- Email: [email protected]
Magnetic Resonance Neurography of Traumatic and Nontraumatic Peripheral Trigeminal Neuropathies
Accepted Manuscript Magnetic Resonance Neurography of Traumatic and Non-Traumatic Peripheral Trigeminal Neuropathies John R. Zuniga, DMD, MS, PhD, Robert V. Walker, DDS, Chair in Oral and Maxillofacial Surgery, Professor, Cyrus Mistry, DDS, MD, Chief Resident, Igor Tikhonov, DDS, MD, Chief Resident, Riham Dessouky, MD, Fellow, Avneesh Chhabra, MD, Associate Professor PII:
S0278-2391(17)31426-X
DOI:
10.1016/j.joms.2017.11.007
Reference:
YJOMS 58043
To appear in:
Journal of Oral and Maxillofacial Surgery
Received Date: 13 September 2017 Revised Date:
8 November 2017
Accepted Date: 8 November 2017
Please cite this article as: Zuniga JR, Walker RV, Mistry C, Tikhonov I, Dessouky R, Chhabra A, Magnetic Resonance Neurography of Traumatic and Non-Traumatic Peripheral Trigeminal Neuropathies, Journal of Oral and Maxillofacial Surgery (2017), doi: 10.1016/j.joms.2017.11.007. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
RI PT
ACCEPTED MANUSCRIPT
M AN U
SC
MAGNETIC RESONANCE NEUROGRAPHY OF TRAUMATIC AND NON-TRAUMATIC PERIPHERAL TRIGEMINAL NEUROPATHIES
John R. Zuniga DMD, MS, PhD, Robert V. Walker DDS Chair in Oral and Maxillofacial Surgery, Professor, Departments of Surgery and Neurology & Neurotherapeutics, University of Texas Southwestern Medical Center at Dallas, Texas Cyrus Mistry DDS, MD, Chief Resident, Oral and Maxillofacial Surgery, University of Texas Southwestern Medical Center at Dallas, Texas
TE D
Igor Tikhonov DDS, MD, Chief Resident, Oral and Maxillofacial Surgery, University of Texas Southwestern Medical Center at Dallas, Texas Riham Dessouky MD, Fellow, Radiology, University of Texas Southwestern Medical Center at Dallas, Texas
EP
Avneesh Chhabra MD, Associate Professor, Radiology & Orthopedic Surgery University of Texas Southwestern Medical Center at Dallas, Texas
All Correspondence to:
AC C
John R. Zuniga DMD, MS, PhD Robert V. Walker DDS Chair in Oral and Maxillofacial Surgery Professor, Departments of Surgery and Neurology & Neurotherapeutics University of Texas Southwestern Medical Center at Dallas Dallas, Texas Email: [email protected]
ACCEPTED MANUSCRIPT
ABSTRACT Purpose: The clinical neurosensory testing (NST) is currently the gold-standard for the diagnosis of traumatic and non-traumatic peripheral trigeminal neuropathies (PTN), but exhibits both false positive and negative results when
RI PT
compared to surgical findings and frequently delays treatment decisions. We tested the hypothesis that magnetic resonance neurography (MRN) of PTN can serve as a diagnostic modality by correlating NST, MRN and surgical findings.
SC
Materials and Methods: Sixty patients with traumatic and non-traumatic PTN of varying etiologies and Sunderland classifications underwent NST followed by MRN on 1.5T and 3.0 T scanners. The protocol included 2D and 3D imaging, including diffusion imaging and isotropic 3D PSIF. The MRN findings were read by two readers in
M AN U
consensus of clinical findings but blinded to the side of abnormality. The MRN results were summarized using Sunderland Classification. In 26 patients, surgery was performed and Sunderland classification was assigned based on surgical photos. Agreement between MRN and NST/Surgical classification was evaluated using kappa statistics. Pearson’s Correlation Coefficient (PCC) was used to assess the correlation between continuous measurements of
TE D
MRN/NST and surgical classification.
Results: Nineteen males and 41 females, mean age 41, ranging 12 to 75, with 54 complaints of altered sensation of the lip/chin/or tongue, including 16 with neuropathic pain and 4 with no neurosensory complaint were included.
EP
Third molar surgery (n=29) represented the most common cause of traumatic PTN. Assuming one nerve abnormality per patient, the lower class was accepted, a kappa of 0.57 was observed between MRN and NST
AC C
classification. A kappa of 0.5 existed between MRN and surgical findings with a PCC of 0.67. Conclusions: MRN anatomically maps PTN and stratifies the nerve injury and neuropathies with moderate to good agreement with NST and surgical findings for clinical use.
ACCEPTED MANUSCRIPT
INTRODUCTION Peripheral trigeminal neuropathies (PTN) include the loss of general sense, special sense and/or development of neuropathic pain in the oral and facial distributions of the trigeminal nerve and are well
RI PT
recognized outcomes of traumatic (i.e., third molar removal, implant placement, etc.) and non-traumatic (i.e., neoplastic, non-invasive endodontic, atypical facial pain, dental injection, etc.) causations. The responsibility to each patient is to provide a timely diagnosis, reasonable prognosis, accurate monitoring
SC
of recovery or lack of and evidence based treatment recommendation for the symptoms of the
trigeminal neuropathy. This is no more important than for trigeminal neuropathies that would benefit
M AN U
by microsurgical repair since earlier intervention is clearly related to better outcome1,2,3 and thus, early diagnosis would be preferred. The data required to accomplish these tasks has been relegated to clinical chairside neurosensory tests (NST). For diagnostic purposes the three-level drop-out algorithm was published in 19924 , which currently serves as the gold standard to diagnose PTN after its statistical power was related to surgical findings in 19985. For accurate monitoring of recovery and outcome of
TE D
surgical repair, the Medical Research Council Scale (MRCS) for sensory recovery is currently the gold standard to identify functional sensory recovery6,7. Both NST and MRCS rely on patient response to stimulus and operator experience in stimulus presentation and interpretation6,8. In addition, the results
EP
from these tests are only accurate when delivered 1 month after the injury since they cannot distinguish
AC C
levels of injury before 1 month. When the three-level drop-out algorithm was corroborated with surgical findings, this NST was shown to exhibit high positive (PPV=95%) and negative (NPV=100%) predictive values for lingual nerve (LN) injuries and moderate PPV (77%) and NPV (60%) for inferior alveolar nerve (IAN) injuries. The NPV of 60% means that NST is less efficient in ruling out an IAN injury resulting in a false negative rate of 40%. In addition, for both IAN and LN, at higher sensory impairment scores, the NST tends to overestimate the degree of nerve injury and, at lower sensory impairment scores, NST tends to underestimate the degree of nerve injury. Furthermore, patients with different
ACCEPTED MANUSCRIPT
degrees of nerve injuries may show similar NST scores and these scores can vary with differences in age, duration and cause of injury9. Thus, although clinically useful in providing diagnosis, prognosis, monitoring and treatment options for patients with PTN, the NST clearly suffers from the following: 1.
RI PT
Delays treatment of higher class injury patients who would likely benefit by earlier intervention; 2. Does not adequately delineate the anatomy and location of the injury; and 3. Can overestimate and underestimate the level of injury, especially if confounding factors are present.
SC
Magnetic resonance neurography (MRN) is an imaging dedicated to the peripheral nerves and provides a non-invasive foot-print of neural anatomy and resolves the intraneural architecture in multiple
M AN U
orthogonal planes10,11,12,13. On MRN of peripheral nerves, neuropathies demonstrate imaging alterations of nerve caliber, intraneural T2 signal intensity ratio (SIR) and altered diffusion characteristics, such as increased apparent diffusion coefficient (ADC) and decreased fractional anisotropy (FA) that correlate with axonal degeneration and demyelination14,15,16,17. In initial studies, MRN of injured and non-injured
TE D
PTN was shown to improve diagnostic information, thereby impacting clinical management and there was moderate to excellent correlation with intraoperative findings18,19,20. The hypotheses that drove this study were two- fold. First, MRN can differentiate normal from
EP
abnormal-injured and non-injured peripheral trigeminal nerves. Second, MRN nerve injury
AC C
classifications based on Sunderland grading system21 correlates with NST grading and surgical findings.
MATERIALS AND METHODS
This was a retrospective evaluation of a case series which included 60 patients from 3/2015 to 5/2017 with traumatic and non-traumatic PTN of varying etiologies and Sunderland classifications who
ACCEPTED MANUSCRIPT
underwent MRN evaluation in addition to clinical NST and, in some cases, trigeminal nerve surgery. This study was conducted with approval from the UTSW institutional review board (IRB) following Health
RI PT
Insurance Portability and Accountability (HIPPA) guidelines. Informed consent was waived. Patient Evaluations
Inclusion criteria for data analysis included those patients seen in the Division of Oral and Maxillofacial
SC
Surgery with a neurosensory complaint who underwent clinical NST and agreed to MRN evaluation for a suspected PTN in the second or third division of the trigeminal nerve. Patient age, gender, trigeminal
M AN U
nerve(s) injury by location and laterality, chief complaint, sensory symptoms, pain descriptions (when present), mechanism of injury, and previous treatment were recorded. For purposes of testing the hypothesis, there were two groups of patients examined, the first were those who had known traumatic injuries of the trigeminal nerve (i.e., third molar removal, dental implant placement, tumor resection, facial trauma) and the second, those who did not have known traumatic injuries (i.e., dental injection,
TE D
non-invasive endodontic treatment, atypical facial pain, neoplastic). Clinical NST was conducted to determine the level of sensory impairment defined by none, mild, moderate, severe and complete previously described4 and functional sensory grading by Grade S0, S1,
EP
S2, S2+, S3, S3+, and S4 on the Medical Research Council Scale (MRCS) as previously described6,7,12 . The
AC C
clinical NST was performed by one examiner (JRZ) prior to MRN evaluation and any surgical intervention. The results of level of sensory impairment NST were scored 4 (normal), 3 (mild impairment), 2 (moderate impairment), 1 (severe impairment) and 0 (complete anesthesia) while the MRCS grading was scored by numerical value in parallel from 0 (no sensation) to 4 (superficial pain and touch without hyperesthesia and static 2-point discrimination of 2-6mm indicating complete sensation). A Sunderland classification was provided from this data from Class I (S3+, S4/ NST normal (4) by 3 months), Class II (S3+, S4/NST normal (4) by 6 months), Class III (S2, S2+, S3/ NST mild (3) or moderate (2) by or greater
ACCEPTED MANUSCRIPT
than 6 months), Class IV (S1, S2, S2+/NST moderate (2) severe (1) and Class V (S0, S1/NST complete (0) by or greater than 6 months) as described in Table 1.
