3 MONTHS
ZUNIGA,
CHEN,
AND
PHILLIPS
receptor cells. In animals, experimental sectioning of taste nerves resulted in taste pore degeneration; although some taste pores persisted.x~9~23*24In animals, residual taste pores were either dysmorphic or atrophic, being 20% to 50% smaller than normal taste pores.23,24 In our study, the difference in the density of fungiform pores and the ratio of pores per papilla between sides of the anterior tongue was significant; thus, like animals, determining the reduction or loss of normal fungiform taste pore density and ratio of pores per papilla in humans will provide positive identification of a damaged lingual nerve when there exists no other possible reason for atrophy (eg, previous chorda-tympani nerve injury secondary to middle ear surgery, etc). Unlike fungiform taste pores, the nerve dependence of fungiform papillae for normal anatomic maintenance remains controversial. In animals, fungiform papillae undergo varying degrees of degenerative changes after taste nerve section; however, anatomically normal papillae persist.8,9,25 In our study, the difference in the density of fungiform papillae between sides of the anterior tongue was significant but marginal; thus, identifying fungiform papillae by staining or by gross examination of the anterior tongue sides will not be a sensitive test for all lingual nerve transection injuries because Class IV and V injuries were found when there was little difference between sides presurgically (false negative). However, the loss of grossly identifiable fungiform papillae on the injured side of the anterior tongue (eg, ipsilateral atrophy) will provide positive identification of lingual nerve sectioning. Interestingly, like animal studies, our findings demonstrated that residual fungiform papillae were generally flat, small, pale, and dysmorphic in appearance,24 and occasionally associated with filiform-like spines.25 Presurgical staining and counting of fungiform taste receptors required the use of videomicroscopy because the normal range of taste pore and papilla size was outside the range for gross visualization and the normal fasciculation of the tongue epithelium by the musculature inhibited in situ counting accuracy. The method will generally preclude routine office examination; thus, live videomicroscopy of the tongue will require dedicated equipment and personnel if accurate and objective testing for dystrophic ageusia are demanded by the clinician. Additionally, we currently use the triphenylmethane dye, FD & C Blue No. 1, as the epithelial stain because methylene blue has been shown to evoke mutagenic changes in bacteria DNA when combined with visible light. The detection of significant differences in taste sensitivity on the anterior tongue (spatial testing) is recommended for the clinical assessment of lingual nerve injuries because the finding of ageusia was universal. Like previous studies, applying different taste stimuli to the anterior tongue using different stimulus delivery systems (eg, filter paper discs, cotton-tipped applica-
11 tors, solution-soaked paint brush) determined the reduction (hypogeusia) or loss (ageusia) of taste on the ipsilateral side after nerve section in humans.3Z435.7S13-‘4 In our study, none of the patients was able to detect or scale suprathreshold concentrations of citric acid applied to a closed spatially matched flow chamber attached to the anterior tongue; they all demonstrated sour ageusia. Although not tested, Class IV and V lingual nerve-injured patients would be expected to demonstrate nonspecific ageusia because previous studies reported loss of salt, sweet, and bitter sensitivity on the ipsilateral anterior tongue.4,5 In contrast, presurgical whole-mouth citric acid detection threshold and magnitude scaling were normal. Our findings are similar to those reported in human studies after experimental lingual nerve anesthesia26 and lingual nerve injury.5.‘4 Thus, humans may be unaware of discrete losses of taste even after the loss of extensive lingual nerve input, making tests of whole-mouth taste function uninformative. The demonstrated impairment of somatosensory capacities on the anterior tongue in patients with Class IV and V injuries of the lingual nerve are in agreement with most studies.‘,* Our results concur that the measured level of sensory impairment in Class IV and V injured patients ranked between severe and complete (anesthesia) and were generally associated with a radiating trigger from the ipsilateral retromolar trigone to the anterior tongue. Patients consistently scored oral functional dimensions as more important than psychosocial dimensions (viz, expectancy ratings) before microsurgical intervention. Chewing was considered a top priority presurgically, followed closely by eating, feeling, taste, and pain; the lowest priority rating was given to speech. These findings are similar to previously reported samples of mixed (eg, inferior alveolar and lingual nerve) nerve-injured patient surveys2i and demonstrate that patients with Class IV and V lingual nerve injuries are aware of deficits in oral function, including taste. The serial examination of 10 postrepair patients and the positive control patient (no repair) for 1 year demonstrated that a repaired lingual nerve transection in humans can result in the recovery of taste, the regeneration of fungiform taste receptors, and the recovery of some sensory function. The recovery of taste sensitivity on the anterior tongue was coincident with the regeneration of identifiable fungiform taste buds. In animal studies, reinnervation of the epithelium by repaired sectioned lingual nerves resulted in the reappearance of fungiform taste pores,‘.” an increase in the number and size of fungiform papillae,8,23 and recorded whole nerve response to tastants applied to the anterior tongue.83’0 Our findings indicated that the characteristics of taste nerve regeneration described in animals are very similar in humans. In animals, the regeneration of taste buds and the recovery of taste response occurred
12 only by homologous reinnervation of axons from taste nerves (eg, chorda-tympani, glossopharyngeal) and not by nontaste nerves (eg, trigeminal, hypoglossal).27 Our findings suggest that the regeneration of taste buds and the recovery of taste sensitivity was taste nerve dependent, because the positive control subject did not demonstrate any change in taste bud anatomy or response over the same period, and four patients recovered sensory function without changes in taste bud anatomy or response. The effect of age, gender, location, classification, duration of injury, or repair on taste bud anatomy or response was not assessed because the sample size was too small to draw definitive conclusions. Hillerup et all3 reported that 3.7 (1.1 to 4.6) years after repair, five of six patients (80%) detected sweet, sour, salty, and bitter stimulants applied to the repaired side of the anterior tongue. Wiethijlter et a128reported that only 3 of 24 patients (12%) and Hesseling et a129 reported that only 2 of 17 (11%) patients demonstrated the recovery of some taste sensitivity after lingual nerve repair. Five of 10 patients (50%) regenerated taste buds and recovered some taste function in our study. The disparate reports of the percentage of patients demonstrating taste recovery after lingual nerve repair may be attributable, in part, to fascicular distributions noted by several authors.12x3’ In cats, Holland and Robinson3’ demonstrated that at the retromolar trigone the lingual nerve axon counts were 2.7 times more than chorda-tympani nerve axons and the distribution of small fibers was toward the chorda-tympani nerve. In human cadaver studies, Girod et al” demonstrated that the chorda-tympani was monofascicular proximal to the retromolar trigone, but then distributed widely within the lingual nerve proper. This study agrees with the proposal espoused by Girod et all2 that the recovery of taste function after lingual nerve repair may be attributable, in part, to anatomic alignment of taste fibers achieved by microsurgery. Interestingly, 90% of patients recovered sensory function, and only 50% recovered some taste function, which points out that the mathematical probability of sensory fiber regeneration may be greater than taste fiber regeneration, in part because of the geometry of fibers aligned by microsurgery. Further studies will be conducted to test this hypothesis. The finding that somatosensory reinnervation of the anterior tongue in humans was substantially better than presurgical conditions, but incomplete, was consistent with previous reports.2,3’ In 90% of cases, the level of sensory impairment decreased from severe/complete sensory impairment (mean score of 4.75) to mild sensory impairment (mean score of 2.1). The recovery of somatosensory sensitivity was nerve dependent because the positive control patient did not demonstrate any change in somatosensory sensitivity over the same period.
