- Email: [email protected]
Trigeminal ganglion cell response to mental nerve transection and repair in the rat
] Oral
Moxillofoc
57:427-437,
Surg 1999
Trigeminal Ganglion Cell Response to Mental Nerve Transection and Repair in the Rat John R. Zuniga,
DMD, PhD*
Purpose: Animal studies have suggestedthat peripheral nerve transection results in substantial loss of ganglion cells and the selective survival of cells based on size. The implications are that subsequent repair of peripheral nerve injuries will be determined by the numerical density and character of the surviving cells. The purpose of this study was twofold: First, to determine the effect of mental nerve transection without repair on trigeminal ganglion cell density and morphology in adult rats, and second, to determine the variation of trigeminal ganglion cell density and morphology after immediate and delayed repair. and Methods: In the first part of the study, 12 adult male Sprague-Dawleyrats had their mental nerves exposed bilaterally (n = 24). Twelve mental nerveswere then transected and prevented from regenerating, and the remaining 12 nerves were uninjured. Ninety and 180 days after transection or shamsurgery, the trigeminal gangliawere serially cut into 5 pm longitudinal sectionsalong the dorsoventral axis. The volume and volume density of the mandibuk mental subdivision containing sensorycells was determined at each section level with pointcounting methods. The numerical density and total number of cellswas estimatedon the same section, using an unbiasedthree-dimensionalstereologicalprobe, the disector. Cell sizeand shapedeterminants were estimatedusingthe disector and computerized planimetry. In the secondpart of the study, six rats had the mental nerves transectedbilaterally and immediately repaired by microscopic sutures.In six additional rats, the repair was delayed for 90 days. In both groups, the trigeminal gangliawere serially cut at 30, GO,and 90 days post-repair and stereologic estimatesof numerical density and histomorphometry were examined usingthe disector and computed planimetry. Materials
In the trigeminal ganglia of the 12 sham-operated animals, the mean number of cells was 20.6 X lo3 (-t 2.9 X 103). After nerve section, the mean number of cells was 10.88 X lo3 (2 0.9 X 103) representing a 47% reduction. The mean volume of the mandibular subdivision cells in the ganglia of the sham surgery animals was 0.3 mm3 (20.05) and 0.22 mm3 (tO.04) in nerve-sectioned ganglia, a 38% difference. There were no ganglia cell size or shape differences between the two groups. The mean number of cells in the ganglia of immediately repaired nerves was 10.66 X 103(t 1.1 X 103) and it was 12.45 X 103 (2 0.9 X 103) after delayed repair. The numerical density was significantly lessthan in the sham surgery ganglia but not different from that of the transection/no repair ganglia. The weighted mean reference volume of the mandibular subdivision after immediate and delayed repair was similar and was significantly greater than the transection/no repair group, but not different from the sham surgery group. The cell size was slightly larger in delayed-repair ganglia compared with immediate-repair ganglia, but the differences were not significant. There were no significant differences in any of the stereologic estimates when analyzed acrosstreatment time. Results:
The results of this study agree with previous reports that peripheral nerve transection produces a substantial loss of nerve cells within specified regions of sensory ganglions. However, the results conflict with evidence that cells survive transection based on size and shape. These findings also indicate that in the adult rat the substantial loss of nerve cells was unaltered by the reconnection of their severed axons. Neither immediate or *Professor, Department of Oral and Maxillofacial Surgery, Univerdelayed repair of the transected nerve altered sity of North Carolina at Chapel Hill, School of Dentistry, Chapel the spectrum of surviving cells based on size or Hill, NC. shape. The reestablishment of the mean referSupported in part by NIBNIDR grant DE08272. ence volume of the mandibular subdivision after Address correspondence and reprint requests to Dr Zuniga: Professor, section and repair suggeststhat demands made Oral and Maxillofacial Surgery, UNC, School of Dentistry, CFW7450, on regenerating axons appear to result in the Chapel Hill, NC 27599.7450; e-mail: J_ZUNIGA@Dentisxry,UNC.edu restoration of ganglionic volume normally lost 6 1999 American Association of Oral and Maxillofacial Surgeons after axotomy, probably the result of axonal 0278.2391/99/5704-0012$3.00/0 branching or supporting cell proliferation. Conclusions:
427
428 The application of microsurgical techniques to repair inferior alveolar and lingual nerve injuries and recover sensory and special sensory function in humans has been demonstrated by many authors.‘s2 These reports suggest that recovery of sensory function perceived by patients or demonstrated by objective tests are rarely complete, but the results are always better than those in patients with nerve injuries without repair. Without question, the reports propose that early surgical repair produces a better outcome. Because microsurgery of the inferior alveolar and lingual sensory nerves is actually surgery on their cells, examining the effects of peripheral nerve injury with or without repair on ganglion cells should be performed to help explain the reported clinical results. It has been known for many years that transection of a peripheral nerve provides a signal that elicits a cell body response that induces changes in protein synthesis, alteration in cell organelles, and cell volume changes that lead to retrograde cell death in adult animals. Previous estimates have suggested that up to 30% or 40% of adult rat sensory ganglion cells die by 160 days after nerve section,3-5 with reported ranged from 15%6 to 55.6%.’ When the size distribution of surviving sensory neurons were examined, many studies reported a preferential and substantial loss of larger cells and a shift in the ganglion size spectrum toward smaller sizes.5,7’9 Because it is generally believed that how an animal senses, perceives, and behaves depends in part on the structured organization and on the number of elements that make up its nervous system, understanding any process that may ultimately affect the number and selectivity of nerve cells is of great importance. For example, Jacquin et allo showed that the decrease in the number of recorded cells in adult trigeminal ganglion after infraorbital nerve section at birth was associated with an absence or a decrease in the number of responding cells when the cutaneous fields normally supplied by the nerve were tested. Furthermore, they showed a shift toward greater numbers of cells that responded to noxious stimuli. If evidence indicates that smaller sensory cells are associated with relaying pain information to more central sites and preferentially survive trigeminal nerve injury, then important neuroanatomic correlates can be made to explain the occurrence of painful dysesthesias (eg, allodynia, hyperpathia, anesthesia dolorosa, sympathetic mediated pain). Most clinicians now agree that microsurgical repair of transected nerves is required in adult patients. This is believed to be true because, although nerve cells do not divide in adults, the remaining cells preserve the potential to repair themselves. In animals, this potential is realized when regeneration is not impeded or the severed ends are repaired using microsurgical
CELL
RESPONSE
TO
NERVE
SECTION
AND
REPAIR
sutures. In adult rats, a direct correlation between cell number, size, volume, and the size of the target after regeneration has been suggested.9,” The regeneration of severed axons in animals may be robust and nearly complete; however, contrary evidence has shown that degeneration of nerve cells was not influenced by the restoration of peripheral contacts and that the same magnitude of cell loss and volume change was independent of the extent of regeneration.5,12 In humans, when stimulation within the skin areas supplied by repaired peripheral sensory limb nerves were examined 5 to 20 years after repair, almost universally, the restitution of sensory function remained deficient.13,14 Similarly, although improved, recovery of tactile, pain, thermal, and taste sensibility to facial and oral sensory fields was never complete in patients who underwent suture repair of inferior alveolar and lingual nerves 1 to 2.5 years previously.1,2 When peripheral microneurography was used to record single units from reinnervated skin of the human hand, Mackel et all5 and Ochs et all6 independently reported reduced receptive field size and sensitivity to mechanical, thermal, and noxious stimuli and suggested that reduced sensitivity was in part the result of reduced density of reinnervating fibers. Whether volumetric or histomorphometric changes in the sensory cell bodies parallel reduced behavioral and physiologic responses after nerve repair remains unknown. The primary purpose of this study was to measure the effect of peripheral nerve transection and repair on trigeminal nerve cells in the adult animal. The mental nerve in the rat was chosen based on previous studies and information of the somatotopic organization of the mental nerve in the trigeminal ganglion of the rat. The study was divided into two parts. The aim of the first part was to describe the diversity of morphometric alterations that modify trigeminal sensory ganglion cell number, size, and shape in response to sectioning of the mental nerve. The aim of the second part was to examine the effects of microsuture repair of the mental nerve on cell histomorphometry and to compare the results with the findings in the first part. Additionally, to simulate clinical conditions in which repair is often delayed and accompanied by axon damming and scar formation before repair, results after immediate repair were compared with delayed (3 month) repair.
Materials
and
Methods
Twenty-four adult male Sprague-Dawley rats weighing between 200 and 250 g were used in this study (Charles River Laboratories, Burlington, NC). All animals were maintained on a 12 : 12 hour light/dark cycle in a macro-environment-controlled rodent room and
JOHN
429
R. ZUNIGA
received water and rat chow ad libitum before and after surgery. For surgery, each rat was anesthetized with an intraperitoneal injection of sodium pentobarbital (Nernbutal, Abbott, North Chicago, IL), 40 mg/kg. The mental nerves were exposed bilaterally in each rat by a submandibular transcutaneous approach, using an aseptic technique that has been previously described” (Fig 1). Each surgery was performed using a Zeiss (Carl Zeiss Inc, Thornwood, NY) surgical microscope. SURGERY AND DENERVATION In the first part of the study, the mental nerves of six animals were transected with surgical scissors 3 mm distal to the mental foramen, and a 5-mm segment of nerve was removed (transection/no repair group). In six additional animals the mental nerves were exposed and the wound immediately closed (sham surgery group). Ninety days later, three animals from the sham surgery and three from the transection/no repair group were deeply anesthetized with sodium pentobarbital, 100 mg/kg administered intraperitoneally. In each animal, transcardiac perfusion was performed with ice-cold 0.9% saline and 2% paraformaldehyde/1.25% glutaraldehyde in 0.01 mol/L phosphatebuffered saline (pH 7.6). The remaining sham and transection/no repair animals were killed in similar fashion 180 days from the initial surgery. The trigeminal ganglion was carefully removed on both sides along with the mandibular nerve trunk, and gross measurements of the width and thickness were recorded. The ganglia were dehydrated in ascending concentrations of alcohol, passed through xylene into paraffin, and serially sectioned at 5 pm along the longitudinal axis. In serial order, from dorsal to ventral surfaces, paraffin sections were platted on subbed slides, dewaxed, and stained in a 1% cresyl violet solution. The width of the broadest section and thickness of each ganglion were recorded.