RI PT
When trigeminal nerve surgery was performed, a single surgeon (JRZ) performed all the surgeries. Surgical findings were recorded by consensus by the surgeons after direct microscopic inspection of the injury site using high (40x) magnification. Operative photographs were taken twice of the same site at high and low magnification by someone other than the operating surgeon. Calipers or a ruler was used
SC
to determine nerve thickness of the proximal and distal ends when a transection injury was required as well as the size of the gap in millimeters, and the size of neuroma when applicable. Based on intra-
M AN U
operative inspection described above the Sunderland classification of the injury was determined by the surgeon and research team in consensus as outline in Table 1. Magnetic Resonance Neurography Evaluations
All MRN studies were performed without contrast for a total imaging time of 50 minutes in supine
TE D
position with a 1.5 T (Siemens, Avanto, Erlangen, Germany) or 45 minutes on a 3T (Achieva, Ingenia, Philips, Best, Netherlands) scanner with a multichannel head coil placed over the head and neck region. The protocol included 2D (dimensional) and 3D T1W, T2 SPAIR (spectral adiabatic inversion recovery)
EP
imaging, diffusion imaging and isotropic 3D PSIF (0.9 mm voxel) as well 3D inversion recovery turbo spin
AC C
echo (3D IR TSE). Thick slab maximum intensity projections (MIP) were obtained in nerve longitudinal planes (coronal and oblique sagittal) on independent workstations (Aquarius, Tera Recon, FosterCity, CA, USA). Perineural fat planes were assessed for presence of scarring or mass lesion. Three-dimensional CISS (constructive interference in steady state) sequence was employed in the protocol to exclude any intracranial neural pathology and the axial T1W and T2 SPAIR sequences were extended to the corpus callosum to the chin to exclude any intranuclear or adjacent intra-cranial pathology. The protocol sequence parameters were as follows: coronal 3D PSIF (0.8-0.9mm isotropic; time of acquisition, TA-6
ACCEPTED MANUSCRIPT
minutes), coronal 3D IR TSE (1.5 mm isotropic; TA- 6 to 7 minutes), axial diffusion tensor imaging, DTI (b50,800,1000; 2mm isotropic, 15 directions of interrogation; TA-8 minutes), axial T1W (3.5 mm; TA – 4 to 5 minutes), and axial T2 SPAIR (3.5mm: TA – 4 to 5 minutes). The MRN findings of nerve signal and
RI PT
caliber alteration were read by two musculoskeletal radiologists in consensus in the light of the clinical findings but blinded to the side of abnormality as part of routine patient care. A Sunderland
classification was given by the radiologists blinded to the NST grading based on the qualitative imaging
SC
criteria described in Table 1.
M AN U
Statistical Analysis
Descriptive statistics were used to compare the demographic data, and Sunderland classification on NST and MRN between traumatic and non-traumatic patient populations. In the traumatic patient population, Kappa analysis was performed to test correlations of Sunderland classification on independently performed NST, MRN and surgical classifications. Pearson’s correlation coefficient was
TE D
used to determine the association of differences in continuous measurements in abnormal versus normal nerves of MRN/NST and surgical findings. Agreement was classified as excellent (>0.80), good
RESULTS
EP
(0.61-0.80), moderate (0.41- 0.60), fair (0.20- 0.40) and poor (<0.20). Type 1 error was set at 0.05.
AC C
Patient population and NST:
This case series was composed of 60 patients with trigeminal injuries involving the inferior alveolar nerve (IAN) and lingual nerve (LN) who underwent clinical NST and MRN examination between 3/2015 to 5/2017, and microsurgery on 26 patients between 4/2015 and 5/2017. There were 19 males, 41 females with a mean age of 43 ± 15 years, median of 39 and range from 12 to 75 years of age. The chief complaint was numbness in 25 cases, pain in 16, diminished or lost taste in 6, abnormal diminished sensation in 9, suspected nerve injury in 2 and pre-procedural check-up in 2. The distribution of cases
ACCEPTED MANUSCRIPT
by mechanism of injury showed that third molar extraction was the common source in 47% of patients (n=28) followed by dental implant 18% (n=11), mandibular pathology surgery 13% (n=8), unknown 10% (n=6) while the remaining mechanisms represented less than 1% incidence per group. By nerve
RI PT
distribution, there were 40 patients with IAN sensory disorders and 20 LN disorders. For purposes of further discussion, there were 48 patients who had traumatic injuries to the IAN and LN (Table 2) and 12 with non-traumatic injuries to the IAN and LN (Table 3). Traumatic injuries are listed in Table 2 under
SC
third molar extraction, dental implant, mandibular surgery and orthognathic surgery. Non-traumatic injuries included three categories of etiologies, they were: dental injection, non-invasive endodontic
M AN U
treatment, and unknown IAN sensory complaints (3 with no known event, 2 incisional biopsy of mucosa and 1 maxillary routine dental extraction). Within the traumatic injury population, there were differences between the IAN and LN in NST level of impairment with 79% LN injuries grade S0 to S2 and NST level IV or more (15/19) whereas 79% IAN injuries grade S3 to S4 and NST level III or less (22/28).
TE D
Neuropathic pain was present in the clinical and NST examination in 45% of IAN (13/28) and 15% of LN (3/19) and the duration of injury to clinical examination reported was 13.61 months (range 0 to 49) for IAN and 5.5 months (range 1 to 45) for LN. Within the non-traumatic injury population, which includes
EP
11 IAN and 1 LN injury, the distribution was Grade 3 to 4 in 100% (12/12) and level III or less in 100% (12/12). Neuropathic pain was present in the clinical and NST examination in 50% (6/12) and the
AC C
duration of injury to clinical examination reported was 4.87 months (range 0.5 to 24). An example of a non-traumatic PTN event is of a 64 y/o female with moderate sensory impairment/S2 and neuropathic pain (allodynia, hyperpathia and burning pain) 2 months following traditional endodontic procedure on tooth #18 for chronic pulpitis. Glass ionomer and GP pulp canal sealer that was irrigated with sodium hypochlorite was reported to have been performed for the treatment of tooth #18. Figure 1 shows a coronal MRN vessel-suppressed view of the right and left inferior alveolar nerve in this case and an axial DTI image showing hyperintensity and enlargement of the left IAN in the area near tooth #18. Figure 2
ACCEPTED MANUSCRIPT
shows the surgical findings demonstrating an intact nerve with intense inflammatory changes. Classification of traumatic nerve injuries on NST, MRN and Surgery:
RI PT
Twenty-six patients underwent microdissection, 10 IAN and 16 LN. Of these 26, 1 patient had a nontraumatic category injury to the IAN associated with endodontic therapy with a NST level of 2 and MRCS grade S3 with neuropathic pain. Table 4 shows the Sunderland classification for each patient based on NST, MRN and surgery. Figure 3 shows coronal view of 3D MRN of the right and left lingual nerve of a
SC
patient with severe sensory impairment/S1 of the left tongue with a trigger and dystrophic ageusia on the same side. Figure 4 shows the surgical findings of a large neuroma of the left lingual nerve. Figure 5
M AN U
shows a coronal 3D MRN of the right and left IAN and right and left LN in 5 months after dental implant placed posterior left mandible with moderate impairment/S2 with neuropathic pain (hyperalgesia). The MRN findings were described as left IAN Class II injury with mild thickening and hyperintense signaling with perineural fibrosis with no discontinuity or neuroma identification. Figure 6A shows the IAN in
TE D
relationship to the dental implant with no discontinuity or neuroma formation. Figure 6B shows higher magnification of the point of injury showing internal fibrosis. There were 6 indeterminate classifications using NST, 5 MRN and 1 surgical where the examiner, radiologist and surgeon could not decide between
EP
class injuries. Assuming one nerve abnormality per patient, and when classification was undetermined, the lower class was accepted. In 2 cases, the MRN could not classify the nerve injury due to artifact
AC C
and/or motion. The distribution of Sunderland classification by nerve and their ratio of correct to incorrect NST and MRN compared to surgical findings is presented in Table 5. There were more Class IV and V injuries in LN injuries than IAN and more Class II and III in IAN injuries than LN as has been previously shown5. For a LN Class IV injury the NST and MRN were accurate in predicting surgical findings and for IAN Class II injuries MRN was more accurate than NST. Except for two exceptions, NST overestimated the class injury in Class II IAN injuries, both NST and MRN underestimated the surgical findings. This study showed kappa of 0.57 and 0.5 between MRN and NST classifications and MRN and
ACCEPTED MANUSCRIPT
surgical classifications. The Pearsons’ Correlation Coefficient was 0.67 between MRN and surgical findings.
RI PT
Post-repair phase imaging There were three patients who had post-repair MRN imaging to address postoperative condition
changes or concerns. Two LN and one IAN and in each case, a neurorraphy with nerve allograft with
SC
protector/connector had been performed for Class IV injuries. In the LN cases, a trigger pain was
reported to have recurred at the surgical site. In the first case, the repair was performed 17 months
M AN U
prior and there was no trigger found and NST level was normal/S4 recovery was recorded. An MRN was conducted the same day and the findings were “postoperative changes (mild prominent) in right lingual nerve with no focal neuroma”. No interventions were recommended and none requested. In the second case, the neurorraphy with allograft repair was done 6 months previously and a trigger was present with moderate impairment/S2 recovery on NST. An MRN was conducted the same day and the findings were
TE D
“ovoid T2 hyperintensity along left lingual nerve consistent with neuroma”. A repeat neurorraphy was performed and a neuroma was found between the proximal and distal repairs (Figures 7) with atrophy of the distal nerve indicating the neuroma was in the previously placed allograft. The neuroma was
EP
removed and another allograft was placed. The single post-repair IAN case involved neuropathic pain and sensory deficit following the repositioning of the left IAN for placement of dental implants. A
AC C
neuroma was located within the mental nerve and removed and an allograft was placed at the initial nerve repair surgery. Four months postoperative, there was spreading numbness and pain. A non-OMS provider suggested either trigeminal neuralgia or recurrence of the neuroma due to “broken graft”. There was no trigger and NST was normal /S4. An MRN was performed reporting “postoperative changes of left alveolar nerve demonstrating continuity with distal nerve branches and no focal neuroma”. No surgical interventions were provided and NST remained normal/S4 and spreading symptoms stopped and all medications were discontinued.