REGENERATION
OF
LINGUAL
NERVE
IN HUMANS
When averaged across time, the perception of change by surgery was perceived as an improvement, but, in all areas, patients reported less of an improvement than expected. This was especially true for functional scores, the areas of greatest concern for the patients presurgically. It was possible that these findings point out that presurgical expectations tend to be higher than what the patient actually experiences. However, patients generally rated the mean global satisfaction with postrepair outcome as progressive from “poor to nil” 1 month postrepair to “fair to good” at 1 year, indicating that they were astute enough to recognize the progressive recovery of somatosensory and chemosensory sensitivity over time. Patients who demonstrated recovery of taste sensitivity were more likely to report subjective improvement than those who demonstrated no recovery of taste sensitivity. In conclusion, we demonstrated that Class IV and V injuries of the lingual nerve in humans result in a sensorineural taste disorder characteristic of dystrophic ageusia of the anterior tongue, severe or complete sensory impairment of the anterior tongue, and patientperceived deficits in oral functional dimensions. The detection of differences in taste sensitivity on the anterior tongue and the staining and counting of fungiform taste receptors are recommended to diagnose dystrophic ageusia and positively identify Class IV and V lingual nerve injuries in patients presurgically. Dedicated equipment and trained personnel are required to perform spatial taste tests and stain lingual epithelium for counting taste buds. This study documented the regeneration of fungiform taste receptors and the recovery of taste in humans after lingual nerve repair. One-year postrepair tests demonstrated that 90% of patients recovered some somatosensory sensitivity, and 50% of patients recovered some or all of their chemosensory sensitivity. The results confirm that patients with Class IV and V lingual nerve injuries benefit by surgical intervention. Acknowledgment The authors and Mohamed
thank Robert Englehardt, DDS, Teresa Tucker, Malik, ES, for their technical support.
RN,
References 1. Mozary PG, Middleton RA: Microsurgical reconstruction of the lingual nerve. J Oral Maxillofac Surg 42:415, 1984 2. Zuniga JR: Multimodal scaling of sensory recovery after microsurgical repair. J Oral Maxillofac Surg 48:85, 1990 3. Bull TR: Taste and the chorda tympani. J Laryngol Otol79:479, 1965 4. Grant R, Miller S, Simpson D, et al: The effect of chorda tympani section on ipsilateral and contralateral salivary secretion and taste in man. J Neurol Neurosurg Psychiatry 52: 1058, 1989 5. Abrahams H, Travers S, Travers J, et al: Chemosensory adaptation following chorda-lingual nerve injury. J Oral Maxillofac Surg 51:180, 1993
SBREN
13
HILLERUP
6. Ogden GR: Atrophy of fungiform papillae following lingual nerve damage: A poor prognosis? Br Dent J 167:332, 1985 7. Zuniga JR, Chen N, Miller IJ: Effects of chorda-lingual nerve injury and repair on human taste. Chem Senses 19:657, 1994 8. Cheal M, Oakley B: Regeneration of fungiform taste buds: Temporal and spatial characteristics. J Comp Neurol 172:609, 1977 9. Robinson PP, Winkles PA: The number and distribution of fungiform papillae and taste buds after lingual nerve injuries in cats. Arch Oral Biol 36:885, 1991 10. Robinson PP: The reinnervation of the tongue and salivary glands after lingual nerve injuries in cats. Brain Res 483:259, 1989 11. Hausaman JE: Zur indikation der mikronervenchirurgie, in Schwenzer N, Pfiefer G (eds): Fortschritte der kiefer-und gesichtschirurgie band XXVIII. Experimentelle mund-kiefergesichtschirurgie, mikrochirurgische eingriffe. New York, NY, Thieme, 1983, pp 163-167 12. Girod SC, Neukam FW, Girod B, et al: The fascicular structure of the lingual nerve and the chorda tympani: An anatomical study. J Oral Maxillofac Surg 47:607, 1989 13. Hillerup S, Hjorting-Hansen E, Reumert T: Repair of the lingual nerve after iatrogenic injury: A follow-up study of return of sensation and taste. J Oral Maxillofac Surg 52: 1028, 1994 14. Baker SB, Foote JW, DiNick V: Gustatory recovery in microsurgical repair of the lingual nerve. J Oral Maxillofac Surg 53:137, 1995 15. Zuniga JR, Davis SD, Englehardt RA, et al: Taste performance on the anterior human tongue varies with fungiform taste bud density. Chem Senses 18:449, 1993 16. Stevens JC, Marks LE: Cross-modality matching functions generated by magnitude estimation. Percept Psychophys 27:379, 1980 17. Miller IJ, Reedy FE: Quantification of fungiform papillae and taste pores in living human subjects. Chem Senses 15:281, 1990 18. Zuniga JR, Essick GK: A contemporary approach to the clinical evaluation of trigeminal nerve injuries. Oral Maxillofac Surg Clin North Am 4:353, 1992 19. Essick GK: Comprehensive clinical evaluation of perioral sensory function. Oral Maxillofac Surg Clin North Am 4:503, 1992
J Oral Maxillofac 55:13-14, 1997
20. Ostler S, Kiyak HA: Treatment expectations versus outcomes among orthognathic surgery patients. Int J Adult Orthod Orthognath Surg 6:247, 1991 21. Zuniga JR: Perceived expectation, outcome, and satisfaction of microsurgical nerve repair. J Oral Maxillofac Surg 49:77, 1991 22. Bartoshuk JM, Rifkin B, Marks LE, et al: Taste and aging. J Gerontol 41:51, 1986 23. Hard af Segerstad C, Hellekant G, Farbman AI: Changes in number and morphology of fungiform taste buds in rat after transection of the chorda tympani or chorda lingual nerve. Chem Senses 14:335, 1989 24. Oliver SD, Whitehead MC: Morphometry and cellular dynamics of denervated fungiform taste buds in the hamster. Chem Senses 17:529, 1992 25. Oakley B: Neuronal-epithelial interactions in mammalian gustatory epithelium, in Bock GR, Whelan J (eds): Regeneration of Vertebrate Sensory Receptor Cells. Chichester, England, Wiley and Sons, 1991, pp 277-293 26. Catalanotto FA, Bartoshuk LM, Ostrum KM, et al: Effects of anesthesia of the facial nerve on taste. Chem Senses 18:461, 1993 27. State FA, Hamed MS, El-Hashash MK, et al: Trophic specificity of the gustatory fibers upon taste bud regeneration. Acta Anat 113:196, 1982 28. Wietholter H, Riediger D, Ehrenfeld M, et al: Ergebnisse der mikrochirurgie sensibler peripherer iiste des nervus mandibularis, in Schwenzer N, Pfiefer G (eds): Fortschritte der Kieferund Gesichtschirurgie. Band XXXV. Mikrochirurgie in der Mund-, Kieferund Gesichtschirurgie. New York, NY, Thieme, 1990, pp 128-134 29. Hesseling KH, Reich RH, Hausaman JE, et al: Langzeitergebnisse nach mikrochirurgischer nervenrekonstruktion im KopfHals-Bereich, in Schwenzer N, Pfiefer G (eds): Fortschritte der Kiefer-und Gesichtschirurgie. Band XXXV. Mikrochirurgie in der Mund-Kieferund Gesichtschimrgie. New York, NY, Thieme, 1990, pp 134-138 30. Holland GR, Robinson PP: Axon populations in cat lingual and chorda tympani nerves. J Dent Res 71:1468, 1992 31. LaBanc JP, Gregg JM: Trigeminal nerve injuries: Basic problems, historical perspectives, early successes and remaining challenges. Oral Maxillofac Surg Clin North Am 4:277, 1992
Surg
Discussion Chemosensory and Somatosensory Regeneration After Lingual Nerve Repair in Humans S&en Hillerup,
DOS, PhD, Dr Odont
Copenhagen
University
This
County
study
on chemosensory,
Hospital
Gloxtrup, somatosensory,
Glostmp,
Denmark
and epithe-
ha1 regeneration after lingual nerve repair deals with an important issue and contributes original results based on sophisticated test methods. It also demonstrates some of the pitfalls associated with this type of clinical research. One of these relates to the authors’ desire to disseminate as much experience as possible thus yielding an overflow of information. The purpose of study was to measure the effect of injury and subsequent repair in patients with partial and complete transection of the lingual nerve based on a number of variables. However, the quality of the sample is compromised
by being composed of cases of both partial and complete lingual nerve transection. This leaves two important questions open, of which only the first one can be answered: 1. What was the basis of diagnosis of partial and complete transection? Partial transection is a broad classification, theoretically including a break in the continuity of only one fascicle to a break in all fascicles but one. 2. What was the potential for restitution by spontaneous healing of the seven cases with only partial transection had they not been repaired? Moreover, tional
sural
three cases were managed with an interposinerve
end anastomosis.
graft,
whereas
seven
This inconsistency
taken into consideration, or in the discussion.