FIGURE 1. Schematic diagram of the mental nerve-containing region (shaded) of the mandibular (Md) subdivision in the triqeminal ganglion in relation to the maxtllarv /) (Maxi and oohthalmic (OohlI branches, and the trigeminal root (Vrt). Also shown is the sensory branch of the mandibular nerve as it exits foramen ovale (FO), enters the mandibular foramen (Md?), and exits the mental foramen (MF). The surgical site (SS) was 3 mm distal to its exit from the MF.
SURGICAL REPAIR In the second part of the study, the surgical preparation of the rat mental nerves was performed in the same manner as previously described. In six rats, the mental nerves were reanastomosed with two 10-O Ethilon monofilament sutures (Johnson and Johnson, NJ) placed epineurially (immediate repair group). In the remaining six rats, the mental nerves were transected, and no attempt was made to reanastomose the cut ends. Three months later, the cut ends were relocated and reanastomosed (delayed repair group). Two animals from each group were killed at 3@, 60-, and 90&y intervals after their respective nerve repair dates. SAMPLING SCHEME A systematic, uniform, random sampling procedure described by Gundersen and Jensen18 was used to sample an equal fraction of each ganglion. To ensure that all the regions of the ganglion had an equal chance of being examined, the position of the first section was taken between 200 and 300 nm from the dorsal surface, where the first section (ie, section 0) was tangent to the dorsal surface and parallel to the long axis of the ganglion with the mandibular trunk positioned laterally (Fig 2). Two adjacent sections at 10 equally spaced (ie, 2 every 75 pm) intervals along the dorsoventral axis were selected and analyzed. The first sections were the 5th and 6th section from section 0, and the last were the 140th and 141st sections. Cell bodies located just posterior and lateral throughout the dorsoventral axis of the entering mandibular nerve trunk were examined in each section using photomicrographs taken through an Olympus Microscope. Previous studies have shown that mental sensory cell bodies are somatotopically organized to the entering mandibular nerve and constitute the majority of cells in the test areas examined.17,‘9 Nerve cell bodies were defined as cell profiles with clearly defined cytoplasm and a well-demarcated nucleus with diffuse or patterned chromatin distribution.
in V ganglion
430
CELL
RESPONSE
TO
NERVE
SECTION
AND
REPAIR
A 0
,’ ‘;
FIGURE 2. A series of 10 photomicrographs of 1% cresyl violetstained 5urn-thick sections of the trigeminal ganglion. The series beqins with the most dorsal section (section a) and then every section 7.5 pm ventral to the previous one [sections b through j]. The large arrows point out the posterior (Pi and ventral (V) direction of each section, and the small arrows point out the stained mandibular cells in relationship to the entering mandibular nerve trunk (Md in sections c through h). Note that the number of cells per unit area appears to be greater in the most ventral sections [sections I and i) than in the most dorsal sections (sections a, b, c). Calibration bar = 1 mm.
STEREOLOGICAL
PROCEDURES
Previously described and validated stereologic procedures to estimate neuronal cell bodies were used in these studies.1sJ@22 The cell volume of the mandibular
subdivision of the trigeminal ganglion was measured to determine the reference volume in which cell density and other stereologic estimates were performed. Cross-sectional profiles of the ganglia were
431
JOHN R. ZUNIGA identified, and the area and perimeter were measured from photomicrographs (final magnification, X 56) with a digitizing tablet interfaced to a microcomputerbased data acquisition and morphometric analysis system (MORPHOMETER, Woods Hole Educational Assoc, Woods Hole, MA) The total nuclear volume, V,,, (ie, reference volume), was obtained from the sum of the areas (a,, to a,,) of the mandibular subdivision within the 10 parallel sections through the ganglion multiplied by the perpendicular distance between sections. Assuming that heterogeneity of mental sensory cell density exists throughout the dorsoventral axis of each ganglion, cell volume density or,> was determined in photomicrographs (final magnification, X550) by point counting as described by WeibeLzO A coherent test system with 20 main test points, grid area of 2.45 cm2, and line length of 40 cm was used for point counting. Cell volume density was determined by binomial estimates that the number of points falling on cell profiles were a percentage of the total number of points within the test area. The number of cells per unit volume, numerical density (N,), was determined by using the disector,2* where the number of cell profiles appearing in one section but not in the corresponding area of the adjacent section was measured. In practice, photomicrographs (final magnitication, X550) of two serial sections of cells in the mandibular region were superimposed with the aid of fiduciary landmarks (eg, capillaries). The number of cell profiles appearing in the first photomicrograph but not in the second was recorded. The sample area was determined by digital planimetry, and the distance between the samples was determined by measuring the sections from the blocked ganglia of known thickness with a stereotaxic device described by Gundersenz2 The intraganglion section thickness coefficient of variance was 5%, and the interganglion coefficient of variance was 8.5%. The very small intrasection and intersection variation in thickness was theoretically inconsequential, and 5-ym section thickness was considered accurate for all the mathematical procedures. To increase the efficiency of the estimate of N, by twofold, both sections of the dissector were used. The total number of cells per ganglia was estimated as the multiple of N, and Vref. Cell size was measured using the dissector to estimate the volume-weighted mean volume distribution of cell size (VJ by estimating the number of points in the coherent test system hitting cell profiles multiplied by the total number of points within the reference area. Cell shape was determined with digital planimetry by measuring and comparing two populations of straight-line segments along two axes. The mean cell profile length (L,) was determined by measuring the major axis (eg, mean width) of each profile. The axial ratio was then determined by measuring the widest diameter perpendicular to the major
axis. The ratio was used to determine volume-equivalent cell shape.
the mean
TISSUE CORRECTION FACTOR Estimated changes in ganglion volume during fixation and tissue processing was 1.55%. This was obtained by determining the width and thickness of ganglia after perfusion and after paraBin embedding/ staining. The tissue fixation correction factor was estimated from the volume ratio of fresh ganglia from age-, sex-, and weight-matched rats to ganglia after perfusion and immersion fixation. In both techniques, fixation increased ganglion volume an average of 5.87%. Postfixation paraffin embedding, dehydration, and staining decreased volume an average of 8.85%. The tissue correction factor was the mean of the volume change and the percentage of residual volume, and was one minus the tissue correction factor. STATISTICALANALYSIS Preliminary surface and numerical density data of eight ganglia, four from the sham surgery and four from transected/no repair group, were subjected to nested analysis of variance (ANOVA). The calculated variance of each sampling level showed that 20 sections per ganglia were sufficient for detecting statistical differences between the study groups at a significant level of 1%. Two general statistical methods were used in the study. All the values given represent the mean and standard deviation of the mean (SD). Assuming that the sample and subsample size were equal (eg, the number of animals, ganglia, and micrographs were equal and the area and volume of the sample areas were equal within and between groups), a two-way ANOVA with the area, perimeter, number, and volume of the sample as the weighting factor was used to determine the statistical significance based on surgical treatment group and time after surgery. When ANOVA indicated a statistical difference, Neuman-Keul’s multiple-range test was performed. In the first part of the study, separate multivariate ANOVA was performed at each of the two times after surgery and sites with 10 variables were completely crossed for V, to determine dorsoventral cell volume density differences between and within groups. When multivariate analysis of variance (MANOVA) indicated a statistical difference, the Hotelling-Lawley trace was performed to identify significant differences between section intervals. In the second part of the study, separate MANOVA tests were performed to determine differences in V, within and between repair groups. Three times after repair and 10 variable levels according to serial section were completely crossed. When MANOVA indicated a statistical difference, the Hotelling-Lawley trace was performed. Comparisons were made between results obtained
432 after immediate repair and results obtained from the 90-day sham and transection/no repair surgery groups. Comparisons were made between results obtained after delayed repair and results obtained from the 180-day sham and transection/no repair surgery groups. For all tests, significance was assumed at the P < .05 level.
Results SAMPLE SIZE AND STATISTICAL POWER In this study, morphometric data was collected from 7,148 cell bodies from 24 animals, 48 ganglia, and 480 sections using the systematic sampling scheme. Sample and subsample size were determined by choosing a significance of 1% for statistical tests and performing preliminary nested ANOVA of V, values within and between the four groups. The results showed that the magnitude of betweenobservation variance was greater among photographs within each ganglion, than among ganglia within animals; the least variance was between animals. A substantial difference between the repaired and control means existed, whereas no differences were detected between the repaired groups. Thus, the sample size was considered adequate, and the coefficient of error of the systematic sampling scheme was calculated to be .087. STEREOLOGICAL OBSERVATIONS The total mean volume estimate of the mandibular subdivision of the trigeminal ganglion (V& of the sham surgery animals was 0.306 2 0.05 mm3. Mental nerve transection without repair decreased Vref to 0.213 k 0.06 at 90 days and 0.22 5 0.05 mm3 at 180 days. This represented an average 38% dilference, which was statistically significant (P < .003; Table 1). The V,,, differences between treatment times within groups were not significant (P = SlS). The mean number of trigeminal ganglion cells per unit volume (numerical density, NJ was 3.53 t 0.3 X 10’ urn-3 in the sham surgery group ganglia and 2.2 + 0.1 X 10’ urn-’ in transection/no repair ganglia. The mean number of cells (N) was 20.6 ? 3 X 103 in the sham surgery group ganglia and 10.88 k 0.92 X 103 in transection/no repair ganglia. This 47% difference was significant between groups (P < .OOl), but there was no difference for treatment time within groups (P = .517 and P = .~os, respectively). Table 2 lists the mean cellular volume density (VJ differences recorded for all the trigeminal ganglia. There was no significant difference in the mean V, between the sham and transection/no repair group (P = .286) or between treatment times within each group (P = .942 and P = .862, respectively). When
CELL
RESPONSE
TO NERVE
V ref Groups* Sham surgery 90 POD 180 POD
Axotomy =%ery 90 POD 180 POD
SECTION
AND
REPAIR
N”
N
(mm3)
(X 10’ m-3)
12 6 6
.3 (.05)t .3 (.o6)t .31 (.o4)t
3.53 (.29)* 3.63 (.35)f 3.3 (.09)*
20.6 20.9 20.2
12 6 6
.22 (.04)t .21 (.o6)t .22 (.05)t
2.2 (.09)* 2.36 (.19)$ 2.09 (.15)$
10.88 (.92)* 11.17 (1.13)* 10.6 (.74)$
(2.9)$ (3.5)$ (1.65)$
NOTE. Values are mean f SD. *Groups are the summation of treatment at 90 and 180 postoperative days (POD). tDifference in mean values between groups are significant at the P < .05 level. #Difference in mean values between groups are significant at the P < .OOl level.