ACCEPTED MANUSCRIPT
DISCUSSION Magnetic resonance imaging of the trigeminal nerve can map the peripheral IAN and LN nerves and
moderate to good agreement with NST and surgical findings.
RI PT
stratify nerve injuries providing clinically useful information about the presence or absence of PTN with
Radiologic assessment of disorders involving the trigeminal nerve has been discussed as potentially valuable in the preinjury, post-injury, and post-repair phase as an objective, noninvasive modality22.
SC
Early MRN studies used T1 weighted MR sequences in the 2D23 and 3D fast SPGR with fat suppression18
M AN U
and recently 3DAC-PROPELLER to show IAN morphologic changes described as endoneural fluid edema and connective tissue overgrowth (perineural fibrosis). Terumitsu et al confirmed their findings with intraoperative and histopathologic findings18,19. Chhabra et al24,25 applied 3D, steady state and diffusion based T2 weighted MRN techniques for high resolution demonstration of peripheral nerves. It was recently published that 1.5 T MRN examinations of the IAN and LN produces results that correlated with
TE D
clinical and surgical findings and there was a significant impact on the clinical management and moderate to excellent correlation with surgical findings20. Traumatic injuries (crush, laceration and transection) and non-traumatic injuries (neurofibroma,
EP
radiation exposure) of non-trigeminal peripheral nerves (peroneal, brachial plexus, sciatic) result in
AC C
increased signal on T2-weighted and inversion recovery pulse sequences at the sites of injury26. In rat sciatic nerve studies, MRN signal intensity correlated with loss of function after crush injury and normal signal return with functional recovery27 and crushed and cut nerve prevented from regenerating demonstrated increased T2 and inversion recovery nerve signal that were correlated with myelin degeneration26. Because there appears to be a direct correlation with peripheral nerve pathology and MRN T2 and diffusion signals, this study adds to the known literature supporting MRN in PTN, both traumatic and non-traumatic in the pre-injury, post-injury and post –repair phases. In this case series,
ACCEPTED MANUSCRIPT
3D PSIF sequences on MRN showed IAN and LN and its smaller terminal nerve branches (mental and lingual) in their entirety from corpus callosum to mentum/lip and tongue tip due to fat and vascular signal suppression and superior resolution (0.9mm isotropic). While 3D PSIF shows the nerve selectively,
RI PT
the 3D IR TSE does not suppress the venous signal28. Increased T2 SIR and caliber of the injured nerve on both of these sequences served as imaging markers for PTN as in other studies20,25,26
Non-traumatic injuries of the IAN and LN were generally associated with lower levels on NST and pain
SC
was a predominant symptom and the MRN findings ranged from normal to Sunderland Class II/III with perineural fibrosis, mild thickening, entrapment and in one case, compression by adjacent benign
M AN U
pathology. Traumatic injuries of the IAN and LN were generally associated with higher levels on NST and MRCS grading with sensory alteration with or without neuropathic pain were predominant symptoms and the MRN findings ranged from Sunderland Class II to V with no normal findings. These findings suggest that MRN can distinguish different degrees of trigeminal neuropathy ranging from
TE D
compression/entrapment in non-traumatic causations to compression/partial transection/transection and neuroma formation commonly found in traumatic injuries. MRN can also demonstrate normal IAN and LN anatomy and normal intermediate signal thereby supporting differential diagnosis from non-
EP
neural pain/sensory disorder conditions. None of the cases showed intracranial neural abnormality and the neural lesions on diffusion imaging were conspicuously similar to T2-weighted abnormalities with
AC C
good background signal suppression.
We demonstrated that MRN had moderate to good correlations with the NST levels/MRCS grading scores and the surgical findings. The Sunderland classification measured before surgery via NST (Table 4) matched with the surgical findings (n=26) in 58% of cases (15/26) and overestimated the surgical findings in 7% of cases (2/26) and underestimated the surgical findings in 35% of cases (9/26). The comparisons with MRN findings are similar when the two unclassified cases were not included. The
ACCEPTED MANUSCRIPT
MRN classification matched the surgical findings in 58% of cases (14/24) and overestimated the surgical findings in 4% of cases (1/24) and underestimated the findings in 37% of cases (9/24).
RI PT
In three cases, post-repair MRN provided valuable information about the status of the IAN and LN repairs which affected treatment decision and management. In one case, a recurrent neuroma was identified on MRN prompting surgical intervention verified at the time of surgery. In two post-repair
identified on MRN obviating the need for surgical intervention.
SC
scenarios, continuity of the nerve without neuroma formation following neurorraphy with allograft was
M AN U
This study has some limitations, including lack of blinding and controls. However, for the neuropathic cases, since the presentation was unilateral, the contralateral nerve caliber and signal served as an adequate internal control. The same surgeon performing the NST and clinical Sunderland classification and the surgery had access to the MRN findings before surgery and the radiologists had access to the surgeons’ clinical findings when the MRN was interpreted. In future studies, a larger sample in a
TE D
prospective fashion with the surgery and clinical examination blinded to the MRN study and inclusion of non-neuropathic cases will be done to address the above limitations. We did not analyze diffusion imaging parameters or tractography apart from routine analysis of signal abnormality on diffusion
EP
imaging. Finding fractional anisotropy and diffusion coefficient alterations in neuropathy can aid in
AC C
further quantitation of neuropathy and will be a subject of future study. This study demonstrated the current applications of MRN in the evaluation and management of nontraumatic and traumatic PTN. The strength of MRN in PTN management is the ability to provide noninvasive objective information of the IAN and LN that can distinguish normal from different degrees of neuropathy in the pre-injury, post-injury and post-repair phases. The correlation with surgical findings was moderate to good and if a randomized clinical trial can confirm this relationship with pre-surgical NST than new strategies in the management of PTN of the IAN and LN should lead to quicker detection
ACCEPTED MANUSCRIPT
of non-recoverable injuries earlier than the current protocols and without dependence on patient’s response to clinical stimuli during NST testing which should lead to improved outcomes of repair. Likewise, earlier detection of recoverable injuries should lead to risk reduction of unnecessary surgical
RI PT
intervention.
Acknowledgments
SC
Dr. Zuniga is a consultant for AxoGen Inc, Daiichi Sankyo/Charleston Labs and Merz USA, none
contributed to this manuscript. Dr. Chabbra is a consultant for ICON medical and holds book royalties
M AN U
with Jaypee, Wolters, none contributed to this manuscript. REFERENCES
1. Bagheri SC, Meyer RA, Khan HA et al: Retrospective measures of microsurgical repair of 222 lingual nerve injuries. J Oral Maxillofac Surg 68:715, 2010.
TE D
2. Brooks DN, Weber RV, Chao JD, et al: Processed nerve allografts for peripheral nerve reconstruction: A multicenter study of the utilization and outcome in sensory, mixed and motor nerve reconstructions. Microsurgery 32:1, 2012.
EP
3. Zuniga JR: Sensory outcomes after reconstruction of lingual and inferior alveolar nerve discontinuities using processed nerve allograft- A case series. J Oral Maxillofac Surg, 73:734,
AC C
2015.
4. Zuniga JR, Essick G: A contemporary approach to the clinical evaluation of trigeminal nerve injuries. OMFS Clin North Amer, 4:353, 1992.
5. Zuniga JR, Meyer RA, Gregg JM, et al: The accuracy of clinical neurosensory testing for nerve injury diagnosis. J Oral Maxillofac Surg, 56:2, 1998 6. Bagheri SC, Meyer RA, Cho SH et al: Microsurgical repair of the inferior alveolar nerve: success rate and factors that adversely affect outcome. J Oral Maxillofacial Surg 70: 1978, 2012.
ACCEPTED MANUSCRIPT
7. Mackinnon SE, Dellon AL: Results of Nerve Repair and Grafting. Surgery of Peripheral Nerve, 1st ed. New York: Thieme Medical, 1988. 8. Dodson TB, Kaban LB: Recommendations for management of trigeminal nerve defects based on a
RI PT
critical appraisal of the literature. J Oral Maxillofac Surg 55:1380, 1997.
9. Robinson PP. Observations on the recovery of sensation following inferior alveolar nerve injuries. Br J Oral Maxillofac Surg 26:177, 1988
SC
10. Chhabra A, Thawait GK, Doldatos T, et al: High-resolution 3T MR neurography of the brachial plexus and its branches, with emphasis on 3D imaging. Am J Neuroradiol, 34:486, 2013
M AN U
11. Baumer P, Weiler M, Nedszus M, et al: Somatotopic fascicular organization of the human sciatic nerve demonstrated by MR neurography. Neurology 84:1782, 2015 12. Cho Sims G, Boothe E, Joodi R, et al: 3D MR Neurography of the lumbosacral plexus: obtaining optimal images for selective longitudinal nerve depiction. Am J Neuroradiol 37:2158, 2016
TE D
13. Manoliu A, Ho M, Nanz D, et al: MR Neurographic orthopantomogram: Ultrashort echo-time imaging of mandibular bone and teeth complimented with high-resolution morphological and functional MR Neurography. J Magn Reson Imaging 44: 393, 2016
EP
14. Bendszus M, Wessig C, Solymosi L, et al: MRI of peripheral nerve degeneration and regeneration: correlation with electrophysiology and histology. Exp Neurol 188: 171, 2004
AC C
15. Li X, Shen J, Chen J, et al: Magnetic resonance imaging evaluation of acute crush injury of rabbit sciatic nerve: correlation with histology. Can Assoc Radiol J 59: 123, 2008
16. Liu V, Li J, Butzkeuven H, et al: Microstructural abnormalities in the trigeminal nerves of patients with trigeminal neuralgia revealed by multiple diffusion metrics. Eur J Radiol 82:783, 2013
17. Shen J, Zhou CP, Zhong XM, et al: MR Neurography: T1 and T2 measurements in acute peripheral nerve traction injury in rabbits. Radiology 254:729, 2010
ACCEPTED MANUSCRIPT
18. Terumitsu M, Seo K, Matsuzawa H, et al: Morphologic evaluation of the inferior alveolar nerve in patients with sensory disorders by high-resolution 3D volume rendering magnetic resonance neurography on a 3.0-T system. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 111:95, 2011
RI PT
19. Terumitsu M, Hatsuzawa H, Seo K, et al: High-contrast high-resolution imaging of posttraumatic mandibular nerve by 3DAC-PROPELLER magnetic resonance imaging: correlation with the severity of sensory disturbance. Oral Surg Oral Med Oral Pathol Oral Radiol 124:85, 2017.