either
cases
had an end-to-
is not referred
in the presentation
to or
of results,
The method of testing sour taste sensitivity using a closed
Surg
Chemosensory and Somatosensory Regeneration After Lingual Nerve Repair in Humans JOHN
R. ZUNIGA, DMD, PHD,” AND CEIB L. PHILLIPS,
NING
CHEN,
MD,t
MPH, PHD*
Purpose: The objective of these studies was to measure the impact of Class IV and V lingual nerve injuries on taste sensitivity and taste receptor density of the anterior tongue before and after microneurosurgical repair. Materials and Methods: Citric acid detection threshold and suprathreshold magnitude response were measured on the anterior tongue in 12 adult volunteers with unilateral lingual nerve Class IV or V injuries. The right and left sides of the anterior tongue were tested at each session to assess the effect of nerve damage before and 1,3,6, and 12 months after repair. Whole-mouth threshold and suprathreshold scales of citric acid taste intensity were measured. The level of sensory impairment was scaled at each test session using a clinical neurosensory test algorithm. Finally, patients completed an 11 -item instrument survey that queried the patient’s perceived expectations of surgery on sensory, taste, and general health before surgery and the patient’s perceived outcome of surgery at each postrepair session. The patient’s perceived global satisfaction of surgery was also assessed. Results: All 12 patients failed to detect and scale citric acid and had complete or severe sensory impairment on the injured side of the anterior tongue. One year after repair, 50% of the patients demonstrated a substantial increase in the number of fungiform papillae, pores, and ratio of pores/papilla at the same time as they demonstrated the ability to detect and scale citric acid. After repair, patients perceived the greatest improvements in the categories of eating, chewing, feeling, and taste, and the least in speech. Conclusion: Lingual nerve repair may result in significant changes in somatosensory and chemosensory function and taste bud anatomy on the anterior tongue over time.
Complete or partial transection injuries of the lingual nerve occur with third molar odontectomy, reconstructive or ablative oral and maxillofacial surgery, mandibular orthognathic surgery, and trauma. When the lingual nerve has been sectioned, patients report varying degrees of sensory and taste dysfunction, with very little chance for spontaneous regeneration.i3’ In humans, the permanent loss of taste sensitivity on the ipsilateral side of the anterior tongue was universal on testing up to 5 years after transection injury.3’5 Case reports of atrophy of fungiform taste buds on the ipsilateral side of the anterior tongue after transection of the lingual nerve have been reported in the literature.3.6 We recently reported the absence of taste perception and reduction in the number of fungiform papillae and pores on the ipsilateral side of the anterior tongue after
* Associate Professor, Department of Oral and Maxillofacial Surgery, The University of North Carolina at Chapel Hill, Chapel Hill, NC. TFormerly, Visiting Research Fellow, Department of Oral and Maxillofacial Surgery, University of North Carolina at Chapel Hill, Chapel Hill, NC; Currently Vice-Chairman, Department of Oral and Maxillofacial Surgery, Nanjing Medical Center, Nanjing, Jiangsu, P.R.C. $ Research Professor, Department of Orthodontics, University of North Carolina at Chapel Hill, Chapel Hill, NC. Supported by National Institute of Dental Research grant DE10141 to JRZ. Address correspondence and reprint requests to Dr Zuniga: Department of Oral and Maxillofacial Surgery, University of North Carolina at Chapel Hill, School of Dentistry, Campus Box 7450, Chapel Hill, NC 27599-7450. 0 1997 American
Association
of Oral and Maxillofacial
Surgeons
0278-2391/97/5501-0001$3.00/O
2
Z&VGA, CHEN, AND PHILLIPS the partial or complete transection of the lingual nerve in 10 patients.7 Together, the published literature suggests that, in addition to sensory loss, patients will demonstrate loss or decrease in taste perception and taste bud density after partial or complete transection lingual nerve injuries attributable, in part, to the loss of the chorda-tympani contribution to the anterior tongue. If so, then the ability to measure the reduction in taste receptor density or taste perception on the anterior tongue may provide additional objective modalities to diagnose lingual nerve injuries during patient evaluation. Experimental studies in animals have shown that the suture anastomosis of the completely transected lingual nerve results in reformation of fungiform taste budsBs9 and the return of responses to taste stimulants applied to the anterior tongue.” In fact, regeneration of taste nerves in rodents has been characterized as being robust. Unfortunately, there is very little known about the recovery of taste after lingual nerve regeneration in humans. Despite proposals that restoration of taste function on the anterior tongue after lingual nerve repair was unlikely in humans,“.” recent reports have demonstrated the recovery of some taste function after lingual nerve repair.7”3X’4 Hillerup et all3 reported that 3.7 (1.1 to 4.6) years after repair, five of six patients detected sweet, sour, salty, and bitter stimulants applied to the repaired side of the anterior tongue. We recently demonstrated an increase in the number of regenerating fungiform taste receptors within 6 months of repair that was coincident with the recovery of sour taste detection and intensity scaling on the anterior tongue.7 Thus, recent studies are consistent with the prospect that, like animals, human taste receptors may regenerate and taste perception may recover after repair of sectioned lingual nerves. The primary purpose of this study was to measure the effect of injury and subsequent repair of the lingual nerve on taste in a clinical trial of patients electing surgery. The first objective of this study was to measure the effects of partial and complete transection of the lingual nerve on taste perception and fungiform taste receptor density on the anterior tongue. The value of distinguishing taste perception and receptor density on the injured and uninjured sides of the anterior tongue was to determine the diagnostic acumen of taste tests and provide baseline information about the regeneration of taste after repair. The second objective was to measure taste perception and fungiform taste receptor density on the repaired side of the anterior tongue for up to 1 year. Finally, different somatosensory modalities were measured to determine the relative contributions of the trigeminal nerve during regeneration of the lingual epithelium, and the relationship between the level of reinnervation of lingual nerve function and the maturation of taste buds and taste function.
Patients
and Methods
PATIENT SELECTION AND STUDY DESIGN Twelve paid volunteers were recruited to participate in this study. Each patient’s age, gender, and description of their lingual nerve injury and repair are provided in Table 1. Volunteers were nine women and three men (ages 23 to 45 years; mean age, 32) who satisfied the inclusion criteria of a healthy patient between the ages of 16 and 45 years, of any gender or race, with a unilateral Class IV or V lingual nerve injury of less than 1 year’s duration, with no historical or clinical evidence of neuropathic pain. Each patient was screened for metabolic, neurologic, and psychiatric disorders and were free of medications before testing. The average length of time from injury to repair was 5.25 ? 2.9 (standard deviation) months. The lingual nerve was injured during odontectomy in 10 cases and mandibular sagittal osteotomy in two cases. There were five right-sided and seven left-sided injuries. Each patient underwent a standardized surgical procedure, which included the transoral dissection of the lingual nerve in the paralingual sulcus of the posterior mouth, the transection of the nerve at the injured site to remove neuromatous tissue for biopsy, and a stumpto-stump coaptation using microsutures in eight cases or a single cable interpositional sural nerve graft in three cases. In one case, the lingual nerve was not repaired because of the extent of the injury, and this patient (no. 203) served as a positive control for repair. Using Seddon’s classification, a neurotmesis-type injury was confirmed by surgical findings and biopsy in every case. Using Sunderland’s classification, there were seven cases of Class IV (ie, partial transection) injury and five cases of Class V (ie, complete transection) injury. Each patient reviewed and signed an informed consent and received incentive payments for testing before and 1, 3, 6, and 12 months after repair. For each session, patients were requested not to eat or drink anything except water for at least 2 hours before testing. The studies were performed in a dental chair in a quiet room in the Clinical Taste Research Laboratory at the University of North Carolina at Chapel Hill, School of Dentistry. Each session required 2 hours of testing. TASTE SENSITIVITY AND TASTE RECEPTOR ANATOMY Detection of differences in sour taste sensitivity on the anterior tongue (ie, spatial testing) was performed with citric acid using a closed flow chamber delivery system previously described.” Briefly, citric acid solutions or distilled water were delivered to a closed spatially matched flow chamber attached to the injured and uninjured sides of the anterior tongue (1 cm from the anterior extent of the outstretched tongue and 0.5
4
REGENERATION
Table
1.