regional variations in V, were examined, a significant difference between dorsal and ventral sections existed in all of the groups (P < .OOl), whereas ventral sections (>900 urn from the dorsal surface) contained greater densities of cell profiles compared with dorsal surface sections. No alteration in the regional distribution of cellular volume density between groups was apparent. The V,,, was 0.278 k 0.06 mm3 90 days after immediate repair, and 0.274 -+ 0.07 mm3 after delayed repair (Fig 3). The difference between the surgical repair groups (P = 924) and within group treatment times (P = ,731) was not statistically different. The difference between the sham operated and repair groups was not statistically significant (P = .19S). The
Section
Number*
Nt
1
10
2 i
12 12
2
12
ShamSurgery Axotomy Surgery
7
12
.21 (.02) .22 (.07) .26 (.08) .25 (al) .32 (.06) .26 (.06)$ .35 (.11)-i
.17 (.06) .20 (.06) .23 .22 (.09) (.09) .25 (.03)* .26 c.07)
8
12
.28 (.OS)
.25 (.Ol) .23 (.04)
9 10
11 IO
.38 (.os)* .46 (.I I)*
.30 (.13)$ .41 (.ll)*
NOTE. Values are mean k SD. *Assigned section number along the dorsoventral axis, beginning 250 f. 50 pm from the dorsal surface (#l) and every 75 pm thereafter. tNumber of histologic sections examined per group when both 90 and 180 postoperative day sections are combined. *Mean values between the section intervals in each group are significant at the P < .05 level.
JOHN
433
R. ZUMGA
FIGURE
3. Histograms of stereologic data of the mandibular subdivision of the trigeminal ganglion containing mental sensory cells. A, V,r means (*SD) 30, 60, and 90 da s after immediate repair do not dif Yer from each other or when compared to the sham control, but are significantly greater when compared to the transection/no repair group (*P 5 0.051. 5, Vref means (+SD) 30, 60, and 90 days after delayed repair showing similarity to A. C, N, means (2 SD] 30,60, and 90 days after immediate repair do not differ from each other or when compared with the transection/no repair group, but are significantly less than in the sham control[*P< .OOl).D,N,means [*SD) 30, 60, and 90 days after delayed repair showing similarity tot.
30
60
90
SHAM
REPAIR
difference between the repaired groups and the transection/no repair group ganglia was significant (P = .017). The mean ganglion N, was 2.07 + 0.32 X 10’ un--3 after immediate repair and 2.55 k 0.24 X 10’ prnP3 after delayed repair (Fig 3). There was no significant difference between repair groups (P = ,132) and there was no difference within groups when N, was compared 30, 60, and 90 days after surgery (P = .45). When the N, of both repair groups was compared with the sham and transection/no repair control groups, ANOVA showed a significant difference (P < .OOl>. Immediate and delayed repair group N, and 90- and 180-day sham operated group differences were significant (P < .OOl>.There was no significant difference between the repair group ganglia and the 90- and 180-day transection/no repair control ganglia (P = .373). The mean weighted number of cells (N) was 10.66 + 1.1 X 103 in the immediate repair and 12.45 + 1 X lo3 in delayed repair ganglia. No statistical difference between these groups was apparent. The mean V, for the entire trigeminal ganglion was 0.3 +- 0.1 after immediate repair of the mental nerve and 0.284 t 0.1 after delayed repair. The difference was not significant (P = .542). Also, no difference was detected between 30-, GO-,and 90-day postrepair ganglia within groups. When compared with sham and transection/no repair control rats, the V, for the entire ganglion in both repair groups was not statistically different (P = ,149).
30
60
90
30
60
90
SHAM
NO
RENpOAlR
SHAM
NO REPAIR
Cell size was determined using the numberweighted mean volume (L7J distribution, which is illustrated for the sham, transection/no repair, and repair surgery groups in Table 3. The mean V,, was 373.3 t 62 m3 in the sham, 422.5 k 140 urn3 in the transection/no repair, 359 + 89 urn3in the immediate repair, and 474.4 ? 151 urns in the delayed repair ganglia. There was no significant difference between sham and transection/no repair control group V,, (P = .48), and there was no difference between treatment times within the control groups. There was also
Group* Immediate
Postoperative repair
Sham Section/no repair Delayed repair
Sham Section/no
repair
90 90 90 30 60 90 180 180
Davs
V, (um3) 369.14 392.26 310.69 311.53 430.53 506.9 472.2 444.18 430.5 414.7
(161.3) (222) (184.5) (66) (127) (138.8) (86) (20.8) (127) (111)
NOTE. Values are the mean i SD. values at 30, 60, and 90 postoperative days in both groups, and at 90 and 180 days after sham or section/no surgery as control.
repair repair
434
CELL RESPONSE TO NERVE SECTION AND REPAIR
no significant difference between surgical repair groups and their respective sham and transection/no repair control V,, (P = ,163 and P = .189, respectively), as well as when weighted across time within repair groups (P = 556 and P = .659, respectively). Figure 4 contrasts the mean profile size frequency distributions in the sham, transection/no repair, immediate, and delayed repair group ganglia. There was no shift in the spectrum of surviving cells based on size, and cell size frequency distributions were similar to that of the sham and transection/no repair trigeminal ganglia. Cell shape was determined by the mean weighted profile length and axial ratio for each group, which is illustrated in Figure 4. The mean diameter of cell profiles (L& of the sham control ganglia was 27.8 fi 4.3 urn, 26.6 2 5.8 pm in the transection/no repair ganglia, 28.3 + 3.4 urn in immediate repair ganglia, and 24.3 2 5 urn in the delayed repair ganglia. There was no significant Merence between the control groups and repair groups (P = .117), within control
50 9 0 fL: 40 2 t 30 0
No. of neurons = 2075 Mean diameter = 27.78pm Std. dev. = 4.3pm
group times (P = .602), or within repair group times (P = .313). The mean axial ratio was 0.5 t 0.06 in the sham control ganglia, 0.48 k 0.08 in the transection/no repair control ganglia, 0.48 -I 0.1 in immediate repair control ganglia, and 0.5 2 0.06 in delayed repair ganglia. There was no difference between control groups and repair groups (P = .593), within control group times (P = .369), or within repair group times (P = 2453).
Discussion This study supports previous evidence that trigeminal sensory ganglion volume and cell density are substantially reduced after peripheral nerve transection. However, in contrast to other reports, the current study suggests that trigeminal ganglion cell size frequency distributions are not disturbed by peripheral nerve transection. Also, this study suggests that trigeminal ganglion cell loss is the natural and irreversible consequence of transection of the mental
No. of neurons = ‘2138 Mean diameter = 26.60pm Std. dev. = 5.8ym
: 20 2 is 0 ‘0 511 0
0 102030405060708090
0 102030405060708090
PROFILE DIAMETER
50 s 2 3 t 0
No. of neurons = 2422 Mean diameter = 28.34ym Std. dev. = 3.4pm
C
(pm)
No. of neurons = 2614 Mean diameter = 24.29ym Std. dev. = 5.1pm
40 30
% s
20
3
IO
s
0 0 102030405060708090
0 102030405060708090
PROFILE DIAMETER
( m)
D
FIGURE 4. Histograms showing freauencv distribution of orofile diameters in’ the mandibular subdivi’ sion of the trigeminal ganglion after [A) sham surgery; (B) transection/no repair surgery; (C) immediate repair; and (D) delayed repair. Each column represents the percentage of cell profiles that fell within a size interval, where size was Ad [d = 10 km). There was no significant difference between the frequency distribution of cell profile size within or between
JOHN R. ZUNIGA nerve, because the reduction in cell density was unaltered by surgical repair. Compared with sham surgery rats, the cellular volume of the mandibular subdivision of the trigeminal ganglion was significantly reduced after mental nerve transection. It can be inferred from this finding that mental nerve transection in the adult rat may reduce the nuclear volume by 38% in 90 days. This volume reduction was accompanied by a substantial (247%) reduction in numerical cell density without any modification in the regional distribution of cells. Because the somatotopic distribution of cells within the mandibular subdivision of the ganglia was not altered, it is likely that the reduction in the nuclear volume and number of mental sensory cells was probably the result of cell death after mental nerve transection. Cell numbers in the dorsal root and trigeminal sensory ganglia decreased after peripheral nerve section in the adult rat,4,5,7,9,23adult cats3 adult rats after neonatal injury,‘O and in adult mice after neonatal injury. 24 Thus, significant cell death in sensory ganglia is a natural and inconsequential result of peripheral nerve injury. The mean cell loss up to 180 days after transection was 47% in this study but may range from 15% to 56% based on other studies. This finding in this study differed from reports that sensory cells survive peripheral nerve section based on size, and that smaller sized cells survive.5~7-9 Jacquin et allo and Renehan et alz5 reported that the number of responding cells within the ophthalmomaxillary division of the trigeminal ganglion 60 days after section at birth or adulthood was significantly reduced. When compared to normal, the number of nociceptive units was increased. EnIiejian et alz6 reported that infraorbital nerve injury in rats at birth resulted in an increase in the percentage of substance-P cell concentration in the adult trigeminal ganglion. Taken together, these studies suggest that peripheral section of trigeminal nerves results in the preferential survival of select ganglion cells that are small and transmit pain information. However, the results of the current study suggest that any hypothesis about functionally selective cell death cannot be explained by shared properties, which include cell position, size, or shape. However, cells of similar type also can be classed together regardless of whether they have homogenous or analogous structural morphologies. For example, this study does not eliminate functional selective cell death based on electrophysiologic biochemical, or projectionally similar properties. Even when the location and extent of the mental nerve transection injury was matched with the age and gender of control rats, the degree of cell loss in transection/no repair ganglia was not different from that in ganglia 30, 60, and 90 days after either immediate or delayed repair. Thus, for mental nerve