SC
20. Cox, B, Zuniga JR, Panchal N, et al: Magnetic Resonance Neurography in the Management of Peripheral Trigeminal Neuropathy: experience in a tertiary care center. Eur Radiol, 10:3392, 2016
M AN U
21. Sunderland S: A classification of peripheral nerve injuries producing loss of function. Brain 74:491, 1951.
22. Miloro, M: Radiologic Assessment of the trigeminal nerve. Oral Maxillofac Surg Clin North Amer. 13:315, 2001
TE D
23. Filler AG, Kliot M, Hayes CE, et al: Application of magnetic resonance neurography in the evaluation of patients with peripheral nerve pathology. J Neurosurg 85: 299, 1996 24. Chhabra A, Soldatos T, Subhawong TK, et al: The application of three-dimensional diffusion –
EP
weighted PSIF technique in peripheral nerve imaging of the distal extremities. J Magn Reson Imaging. 34:962, 2011.
AC C
25. Chhabra A, Carrino J: Current MR neurography techniques and whole-body MR neurography. Semin Musculoskeletal Radiol. 19:79, 2015.
26. Grant GA, Britz GW, Goodkin R, et al: The utility of magnetic resonance imaging in evaluating peripheral nerve disorders. Muscle Nerve 25: 314, 2002.
27. Cudlip SA, Howe FA, Phil D, et al: Magnetic resonance neurography of peripheral nerve following experimental crush injury and correlation with functional deficit. J Neurosurg 96:755, 2002.
ACCEPTED MANUSCRIPT
28. Chhabra A, Flammang A, Padua A, et al: Magnetic resonance neurography: technical
RI PT
conisderations. Neuroimaging Clin N Am 24:67, 2014.
FIGURE LEGEND
SC
Figure 1: Coronal 3D PSIF MIP image shows the diffusely prominent and hyperintense left IAN (large arrows) as compared to normal intermediate signal of right IAN (small arrows). Corresponding axial
M AN U
DTI image (b=600 map) confirms the conspicuous abnormal enlargement and signal alteration of the left IAN.
Figure 2: Intraoperative high magnified photo of the left IAN resting on a neuropatty platform with suction catheter. There is continuity of the IAN but there is intense erythema of the IAN proper in
TE D
the area near tooth #18 (small arrow) while the distal continuation is normal (large arrow), with a clear border seen between abnormal and normal Figure 3: Coronal view of MRN 3D MIP of the right and left LN and IAN in a patient with severe
EP
sensory impairment/S1 of the left tongue with a trigger and dystrophic ageusia. The right LN (small arrows) and right and left IAN are shown and are normal in size and location. The double arrows
AC C
point out an abnormal left LN with enlargement of the caliber size ending in a focal mass of hyperintense tissue consistent with a neuroma. Figure 4: Intraoperative high magnified photo of the left LN demonstrated in Figure 3 resting on a neuropatty platform with suction catheter. The double arrows point to abnormal neural tissue which was consistent with a neuroma in continuity with enlargement of the proximal nerve tissue in the field posterior and lateral to the lower second molar.
ACCEPTED MANUSCRIPT
Figure 5: Coronal view of MRN right and left IAN and LN in a patient with moderate impairment/S2 with neuropathic pain of the left IAN 5 months after dental implant placement. Coronal 3D PSIF MIP
normal intermediate signal and caliber of right IAN (small arrows).
RI PT
image shows the mildly diffusely prominent and hyperintense left IAN (large arrows) as compared to
Figure 6A: Intraoperative photo of the left IAN exposed via a lateral cortical window showing the base of the dental implant (small arrow) in proximity to the IAN (large arrow) with thickening of the
SC
nerve at the contact site with no discontinuity or neuroma formation
M AN U
Figure 6B: Intraoperative high magnified photo of the right IAN resting on a neuropatty platform at the point of contact with the dental implant showing internal fibrosis (arrow). Figure 7: Intraoperative high magnified photo of the right LN resting on a neuropatty platform showing a large neuroma in continuity (large arrow) within an allograft placed 6 months prior with atrophy of the distal nerve. Retained microsuture is seen proximal and distal to the neuroma (small
TE D
arrows) indicating the neuroma formation was within the allograft and not due to disruption of the
AC C
EP
prior repair. These findings corroborate the MRN findings.
RI PT
ACCEPTED MANUSCRIPT
Table 1. Sunderland Nerve Classification by Clinical NST/MRCS Grade, Surgical findings and MRN Imaging Clinical NST level and MRCS Grade Description
Surgical by Direct inspection
MR Neurography
I
Normal (4)/ S3+ or S4 by 3 months
Anatomic: Homogenous mild increased T2 signal of nerve
II
Normal (4)/ S3+ or S4 by 6 months
III
Mild (3) or Moderate (2)/S2, S2+, S3 by 6 months or more
IV
Moderate (2) or Severe (1)/S1, S2, S2+ by 6 months or more
Intact with no internal or external fibrosis, normal mobility and neuroarchitecture (visualized fascicles and fanconi bands) Intact with no internal fibrosis. External fibrosis, restricted, mobility but neuroarchitecture intact (visualized fascicles and fanconi bands once external scar removed) Intact with both internal and external fibrosis, restricted mobility and disturbance of neuroarchitecture (abnormal fascicle patterns and/or fanconi bands not visible) Partial transected nerve but some amount of distal nerve present with or without neuroma in continuity
M AN U
TE D
EP
Severe (1) or Complete (0)/S0, S1 by 6 months or more
AC C
V
SC
Sunderland Classification
Complete transected nerve with or without amputation neuroma
Anatomic: Homogenous increased T2 signal of nerve and mild nerve thickening or constriction. Perineural fibrosis
Anatomic: Homogenous increased T2 signal of nerve and moderate thickening or constriction. Perineural fibrosis
Anatomic: Heterogeneous T2 signal of nerve and neuroma in continuity. Perineural and intraneural fibrosis. Anatomic: Discontinuous nerve with end-bulb neuroma
RI PT
ACCEPTED MANUSCRIPT
Table 2. Distribution of Traumatic Peripheral Trigeminal Neuropathies by Nerve and Etiology Third Molar
Dental Implant
Benign Pathology
Orthognathic
IAN
10
11
7
1
LN
18
0
SC
Nerve
M AN U
0
0
Table 3. Distribution of Non-Traumatic Peripheral Trigeminal Neuropathies by Nerve and Etiology Dental Injection
Non-Invasive Endodontics
Unknown*
IAN
1
4
6
LN
1
0
0
TE D
Nerve
AC C
EP
*complaint within the correct nerve distribution but the event unknown or beyond/not involving correct distribution of nerve
RI PT
ACCEPTED MANUSCRIPT
Table 4. Traumatic Trigeminal Neuropathies by Sunderland Classification on NST, MRN and Surgical
NST
MRN
Right LN
IV
IV
Left IAN
III
II
Left IAN
II
Right IAN
V
Left IAN
II
Right LN
IV
M AN U
Injured Nerve
Surgery
SC
Findings in Case Series
IV II II
Unclassified
II
II
II
IV
IV
III
III
III
III
II/III
III
IV/V
unclassified
IV
IV/V
IV
IV
III
III
III
Right LN
II/III
II/III
III
Right IAN
II
II
II
Left LN
V
V
V
Left LN
IV
IV
V
Left LN
IV/V
IV/V
IV
Left IAN Right LN Right LN
AC C
Right IAN
EP
Right LN
TE D
II
ACCEPTED MANUSCRIPT
II
III/IV
IV
Right LN
II/III
II/III
III
Right LN
IV/V
IV
V
Left LN
IV/V
IV
V
Left IAN
II
II
Left LN
IV
V
Left LN
IV
IV
Left LN
IV
IV
Left LN
III/IV
IV
Right IAN
II/III
III
IV
SC
IV
M AN U
TE D EP AC C
RI PT
Right IAN
II/III
IV V III/IV
RI PT
ACCEPTED MANUSCRIPT
Table 5. The Distribution of Correct/Incorrect Sunderland Classification by NST and MRN for Nerve injury Found at Surgery
5
NST MRN %Correct/Incorrect %Correct/Incorrect 20% 20%
LN
IV
8
100%
LN
III
3
LN
II
0
IAN
V
0
IAN
IV
IAN
III
IAN
II
M AN U
SC
N
LN
Sunderland Classification V
100% (*one unclassified)
33%
33%
n/a
n/a
n/a
n/a
1
0%
0%
4
50%
50%
5
60%
100% (*one unclassified)
TE D
AC C
EP
Nerve Injured
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Ms. Ref. No.: JOMS-D-17-01110 Title: Magnetic Resonance Neurography of Traumatic and Non-Traumatic Peripheral Trigeminal Neuropathies
RI PT
REVISION COVER LETTER
SC
Carmen Hupp JOMS Editorial Office Journal of Oral and Maxillofacial Surgery
AC C
EP
TE D
M AN U
Thank you very much for the opportunity to submit revisions to the article noted above for publication in JOMS. My co-authors and I agree that the recommended comments were necessary to improve the quality of the submission. The changes were identified in blue color. We have addressed the comments as follows: 1. Please cite all references in numerical order in the text of the manuscript (Reference #9 appears to be cited before #8) a. Reference #9 and #8 were corrected in the text and in the Reference section in the correct numerical order as recommended. All of the references were checked and found to be in correct numerical order 2. Please cite all tables in numerical order in the text of the manuscript (Table 6 does not appear to be cited) a. Table 6 was found to be incorrectly numbered in the original submission and should have been Table 5 which is correctly referenced to in Results/subsection Classification of traumatic nerve injuries on NST, MRN and Surgery, first paragraph. Table 6 in the revision is labeled Table 5. 3. Please add a disclosure statement at the end of the manuscript for Drs. Zuniga and Chhabra (as noted on the AAOMS Disclosure Form) a. Drs. Zuniga and Chhabra consultant and book royalty disclosures were noted at the end of the manuscript under ACKNOWLEDGMENTS. Note that there was no contribution by the disclosed activities to the study conducted or its’ publication. Thank you for the opportunity to submit the revision and look forward to its’ acceptance and publication. Sincerely
John R. Zuniga DMD, MS, PhD Robert V. Walker DDS Chair in Oral and Maxillofacial Surgery Professor, Departments of Surgery and Neurology & Neurotherapeutics University of Texas Southwestern
S0278-2391(17)31426-X
DOI:
10.1016/j.joms.2017.11.007
Reference:
YJOMS 58043
To appear in:
Journal of Oral and Maxillofacial Surgery
Received Date: 13 September 2017 Revised Date:
8 November 2017
Accepted Date: 8 November 2017
Please cite this article as: Zuniga JR, Walker RV, Mistry C, Tikhonov I, Dessouky R, Chhabra A, Magnetic Resonance Neurography of Traumatic and Non-Traumatic Peripheral Trigeminal Neuropathies, Journal of Oral and Maxillofacial Surgery (2017), doi: 10.1016/j.joms.2017.11.007. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
RI PT
ACCEPTED MANUSCRIPT
M AN U
SC
MAGNETIC RESONANCE NEUROGRAPHY OF TRAUMATIC AND NON-TRAUMATIC PERIPHERAL TRIGEMINAL NEUROPATHIES
John R. Zuniga DMD, MS, PhD, Robert V. Walker DDS Chair in Oral and Maxillofacial Surgery, Professor, Departments of Surgery and Neurology & Neurotherapeutics, University of Texas Southwestern Medical Center at Dallas, Texas Cyrus Mistry DDS, MD, Chief Resident, Oral and Maxillofacial Surgery, University of Texas Southwestern Medical Center at Dallas, Texas
TE D
Igor Tikhonov DDS, MD, Chief Resident, Oral and Maxillofacial Surgery, University of Texas Southwestern Medical Center at Dallas, Texas Riham Dessouky MD, Fellow, Radiology, University of Texas Southwestern Medical Center at Dallas, Texas
EP
Avneesh Chhabra MD, Associate Professor, Radiology & Orthopedic Surgery University of Texas Southwestern Medical Center at Dallas, Texas
All Correspondence to:
AC C
John R. Zuniga DMD, MS, PhD Robert V. Walker DDS Chair in Oral and Maxillofacial Surgery Professor, Departments of Surgery and Neurology & Neurotherapeutics University of Texas Southwestern Medical Center at Dallas Dallas, Texas Email: [email protected]
ACCEPTED MANUSCRIPT
ABSTRACT Purpose: The clinical neurosensory testing (NST) is currently the gold-standard for the diagnosis of traumatic and non-traumatic peripheral trigeminal neuropathies (PTN), but exhibits both false positive and negative results when
RI PT
compared to surgical findings and frequently delays treatment decisions. We tested the hypothesis that magnetic resonance neurography (MRN) of PTN can serve as a diagnostic modality by correlating NST, MRN and surgical findings.