Patient
and Lingual
Nerve
Study
OF LINGUAL
NERVE IN HUMANS
Population Lingual Nerve Demographics
Patient No.
Gender
-4s W
Locatron
201
F F F M M F M F F F F F
27 26 29 23 44 26 45 27 29 43 43 26
Left Right Left Right Left Left Left Right Left Right Right Left
202 203 204 205 206 207 208 209 210 211 212
Classification”
V IV V IV IV V V IV V IV IV IV
Duration
(mo)
9 2 2
11 4 3 3 3 6 6 8 6
C!aLR
Repair
Odontectomy? Odontectomy Odontectomy Odontectomy Odontectomy Odontectomy Odontectomy Odontectomy BSSO BSSO Odontectomy Odontectomy
Graft Direct No repair$ Direct Direct Graft Direct Direct Graft Direct Direct Direct
* Sunderland partial (IV) or complete (V) transection. 7 Third molar extraction surgery. $ Positive control for repair.
cm from the midline). The internal dimensions of each chamber were identical (1,250 ,uL capacity: internal diameter = 14 mm, area = 1.54 cm2, perimeter = 44 mm) so that each tested area was spatially matched for each subject and for each session. Solutions were delivered to the chamber by an electric pump (Masterflex 7550-20, Cole Palmer Inst. Co, Chicago 11) at a flow rate of 1.5 mL/sec. Sour taste detection thresholds were measured with a temporal two-alternative forcedchoice procedure using the two-down one-up rule and an adaptation to the modified staircase.15 The citric acid stimuli were 16 concentrations of citric acid (J.T. Baker, #0122-01) that differed by 0.5 log unit serial dilutions (1 X 10e7 to 3 mol/L). Each citric acid stimulus was dissolved in room temperature distilled water. Room temperature distilled water served as the null stimulus and the rinsing solution. Citric acid and null stimuli were presented for 10 seconds; the intertrial interval was 30 seconds, during which the chamber was rinsed with distilled water before presentation of the next pair of stimuli. Scaling of citric acid taste intensity was accomplished with a modification of the magnitude matching procedure described by Stevens The magnitude responses to varying and MarksI brightnesses of a light source were used to normalize the magnitude response to taste stimuli. Delivery of visual stimuli was provided by a microcomputer-controlled tungsten lamp mounted inside a small windowed metallic box placed 30 cm in front and slightly to the left of the patient. Five levels of brightness were defined by approximately equal logarithmic increments of voltage over the range of 2.2 to 9.0 V. The citric acid stimuli were five concentrations differing by 0.5 log units of molar concentrations (0.003, 0.01, 0.03, 0.1,0.3 mom), presented in random order and in triplicate. After completion of each spatially matched citric
acid detection and scaling procedure, 5 mL 0.4% methylene blue was passed through the test chamber for 10 seconds. In humans, fungiform papillae are unstained by methylene blue, and are up to 1 cm in diameter, whereas pores are stained by methylene blue and are 10 to 50 pm in diameter.17 After staining, the chamber was removed, and images of the fungiform papillae and pores were recorded with a Zeiss OMPI dissecting microscope onto videotape (S-VHS-RT 120) with a solid-state color video camera (VK-C350, Hitachi, Lyndhurst, NJ). The patient and investigator viewed the images on a 20” color video monitor (CT-2010Y, Panasonic Comm. and Systems Co, Norcoss, GA). Videotapes of stained fungiform taste buds were reviewed using an S-VHS videoprinter (VP-1500, Codonics Inc, Middleburg Heights, OH). Fungiform papillae were identified and counted on images recorded at 40x, and pores were identified and counted from images recorded at 110X. After completing the citric acid spatial tests and videomicroscopy on both sides of the anterior tongue, the citric acid detection threshold and magnitude matching procedures were performed with a sip-and-spit method to examine taste sensitivity under whole-mouth conditions. SOMATOSENSORY
CAPACITY
AND IMPAIRMENT
SCALES
The level of somatosensory capacity on the anterior tongue was measured using a previously described clinical neurosensory test algorithm.‘* Three levels of testing were performed on the injured and uninjured sides of the anterior tongue at each session. Standard sensory procedures and psychophysical methodologies described by Essick” were used for testing. At level A testing, an evaluation of direction sensitivity using a No. 0 camel hair brush and two-point discrimination using a Boley’s gauge was performed. At level B test-
ZUNIGA,
CHEN,
AND
PHILLIPS
ing, an evaluation of contact detection using Semmes Weinstein pressure anesthesiometers (Stoeling Co, Houston, TX) was performed. At level C testing, an evaluation of thermal pain sensitivity using a computer-controlled thermode, built at University of North Carolina from a prototype developed at the Neurobiology and Anesthesiology Branch, National Institute of Dental Research, was performed. The objective of this evaluation was to grade the level of impairment in sensory capacity as normal, mild, moderate, severe, or complete. An abnormal response indicated that the test site value was less than that of the control site or published “upper normative limits,” and a normal response indicated that the test and control sites exhibited comparable values within published normative limits. An abnormal response at levels A and B and no response at level C was graded as complete sensory impairment (viz, anesthetic). An abnormal response at levels A and B and an elevated response at level C was graded a severe sensory impairment. An abnormal response at levels A and B and normal response at level C was graded a moderate sensory impairment. An abnormal response at level A and a normal response at level B was graded a mild sensory impairment. Finally, a normal response at level A was graded as normal (note that “normal” refers to the patient’s sensory capacity and not his or her report of altered sensation) . PATIENT-REPORTED
EXPECTATIONS
AND
OUTCOMES
Patient-reported outcome variables were assessedat each sessionusing instrumented surveys described by Ostler and Kiyak” and used previously by uszl to measure expectations of surgical outcomes versus actual surgical experiences and problems with oral function. Figure 1 illustrates the 1l-item expectancies questionnaire distributed to the patient before surgery by a member of the surgical team, who explained the form. The 11-item survey measured two dimensions: oral function dimensions, including eating, speech, taste, feeling, pain, and chewing; and psychosocial dimensions, including appearance, health, self-esteem, work performance, and a category called “other.” These items were accompanied by a 7-point response scale indicating the expected changes in each item in the list. A parallel measure was distributed to the patients at each postrepair examination to assessthe perception of change on a scale of “much worse” to “much better.” The postrepair outcome scale was delivered personally to the patient before any examination, completed in private by the patient, and returned to a member of the team who filed the results in private. The postrepair outcome scale used the same 11-item 7point response scale so that patients could indicate how each item had been changed when compared with presurgical conditions. Finally, a global satisfaction
Expectancies Whenever we undergo any medical or dental treatment we usually expect the treatment to result in some change in our physical and psychological health. We may expect these changes to be positive or negative. You are probably anticipating some change after your nerve surgery. For each of the areas listed below, circle the number which best describes your expectations, f?om -3 @iI1 become much worse) to 0 (no change), to +3 (will become much better). Will be much worse Eating Speech Tasting Feeling Pain Chewing General appearance General health Feeling about self Performing at work Other (Specify _)
-3 -3 -3 -3 -3 -3 -3 -3 -3 -3 -3
Will be much better
No change -2 -2 -2 -2 -2 -2 -2 -2 -2 -2 -2
-1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1
0 0 0 0 0 0 0 0 0 0 0
+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1
+2 +2 +2 +2 +2 +2 Jr2 +2 +2 +2 +2
+3 +3 +3 +3 +3 +3 +3 +3 +3 +3 +3
FIGURE 1. Expectancy questionaire used to survey patient’s motivation for surgery and perceived deficit(s) of six functional and five psychosocial dimensions. A postsurgical scale questionaire was collected at each postrepair examination.