435 transection injuries in adult animals, cell loss was a universal event that was an irreversible consequence of injury and not affected by nerve repair. Other studies have shown that sensory ganglion cell death is not influenced by the regeneration or inhibition of regeneration of peripheral nerves. For example, Ransonz7 first showed in cervical ganglia that immediate repair of the peripheral nerve did not restore lost cells. Cavanaugh12 repeated these results in intercostal nerves of adult rats 130 days after injury. Schmalbruch5 showed that the regeneration of the sciatic nerve did not change the extent of thoracic ganglion cell death in adult rats. Thus, trigeminal ganglion cell death is independent of the restoration and possibly even the extent of axonal regeneration because these are postmitotic events, and the cells are not capable of proliferation after a transection injury of the peripheral part of the nerve. When the mental nerve was transected and repair was delayed, the mean number of cells and cell size were always greater than after immediate repair. This was true when the mean N, and V, of the delayed repair ganglia were compared with the mean N, and V, of the immediate repair ganglia at 30, 60, and 90 postrepair days. However, the difference within or between the immediate and delayed repair groups was not significant. These trends may represent conditioning, which is characterized as an increase in the rate and content of axonal outgrowth and an increase in ganglion cell amino acid incorporation and cellular protein synthesis after repeated peripheral nerve injury.28,29 Confined to only peripheral nerve injuries, conditioning responses may account for the increase in the number of regenerating fibers and the increase in ganglion cell size. Because delayed nerve repair should be considered a repetitive conditioning injury, the modified cell responses may be the basis for the increased mean N, and V, in the delayed repair group when compared with the immediate repair group. When Murphy et aP” carefully examined cells in the trochlear nucleus after the regeneration of peripheral contact after a transection injury, they reported an inverse relationship between the number of lost cells and the size of the regenerated cells and speculated that the number of regenerated fibers determined the size of the surviving cells. Although there was no relationship detected between the number of lost cells and the size or shape of the surviving cells in this study, there was an inverse relationship between the numerical density and nuclear volume of the ganglia in the nerve repair groups. The results are proposed to be caused by the increase in the number of regenerating fibers after repair. Evidence to support this hypothesis derives from the following: 1) the significant differences in V,,, between ganglia in the sham and transection/no repair groups, and between the ganglia
436 in the transection/no repair and immediate and delayed repair groups; 2) no significant difference in V,, between ganglia in the sham and immediate or delayed repair groups; 3) no significant difference in N, in the ganglia in the transection/no repair and immediate and delayed repair groups; and 4) the total and regional variation in V, was similar in the ganglia of all the groups studied. Two morphologic changes that can explain this hypothesis are that significant branching of injured axons occurs during regeneration and nerve trunks reestablish their normal size during regeneration by the combined effect of proliferating axons and regeneration of their supporting cells.sl Thus, the response of regenerating peripheral mental nerve fibers and the number of ganglion cells sending out nerve processes to compensate for lost cells may be a major determinant of restoring nuclear volume after repair without increasing the numerical density. CLINICAL CORRELATION When sensory capacity was tested within the mental nerve distribution in humans after inferior alveolar nerve suture repair 1 to 2.5 years previously, responses to tactile and pain stimuli improved but was never fully restored. lo Other clinical investigators have reported similar findings after suture repair in peripheral limb nerves. ls~* Because sensory ganglion cells are postmitotic and incapable of dividing to compensate for lost cells, the number of damaged cells remaining must be capable of surviving and able to derive potential for regeneration if their axons are successful in reestablishing contact with lost receptor sites. It may be hypothesized that the demands made on remaining sensory ganglion cells after the interaction with target sites may be made through some signal to induce a series of structural changes (eg, axon branching) that attempts to restore and maintain the original volume of fiber projections to the target site. Thus, the volume of both the nerve trunks and the ganglion are restored. However, the efficiency of this response is limited by, among others, the natural consequence of the distance and extent of the injury. For example, axon branching is less extensive after nerve crush than after nerve section, but cell death is more extensive after nerve transection than after nerve crush.ll Thus, because most evidence in this and other studies indicated that cell death was independent of regeneration, the number of cells available to regenerating fibers is limited, and one can speculate that the decreased N, is one neuroanatomic correlate to the results of similar clinical studies in which successful surgical repair of transected nerves rarely completely restored sensory function. The morphometric changes in cell size and numerical density after delayed repair compared with imme-
CELL
RESPONSE
TO NERVE
SECTION
AND
REPAIR
diate repair were interpreted to represent a conditioned, primed, cellular response. If evidence reported by others28,29 that conditioning lesions enhance regeneration of sensory cells, then delaying repair is in effect carrying out a conditioning lesion. However, it is certainly unclear whether conditioning lesions have any clinical role in the management of trigeminal injuries because the number of lost cells was still significantly less than after sham surgery, and the increase was not significantly different from what occurred with immediate repair. Further studies are needed to provide more information on the optimal timing of repair and its relationship to modifying cellular responses for maximum regeneration.
References 1. Robinson PP: Observations on the recovery of sensation following inferior alveolar nerve injuries. Br J Oral Maxillofac Surg 26:177, 1988 2. Zuniga JR, Chen N, Phillips CL: Chemosensory and somatosensory regeneration after lingual nerve repair in humans. J Oral Maxillofac Surg 55:2, 1997 3. Risling M, Hildebrand C, Remahl S: Effects of sciatic nerve resection of L7 spinal roots and dorsal root ganglia in adult cats. Exp Neurol82:568, 1983 4. Ygge J, Aldskoguis H: Intercostal nerve transection and its effects on the dorsal root ganglion: A quantitative study on thoracic ganglion cell numbers and sized in the rat. Exp Brain Res 55:402, 1984 5. Schmalbruch H: Loss of sensory neurons after sciatic nerve section in the rat. Anat Ret 219:323, 1987 6. Arvidsson J, Ygge J, Grant G: Cell loss in lumbar dorsal root ganglia and transganglionic degeneration after sciatic nerve section in the rat. Brain Res 373:15, I986 7. Peyronnard JM, Charron L, LaVoie J, et al: Differences in horseradish peroxidase labeling of sensory, motor and sympathetic neurons following chronic axotomy of the rat sural nerve. Brain Res 364: 137, 1986 8. AIdskoguis H, Risling M: Effects of sciatic neurectomy on neuronal number and size distribution in the L7 ganglion of kittens. Exp Neurol74:597, 1981 9. Rich KM, Disch SP, Eichler ME: The influence of regeneration and nerve growth factor on the neuronal cell body reaction to injury. J Neurocytol l&569, 1989 10. Jacquin MF, Renehan WE, Klein BG, et al: Functional consequences of neonatal infraorbital nerve section in rat trigeminal ganglion. J Neurosci 6:3706, 1986 11. Lieberman AR: The axon reaction: A review of the principle features of perikaryal response to axon injury. Int Rev Neurobiol14:49, 1974 12. Cavanaugh MW: Quantitative effects of the peripheral innervation area of nerves and spinal ganglion cells. J Comp Neurol 94:181, 1951 13. Buchthal F, Kuhl V: Nerve conduction, tactile sensibility and the electromyogram after suture or compression of peripheral nerve: A longitudinal study in man. J Neurol Neurosurg Psychiatry 42:436,1979 14. Hallin RG, Wiesenfeld 2, Lindblom U: Neurophysiological studies on patients with sutured median nerves: Faulty localization after nerve regeneration and its physiological correlates. Exp Neural 73:90, 1981 15. Mackel R, Kunesch E, Waldhor F, et al: Reinnervation of mechanoreceptors in the human glaborous skin following peripheral nerve repair. Brain Res 268:49, 1983 16. Ochs G, Shenk MM, Struppler A: Painful dysesthesia following peripheral nerve injury: A clinical and electrophysiological study. Brain Res 496:228, 1989
JOHN
R. ZUNIGA
AK: Regenerative organization of 17. Zuniga JR, Pate JD, Hegtvedt the trigeminal ganglion following mental nerve section and repair in the adult rat. J Comp Neurol295:548, 1990 18. Gundersen HJG, Jensen EB: Stereological estimation of the volume-weighted mean volume of arbitrary particles observed on random sections. J Microsc 138:127, 1985 organization of the trigemi19. Gregg JM, Dixon AD: Somatotopic nal ganglion in the rat. Arch Oral Biol l&487, 1973 20. Weibel ER: Point counting methods, in Weibel ER (ed): Stereological Methods. New York, NY, Academic Press, 1979, pp 101-203 of number and sizes of 21. Sterio DC: The unbiased estimation arbitrary particles using the disector. J Microsc 134:127, 1984 22. Gundersen HJG: Stereology of arbitrary particles. J Microsc 143:3, 1986 23. Aldskoguis H, Arvidsson J: Nerve cell degeneration and death in the trigeminal ganglion of the adult rat following peripheral nerve transection. J Neurocytol7:229, 1978 24. Savy C, Margules S, Farkas-Bargeton E, et al: A morphometric study of mouse trigeminal ganglion after unilateral destruction of vibrissae follicles at birth. Brain Res 217:265, 1981
437 25. Renehan WE, Klein BG, Chiaia NL, et al: Physiological and anatomical consequences of infraorbital nerve transection in the trigeminal ganglion and trigeminal spinal tract in the adult rat. J Neurosci 9:548, 1989 26. Entiejian HJ, Chiaia NL, Macdonald GJ, et al: Neonatal transection alters the percentage of substance-P positive trigeminal ganglion cells that contribute axons to the regenerate infraorbital nerve. Somatosen Motor Res 6:537, 1989 27. Ranson SW: Retrograde degeneration in the spinal nerves. J Comp Neurol16:265, 1906 28. McQuarrie IG: Effect of a conditioning lesion on axonal sprout formation at nodes of ranvier. J Comp Neurol231:239, 1985 29. Jeng C-B, Jang HM, Bear HM, et al: Conditioning lesions of peripheral nerves change regenerated axon numbers. Brain Res 457163, 1988 30. Murphy EH, Brown J, Iannozzelli PG, et al: Regeneration and soma size changes following axotomy of the trochlear nerve. J Comp Nemo1 292:524, 1990 3 1. HuIsebosch CE, CoggeshalI RE: Sprouting of dorsal root axons. Brain Res 224:170, 1981
Moxillofoc
57:427-437,
Surg 1999
Trigeminal Ganglion Cell Response to Mental Nerve Transection and Repair in the Rat John R. Zuniga,
DMD, PhD*