SC
Materials and Methods: Sixty patients with traumatic and non-traumatic PTN of varying etiologies and Sunderland classifications underwent NST followed by MRN on 1.5T and 3.0 T scanners. The protocol included 2D and 3D imaging, including diffusion imaging and isotropic 3D PSIF. The MRN findings were read by two readers in
M AN U
consensus of clinical findings but blinded to the side of abnormality. The MRN results were summarized using Sunderland Classification. In 26 patients, surgery was performed and Sunderland classification was assigned based on surgical photos. Agreement between MRN and NST/Surgical classification was evaluated using kappa statistics. Pearson’s Correlation Coefficient (PCC) was used to assess the correlation between continuous measurements of
TE D
MRN/NST and surgical classification.
Results: Nineteen males and 41 females, mean age 41, ranging 12 to 75, with 54 complaints of altered sensation of the lip/chin/or tongue, including 16 with neuropathic pain and 4 with no neurosensory complaint were included.
EP
Third molar surgery (n=29) represented the most common cause of traumatic PTN. Assuming one nerve abnormality per patient, the lower class was accepted, a kappa of 0.57 was observed between MRN and NST
AC C
classification. A kappa of 0.5 existed between MRN and surgical findings with a PCC of 0.67. Conclusions: MRN anatomically maps PTN and stratifies the nerve injury and neuropathies with moderate to good agreement with NST and surgical findings for clinical use.
ACCEPTED MANUSCRIPT
INTRODUCTION Peripheral trigeminal neuropathies (PTN) include the loss of general sense, special sense and/or development of neuropathic pain in the oral and facial distributions of the trigeminal nerve and are well
RI PT
recognized outcomes of traumatic (i.e., third molar removal, implant placement, etc.) and non-traumatic (i.e., neoplastic, non-invasive endodontic, atypical facial pain, dental injection, etc.) causations. The responsibility to each patient is to provide a timely diagnosis, reasonable prognosis, accurate monitoring
SC
of recovery or lack of and evidence based treatment recommendation for the symptoms of the
trigeminal neuropathy. This is no more important than for trigeminal neuropathies that would benefit
M AN U
by microsurgical repair since earlier intervention is clearly related to better outcome1,2,3 and thus, early diagnosis would be preferred. The data required to accomplish these tasks has been relegated to clinical chairside neurosensory tests (NST). For diagnostic purposes the three-level drop-out algorithm was published in 19924 , which currently serves as the gold standard to diagnose PTN after its statistical power was related to surgical findings in 19985. For accurate monitoring of recovery and outcome of
TE D
surgical repair, the Medical Research Council Scale (MRCS) for sensory recovery is currently the gold standard to identify functional sensory recovery6,7. Both NST and MRCS rely on patient response to stimulus and operator experience in stimulus presentation and interpretation6,8. In addition, the results
EP
from these tests are only accurate when delivered 1 month after the injury since they cannot distinguish
AC C
levels of injury before 1 month. When the three-level drop-out algorithm was corroborated with surgical findings, this NST was shown to exhibit high positive (PPV=95%) and negative (NPV=100%) predictive values for lingual nerve (LN) injuries and moderate PPV (77%) and NPV (60%) for inferior alveolar nerve (IAN) injuries. The NPV of 60% means that NST is less efficient in ruling out an IAN injury resulting in a false negative rate of 40%. In addition, for both IAN and LN, at higher sensory impairment scores, the NST tends to overestimate the degree of nerve injury and, at lower sensory impairment scores, NST tends to underestimate the degree of nerve injury. Furthermore, patients with different
ACCEPTED MANUSCRIPT
degrees of nerve injuries may show similar NST scores and these scores can vary with differences in age, duration and cause of injury9. Thus, although clinically useful in providing diagnosis, prognosis, monitoring and treatment options for patients with PTN, the NST clearly suffers from the following: 1.
RI PT
Delays treatment of higher class injury patients who would likely benefit by earlier intervention; 2. Does not adequately delineate the anatomy and location of the injury; and 3. Can overestimate and underestimate the level of injury, especially if confounding factors are present.
SC
Magnetic resonance neurography (MRN) is an imaging dedicated to the peripheral nerves and provides a non-invasive foot-print of neural anatomy and resolves the intraneural architecture in multiple
M AN U
orthogonal planes10,11,12,13. On MRN of peripheral nerves, neuropathies demonstrate imaging alterations of nerve caliber, intraneural T2 signal intensity ratio (SIR) and altered diffusion characteristics, such as increased apparent diffusion coefficient (ADC) and decreased fractional anisotropy (FA) that correlate with axonal degeneration and demyelination14,15,16,17. In initial studies, MRN of injured and non-injured
TE D
PTN was shown to improve diagnostic information, thereby impacting clinical management and there was moderate to excellent correlation with intraoperative findings18,19,20. The hypotheses that drove this study were two- fold. First, MRN can differentiate normal from
EP
abnormal-injured and non-injured peripheral trigeminal nerves. Second, MRN nerve injury
AC C
classifications based on Sunderland grading system21 correlates with NST grading and surgical findings.
MATERIALS AND METHODS
This was a retrospective evaluation of a case series which included 60 patients from 3/2015 to 5/2017 with traumatic and non-traumatic PTN of varying etiologies and Sunderland classifications who
ACCEPTED MANUSCRIPT
underwent MRN evaluation in addition to clinical NST and, in some cases, trigeminal nerve surgery. This study was conducted with approval from the UTSW institutional review board (IRB) following Health
RI PT
Insurance Portability and Accountability (HIPPA) guidelines. Informed consent was waived. Patient Evaluations
Inclusion criteria for data analysis included those patients seen in the Division of Oral and Maxillofacial
SC
Surgery with a neurosensory complaint who underwent clinical NST and agreed to MRN evaluation for a suspected PTN in the second or third division of the trigeminal nerve. Patient age, gender, trigeminal
M AN U
nerve(s) injury by location and laterality, chief complaint, sensory symptoms, pain descriptions (when present), mechanism of injury, and previous treatment were recorded. For purposes of testing the hypothesis, there were two groups of patients examined, the first were those who had known traumatic injuries of the trigeminal nerve (i.e., third molar removal, dental implant placement, tumor resection, facial trauma) and the second, those who did not have known traumatic injuries (i.e., dental injection,
TE D
non-invasive endodontic treatment, atypical facial pain, neoplastic). Clinical NST was conducted to determine the level of sensory impairment defined by none, mild, moderate, severe and complete previously described4 and functional sensory grading by Grade S0, S1,
EP
S2, S2+, S3, S3+, and S4 on the Medical Research Council Scale (MRCS) as previously described6,7,12 . The
AC C
clinical NST was performed by one examiner (JRZ) prior to MRN evaluation and any surgical intervention. The results of level of sensory impairment NST were scored 4 (normal), 3 (mild impairment), 2 (moderate impairment), 1 (severe impairment) and 0 (complete anesthesia) while the MRCS grading was scored by numerical value in parallel from 0 (no sensation) to 4 (superficial pain and touch without hyperesthesia and static 2-point discrimination of 2-6mm indicating complete sensation). A Sunderland classification was provided from this data from Class I (S3+, S4/ NST normal (4) by 3 months), Class II (S3+, S4/NST normal (4) by 6 months), Class III (S2, S2+, S3/ NST mild (3) or moderate (2) by or greater
ACCEPTED MANUSCRIPT
than 6 months), Class IV (S1, S2, S2+/NST moderate (2) severe (1) and Class V (S0, S1/NST complete (0) by or greater than 6 months) as described in Table 1.