questionnaire was delivered to the patient at each postrepair examination that required the patient to assign a numerical weight ranging from 0 to 4 (nil = 0, poor = 1, fair = 2, good = 3, excellent = 4) to indicate the overall satisfaction with the postrepair results. This survey was also self-administered in private and before examination. DATA
ANALYSES
The distribution of taste receptor anatomic variables (eg, fungifonn papilla density and ratio of pores, etc) was described by means, standard deviation (SD) and range values. The distribution of somatosensory tests results, the level of impairment, was numerically weighted (eg, normal = 1, mild = 2, moderate = 3, severe = 4, complete = 5) and described by means, SD, and range values. The distribution of scaled patient-reported expectancy and outcome results was numerically weighted (range, -3 to +3) and described by means, SD, and range values. A Wilcoxon signed rank test was used to determine the significant differences in anatomic variables between the injured and uninjured sides of the anterior tongue and to compare the presurgery expectancy score and sensory impairment level with the average postsurgery perception score and sensory level. For each anatomic measure, repeated measures analysis of variance with two within-subject factors (time and side) was used to assessthe pattern of recovery. All of the taste magnitude and detection threshold values were converted to logarithmic scales. At each point, for both the spatially matched and whole-mouth
6
REGENERATION
FIGURE 2. Clinical photographs of the anterior tongue in patients with unilateral lingual nerve transection injuries demonstrating the reduction in the number of grossly identifiable fungiform papillae on the injured side when compared with the uninjured side. The patient in A suffered from a Class V injury to the left lingual nerve. The patient in B suffered from a Class IV injury to the right lingual nerve.
procedures, least squares regression was used to describe the patient’s function between the log molar concentration and the geometric mean of the adjusted log response (each estimate of taste intensity was divided by the geometric mean of the patient’s estimate of brightness) for each side at each test session. The regression coefficient, calculated separately for time and side, was used as the outcome measures for evaluation of taste sensitivity. Repeated measures analysis of variance with two within-subject factors (time and side) was used to assess the pattern of recovery. Level of significance was set at .05. Results EFFECTS
OF INJURY
The findings indicated that Class IV and V injuries of the lingual nerve result in reduced numbers of fungiform papillae and pores on the anterior tongue in humans. Figure 2 shows clinical photographs of the ante-
OF LINGUAL
NERVE
IN HUMANS
rior tongue in two patients before repair, which demonstrate unilateral reductions in fungiform papillae without the aid of vital stains and videomicroscopy. The average number and standard deviation of stained fungiform papillae per chamber on the uninjured side of the anterior tongue was 16.6 t 8.6 (range, 7 to 37) and 10.9 t 6.8 (range, 4 to 23) on the injured side. The difference between the sides of the tongue was significant (P = .012). The average number and standard deviation of fungiform pores per chamber on the uninjured side of the anterior tongue was 50 2 29.6 (range, 21 to 117) and 9.8 + 8.4 (range, 0 to 25) on the injured side. The difference between sides of the tongue was statistically significant (P = ,005). The mean ratio of fungiform pores per papilla on the uninjured side of the anterior tongue was 3 + 0.8 (range, 2 to 4.8) and 0.8 +- 0.4 (range, 0 to 1.3) on the injured side. The difference between sides of the tongue was statistically significant (P = .OOS). Residual fungiform papillae were found on the injured side of the anterior tongue in every case. Figure 3 illustrates residual fungiform papillae in select patients. Residual fungiform papillae were generally located along the midline and anterior tongue tip and were frequently malformed; they were smaller, flatter, paler, polymorphic, and occasionally were filiform-like (ie, spinous processes extending from their centers). Normal-appearing residual papillae were found, but they were generally devoid of identifiable pores. The mean sour detection threshold on the uninjured side of the anterior tongue before repair was 0.97 f 1.2 mol/L (range, 0.008 to 3), which was not different from previously published normative values.15 The mean threshold on the injured side of the anterior tongue was >3 t 0 mol/L. These findings demonstrated that none of the patients could detect the difference between a stimulus of 3 mol/L citric acid and water on the injured side. Citric acid solutions more concentrated than 3 mol/L were not used to avoid burning the tongue mucosa. The mean whole-mouth threshold was 0.0017 -+ 0.002 mol/L (range, 0.00006 to O.Ol), which was not different from previously published normative values.22 Figure 4 illustrates the intensity rating power function curves of increasing suprathreshold concentrations of citric acid under the three conditions. The averaged patient’s responses and citric acid concentrations are plotted on logarithmic axes. The data for responses under the three conditions were extended for visual comparison. The averaged regression coefficient values for whole-mouth conditions was 2 2 0 for y-intercept and 0.4 + 0.2 for slope, which was not different from previously published normative values7 The mean intercept was 1.2 f 0.5 and mean slope was 0.32 I 0.24 on the uninjured side of the anterior tongue before repair, which was not different from previously published normative values. The mean intercept was 0.18 f 034. and the mean slope was
ZUNIGA,
CHEN,
AND
7
PHILLIPS
Table 2. Patient-Perceived Microsurgical Treatment
FIGURE! 3. Videophotographs of the methylene blue-stained anterior tongue in select patients. A, Many identifiable fungiform papillae are seen on the uninjured surface side of the anterior tongue after removal of a flow chamber. B. For comparison, the injured surface side of an anterior tongue contains very few identifiable fungiform papillae. C, Normal fungiform papillae with pores (arrows) on the uninjured side of the anterior tongue are demonstrated at a higher power. D, Residual fungiform papillae on the surface of the injured side of the anterior tongue are shown at a higher power to be smaller, paler, dysmorphic, and devoid of pores when compared with normal fungiform papillae in C. E, Two residual fungiform papillae on the surface of the injured side of the anterior tongue at higher power appear to be filiform-like (spinous).
Dimension
Mean
Speech Feeling Taste Eating Chewing Pain Appearance Health Self Work Other
1.2 2.0 2.0 2.1 2.2 1.9 0.9 0.9 0.4 0.2 0
Motives
for
(SD)
Range
(1.2)
o-+3 +1-+3 -k-t3 +1-f3 o-+3 0-+3 o--t3 o-+3 o-+3 o-+1 0
co.71
(0.5)
(0.6) (0.9)
(1.0)
(2.0) (2.0) (0.9) (0.4) 0
0.01 ? 0.05 on the injured side of the anterior tongue, which was significantly different from the uninjured side before repair (P = .0006 and .0002, respectively). Patients could not detect or scale the increasing concentrations of citric acid solutions on the injured side of the anterior tongue before repair. Before repair, the sensory impairment rating on the uninjured side of the anterior tongue was graded as normal (mean = 1 _f 0) in every case, and there were no subjective reports of altered sensation. The mean level of sensory impairment on the injured side before repair was 4.75 I- 0.4, and the difference between sides was significant (P = .002); nine patients demonstrated complete and three patients demonstrated severe sensory impairment. A positive trigger was elicited from the ipsilateral retromolar trigone in every caseon clinical examination before repair. Table 2 illustrates the rank order of patient-perceived expectancies for surgery. As found in previous studies, oral function dimensions, including improved taste and feeling, were most important at the presurgical assessment,” and
FIGURE 4. Mean intensity ratings of citric acid plotted against molar concentration before repair. The standard error of the normalized logarithmic means are marked. The figure graphs the magnitude responses under whole mouth conditions (m); spatially matched regions of the uninjured side of the anterior tongue (+); and spatially matched regions of the injured side of the anterior tongue (*).
MOLAR CONCENTRATION
(Log)
8
REGENERATION
Table
3.