Purpose: Animal studies have suggestedthat peripheral nerve transection results in substantial loss of ganglion cells and the selective survival of cells based on size. The implications are that subsequent repair of peripheral nerve injuries will be determined by the numerical density and character of the surviving cells. The purpose of this study was twofold: First, to determine the effect of mental nerve transection without repair on trigeminal ganglion cell density and morphology in adult rats, and second, to determine the variation of trigeminal ganglion cell density and morphology after immediate and delayed repair. and Methods: In the first part of the study, 12 adult male Sprague-Dawleyrats had their mental nerves exposed bilaterally (n = 24). Twelve mental nerveswere then transected and prevented from regenerating, and the remaining 12 nerves were uninjured. Ninety and 180 days after transection or shamsurgery, the trigeminal gangliawere serially cut into 5 pm longitudinal sectionsalong the dorsoventral axis. The volume and volume density of the mandibuk mental subdivision containing sensorycells was determined at each section level with pointcounting methods. The numerical density and total number of cellswas estimatedon the same section, using an unbiasedthree-dimensionalstereologicalprobe, the disector. Cell sizeand shapedeterminants were estimatedusingthe disector and computerized planimetry. In the secondpart of the study, six rats had the mental nerves transectedbilaterally and immediately repaired by microscopic sutures.In six additional rats, the repair was delayed for 90 days. In both groups, the trigeminal gangliawere serially cut at 30, GO,and 90 days post-repair and stereologic estimatesof numerical density and histomorphometry were examined usingthe disector and computed planimetry. Materials
In the trigeminal ganglia of the 12 sham-operated animals, the mean number of cells was 20.6 X lo3 (-t 2.9 X 103). After nerve section, the mean number of cells was 10.88 X lo3 (2 0.9 X 103) representing a 47% reduction. The mean volume of the mandibular subdivision cells in the ganglia of the sham surgery animals was 0.3 mm3 (20.05) and 0.22 mm3 (tO.04) in nerve-sectioned ganglia, a 38% difference. There were no ganglia cell size or shape differences between the two groups. The mean number of cells in the ganglia of immediately repaired nerves was 10.66 X 103(t 1.1 X 103) and it was 12.45 X 103 (2 0.9 X 103) after delayed repair. The numerical density was significantly lessthan in the sham surgery ganglia but not different from that of the transection/no repair ganglia. The weighted mean reference volume of the mandibular subdivision after immediate and delayed repair was similar and was significantly greater than the transection/no repair group, but not different from the sham surgery group. The cell size was slightly larger in delayed-repair ganglia compared with immediate-repair ganglia, but the differences were not significant. There were no significant differences in any of the stereologic estimates when analyzed acrosstreatment time. Results:
The results of this study agree with previous reports that peripheral nerve transection produces a substantial loss of nerve cells within specified regions of sensory ganglions. However, the results conflict with evidence that cells survive transection based on size and shape. These findings also indicate that in the adult rat the substantial loss of nerve cells was unaltered by the reconnection of their severed axons. Neither immediate or *Professor, Department of Oral and Maxillofacial Surgery, Univerdelayed repair of the transected nerve altered sity of North Carolina at Chapel Hill, School of Dentistry, Chapel the spectrum of surviving cells based on size or Hill, NC. shape. The reestablishment of the mean referSupported in part by NIBNIDR grant DE08272. ence volume of the mandibular subdivision after Address correspondence and reprint requests to Dr Zuniga: Professor, section and repair suggeststhat demands made Oral and Maxillofacial Surgery, UNC, School of Dentistry, CFW7450, on regenerating axons appear to result in the Chapel Hill, NC 27599.7450; e-mail: J_ZUNIGA@Dentisxry,UNC.edu restoration of ganglionic volume normally lost 6 1999 American Association of Oral and Maxillofacial Surgeons after axotomy, probably the result of axonal 0278.2391/99/5704-0012$3.00/0 branching or supporting cell proliferation. Conclusions:
427
428 The application of microsurgical techniques to repair inferior alveolar and lingual nerve injuries and recover sensory and special sensory function in humans has been demonstrated by many authors.‘s2 These reports suggest that recovery of sensory function perceived by patients or demonstrated by objective tests are rarely complete, but the results are always better than those in patients with nerve injuries without repair. Without question, the reports propose that early surgical repair produces a better outcome. Because microsurgery of the inferior alveolar and lingual sensory nerves is actually surgery on their cells, examining the effects of peripheral nerve injury with or without repair on ganglion cells should be performed to help explain the reported clinical results. It has been known for many years that transection of a peripheral nerve provides a signal that elicits a cell body response that induces changes in protein synthesis, alteration in cell organelles, and cell volume changes that lead to retrograde cell death in adult animals. Previous estimates have suggested that up to 30% or 40% of adult rat sensory ganglion cells die by 160 days after nerve section,3-5 with reported ranged from 15%6 to 55.6%.’ When the size distribution of surviving sensory neurons were examined, many studies reported a preferential and substantial loss of larger cells and a shift in the ganglion size spectrum toward smaller sizes.5,7’9 Because it is generally believed that how an animal senses, perceives, and behaves depends in part on the structured organization and on the number of elements that make up its nervous system, understanding any process that may ultimately affect the number and selectivity of nerve cells is of great importance. For example, Jacquin et allo showed that the decrease in the number of recorded cells in adult trigeminal ganglion after infraorbital nerve section at birth was associated with an absence or a decrease in the number of responding cells when the cutaneous fields normally supplied by the nerve were tested. Furthermore, they showed a shift toward greater numbers of cells that responded to noxious stimuli. If evidence indicates that smaller sensory cells are associated with relaying pain information to more central sites and preferentially survive trigeminal nerve injury, then important neuroanatomic correlates can be made to explain the occurrence of painful dysesthesias (eg, allodynia, hyperpathia, anesthesia dolorosa, sympathetic mediated pain). Most clinicians now agree that microsurgical repair of transected nerves is required in adult patients. This is believed to be true because, although nerve cells do not divide in adults, the remaining cells preserve the potential to repair themselves. In animals, this potential is realized when regeneration is not impeded or the severed ends are repaired using microsurgical
CELL
RESPONSE
TO
NERVE
SECTION
AND
REPAIR
sutures. In adult rats, a direct correlation between cell number, size, volume, and the size of the target after regeneration has been suggested.9,” The regeneration of severed axons in animals may be robust and nearly complete; however, contrary evidence has shown that degeneration of nerve cells was not influenced by the restoration of peripheral contacts and that the same magnitude of cell loss and volume change was independent of the extent of regeneration.5,12 In humans, when stimulation within the skin areas supplied by repaired peripheral sensory limb nerves were examined 5 to 20 years after repair, almost universally, the restitution of sensory function remained deficient.13,14 Similarly, although improved, recovery of tactile, pain, thermal, and taste sensibility to facial and oral sensory fields was never complete in patients who underwent suture repair of inferior alveolar and lingual nerves 1 to 2.5 years previously.1,2 When peripheral microneurography was used to record single units from reinnervated skin of the human hand, Mackel et all5 and Ochs et all6 independently reported reduced receptive field size and sensitivity to mechanical, thermal, and noxious stimuli and suggested that reduced sensitivity was in part the result of reduced density of reinnervating fibers. Whether volumetric or histomorphometric changes in the sensory cell bodies parallel reduced behavioral and physiologic responses after nerve repair remains unknown. The primary purpose of this study was to measure the effect of peripheral nerve transection and repair on trigeminal nerve cells in the adult animal. The mental nerve in the rat was chosen based on previous studies and information of the somatotopic organization of the mental nerve in the trigeminal ganglion of the rat. The study was divided into two parts. The aim of the first part was to describe the diversity of morphometric alterations that modify trigeminal sensory ganglion cell number, size, and shape in response to sectioning of the mental nerve. The aim of the second part was to examine the effects of microsuture repair of the mental nerve on cell histomorphometry and to compare the results with the findings in the first part. Additionally, to simulate clinical conditions in which repair is often delayed and accompanied by axon damming and scar formation before repair, results after immediate repair were compared with delayed (3 month) repair.
Materials
and
Methods
Twenty-four adult male Sprague-Dawley rats weighing between 200 and 250 g were used in this study (Charles River Laboratories, Burlington, NC). All animals were maintained on a 12 : 12 hour light/dark cycle in a macro-environment-controlled rodent room and
JOHN
429
R. ZUNIGA
received water and rat chow ad libitum before and after surgery. For surgery, each rat was anesthetized with an intraperitoneal injection of sodium pentobarbital (Nernbutal, Abbott, North Chicago, IL), 40 mg/kg. The mental nerves were exposed bilaterally in each rat by a submandibular transcutaneous approach, using an aseptic technique that has been previously described” (Fig 1). Each surgery was performed using a Zeiss (Carl Zeiss Inc, Thornwood, NY) surgical microscope. SURGERY AND DENERVATION In the first part of the study, the mental nerves of six animals were transected with surgical scissors 3 mm distal to the mental foramen, and a 5-mm segment of nerve was removed (transection/no repair group). In six additional animals the mental nerves were exposed and the wound immediately closed (sham surgery group). Ninety days later, three animals from the sham surgery and three from the transection/no repair group were deeply anesthetized with sodium pentobarbital, 100 mg/kg administered intraperitoneally. In each animal, transcardiac perfusion was performed with ice-cold 0.9% saline and 2% paraformaldehyde/1.25% glutaraldehyde in 0.01 mol/L phosphatebuffered saline (pH 7.6). The remaining sham and transection/no repair animals were killed in similar fashion 180 days from the initial surgery. The trigeminal ganglion was carefully removed on both sides along with the mandibular nerve trunk, and gross measurements of the width and thickness were recorded. The ganglia were dehydrated in ascending concentrations of alcohol, passed through xylene into paraffin, and serially sectioned at 5 pm along the longitudinal axis. In serial order, from dorsal to ventral surfaces, paraffin sections were platted on subbed slides, dewaxed, and stained in a 1% cresyl violet solution. The width of the broadest section and thickness of each ganglion were recorded.