RI PT
When trigeminal nerve surgery was performed, a single surgeon (JRZ) performed all the surgeries. Surgical findings were recorded by consensus by the surgeons after direct microscopic inspection of the injury site using high (40x) magnification. Operative photographs were taken twice of the same site at high and low magnification by someone other than the operating surgeon. Calipers or a ruler was used
SC
to determine nerve thickness of the proximal and distal ends when a transection injury was required as well as the size of the gap in millimeters, and the size of neuroma when applicable. Based on intra-
M AN U
operative inspection described above the Sunderland classification of the injury was determined by the surgeon and research team in consensus as outline in Table 1. Magnetic Resonance Neurography Evaluations
All MRN studies were performed without contrast for a total imaging time of 50 minutes in supine
TE D
position with a 1.5 T (Siemens, Avanto, Erlangen, Germany) or 45 minutes on a 3T (Achieva, Ingenia, Philips, Best, Netherlands) scanner with a multichannel head coil placed over the head and neck region. The protocol included 2D (dimensional) and 3D T1W, T2 SPAIR (spectral adiabatic inversion recovery)
EP
imaging, diffusion imaging and isotropic 3D PSIF (0.9 mm voxel) as well 3D inversion recovery turbo spin
AC C
echo (3D IR TSE). Thick slab maximum intensity projections (MIP) were obtained in nerve longitudinal planes (coronal and oblique sagittal) on independent workstations (Aquarius, Tera Recon, FosterCity, CA, USA). Perineural fat planes were assessed for presence of scarring or mass lesion. Three-dimensional CISS (constructive interference in steady state) sequence was employed in the protocol to exclude any intracranial neural pathology and the axial T1W and T2 SPAIR sequences were extended to the corpus callosum to the chin to exclude any intranuclear or adjacent intra-cranial pathology. The protocol sequence parameters were as follows: coronal 3D PSIF (0.8-0.9mm isotropic; time of acquisition, TA-6
ACCEPTED MANUSCRIPT
minutes), coronal 3D IR TSE (1.5 mm isotropic; TA- 6 to 7 minutes), axial diffusion tensor imaging, DTI (b50,800,1000; 2mm isotropic, 15 directions of interrogation; TA-8 minutes), axial T1W (3.5 mm; TA – 4 to 5 minutes), and axial T2 SPAIR (3.5mm: TA – 4 to 5 minutes). The MRN findings of nerve signal and
RI PT
caliber alteration were read by two musculoskeletal radiologists in consensus in the light of the clinical findings but blinded to the side of abnormality as part of routine patient care. A Sunderland
classification was given by the radiologists blinded to the NST grading based on the qualitative imaging
SC
criteria described in Table 1.
M AN U
Statistical Analysis
Descriptive statistics were used to compare the demographic data, and Sunderland classification on NST and MRN between traumatic and non-traumatic patient populations. In the traumatic patient population, Kappa analysis was performed to test correlations of Sunderland classification on independently performed NST, MRN and surgical classifications. Pearson’s correlation coefficient was
TE D
used to determine the association of differences in continuous measurements in abnormal versus normal nerves of MRN/NST and surgical findings. Agreement was classified as excellent (>0.80), good
RESULTS
EP
(0.61-0.80), moderate (0.41- 0.60), fair (0.20- 0.40) and poor (<0.20). Type 1 error was set at 0.05.
AC C
Patient population and NST:
This case series was composed of 60 patients with trigeminal injuries involving the inferior alveolar nerve (IAN) and lingual nerve (LN) who underwent clinical NST and MRN examination between 3/2015 to 5/2017, and microsurgery on 26 patients between 4/2015 and 5/2017. There were 19 males, 41 females with a mean age of 43 ± 15 years, median of 39 and range from 12 to 75 years of age. The chief complaint was numbness in 25 cases, pain in 16, diminished or lost taste in 6, abnormal diminished sensation in 9, suspected nerve injury in 2 and pre-procedural check-up in 2. The distribution of cases
ACCEPTED MANUSCRIPT
by mechanism of injury showed that third molar extraction was the common source in 47% of patients (n=28) followed by dental implant 18% (n=11), mandibular pathology surgery 13% (n=8), unknown 10% (n=6) while the remaining mechanisms represented less than 1% incidence per group. By nerve
RI PT
distribution, there were 40 patients with IAN sensory disorders and 20 LN disorders. For purposes of further discussion, there were 48 patients who had traumatic injuries to the IAN and LN (Table 2) and 12 with non-traumatic injuries to the IAN and LN (Table 3). Traumatic injuries are listed in Table 2 under
SC
third molar extraction, dental implant, mandibular surgery and orthognathic surgery. Non-traumatic injuries included three categories of etiologies, they were: dental injection, non-invasive endodontic
M AN U
treatment, and unknown IAN sensory complaints (3 with no known event, 2 incisional biopsy of mucosa and 1 maxillary routine dental extraction). Within the traumatic injury population, there were differences between the IAN and LN in NST level of impairment with 79% LN injuries grade S0 to S2 and NST level IV or more (15/19) whereas 79% IAN injuries grade S3 to S4 and NST level III or less (22/28).
TE D
Neuropathic pain was present in the clinical and NST examination in 45% of IAN (13/28) and 15% of LN (3/19) and the duration of injury to clinical examination reported was 13.61 months (range 0 to 49) for IAN and 5.5 months (range 1 to 45) for LN. Within the non-traumatic injury population, which includes
EP
11 IAN and 1 LN injury, the distribution was Grade 3 to 4 in 100% (12/12) and level III or less in 100% (12/12). Neuropathic pain was present in the clinical and NST examination in 50% (6/12) and the
AC C
duration of injury to clinical examination reported was 4.87 months (range 0.5 to 24). An example of a non-traumatic PTN event is of a 64 y/o female with moderate sensory impairment/S2 and neuropathic pain (allodynia, hyperpathia and burning pain) 2 months following traditional endodontic procedure on tooth #18 for chronic pulpitis. Glass ionomer and GP pulp canal sealer that was irrigated with sodium hypochlorite was reported to have been performed for the treatment of tooth #18. Figure 1 shows a coronal MRN vessel-suppressed view of the right and left inferior alveolar nerve in this case and an axial DTI image showing hyperintensity and enlargement of the left IAN in the area near tooth #18. Figure 2
ACCEPTED MANUSCRIPT
shows the surgical findings demonstrating an intact nerve with intense inflammatory changes. Classification of traumatic nerve injuries on NST, MRN and Surgery:
RI PT
Twenty-six patients underwent microdissection, 10 IAN and 16 LN. Of these 26, 1 patient had a nontraumatic category injury to the IAN associated with endodontic therapy with a NST level of 2 and MRCS grade S3 with neuropathic pain. Table 4 shows the Sunderland classification for each patient based on NST, MRN and surgery. Figure 3 shows coronal view of 3D MRN of the right and left lingual nerve of a
SC
patient with severe sensory impairment/S1 of the left tongue with a trigger and dystrophic ageusia on the same side. Figure 4 shows the surgical findings of a large neuroma of the left lingual nerve. Figure 5
M AN U
shows a coronal 3D MRN of the right and left IAN and right and left LN in 5 months after dental implant placed posterior left mandible with moderate impairment/S2 with neuropathic pain (hyperalgesia). The MRN findings were described as left IAN Class II injury with mild thickening and hyperintense signaling with perineural fibrosis with no discontinuity or neuroma identification. Figure 6A shows the IAN in
TE D
relationship to the dental implant with no discontinuity or neuroma formation. Figure 6B shows higher magnification of the point of injury showing internal fibrosis. There were 6 indeterminate classifications using NST, 5 MRN and 1 surgical where the examiner, radiologist and surgeon could not decide between
EP
class injuries. Assuming one nerve abnormality per patient, and when classification was undetermined, the lower class was accepted. In 2 cases, the MRN could not classify the nerve injury due to artifact
AC C
and/or motion. The distribution of Sunderland classification by nerve and their ratio of correct to incorrect NST and MRN compared to surgical findings is presented in Table 5. There were more Class IV and V injuries in LN injuries than IAN and more Class II and III in IAN injuries than LN as has been previously shown5. For a LN Class IV injury the NST and MRN were accurate in predicting surgical findings and for IAN Class II injuries MRN was more accurate than NST. Except for two exceptions, NST overestimated the class injury in Class II IAN injuries, both NST and MRN underestimated the surgical findings. This study showed kappa of 0.57 and 0.5 between MRN and NST classifications and MRN and
ACCEPTED MANUSCRIPT
surgical classifications. The Pearsons’ Correlation Coefficient was 0.67 between MRN and surgical findings.
RI PT
Post-repair phase imaging There were three patients who had post-repair MRN imaging to address postoperative condition
changes or concerns. Two LN and one IAN and in each case, a neurorraphy with nerve allograft with
SC
protector/connector had been performed for Class IV injuries. In the LN cases, a trigger pain was
reported to have recurred at the surgical site. In the first case, the repair was performed 17 months
M AN U
prior and there was no trigger found and NST level was normal/S4 recovery was recorded. An MRN was conducted the same day and the findings were “postoperative changes (mild prominent) in right lingual nerve with no focal neuroma”. No interventions were recommended and none requested. In the second case, the neurorraphy with allograft repair was done 6 months previously and a trigger was present with moderate impairment/S2 recovery on NST. An MRN was conducted the same day and the findings were
TE D
“ovoid T2 hyperintensity along left lingual nerve consistent with neuroma”. A repeat neurorraphy was performed and a neuroma was found between the proximal and distal repairs (Figures 7) with atrophy of the distal nerve indicating the neuroma was in the previously placed allograft. The neuroma was
EP
removed and another allograft was placed. The single post-repair IAN case involved neuropathic pain and sensory deficit following the repositioning of the left IAN for placement of dental implants. A
AC C
neuroma was located within the mental nerve and removed and an allograft was placed at the initial nerve repair surgery. Four months postoperative, there was spreading numbness and pain. A non-OMS provider suggested either trigeminal neuralgia or recurrence of the neuroma due to “broken graft”. There was no trigger and NST was normal /S4. An MRN was performed reporting “postoperative changes of left alveolar nerve demonstrating continuity with distal nerve branches and no focal neuroma”. No surgical interventions were provided and NST remained normal/S4 and spreading symptoms stopped and all medications were discontinued.