Fungiform
Taste
Receptor
Response
to Microsurgical
OF
* t z 5
l-Year
N
FP*
Fpi-
Fp/FP:
5 5 1
13.4 (8.1) 8.4 (6.2) 15
11.8 (9.1) 7.6 (10) 9
0.8 (0.3) 0.7 (0.5) 0.6
Mean and standard deviation of identifiable fungiform Mean and standard deviation of identifiable fungiform Mean and standard deviation of the ratio of fungiform Difference within the same group and among groups
papillae. pores. pores/papilla. was statistically
psychosocial dimensions were least important. Based on mean scoresper dimension, the most important oral function expectancy among presurgical patients was chewing (2.2 + 0.9) and the least was speech (1.2 k 1.2). EFFECTS
OF REPAIR
Of the 12 patient volunteers, 1 did not complete all the postsurgical testing and was eliminated from postrepair analyses, and 1 did not have a repair performed because of the extent of injury and served as a positive control for repair. Repeated measuresanalysis of variance for the remaining 10 patients indicated that the pattern of difference between the injured and uninjured sides of the tongue was not consistent over time for the number of fungiform taste pores (P = .05), and the spatially matched detection threshold (P = .017), and was marginally significant for the number of fungiform papillae (P = .07). The pattern was consistent for the whole mouth detection threshold (P = .55). Five of the 10 patients demonstrated some recovery of taste function on the repaired side by 1 year and were considered responders. The remaining five patients demonstrated no recovery of taste function on the repaired side by 1 year and were considered nonresponders. The mean number of identifiable fungiform papillae (FP), pores (Fp) and the ratio of pores per papilla (Fp/FP) on the repaired side of the anterior tongue before and 1 year after repair in the five responders, five nonresponders, and positive control patient are tabulated in Table 3. There were measurable differences between the pre- and l-year postrepair mean numbers of FP (p = 5.4), Fp (,u = 23.8), and Fp/PP (II = 1.24) in the responders. By comparison, the pre- and l-year postrepair differences in mean numbers of FP (p = 2.2), Fp (p = 2.2), and Fp/FP (p = 0.36) in the nonrespondersand positive control patient was less. The data point out that neither the nonrespondersnor the positive control patient demonstrated a significant change in fungiform receptor density over time; however, the responders demonstrated an in-
significant
NERVE
IN HUMANS
Repair
Presurgical
Responders Nonresponders Positive control
LINGUAL
FP
18.8 (9.5)s 10.6 (5.3) 22
Postsurglcal FP
38.8 (19)§ 9.8 (6.9) 19
Fp/FP
2.1 (0.1)s 0.9 (0.4) 0.8
(P < .OOOl)
crease in fungiform receptor density over time. Figure 5 illustrates fungiform papillae and pores stained before and 1 year after repair in a responder patient. The increase in the density of fungiform papillae and pores are demonstrated, as well asthe detectable differences in receptor morphology over time. The mean citric acid detection threshold decreasedover time on the repaired side in the responders (1.3 t 1.5 mol/L) by 1 year.
FIGURE 5. Videophotographs of the injured sides of the anterior tongue before and after repair in a patient who recovered somatosensory and chemosensory function. A, The injured side of the anterior tongue before repair at low magnification, showing the reduced number of identifiable fungiform papillae. B, For comparison, the repaired side of the anterior tongue 1 year after repair, showing me increase in identifiable fungiform papillae. C, The injured side of the anterior tongue before repair in A at a higher magnification. D, For comparison with C, the repaired side of the anterior tongue 1 year after repair. E, The injured side of the anterior tongue before repair at high magnification, showing residual fungiform papillae. F, For comparison with E, the repaired side 1 year later, showing regenerate fungiform papillae with identifiable pores (arrows).
ZUNIGA,
CHEN,
AND
PHILLIPS
wx.4R~ENlR4noN,~~
IOURWC~WITIoN(t4,
FIGURE 6. Mean intensity ratings of citric acid plotted against molar concentration after repair over time. In each figure, magnitude scaling under whole-mouth conditions are indicated by 0, scaling of the uninjured side of the anterior tongue are indicated by q , and scaling on the repaired side are indicated by n . The test sessions are indicated as A, presurgical, B, 1 month postrepair, C, 3 months postrepair, D, 6 months postrepair, and /Z, 12 months postrepair.
In three of the five responders, the mean threshold difference between the injured/repaired (0.15 t 0.04 mol/L) and the uninjured sides (0.16 5 0.2 mol/L) was nearly identical at 1 postrepair year. Neither the five nonresponders nor positive control patient could detect 3 mol/L of citric acid applied to the repaired side of the tongue after 1 year. The pattern of difference between the regression coefficients of the injured and uninjured sides was not consistent over time (P = .04). There was a statistically significant difference in uninjured and injured regression coefficients at all times but the 6-month interval (p = 0.07). At all other times, the mean difference was greater than 0.23 (P 5 .Ol). Figure 6 demonstrates that responders were able to detect and scale citric acid on the uninjured side of the anterior tongue and whole mouth both before and at each session after repair. Figure 6 also demonstrates that responders were unable to scale citric acid before repair, but demonstrated the progressive increase in the ability to detect and scale citric acid on the repaired side over time. Two of the five responders were first able to scale citric acid stimuli 3 months after repair; two after 6 months, and one at 1 year. By 1 year, three of the five responders were able to scale the whole range of stimuli concentrations, whereas two scaled the weaker concentrations of citric acid as less intense than the same citric acid concentrations applied to the uninjured side. Neither the nonresponders nor the positive control subject could scale citric acid on the repaired side over the same time. Nine of 10 patients (90%) demonstrated an increase in somatosensory capacity on the repaired side of the
anterior tongue by 1 year. The mean score for level of sensory impairment of the nine patients was 2.1 2 0.9 at the l-year postrepair visit. The difference between the presurgical score (4.75 2 0.45) and l-year postrepair score was statistically significant (P = .OOOl). The distribution of sensory impairment levels ranged from normal to moderate impairment; the mean level of sensory impairment was first significantly different from presurgical levels at 3 months. One of 10 patients (10%) was unable to detect either somatosensory or chemosensory stimuli on the anterior tongue at the lyear postrepair examination. The positive control subject was unable to detect either somatosensory or chemosensory stimuli on the ipsilateral anterior tongue over the same period. Figure 7 illustrates the 12-month postrepair mean outcome score versus the mean presurgical expectation score for each functional and psychosocial dimension queried. Analysis of patient’s expectations and actual experiences with microsurgery showed that the functional and psychosocial expectations were achieved through surgery; in all areas, patients reported minimal to moderate fulfillment. The ranking of postrepair outcome scores was very similar to presurgical expectation scores, for which functional scores were more important and achieved greater scores than psychosocial scores. Among the functional outcomes scores, the greatest improvement was noted for chewing (1.1 + 0.8) and the least for speech (0.2 f 0.7). Among the patients who demonstrated taste recovery, the category of taste was scored higher (1.0 + 1) than by nonresponders (0.2 I 0.6) or the positive control patient (0). Among patients who demonstrated
10
REGENERATION
MUCH BElTER
NO
3
OF
LINGUAL
IN HUMANS
5 1% ;
FIGURE 7. Mean ratings of presurgical expectations score (0) vs 12-month postrepair outcomes score q for each functional (speech, feeling, taste, eating, chewing, and pain) and psychosocial (appearance, health, self, work, and other) dimension queried. The standard deviations of the means are marked.
0
CHANGE
0
EXPECTATIONS 12 MONTH
MUCH WORSE
NERVE
SCORE
OUTCOME
SCORE
m3
sensory recovery, the category of feeling was scored higher (1.3 t 1.2) than for the positive control patient (-1) or the patient without sensory recovery (0). Figure 8 shows graphic plots of the global satisfaction scores of the surgical outcomes by the patients at the different postrepair examinations. The highest scores were recorded at the l-year postrepair examination and the least at the l-month postrepair examination, suggesting there was a relationship between the progressive improvement in surgical outcome and global satisfaction. Discussion The reduction in the number of identifiable fungiform papillae, pores, and the ratio of pores per papilla, and the absenceof citric acid detection and magnitude
EXCELLENT
GOOD
FIGURE 8. Mean rating of patient’s satisfaction score with surgery at each postrepair examination. The standard deviations of the means are marked.