FIGURE 1. Schematic diagram of the mental nerve-containing region (shaded) of the mandibular (Md) subdivision in the triqeminal ganglion in relation to the maxtllarv /) (Maxi and oohthalmic (OohlI branches, and the trigeminal root (Vrt). Also shown is the sensory branch of the mandibular nerve as it exits foramen ovale (FO), enters the mandibular foramen (Md?), and exits the mental foramen (MF). The surgical site (SS) was 3 mm distal to its exit from the MF.
SURGICAL REPAIR In the second part of the study, the surgical preparation of the rat mental nerves was performed in the same manner as previously described. In six rats, the mental nerves were reanastomosed with two 10-O Ethilon monofilament sutures (Johnson and Johnson, NJ) placed epineurially (immediate repair group). In the remaining six rats, the mental nerves were transected, and no attempt was made to reanastomose the cut ends. Three months later, the cut ends were relocated and reanastomosed (delayed repair group). Two animals from each group were killed at 3@, 60-, and 90&y intervals after their respective nerve repair dates. SAMPLING SCHEME A systematic, uniform, random sampling procedure described by Gundersen and Jensen18 was used to sample an equal fraction of each ganglion. To ensure that all the regions of the ganglion had an equal chance of being examined, the position of the first section was taken between 200 and 300 nm from the dorsal surface, where the first section (ie, section 0) was tangent to the dorsal surface and parallel to the long axis of the ganglion with the mandibular trunk positioned laterally (Fig 2). Two adjacent sections at 10 equally spaced (ie, 2 every 75 pm) intervals along the dorsoventral axis were selected and analyzed. The first sections were the 5th and 6th section from section 0, and the last were the 140th and 141st sections. Cell bodies located just posterior and lateral throughout the dorsoventral axis of the entering mandibular nerve trunk were examined in each section using photomicrographs taken through an Olympus Microscope. Previous studies have shown that mental sensory cell bodies are somatotopically organized to the entering mandibular nerve and constitute the majority of cells in the test areas examined.17,‘9 Nerve cell bodies were defined as cell profiles with clearly defined cytoplasm and a well-demarcated nucleus with diffuse or patterned chromatin distribution.
in V ganglion
430
CELL
RESPONSE
TO
NERVE
SECTION
AND
REPAIR
A 0
,’ ‘;
FIGURE 2. A series of 10 photomicrographs of 1% cresyl violetstained 5urn-thick sections of the trigeminal ganglion. The series beqins with the most dorsal section (section a) and then every section 7.5 pm ventral to the previous one [sections b through j]. The large arrows point out the posterior (Pi and ventral (V) direction of each section, and the small arrows point out the stained mandibular cells in relationship to the entering mandibular nerve trunk (Md in sections c through h). Note that the number of cells per unit area appears to be greater in the most ventral sections [sections I and i) than in the most dorsal sections (sections a, b, c). Calibration bar = 1 mm.
STEREOLOGICAL
PROCEDURES
Previously described and validated stereologic procedures to estimate neuronal cell bodies were used in these studies.1sJ@22 The cell volume of the mandibular
subdivision of the trigeminal ganglion was measured to determine the reference volume in which cell density and other stereologic estimates were performed. Cross-sectional profiles of the ganglia were
431
JOHN R. ZUNIGA identified, and the area and perimeter were measured from photomicrographs (final magnification, X 56) with a digitizing tablet interfaced to a microcomputerbased data acquisition and morphometric analysis system (MORPHOMETER, Woods Hole Educational Assoc, Woods Hole, MA) The total nuclear volume, V,,, (ie, reference volume), was obtained from the sum of the areas (a,, to a,,) of the mandibular subdivision within the 10 parallel sections through the ganglion multiplied by the perpendicular distance between sections. Assuming that heterogeneity of mental sensory cell density exists throughout the dorsoventral axis of each ganglion, cell volume density or,> was determined in photomicrographs (final magnification, X550) by point counting as described by WeibeLzO A coherent test system with 20 main test points, grid area of 2.45 cm2, and line length of 40 cm was used for point counting. Cell volume density was determined by binomial estimates that the number of points falling on cell profiles were a percentage of the total number of points within the test area. The number of cells per unit volume, numerical density (N,), was determined by using the disector,2* where the number of cell profiles appearing in one section but not in the corresponding area of the adjacent section was measured. In practice, photomicrographs (final magnitication, X550) of two serial sections of cells in the mandibular region were superimposed with the aid of fiduciary landmarks (eg, capillaries). The number of cell profiles appearing in the first photomicrograph but not in the second was recorded. The sample area was determined by digital planimetry, and the distance between the samples was determined by measuring the sections from the blocked ganglia of known thickness with a stereotaxic device described by Gundersenz2 The intraganglion section thickness coefficient of variance was 5%, and the interganglion coefficient of variance was 8.5%. The very small intrasection and intersection variation in thickness was theoretically inconsequential, and 5-ym section thickness was considered accurate for all the mathematical procedures. To increase the efficiency of the estimate of N, by twofold, both sections of the dissector were used. The total number of cells per ganglia was estimated as the multiple of N, and Vref. Cell size was measured using the dissector to estimate the volume-weighted mean volume distribution of cell size (VJ by estimating the number of points in the coherent test system hitting cell profiles multiplied by the total number of points within the reference area. Cell shape was determined with digital planimetry by measuring and comparing two populations of straight-line segments along two axes. The mean cell profile length (L,) was determined by measuring the major axis (eg, mean width) of each profile. The axial ratio was then determined by measuring the widest diameter perpendicular to the major
axis. The ratio was used to determine volume-equivalent cell shape.
the mean
TISSUE CORRECTION FACTOR Estimated changes in ganglion volume during fixation and tissue processing was 1.55%. This was obtained by determining the width and thickness of ganglia after perfusion and after paraBin embedding/ staining. The tissue fixation correction factor was estimated from the volume ratio of fresh ganglia from age-, sex-, and weight-matched rats to ganglia after perfusion and immersion fixation. In both techniques, fixation increased ganglion volume an average of 5.87%. Postfixation paraffin embedding, dehydration, and staining decreased volume an average of 8.85%. The tissue correction factor was the mean of the volume change and the percentage of residual volume, and was one minus the tissue correction factor. STATISTICALANALYSIS Preliminary surface and numerical density data of eight ganglia, four from the sham surgery and four from transected/no repair group, were subjected to nested analysis of variance (ANOVA). The calculated variance of each sampling level showed that 20 sections per ganglia were sufficient for detecting statistical differences between the study groups at a significant level of 1%. Two general statistical methods were used in the study. All the values given represent the mean and standard deviation of the mean (SD). Assuming that the sample and subsample size were equal (eg, the number of animals, ganglia, and micrographs were equal and the area and volume of the sample areas were equal within and between groups), a two-way ANOVA with the area, perimeter, number, and volume of the sample as the weighting factor was used to determine the statistical significance based on surgical treatment group and time after surgery. When ANOVA indicated a statistical difference, Neuman-Keul’s multiple-range test was performed. In the first part of the study, separate multivariate ANOVA was performed at each of the two times after surgery and sites with 10 variables were completely crossed for V, to determine dorsoventral cell volume density differences between and within groups. When multivariate analysis of variance (MANOVA) indicated a statistical difference, the Hotelling-Lawley trace was performed to identify significant differences between section intervals. In the second part of the study, separate MANOVA tests were performed to determine differences in V, within and between repair groups. Three times after repair and 10 variable levels according to serial section were completely crossed. When MANOVA indicated a statistical difference, the Hotelling-Lawley trace was performed. Comparisons were made between results obtained
432 after immediate repair and results obtained from the 90-day sham and transection/no repair surgery groups. Comparisons were made between results obtained after delayed repair and results obtained from the 180-day sham and transection/no repair surgery groups. For all tests, significance was assumed at the P < .05 level.
Results SAMPLE SIZE AND STATISTICAL POWER In this study, morphometric data was collected from 7,148 cell bodies from 24 animals, 48 ganglia, and 480 sections using the systematic sampling scheme. Sample and subsample size were determined by choosing a significance of 1% for statistical tests and performing preliminary nested ANOVA of V, values within and between the four groups. The results showed that the magnitude of betweenobservation variance was greater among photographs within each ganglion, than among ganglia within animals; the least variance was between animals. A substantial difference between the repaired and control means existed, whereas no differences were detected between the repaired groups. Thus, the sample size was considered adequate, and the coefficient of error of the systematic sampling scheme was calculated to be .087. STEREOLOGICAL OBSERVATIONS The total mean volume estimate of the mandibular subdivision of the trigeminal ganglion (V& of the sham surgery animals was 0.306 2 0.05 mm3. Mental nerve transection without repair decreased Vref to 0.213 k 0.06 at 90 days and 0.22 5 0.05 mm3 at 180 days. This represented an average 38% dilference, which was statistically significant (P < .003; Table 1). The V,,, differences between treatment times within groups were not significant (P = SlS). The mean number of trigeminal ganglion cells per unit volume (numerical density, NJ was 3.53 t 0.3 X 10’ urn-3 in the sham surgery group ganglia and 2.2 + 0.1 X 10’ urn-’ in transection/no repair ganglia. The mean number of cells (N) was 20.6 ? 3 X 103 in the sham surgery group ganglia and 10.88 k 0.92 X 103 in transection/no repair ganglia. This 47% difference was significant between groups (P < .OOl), but there was no difference for treatment time within groups (P = .517 and P = .~os, respectively). Table 2 lists the mean cellular volume density (VJ differences recorded for all the trigeminal ganglia. There was no significant difference in the mean V, between the sham and transection/no repair group (P = .286) or between treatment times within each group (P = .942 and P = .862, respectively). When
CELL
RESPONSE
TO NERVE
V ref Groups* Sham surgery 90 POD 180 POD
Axotomy =%ery 90 POD 180 POD
SECTION
AND
REPAIR
N”
N
(mm3)
(X 10’ m-3)
12 6 6
.3 (.05)t .3 (.o6)t .31 (.o4)t
3.53 (.29)* 3.63 (.35)f 3.3 (.09)*
20.6 20.9 20.2
12 6 6
.22 (.04)t .21 (.o6)t .22 (.05)t
2.2 (.09)* 2.36 (.19)$ 2.09 (.15)$
10.88 (.92)* 11.17 (1.13)* 10.6 (.74)$
(2.9)$ (3.5)$ (1.65)$
NOTE. Values are mean f SD. *Groups are the summation of treatment at 90 and 180 postoperative days (POD). tDifference in mean values between groups are significant at the P < .05 level. #Difference in mean values between groups are significant at the P < .OOl level.