ACCEPTED MANUSCRIPT
DISCUSSION Magnetic resonance imaging of the trigeminal nerve can map the peripheral IAN and LN nerves and
moderate to good agreement with NST and surgical findings.
RI PT
stratify nerve injuries providing clinically useful information about the presence or absence of PTN with
Radiologic assessment of disorders involving the trigeminal nerve has been discussed as potentially valuable in the preinjury, post-injury, and post-repair phase as an objective, noninvasive modality22.
SC
Early MRN studies used T1 weighted MR sequences in the 2D23 and 3D fast SPGR with fat suppression18
M AN U
and recently 3DAC-PROPELLER to show IAN morphologic changes described as endoneural fluid edema and connective tissue overgrowth (perineural fibrosis). Terumitsu et al confirmed their findings with intraoperative and histopathologic findings18,19. Chhabra et al24,25 applied 3D, steady state and diffusion based T2 weighted MRN techniques for high resolution demonstration of peripheral nerves. It was recently published that 1.5 T MRN examinations of the IAN and LN produces results that correlated with
TE D
clinical and surgical findings and there was a significant impact on the clinical management and moderate to excellent correlation with surgical findings20. Traumatic injuries (crush, laceration and transection) and non-traumatic injuries (neurofibroma,
EP
radiation exposure) of non-trigeminal peripheral nerves (peroneal, brachial plexus, sciatic) result in
AC C
increased signal on T2-weighted and inversion recovery pulse sequences at the sites of injury26. In rat sciatic nerve studies, MRN signal intensity correlated with loss of function after crush injury and normal signal return with functional recovery27 and crushed and cut nerve prevented from regenerating demonstrated increased T2 and inversion recovery nerve signal that were correlated with myelin degeneration26. Because there appears to be a direct correlation with peripheral nerve pathology and MRN T2 and diffusion signals, this study adds to the known literature supporting MRN in PTN, both traumatic and non-traumatic in the pre-injury, post-injury and post –repair phases. In this case series,
ACCEPTED MANUSCRIPT
3D PSIF sequences on MRN showed IAN and LN and its smaller terminal nerve branches (mental and lingual) in their entirety from corpus callosum to mentum/lip and tongue tip due to fat and vascular signal suppression and superior resolution (0.9mm isotropic). While 3D PSIF shows the nerve selectively,
RI PT
the 3D IR TSE does not suppress the venous signal28. Increased T2 SIR and caliber of the injured nerve on both of these sequences served as imaging markers for PTN as in other studies20,25,26
Non-traumatic injuries of the IAN and LN were generally associated with lower levels on NST and pain
SC
was a predominant symptom and the MRN findings ranged from normal to Sunderland Class II/III with perineural fibrosis, mild thickening, entrapment and in one case, compression by adjacent benign
M AN U
pathology. Traumatic injuries of the IAN and LN were generally associated with higher levels on NST and MRCS grading with sensory alteration with or without neuropathic pain were predominant symptoms and the MRN findings ranged from Sunderland Class II to V with no normal findings. These findings suggest that MRN can distinguish different degrees of trigeminal neuropathy ranging from
TE D
compression/entrapment in non-traumatic causations to compression/partial transection/transection and neuroma formation commonly found in traumatic injuries. MRN can also demonstrate normal IAN and LN anatomy and normal intermediate signal thereby supporting differential diagnosis from non-
EP
neural pain/sensory disorder conditions. None of the cases showed intracranial neural abnormality and the neural lesions on diffusion imaging were conspicuously similar to T2-weighted abnormalities with
AC C
good background signal suppression.
We demonstrated that MRN had moderate to good correlations with the NST levels/MRCS grading scores and the surgical findings. The Sunderland classification measured before surgery via NST (Table 4) matched with the surgical findings (n=26) in 58% of cases (15/26) and overestimated the surgical findings in 7% of cases (2/26) and underestimated the surgical findings in 35% of cases (9/26). The comparisons with MRN findings are similar when the two unclassified cases were not included. The
ACCEPTED MANUSCRIPT
MRN classification matched the surgical findings in 58% of cases (14/24) and overestimated the surgical findings in 4% of cases (1/24) and underestimated the findings in 37% of cases (9/24).
RI PT
In three cases, post-repair MRN provided valuable information about the status of the IAN and LN repairs which affected treatment decision and management. In one case, a recurrent neuroma was identified on MRN prompting surgical intervention verified at the time of surgery. In two post-repair
identified on MRN obviating the need for surgical intervention.
SC
scenarios, continuity of the nerve without neuroma formation following neurorraphy with allograft was
M AN U
This study has some limitations, including lack of blinding and controls. However, for the neuropathic cases, since the presentation was unilateral, the contralateral nerve caliber and signal served as an adequate internal control. The same surgeon performing the NST and clinical Sunderland classification and the surgery had access to the MRN findings before surgery and the radiologists had access to the surgeons’ clinical findings when the MRN was interpreted. In future studies, a larger sample in a
TE D
prospective fashion with the surgery and clinical examination blinded to the MRN study and inclusion of non-neuropathic cases will be done to address the above limitations. We did not analyze diffusion imaging parameters or tractography apart from routine analysis of signal abnormality on diffusion
EP
imaging. Finding fractional anisotropy and diffusion coefficient alterations in neuropathy can aid in
AC C
further quantitation of neuropathy and will be a subject of future study. This study demonstrated the current applications of MRN in the evaluation and management of nontraumatic and traumatic PTN. The strength of MRN in PTN management is the ability to provide noninvasive objective information of the IAN and LN that can distinguish normal from different degrees of neuropathy in the pre-injury, post-injury and post-repair phases. The correlation with surgical findings was moderate to good and if a randomized clinical trial can confirm this relationship with pre-surgical NST than new strategies in the management of PTN of the IAN and LN should lead to quicker detection
ACCEPTED MANUSCRIPT
of non-recoverable injuries earlier than the current protocols and without dependence on patient’s response to clinical stimuli during NST testing which should lead to improved outcomes of repair. Likewise, earlier detection of recoverable injuries should lead to risk reduction of unnecessary surgical
RI PT
intervention.
Acknowledgments
SC
Dr. Zuniga is a consultant for AxoGen Inc, Daiichi Sankyo/Charleston Labs and Merz USA, none
contributed to this manuscript. Dr. Chabbra is a consultant for ICON medical and holds book royalties
M AN U
with Jaypee, Wolters, none contributed to this manuscript. REFERENCES
1. Bagheri SC, Meyer RA, Khan HA et al: Retrospective measures of microsurgical repair of 222 lingual nerve injuries. J Oral Maxillofac Surg 68:715, 2010.
TE D
2. Brooks DN, Weber RV, Chao JD, et al: Processed nerve allografts for peripheral nerve reconstruction: A multicenter study of the utilization and outcome in sensory, mixed and motor nerve reconstructions. Microsurgery 32:1, 2012.
EP
3. Zuniga JR: Sensory outcomes after reconstruction of lingual and inferior alveolar nerve discontinuities using processed nerve allograft- A case series. J Oral Maxillofac Surg, 73:734,
AC C
2015.
4. Zuniga JR, Essick G: A contemporary approach to the clinical evaluation of trigeminal nerve injuries. OMFS Clin North Amer, 4:353, 1992.
5. Zuniga JR, Meyer RA, Gregg JM, et al: The accuracy of clinical neurosensory testing for nerve injury diagnosis. J Oral Maxillofac Surg, 56:2, 1998 6. Bagheri SC, Meyer RA, Cho SH et al: Microsurgical repair of the inferior alveolar nerve: success rate and factors that adversely affect outcome. J Oral Maxillofacial Surg 70: 1978, 2012.
ACCEPTED MANUSCRIPT
7. Mackinnon SE, Dellon AL: Results of Nerve Repair and Grafting. Surgery of Peripheral Nerve, 1st ed. New York: Thieme Medical, 1988. 8. Dodson TB, Kaban LB: Recommendations for management of trigeminal nerve defects based on a
RI PT
critical appraisal of the literature. J Oral Maxillofac Surg 55:1380, 1997.
9. Robinson PP. Observations on the recovery of sensation following inferior alveolar nerve injuries. Br J Oral Maxillofac Surg 26:177, 1988
SC
10. Chhabra A, Thawait GK, Doldatos T, et al: High-resolution 3T MR neurography of the brachial plexus and its branches, with emphasis on 3D imaging. Am J Neuroradiol, 34:486, 2013
M AN U
11. Baumer P, Weiler M, Nedszus M, et al: Somatotopic fascicular organization of the human sciatic nerve demonstrated by MR neurography. Neurology 84:1782, 2015 12. Cho Sims G, Boothe E, Joodi R, et al: 3D MR Neurography of the lumbosacral plexus: obtaining optimal images for selective longitudinal nerve depiction. Am J Neuroradiol 37:2158, 2016
TE D
13. Manoliu A, Ho M, Nanz D, et al: MR Neurographic orthopantomogram: Ultrashort echo-time imaging of mandibular bone and teeth complimented with high-resolution morphological and functional MR Neurography. J Magn Reson Imaging 44: 393, 2016
EP
14. Bendszus M, Wessig C, Solymosi L, et al: MRI of peripheral nerve degeneration and regeneration: correlation with electrophysiology and histology. Exp Neurol 188: 171, 2004
AC C
15. Li X, Shen J, Chen J, et al: Magnetic resonance imaging evaluation of acute crush injury of rabbit sciatic nerve: correlation with histology. Can Assoc Radiol J 59: 123, 2008
16. Liu V, Li J, Butzkeuven H, et al: Microstructural abnormalities in the trigeminal nerves of patients with trigeminal neuralgia revealed by multiple diffusion metrics. Eur J Radiol 82:783, 2013
17. Shen J, Zhou CP, Zhong XM, et al: MR Neurography: T1 and T2 measurements in acute peripheral nerve traction injury in rabbits. Radiology 254:729, 2010
ACCEPTED MANUSCRIPT
18. Terumitsu M, Seo K, Matsuzawa H, et al: Morphologic evaluation of the inferior alveolar nerve in patients with sensory disorders by high-resolution 3D volume rendering magnetic resonance neurography on a 3.0-T system. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 111:95, 2011
RI PT
19. Terumitsu M, Hatsuzawa H, Seo K, et al: High-contrast high-resolution imaging of posttraumatic mandibular nerve by 3DAC-PROPELLER magnetic resonance imaging: correlation with the severity of sensory disturbance. Oral Surg Oral Med Oral Pathol Oral Radiol 124:85, 2017.