4
1
3
FAIR
2
POOR
1
NIL
scaling on the injured side of the anterior tongue demonstrated that Class IV and V lingual nerve injuries in humans result in a sensorineuraltaste disorder characteristic of dystrophic ageusia.The progressive increase in the number of identifiable fungiform papillae, pores, and the ratio of pores per papilla along with increased response sensitivity to citric acid stimuli on the repaired side of the anterior tongue in 50% of cases demonstrated that taste may regenerate after repair of sectioned lingual nerves in humans. Our finding that fungiform taste buds atrophy after lingual nerve injury points out that epithelial staining to count the number of taste pores in papillae may serve as a useful quantitative test for the diagnosis of Class IV and V lingual nerve injuries. The basis for staining was that fungiform taste pores are thought to be a classic example of neurotrophically dependent
0
1
6
3 MONTHS
ZUNIGA,
CHEN,
AND
PHILLIPS
receptor cells. In animals, experimental sectioning of taste nerves resulted in taste pore degeneration; although some taste pores persisted.x~9~23*24In animals, residual taste pores were either dysmorphic or atrophic, being 20% to 50% smaller than normal taste pores.23,24 In our study, the difference in the density of fungiform pores and the ratio of pores per papilla between sides of the anterior tongue was significant; thus, like animals, determining the reduction or loss of normal fungiform taste pore density and ratio of pores per papilla in humans will provide positive identification of a damaged lingual nerve when there exists no other possible reason for atrophy (eg, previous chorda-tympani nerve injury secondary to middle ear surgery, etc). Unlike fungiform taste pores, the nerve dependence of fungiform papillae for normal anatomic maintenance remains controversial. In animals, fungiform papillae undergo varying degrees of degenerative changes after taste nerve section; however, anatomically normal papillae persist.8,9,25 In our study, the difference in the density of fungiform papillae between sides of the anterior tongue was significant but marginal; thus, identifying fungiform papillae by staining or by gross examination of the anterior tongue sides will not be a sensitive test for all lingual nerve transection injuries because Class IV and V injuries were found when there was little difference between sides presurgically (false negative). However, the loss of grossly identifiable fungiform papillae on the injured side of the anterior tongue (eg, ipsilateral atrophy) will provide positive identification of lingual nerve sectioning. Interestingly, like animal studies, our findings demonstrated that residual fungiform papillae were generally flat, small, pale, and dysmorphic in appearance,24 and occasionally associated with filiform-like spines.25 Presurgical staining and counting of fungiform taste receptors required the use of videomicroscopy because the normal range of taste pore and papilla size was outside the range for gross visualization and the normal fasciculation of the tongue epithelium by the musculature inhibited in situ counting accuracy. The method will generally preclude routine office examination; thus, live videomicroscopy of the tongue will require dedicated equipment and personnel if accurate and objective testing for dystrophic ageusia are demanded by the clinician. Additionally, we currently use the triphenylmethane dye, FD & C Blue No. 1, as the epithelial stain because methylene blue has been shown to evoke mutagenic changes in bacteria DNA when combined with visible light. The detection of significant differences in taste sensitivity on the anterior tongue (spatial testing) is recommended for the clinical assessment of lingual nerve injuries because the finding of ageusia was universal. Like previous studies, applying different taste stimuli to the anterior tongue using different stimulus delivery systems (eg, filter paper discs, cotton-tipped applica-
11 tors, solution-soaked paint brush) determined the reduction (hypogeusia) or loss (ageusia) of taste on the ipsilateral side after nerve section in humans.3Z435.7S13-‘4 In our study, none of the patients was able to detect or scale suprathreshold concentrations of citric acid applied to a closed spatially matched flow chamber attached to the anterior tongue; they all demonstrated sour ageusia. Although not tested, Class IV and V lingual nerve-injured patients would be expected to demonstrate nonspecific ageusia because previous studies reported loss of salt, sweet, and bitter sensitivity on the ipsilateral anterior tongue.4,5 In contrast, presurgical whole-mouth citric acid detection threshold and magnitude scaling were normal. Our findings are similar to those reported in human studies after experimental lingual nerve anesthesia26 and lingual nerve injury.5.‘4 Thus, humans may be unaware of discrete losses of taste even after the loss of extensive lingual nerve input, making tests of whole-mouth taste function uninformative. The demonstrated impairment of somatosensory capacities on the anterior tongue in patients with Class IV and V injuries of the lingual nerve are in agreement with most studies.‘,* Our results concur that the measured level of sensory impairment in Class IV and V injured patients ranked between severe and complete (anesthesia) and were generally associated with a radiating trigger from the ipsilateral retromolar trigone to the anterior tongue. Patients consistently scored oral functional dimensions as more important than psychosocial dimensions (viz, expectancy ratings) before microsurgical intervention. Chewing was considered a top priority presurgically, followed closely by eating, feeling, taste, and pain; the lowest priority rating was given to speech. These findings are similar to previously reported samples of mixed (eg, inferior alveolar and lingual nerve) nerve-injured patient surveys2i and demonstrate that patients with Class IV and V lingual nerve injuries are aware of deficits in oral function, including taste. The serial examination of 10 postrepair patients and the positive control patient (no repair) for 1 year demonstrated that a repaired lingual nerve transection in humans can result in the recovery of taste, the regeneration of fungiform taste receptors, and the recovery of some sensory function. The recovery of taste sensitivity on the anterior tongue was coincident with the regeneration of identifiable fungiform taste buds. In animal studies, reinnervation of the epithelium by repaired sectioned lingual nerves resulted in the reappearance of fungiform taste pores,‘.” an increase in the number and size of fungiform papillae,8,23 and recorded whole nerve response to tastants applied to the anterior tongue.83’0 Our findings indicated that the characteristics of taste nerve regeneration described in animals are very similar in humans. In animals, the regeneration of taste buds and the recovery of taste response occurred
12 only by homologous reinnervation of axons from taste nerves (eg, chorda-tympani, glossopharyngeal) and not by nontaste nerves (eg, trigeminal, hypoglossal).27 Our findings suggest that the regeneration of taste buds and the recovery of taste sensitivity was taste nerve dependent, because the positive control subject did not demonstrate any change in taste bud anatomy or response over the same period, and four patients recovered sensory function without changes in taste bud anatomy or response. The effect of age, gender, location, classification, duration of injury, or repair on taste bud anatomy or response was not assessed because the sample size was too small to draw definitive conclusions. Hillerup et all3 reported that 3.7 (1.1 to 4.6) years after repair, five of six patients (80%) detected sweet, sour, salty, and bitter stimulants applied to the repaired side of the anterior tongue. Wiethijlter et a128reported that only 3 of 24 patients (12%) and Hesseling et a129 reported that only 2 of 17 (11%) patients demonstrated the recovery of some taste sensitivity after lingual nerve repair. Five of 10 patients (50%) regenerated taste buds and recovered some taste function in our study. The disparate reports of the percentage of patients demonstrating taste recovery after lingual nerve repair may be attributable, in part, to fascicular distributions noted by several authors.12x3’ In cats, Holland and Robinson3’ demonstrated that at the retromolar trigone the lingual nerve axon counts were 2.7 times more than chorda-tympani nerve axons and the distribution of small fibers was toward the chorda-tympani nerve. In human cadaver studies, Girod et al” demonstrated that the chorda-tympani was monofascicular proximal to the retromolar trigone, but then distributed widely within the lingual nerve proper. This study agrees with the proposal espoused by Girod et all2 that the recovery of taste function after lingual nerve repair may be attributable, in part, to anatomic alignment of taste fibers achieved by microsurgery. Interestingly, 90% of patients recovered sensory function, and only 50% recovered some taste function, which points out that the mathematical probability of sensory fiber regeneration may be greater than taste fiber regeneration, in part because of the geometry of fibers aligned by microsurgery. Further studies will be conducted to test this hypothesis. The finding that somatosensory reinnervation of the anterior tongue in humans was substantially better than presurgical conditions, but incomplete, was consistent with previous reports.2,3’ In 90% of cases, the level of sensory impairment decreased from severe/complete sensory impairment (mean score of 4.75) to mild sensory impairment (mean score of 2.1). The recovery of somatosensory sensitivity was nerve dependent because the positive control patient did not demonstrate any change in somatosensory sensitivity over the same period.