regional variations in V, were examined, a significant difference between dorsal and ventral sections existed in all of the groups (P < .OOl), whereas ventral sections (>900 urn from the dorsal surface) contained greater densities of cell profiles compared with dorsal surface sections. No alteration in the regional distribution of cellular volume density between groups was apparent. The V,,, was 0.278 k 0.06 mm3 90 days after immediate repair, and 0.274 -+ 0.07 mm3 after delayed repair (Fig 3). The difference between the surgical repair groups (P = 924) and within group treatment times (P = ,731) was not statistically different. The difference between the sham operated and repair groups was not statistically significant (P = .19S). The
Section
Number*
Nt
1
10
2 i
12 12
2
12
ShamSurgery Axotomy Surgery
7
12
.21 (.02) .22 (.07) .26 (.08) .25 (al) .32 (.06) .26 (.06)$ .35 (.11)-i
.17 (.06) .20 (.06) .23 .22 (.09) (.09) .25 (.03)* .26 c.07)
8
12
.28 (.OS)
.25 (.Ol) .23 (.04)
9 10
11 IO
.38 (.os)* .46 (.I I)*
.30 (.13)$ .41 (.ll)*
NOTE. Values are mean k SD. *Assigned section number along the dorsoventral axis, beginning 250 f. 50 pm from the dorsal surface (#l) and every 75 pm thereafter. tNumber of histologic sections examined per group when both 90 and 180 postoperative day sections are combined. *Mean values between the section intervals in each group are significant at the P < .05 level.
JOHN
433
R. ZUMGA
FIGURE
3. Histograms of stereologic data of the mandibular subdivision of the trigeminal ganglion containing mental sensory cells. A, V,r means (*SD) 30, 60, and 90 da s after immediate repair do not dif Yer from each other or when compared to the sham control, but are significantly greater when compared to the transection/no repair group (*P 5 0.051. 5, Vref means (+SD) 30, 60, and 90 days after delayed repair showing similarity to A. C, N, means (2 SD] 30,60, and 90 days after immediate repair do not differ from each other or when compared with the transection/no repair group, but are significantly less than in the sham control[*P< .OOl).D,N,means [*SD) 30, 60, and 90 days after delayed repair showing similarity tot.
30
60
90
SHAM
REPAIR
difference between the repaired groups and the transection/no repair group ganglia was significant (P = .017). The mean ganglion N, was 2.07 + 0.32 X 10’ un--3 after immediate repair and 2.55 k 0.24 X 10’ prnP3 after delayed repair (Fig 3). There was no significant difference between repair groups (P = ,132) and there was no difference within groups when N, was compared 30, 60, and 90 days after surgery (P = .45). When the N, of both repair groups was compared with the sham and transection/no repair control groups, ANOVA showed a significant difference (P < .OOl>. Immediate and delayed repair group N, and 90- and 180-day sham operated group differences were significant (P < .OOl>.There was no significant difference between the repair group ganglia and the 90- and 180-day transection/no repair control ganglia (P = .373). The mean weighted number of cells (N) was 10.66 + 1.1 X 103 in the immediate repair and 12.45 + 1 X lo3 in delayed repair ganglia. No statistical difference between these groups was apparent. The mean V, for the entire trigeminal ganglion was 0.3 +- 0.1 after immediate repair of the mental nerve and 0.284 t 0.1 after delayed repair. The difference was not significant (P = .542). Also, no difference was detected between 30-, GO-,and 90-day postrepair ganglia within groups. When compared with sham and transection/no repair control rats, the V, for the entire ganglion in both repair groups was not statistically different (P = ,149).
30
60
90
30
60
90
SHAM
NO
RENpOAlR
SHAM
NO REPAIR
Cell size was determined using the numberweighted mean volume (L7J distribution, which is illustrated for the sham, transection/no repair, and repair surgery groups in Table 3. The mean V,, was 373.3 t 62 m3 in the sham, 422.5 k 140 urn3 in the transection/no repair, 359 + 89 urn3in the immediate repair, and 474.4 ? 151 urns in the delayed repair ganglia. There was no significant difference between sham and transection/no repair control group V,, (P = .48), and there was no difference between treatment times within the control groups. There was also
Group* Immediate
Postoperative repair
Sham Section/no repair Delayed repair
Sham Section/no
repair
90 90 90 30 60 90 180 180
Davs
V, (um3) 369.14 392.26 310.69 311.53 430.53 506.9 472.2 444.18 430.5 414.7
(161.3) (222) (184.5) (66) (127) (138.8) (86) (20.8) (127) (111)
NOTE. Values are the mean i SD. values at 30, 60, and 90 postoperative days in both groups, and at 90 and 180 days after sham or section/no surgery as control.
repair repair
434
CELL RESPONSE TO NERVE SECTION AND REPAIR
no significant difference between surgical repair groups and their respective sham and transection/no repair control V,, (P = ,163 and P = .189, respectively), as well as when weighted across time within repair groups (P = 556 and P = .659, respectively). Figure 4 contrasts the mean profile size frequency distributions in the sham, transection/no repair, immediate, and delayed repair group ganglia. There was no shift in the spectrum of surviving cells based on size, and cell size frequency distributions were similar to that of the sham and transection/no repair trigeminal ganglia. Cell shape was determined by the mean weighted profile length and axial ratio for each group, which is illustrated in Figure 4. The mean diameter of cell profiles (L& of the sham control ganglia was 27.8 fi 4.3 urn, 26.6 2 5.8 pm in the transection/no repair ganglia, 28.3 + 3.4 urn in immediate repair ganglia, and 24.3 2 5 urn in the delayed repair ganglia. There was no significant Merence between the control groups and repair groups (P = .117), within control
50 9 0 fL: 40 2 t 30 0
No. of neurons = 2075 Mean diameter = 27.78pm Std. dev. = 4.3pm
group times (P = .602), or within repair group times (P = .313). The mean axial ratio was 0.5 t 0.06 in the sham control ganglia, 0.48 k 0.08 in the transection/no repair control ganglia, 0.48 -I 0.1 in immediate repair control ganglia, and 0.5 2 0.06 in delayed repair ganglia. There was no difference between control groups and repair groups (P = .593), within control group times (P = .369), or within repair group times (P = 2453).
Discussion This study supports previous evidence that trigeminal sensory ganglion volume and cell density are substantially reduced after peripheral nerve transection. However, in contrast to other reports, the current study suggests that trigeminal ganglion cell size frequency distributions are not disturbed by peripheral nerve transection. Also, this study suggests that trigeminal ganglion cell loss is the natural and irreversible consequence of transection of the mental
No. of neurons = ‘2138 Mean diameter = 26.60pm Std. dev. = 5.8ym
: 20 2 is 0 ‘0 511 0
0 102030405060708090
0 102030405060708090
PROFILE DIAMETER
50 s 2 3 t 0
No. of neurons = 2422 Mean diameter = 28.34ym Std. dev. = 3.4pm
C
(pm)
No. of neurons = 2614 Mean diameter = 24.29ym Std. dev. = 5.1pm
40 30
% s
20
3
IO
s
0 0 102030405060708090
0 102030405060708090
PROFILE DIAMETER
( m)
D
FIGURE 4. Histograms showing freauencv distribution of orofile diameters in’ the mandibular subdivi’ sion of the trigeminal ganglion after [A) sham surgery; (B) transection/no repair surgery; (C) immediate repair; and (D) delayed repair. Each column represents the percentage of cell profiles that fell within a size interval, where size was Ad [d = 10 km). There was no significant difference between the frequency distribution of cell profile size within or between
JOHN R. ZUNIGA nerve, because the reduction in cell density was unaltered by surgical repair. Compared with sham surgery rats, the cellular volume of the mandibular subdivision of the trigeminal ganglion was significantly reduced after mental nerve transection. It can be inferred from this finding that mental nerve transection in the adult rat may reduce the nuclear volume by 38% in 90 days. This volume reduction was accompanied by a substantial (247%) reduction in numerical cell density without any modification in the regional distribution of cells. Because the somatotopic distribution of cells within the mandibular subdivision of the ganglia was not altered, it is likely that the reduction in the nuclear volume and number of mental sensory cells was probably the result of cell death after mental nerve transection. Cell numbers in the dorsal root and trigeminal sensory ganglia decreased after peripheral nerve section in the adult rat,4,5,7,9,23adult cats3 adult rats after neonatal injury,‘O and in adult mice after neonatal injury. 24 Thus, significant cell death in sensory ganglia is a natural and inconsequential result of peripheral nerve injury. The mean cell loss up to 180 days after transection was 47% in this study but may range from 15% to 56% based on other studies. This finding in this study differed from reports that sensory cells survive peripheral nerve section based on size, and that smaller sized cells survive.5~7-9 Jacquin et allo and Renehan et alz5 reported that the number of responding cells within the ophthalmomaxillary division of the trigeminal ganglion 60 days after section at birth or adulthood was significantly reduced. When compared to normal, the number of nociceptive units was increased. EnIiejian et alz6 reported that infraorbital nerve injury in rats at birth resulted in an increase in the percentage of substance-P cell concentration in the adult trigeminal ganglion. Taken together, these studies suggest that peripheral section of trigeminal nerves results in the preferential survival of select ganglion cells that are small and transmit pain information. However, the results of the current study suggest that any hypothesis about functionally selective cell death cannot be explained by shared properties, which include cell position, size, or shape. However, cells of similar type also can be classed together regardless of whether they have homogenous or analogous structural morphologies. For example, this study does not eliminate functional selective cell death based on electrophysiologic biochemical, or projectionally similar properties. Even when the location and extent of the mental nerve transection injury was matched with the age and gender of control rats, the degree of cell loss in transection/no repair ganglia was not different from that in ganglia 30, 60, and 90 days after either immediate or delayed repair. Thus, for mental nerve