SC
20. Cox, B, Zuniga JR, Panchal N, et al: Magnetic Resonance Neurography in the Management of Peripheral Trigeminal Neuropathy: experience in a tertiary care center. Eur Radiol, 10:3392, 2016
M AN U
21. Sunderland S: A classification of peripheral nerve injuries producing loss of function. Brain 74:491, 1951.
22. Miloro, M: Radiologic Assessment of the trigeminal nerve. Oral Maxillofac Surg Clin North Amer. 13:315, 2001
TE D
23. Filler AG, Kliot M, Hayes CE, et al: Application of magnetic resonance neurography in the evaluation of patients with peripheral nerve pathology. J Neurosurg 85: 299, 1996 24. Chhabra A, Soldatos T, Subhawong TK, et al: The application of three-dimensional diffusion –
EP
weighted PSIF technique in peripheral nerve imaging of the distal extremities. J Magn Reson Imaging. 34:962, 2011.
AC C
25. Chhabra A, Carrino J: Current MR neurography techniques and whole-body MR neurography. Semin Musculoskeletal Radiol. 19:79, 2015.
26. Grant GA, Britz GW, Goodkin R, et al: The utility of magnetic resonance imaging in evaluating peripheral nerve disorders. Muscle Nerve 25: 314, 2002.
27. Cudlip SA, Howe FA, Phil D, et al: Magnetic resonance neurography of peripheral nerve following experimental crush injury and correlation with functional deficit. J Neurosurg 96:755, 2002.
ACCEPTED MANUSCRIPT
28. Chhabra A, Flammang A, Padua A, et al: Magnetic resonance neurography: technical
RI PT
conisderations. Neuroimaging Clin N Am 24:67, 2014.
FIGURE LEGEND
SC
Figure 1: Coronal 3D PSIF MIP image shows the diffusely prominent and hyperintense left IAN (large arrows) as compared to normal intermediate signal of right IAN (small arrows). Corresponding axial
M AN U
DTI image (b=600 map) confirms the conspicuous abnormal enlargement and signal alteration of the left IAN.
Figure 2: Intraoperative high magnified photo of the left IAN resting on a neuropatty platform with suction catheter. There is continuity of the IAN but there is intense erythema of the IAN proper in
TE D
the area near tooth #18 (small arrow) while the distal continuation is normal (large arrow), with a clear border seen between abnormal and normal Figure 3: Coronal view of MRN 3D MIP of the right and left LN and IAN in a patient with severe
EP
sensory impairment/S1 of the left tongue with a trigger and dystrophic ageusia. The right LN (small arrows) and right and left IAN are shown and are normal in size and location. The double arrows
AC C
point out an abnormal left LN with enlargement of the caliber size ending in a focal mass of hyperintense tissue consistent with a neuroma. Figure 4: Intraoperative high magnified photo of the left LN demonstrated in Figure 3 resting on a neuropatty platform with suction catheter. The double arrows point to abnormal neural tissue which was consistent with a neuroma in continuity with enlargement of the proximal nerve tissue in the field posterior and lateral to the lower second molar.
ACCEPTED MANUSCRIPT
Figure 5: Coronal view of MRN right and left IAN and LN in a patient with moderate impairment/S2 with neuropathic pain of the left IAN 5 months after dental implant placement. Coronal 3D PSIF MIP
normal intermediate signal and caliber of right IAN (small arrows).
RI PT
image shows the mildly diffusely prominent and hyperintense left IAN (large arrows) as compared to
Figure 6A: Intraoperative photo of the left IAN exposed via a lateral cortical window showing the base of the dental implant (small arrow) in proximity to the IAN (large arrow) with thickening of the
SC
nerve at the contact site with no discontinuity or neuroma formation
M AN U
Figure 6B: Intraoperative high magnified photo of the right IAN resting on a neuropatty platform at the point of contact with the dental implant showing internal fibrosis (arrow). Figure 7: Intraoperative high magnified photo of the right LN resting on a neuropatty platform showing a large neuroma in continuity (large arrow) within an allograft placed 6 months prior with atrophy of the distal nerve. Retained microsuture is seen proximal and distal to the neuroma (small
TE D
arrows) indicating the neuroma formation was within the allograft and not due to disruption of the
AC C
EP
prior repair. These findings corroborate the MRN findings.
RI PT
ACCEPTED MANUSCRIPT
Table 1. Sunderland Nerve Classification by Clinical NST/MRCS Grade, Surgical findings and MRN Imaging Clinical NST level and MRCS Grade Description
Surgical by Direct inspection
MR Neurography
I
Normal (4)/ S3+ or S4 by 3 months
Anatomic: Homogenous mild increased T2 signal of nerve
II
Normal (4)/ S3+ or S4 by 6 months
III
Mild (3) or Moderate (2)/S2, S2+, S3 by 6 months or more
IV
Moderate (2) or Severe (1)/S1, S2, S2+ by 6 months or more
Intact with no internal or external fibrosis, normal mobility and neuroarchitecture (visualized fascicles and fanconi bands) Intact with no internal fibrosis. External fibrosis, restricted, mobility but neuroarchitecture intact (visualized fascicles and fanconi bands once external scar removed) Intact with both internal and external fibrosis, restricted mobility and disturbance of neuroarchitecture (abnormal fascicle patterns and/or fanconi bands not visible) Partial transected nerve but some amount of distal nerve present with or without neuroma in continuity
M AN U
TE D
EP
Severe (1) or Complete (0)/S0, S1 by 6 months or more
AC C
V
SC
Sunderland Classification
Complete transected nerve with or without amputation neuroma
Anatomic: Homogenous increased T2 signal of nerve and mild nerve thickening or constriction. Perineural fibrosis
Anatomic: Homogenous increased T2 signal of nerve and moderate thickening or constriction. Perineural fibrosis
Anatomic: Heterogeneous T2 signal of nerve and neuroma in continuity. Perineural and intraneural fibrosis. Anatomic: Discontinuous nerve with end-bulb neuroma
RI PT
ACCEPTED MANUSCRIPT
Table 2. Distribution of Traumatic Peripheral Trigeminal Neuropathies by Nerve and Etiology Third Molar
Dental Implant
Benign Pathology
Orthognathic
IAN
10
11
7
1
LN
18
0
SC
Nerve
M AN U
0
0
Table 3. Distribution of Non-Traumatic Peripheral Trigeminal Neuropathies by Nerve and Etiology Dental Injection
Non-Invasive Endodontics
Unknown*
IAN
1
4
6
LN
1
0
0
TE D
Nerve
AC C
EP
*complaint within the correct nerve distribution but the event unknown or beyond/not involving correct distribution of nerve
RI PT
ACCEPTED MANUSCRIPT
Table 4. Traumatic Trigeminal Neuropathies by Sunderland Classification on NST, MRN and Surgical
NST
MRN
Right LN
IV
IV
Left IAN
III
II
Left IAN
II
Right IAN
V
Left IAN
II
Right LN
IV
M AN U
Injured Nerve
Surgery
SC
Findings in Case Series
IV II II
Unclassified
II
II
II
IV
IV
III
III
III
III
II/III
III
IV/V
unclassified
IV
IV/V
IV
IV
III
III
III
Right LN
II/III
II/III
III
Right IAN
II
II
II
Left LN
V
V
V
Left LN
IV
IV
V
Left LN
IV/V
IV/V
IV
Left IAN Right LN Right LN
AC C
Right IAN
EP
Right LN
TE D
II
ACCEPTED MANUSCRIPT
II
III/IV
IV
Right LN
II/III
II/III
III
Right LN
IV/V
IV
V
Left LN
IV/V
IV
V
Left IAN
II
II
Left LN
IV
V
Left LN
IV
IV
Left LN
IV
IV
Left LN
III/IV
IV
Right IAN
II/III
III
IV
SC
IV
M AN U
TE D EP AC C
RI PT
Right IAN
II/III
IV V III/IV
RI PT
ACCEPTED MANUSCRIPT
Table 5. The Distribution of Correct/Incorrect Sunderland Classification by NST and MRN for Nerve injury Found at Surgery
5
NST MRN %Correct/Incorrect %Correct/Incorrect 20% 20%
LN
IV
8
100%
LN
III
3
LN
II
0
IAN
V
0
IAN
IV
IAN
III
IAN
II
M AN U
SC
N
LN
Sunderland Classification V
100% (*one unclassified)
33%
33%
n/a
n/a
n/a
n/a
1
0%
0%
4
50%
50%
5
60%
100% (*one unclassified)
TE D
AC C
EP
Nerve Injured
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Ms. Ref. No.: JOMS-D-17-01110 Title: Magnetic Resonance Neurography of Traumatic and Non-Traumatic Peripheral Trigeminal Neuropathies
RI PT
REVISION COVER LETTER
SC
Carmen Hupp JOMS Editorial Office Journal of Oral and Maxillofacial Surgery
AC C
EP
TE D
M AN U
Thank you very much for the opportunity to submit revisions to the article noted above for publication in JOMS. My co-authors and I agree that the recommended comments were necessary to improve the quality of the submission. The changes were identified in blue color. We have addressed the comments as follows: 1. Please cite all references in numerical order in the text of the manuscript (Reference #9 appears to be cited before #8) a. Reference #9 and #8 were corrected in the text and in the Reference section in the correct numerical order as recommended. All of the references were checked and found to be in correct numerical order 2. Please cite all tables in numerical order in the text of the manuscript (Table 6 does not appear to be cited) a. Table 6 was found to be incorrectly numbered in the original submission and should have been Table 5 which is correctly referenced to in Results/subsection Classification of traumatic nerve injuries on NST, MRN and Surgery, first paragraph. Table 6 in the revision is labeled Table 5. 3. Please add a disclosure statement at the end of the manuscript for Drs. Zuniga and Chhabra (as noted on the AAOMS Disclosure Form) a. Drs. Zuniga and Chhabra consultant and book royalty disclosures were noted at the end of the manuscript under ACKNOWLEDGMENTS. Note that there was no contribution by the disclosed activities to the study conducted or its’ publication. Thank you for the opportunity to submit the revision and look forward to its’ acceptance and publication. Sincerely
John R. Zuniga DMD, MS, PhD Robert V. Walker DDS Chair in Oral and Maxillofacial Surgery Professor, Departments of Surgery and Neurology & Neurotherapeutics University of Texas Southwestern