REGENERATION
OF
LINGUAL
NERVE
IN HUMANS
When averaged across time, the perception of change by surgery was perceived as an improvement, but, in all areas, patients reported less of an improvement than expected. This was especially true for functional scores, the areas of greatest concern for the patients presurgically. It was possible that these findings point out that presurgical expectations tend to be higher than what the patient actually experiences. However, patients generally rated the mean global satisfaction with postrepair outcome as progressive from “poor to nil” 1 month postrepair to “fair to good” at 1 year, indicating that they were astute enough to recognize the progressive recovery of somatosensory and chemosensory sensitivity over time. Patients who demonstrated recovery of taste sensitivity were more likely to report subjective improvement than those who demonstrated no recovery of taste sensitivity. In conclusion, we demonstrated that Class IV and V injuries of the lingual nerve in humans result in a sensorineural taste disorder characteristic of dystrophic ageusia of the anterior tongue, severe or complete sensory impairment of the anterior tongue, and patientperceived deficits in oral functional dimensions. The detection of differences in taste sensitivity on the anterior tongue and the staining and counting of fungiform taste receptors are recommended to diagnose dystrophic ageusia and positively identify Class IV and V lingual nerve injuries in patients presurgically. Dedicated equipment and trained personnel are required to perform spatial taste tests and stain lingual epithelium for counting taste buds. This study documented the regeneration of fungiform taste receptors and the recovery of taste in humans after lingual nerve repair. One-year postrepair tests demonstrated that 90% of patients recovered some somatosensory sensitivity, and 50% of patients recovered some or all of their chemosensory sensitivity. The results confirm that patients with Class IV and V lingual nerve injuries benefit by surgical intervention. Acknowledgment The authors and Mohamed
thank Robert Englehardt, DDS, Teresa Tucker, Malik, ES, for their technical support.
RN,
References 1. Mozary PG, Middleton RA: Microsurgical reconstruction of the lingual nerve. J Oral Maxillofac Surg 42:415, 1984 2. Zuniga JR: Multimodal scaling of sensory recovery after microsurgical repair. J Oral Maxillofac Surg 48:85, 1990 3. Bull TR: Taste and the chorda tympani. J Laryngol Otol79:479, 1965 4. Grant R, Miller S, Simpson D, et al: The effect of chorda tympani section on ipsilateral and contralateral salivary secretion and taste in man. J Neurol Neurosurg Psychiatry 52: 1058, 1989 5. Abrahams H, Travers S, Travers J, et al: Chemosensory adaptation following chorda-lingual nerve injury. J Oral Maxillofac Surg 51:180, 1993
SBREN
13
HILLERUP
6. Ogden GR: Atrophy of fungiform papillae following lingual nerve damage: A poor prognosis? Br Dent J 167:332, 1985 7. Zuniga JR, Chen N, Miller IJ: Effects of chorda-lingual nerve injury and repair on human taste. Chem Senses 19:657, 1994 8. Cheal M, Oakley B: Regeneration of fungiform taste buds: Temporal and spatial characteristics. J Comp Neurol 172:609, 1977 9. Robinson PP, Winkles PA: The number and distribution of fungiform papillae and taste buds after lingual nerve injuries in cats. Arch Oral Biol 36:885, 1991 10. Robinson PP: The reinnervation of the tongue and salivary glands after lingual nerve injuries in cats. Brain Res 483:259, 1989 11. Hausaman JE: Zur indikation der mikronervenchirurgie, in Schwenzer N, Pfiefer G (eds): Fortschritte der kiefer-und gesichtschirurgie band XXVIII. Experimentelle mund-kiefergesichtschirurgie, mikrochirurgische eingriffe. New York, NY, Thieme, 1983, pp 163-167 12. Girod SC, Neukam FW, Girod B, et al: The fascicular structure of the lingual nerve and the chorda tympani: An anatomical study. J Oral Maxillofac Surg 47:607, 1989 13. Hillerup S, Hjorting-Hansen E, Reumert T: Repair of the lingual nerve after iatrogenic injury: A follow-up study of return of sensation and taste. J Oral Maxillofac Surg 52: 1028, 1994 14. Baker SB, Foote JW, DiNick V: Gustatory recovery in microsurgical repair of the lingual nerve. J Oral Maxillofac Surg 53:137, 1995 15. Zuniga JR, Davis SD, Englehardt RA, et al: Taste performance on the anterior human tongue varies with fungiform taste bud density. Chem Senses 18:449, 1993 16. Stevens JC, Marks LE: Cross-modality matching functions generated by magnitude estimation. Percept Psychophys 27:379, 1980 17. Miller IJ, Reedy FE: Quantification of fungiform papillae and taste pores in living human subjects. Chem Senses 15:281, 1990 18. Zuniga JR, Essick GK: A contemporary approach to the clinical evaluation of trigeminal nerve injuries. Oral Maxillofac Surg Clin North Am 4:353, 1992 19. Essick GK: Comprehensive clinical evaluation of perioral sensory function. Oral Maxillofac Surg Clin North Am 4:503, 1992
J Oral Maxillofac 55:13-14, 1997
20. Ostler S, Kiyak HA: Treatment expectations versus outcomes among orthognathic surgery patients. Int J Adult Orthod Orthognath Surg 6:247, 1991 21. Zuniga JR: Perceived expectation, outcome, and satisfaction of microsurgical nerve repair. J Oral Maxillofac Surg 49:77, 1991 22. Bartoshuk JM, Rifkin B, Marks LE, et al: Taste and aging. J Gerontol 41:51, 1986 23. Hard af Segerstad C, Hellekant G, Farbman AI: Changes in number and morphology of fungiform taste buds in rat after transection of the chorda tympani or chorda lingual nerve. Chem Senses 14:335, 1989 24. Oliver SD, Whitehead MC: Morphometry and cellular dynamics of denervated fungiform taste buds in the hamster. Chem Senses 17:529, 1992 25. Oakley B: Neuronal-epithelial interactions in mammalian gustatory epithelium, in Bock GR, Whelan J (eds): Regeneration of Vertebrate Sensory Receptor Cells. Chichester, England, Wiley and Sons, 1991, pp 277-293 26. Catalanotto FA, Bartoshuk LM, Ostrum KM, et al: Effects of anesthesia of the facial nerve on taste. Chem Senses 18:461, 1993 27. State FA, Hamed MS, El-Hashash MK, et al: Trophic specificity of the gustatory fibers upon taste bud regeneration. Acta Anat 113:196, 1982 28. Wietholter H, Riediger D, Ehrenfeld M, et al: Ergebnisse der mikrochirurgie sensibler peripherer iiste des nervus mandibularis, in Schwenzer N, Pfiefer G (eds): Fortschritte der Kieferund Gesichtschirurgie. Band XXXV. Mikrochirurgie in der Mund-, Kieferund Gesichtschirurgie. New York, NY, Thieme, 1990, pp 128-134 29. Hesseling KH, Reich RH, Hausaman JE, et al: Langzeitergebnisse nach mikrochirurgischer nervenrekonstruktion im KopfHals-Bereich, in Schwenzer N, Pfiefer G (eds): Fortschritte der Kiefer-und Gesichtschirurgie. Band XXXV. Mikrochirurgie in der Mund-Kieferund Gesichtschimrgie. New York, NY, Thieme, 1990, pp 134-138 30. Holland GR, Robinson PP: Axon populations in cat lingual and chorda tympani nerves. J Dent Res 71:1468, 1992 31. LaBanc JP, Gregg JM: Trigeminal nerve injuries: Basic problems, historical perspectives, early successes and remaining challenges. Oral Maxillofac Surg Clin North Am 4:277, 1992
Surg
Discussion Chemosensory and Somatosensory Regeneration After Lingual Nerve Repair in Humans S&en Hillerup,
DOS, PhD, Dr Odont
Copenhagen
University
This
County
study
on chemosensory,
Hospital
Gloxtrup, somatosensory,
Glostmp,
Denmark
and epithe-
ha1 regeneration after lingual nerve repair deals with an important issue and contributes original results based on sophisticated test methods. It also demonstrates some of the pitfalls associated with this type of clinical research. One of these relates to the authors’ desire to disseminate as much experience as possible thus yielding an overflow of information. The purpose of study was to measure the effect of injury and subsequent repair in patients with partial and complete transection of the lingual nerve based on a number of variables. However, the quality of the sample is compromised
by being composed of cases of both partial and complete lingual nerve transection. This leaves two important questions open, of which only the first one can be answered: 1. What was the basis of diagnosis of partial and complete transection? Partial transection is a broad classification, theoretically including a break in the continuity of only one fascicle to a break in all fascicles but one. 2. What was the potential for restitution by spontaneous healing of the seven cases with only partial transection had they not been repaired? Moreover, tional
sural
three cases were managed with an interposinerve
end anastomosis.
graft,
whereas
seven
This inconsistency
taken into consideration, or in the discussion.
either
cases
had an end-to-
is not referred
in the presentation
to or
of results,
The method of testing sour taste sensitivity using a closed