435 transection injuries in adult animals, cell loss was a universal event that was an irreversible consequence of injury and not affected by nerve repair. Other studies have shown that sensory ganglion cell death is not influenced by the regeneration or inhibition of regeneration of peripheral nerves. For example, Ransonz7 first showed in cervical ganglia that immediate repair of the peripheral nerve did not restore lost cells. Cavanaugh12 repeated these results in intercostal nerves of adult rats 130 days after injury. Schmalbruch5 showed that the regeneration of the sciatic nerve did not change the extent of thoracic ganglion cell death in adult rats. Thus, trigeminal ganglion cell death is independent of the restoration and possibly even the extent of axonal regeneration because these are postmitotic events, and the cells are not capable of proliferation after a transection injury of the peripheral part of the nerve. When the mental nerve was transected and repair was delayed, the mean number of cells and cell size were always greater than after immediate repair. This was true when the mean N, and V, of the delayed repair ganglia were compared with the mean N, and V, of the immediate repair ganglia at 30, 60, and 90 postrepair days. However, the difference within or between the immediate and delayed repair groups was not significant. These trends may represent conditioning, which is characterized as an increase in the rate and content of axonal outgrowth and an increase in ganglion cell amino acid incorporation and cellular protein synthesis after repeated peripheral nerve injury.28,29 Confined to only peripheral nerve injuries, conditioning responses may account for the increase in the number of regenerating fibers and the increase in ganglion cell size. Because delayed nerve repair should be considered a repetitive conditioning injury, the modified cell responses may be the basis for the increased mean N, and V, in the delayed repair group when compared with the immediate repair group. When Murphy et aP” carefully examined cells in the trochlear nucleus after the regeneration of peripheral contact after a transection injury, they reported an inverse relationship between the number of lost cells and the size of the regenerated cells and speculated that the number of regenerated fibers determined the size of the surviving cells. Although there was no relationship detected between the number of lost cells and the size or shape of the surviving cells in this study, there was an inverse relationship between the numerical density and nuclear volume of the ganglia in the nerve repair groups. The results are proposed to be caused by the increase in the number of regenerating fibers after repair. Evidence to support this hypothesis derives from the following: 1) the significant differences in V,,, between ganglia in the sham and transection/no repair groups, and between the ganglia
436 in the transection/no repair and immediate and delayed repair groups; 2) no significant difference in V,, between ganglia in the sham and immediate or delayed repair groups; 3) no significant difference in N, in the ganglia in the transection/no repair and immediate and delayed repair groups; and 4) the total and regional variation in V, was similar in the ganglia of all the groups studied. Two morphologic changes that can explain this hypothesis are that significant branching of injured axons occurs during regeneration and nerve trunks reestablish their normal size during regeneration by the combined effect of proliferating axons and regeneration of their supporting cells.sl Thus, the response of regenerating peripheral mental nerve fibers and the number of ganglion cells sending out nerve processes to compensate for lost cells may be a major determinant of restoring nuclear volume after repair without increasing the numerical density. CLINICAL CORRELATION When sensory capacity was tested within the mental nerve distribution in humans after inferior alveolar nerve suture repair 1 to 2.5 years previously, responses to tactile and pain stimuli improved but was never fully restored. lo Other clinical investigators have reported similar findings after suture repair in peripheral limb nerves. ls~* Because sensory ganglion cells are postmitotic and incapable of dividing to compensate for lost cells, the number of damaged cells remaining must be capable of surviving and able to derive potential for regeneration if their axons are successful in reestablishing contact with lost receptor sites. It may be hypothesized that the demands made on remaining sensory ganglion cells after the interaction with target sites may be made through some signal to induce a series of structural changes (eg, axon branching) that attempts to restore and maintain the original volume of fiber projections to the target site. Thus, the volume of both the nerve trunks and the ganglion are restored. However, the efficiency of this response is limited by, among others, the natural consequence of the distance and extent of the injury. For example, axon branching is less extensive after nerve crush than after nerve section, but cell death is more extensive after nerve transection than after nerve crush.ll Thus, because most evidence in this and other studies indicated that cell death was independent of regeneration, the number of cells available to regenerating fibers is limited, and one can speculate that the decreased N, is one neuroanatomic correlate to the results of similar clinical studies in which successful surgical repair of transected nerves rarely completely restored sensory function. The morphometric changes in cell size and numerical density after delayed repair compared with imme-
CELL
RESPONSE
TO NERVE
SECTION
AND
REPAIR
diate repair were interpreted to represent a conditioned, primed, cellular response. If evidence reported by others28,29 that conditioning lesions enhance regeneration of sensory cells, then delaying repair is in effect carrying out a conditioning lesion. However, it is certainly unclear whether conditioning lesions have any clinical role in the management of trigeminal injuries because the number of lost cells was still significantly less than after sham surgery, and the increase was not significantly different from what occurred with immediate repair. Further studies are needed to provide more information on the optimal timing of repair and its relationship to modifying cellular responses for maximum regeneration.
References 1. Robinson PP: Observations on the recovery of sensation following inferior alveolar nerve injuries. Br J Oral Maxillofac Surg 26:177, 1988 2. Zuniga JR, Chen N, Phillips CL: Chemosensory and somatosensory regeneration after lingual nerve repair in humans. J Oral Maxillofac Surg 55:2, 1997 3. Risling M, Hildebrand C, Remahl S: Effects of sciatic nerve resection of L7 spinal roots and dorsal root ganglia in adult cats. Exp Neurol82:568, 1983 4. Ygge J, Aldskoguis H: Intercostal nerve transection and its effects on the dorsal root ganglion: A quantitative study on thoracic ganglion cell numbers and sized in the rat. Exp Brain Res 55:402, 1984 5. Schmalbruch H: Loss of sensory neurons after sciatic nerve section in the rat. Anat Ret 219:323, 1987 6. Arvidsson J, Ygge J, Grant G: Cell loss in lumbar dorsal root ganglia and transganglionic degeneration after sciatic nerve section in the rat. Brain Res 373:15, I986 7. Peyronnard JM, Charron L, LaVoie J, et al: Differences in horseradish peroxidase labeling of sensory, motor and sympathetic neurons following chronic axotomy of the rat sural nerve. Brain Res 364: 137, 1986 8. AIdskoguis H, Risling M: Effects of sciatic neurectomy on neuronal number and size distribution in the L7 ganglion of kittens. Exp Neurol74:597, 1981 9. Rich KM, Disch SP, Eichler ME: The influence of regeneration and nerve growth factor on the neuronal cell body reaction to injury. J Neurocytol l&569, 1989 10. Jacquin MF, Renehan WE, Klein BG, et al: Functional consequences of neonatal infraorbital nerve section in rat trigeminal ganglion. J Neurosci 6:3706, 1986 11. Lieberman AR: The axon reaction: A review of the principle features of perikaryal response to axon injury. Int Rev Neurobiol14:49, 1974 12. Cavanaugh MW: Quantitative effects of the peripheral innervation area of nerves and spinal ganglion cells. J Comp Neurol 94:181, 1951 13. Buchthal F, Kuhl V: Nerve conduction, tactile sensibility and the electromyogram after suture or compression of peripheral nerve: A longitudinal study in man. J Neurol Neurosurg Psychiatry 42:436,1979 14. Hallin RG, Wiesenfeld 2, Lindblom U: Neurophysiological studies on patients with sutured median nerves: Faulty localization after nerve regeneration and its physiological correlates. Exp Neural 73:90, 1981 15. Mackel R, Kunesch E, Waldhor F, et al: Reinnervation of mechanoreceptors in the human glaborous skin following peripheral nerve repair. Brain Res 268:49, 1983 16. Ochs G, Shenk MM, Struppler A: Painful dysesthesia following peripheral nerve injury: A clinical and electrophysiological study. Brain Res 496:228, 1989
JOHN
R. ZUNIGA
AK: Regenerative organization of 17. Zuniga JR, Pate JD, Hegtvedt the trigeminal ganglion following mental nerve section and repair in the adult rat. J Comp Neurol295:548, 1990 18. Gundersen HJG, Jensen EB: Stereological estimation of the volume-weighted mean volume of arbitrary particles observed on random sections. J Microsc 138:127, 1985 organization of the trigemi19. Gregg JM, Dixon AD: Somatotopic nal ganglion in the rat. Arch Oral Biol l&487, 1973 20. Weibel ER: Point counting methods, in Weibel ER (ed): Stereological Methods. New York, NY, Academic Press, 1979, pp 101-203 of number and sizes of 21. Sterio DC: The unbiased estimation arbitrary particles using the disector. J Microsc 134:127, 1984 22. Gundersen HJG: Stereology of arbitrary particles. J Microsc 143:3, 1986 23. Aldskoguis H, Arvidsson J: Nerve cell degeneration and death in the trigeminal ganglion of the adult rat following peripheral nerve transection. J Neurocytol7:229, 1978 24. Savy C, Margules S, Farkas-Bargeton E, et al: A morphometric study of mouse trigeminal ganglion after unilateral destruction of vibrissae follicles at birth. Brain Res 217:265, 1981
437 25. Renehan WE, Klein BG, Chiaia NL, et al: Physiological and anatomical consequences of infraorbital nerve transection in the trigeminal ganglion and trigeminal spinal tract in the adult rat. J Neurosci 9:548, 1989 26. Entiejian HJ, Chiaia NL, Macdonald GJ, et al: Neonatal transection alters the percentage of substance-P positive trigeminal ganglion cells that contribute axons to the regenerate infraorbital nerve. Somatosen Motor Res 6:537, 1989 27. Ranson SW: Retrograde degeneration in the spinal nerves. J Comp Neurol16:265, 1906 28. McQuarrie IG: Effect of a conditioning lesion on axonal sprout formation at nodes of ranvier. J Comp Neurol231:239, 1985 29. Jeng C-B, Jang HM, Bear HM, et al: Conditioning lesions of peripheral nerves change regenerated axon numbers. Brain Res 457163, 1988 30. Murphy EH, Brown J, Iannozzelli PG, et al: Regeneration and soma size changes following axotomy of the trochlear nerve. J Comp Nemo1 292:524, 1990 3 1. HuIsebosch CE, CoggeshalI RE: Sprouting of dorsal root axons. Brain Res 224:170, 1981