Vol. 109 No. 6 June 2010

ORAL AND MAXILLOFACIAL PATHOLOGY Editor: Mark W. Lingen

Patterns of second-opinion diagnosis in oral andmaxillofacial pathologyKyle Jones, BS,a and Richard C. K. Jordan, DDS, PhD, FRCPath,a,b,c San Francisco, CaliforniaUNIVERSITY OF CALIFORNIA

Objectives. Several studies have documented the beneficial effect of second opinions in diagnostic pathology. Amongdisease sites, the head and neck can be a particularly problematic area for pathologists, prompting frequent secondopinions. However, the effect of second opinion requests made by physician pathologists (PPs) to oral andmaxillofacial pathologists (OMPs) has not been well studied and might identify disease and subsites that posediagnostic challenges. The objectives of this study were to study the referral patterns of PPs to a referral center for oraland maxillofacial pathology and to assess changes in diagnosis following second opinion.Study design. We retrospectively reviewed 142 consecutive pathology consultation requests over a 2-year period. Thesubmitted report and matched second opinion report were reviewed to extract predetermined demographic, clinical,and pathologic data. Each diagnosis was reviewed to determine if there was agreement, minor disagreement, or majordisagreement between the original and the second opinion.Results. The most common diagnostic categories sent for second opinion were dysplasia/carcinoma, odontogeniccysts, and odontogenic tumors. In the 135 cases where agreement could be assessed, there were a total of 46 cases(34.1%) with differences in diagnostic opinion. Minor disagreements occurred in 24 cases (17.8%) and majordisagreements in 22 cases (16.3%). Importantly, major disagreements identified here would have resulted in significantdifferences in patient evaluation and management.Conclusions. This study supports the positive impact of second-opinion surgical pathology for lesions in themaxillofacial complex and supports the role of OMPs in subspecialty diagnostic pathology. (Oral Surg Oral Med Oral

Pathol Oral Radiol Endod 2010;109:865-869)

Second opinions play an important role in the practice ofdiagnostic pathology and several studies have docu-mented the cost-effectiveness and positive impact thatthey have on patient management.1-11 For example, Cooket al.12 showed that in 1 year, 31% of pathology cases sentout for review from their institution had either the diag-nosis changed or a definitive diagnosis determined whenthere previously was none. Several studies have shownthat among disease sites, the head and neck can be par-

Supported by National Institutes of Health grants CA095231,T32DE017249, U10CA21661, and T32DE019096 (to R.J.).aDepartment of Orofacial Sciences.bHelen Diller Comprehensive Cancer Center.cDepartment of Pathology.Received for publication Oct 26, 2009; accepted for publication Dec10, 2009.1079-2104/$ - see front matter© 2010 Mosby, Inc. All rights reserved.

doi:10.1016/j.tripleo.2009.12.023

ticularly problematic for pathologists.5,6,13 For this reason,Westra et al.6 stated that the head and neck should beconsidered as a high-risk diagnostic area and that secondopinions should be mandatory for problematic cases.

Oral and maxillofacial pathology, a recognized spe-cialty of dentistry, deals with the nature, identification, andmanagement of diseases affecting the oral and maxillofa-cial regions. Oral and maxillofacial pathologists (OMPs)complete a 3-year accredited training program that pro-vides education in the microscopic and clinical diagnosisof diseases within the oral cavity and maxillofacial com-plex. Although the number of oral biopsies received byphysician pathologists (PPs) in a community setting maybe small, the expertise of OMPs may prove invaluable inassisting PPs with problematic cases. Barrett andSpeight14 showed that even though biopsies taken fromthe oral cavity comprised a small number of the overallbiopsy workload received by PPs at 167 general histopa-

thology departments in England and Wales, PPs still used

865

OOOOE866 Jones and Jordan June 2010

the services of OMPs to reach definitive diagnoses formany challenging cases. Thus, the specialized skills of theOMP fill what can sometimes be an important gap in theexpertise of some PPs.

Although some authors have attempted to determinethe specific subsites within the head and neck that posethe greatest diagnostic challenges to pathologists, sec-ond opinion requests made by PPs to OMPs has notbeen well studied.1,6,14 These patterns could be impor-tant, because they could identify diseases and sites thatpose the greatest diagnostic challenge for PPs whendealing with specimens from the oral and maxillofacialcomplex. Moreover, by comparing the initial diagnosisfrom the referring PP with the second opinion, it maybe possible to assess the potential impact that secondopinions made by OMPs might have on patient man-agement. Therefore, the objectives of the present studywere to study the referral patterns of PPs to a referralcenter for oral and maxillofacial pathology and to as-sess changes in diagnosis following second opinion.

MATERIALS AND METHODSWe retrospectively reviewed 225 consecutive pathol-

ogy consultation requests sent to the senior author forsecond opinion between January 1, 2007, and December31, 2008 (a 24-month period). Of these, there were 79cases reviewed in compliance with the Joint Commissionon Accreditation of Healthcare Organizations policies forpatients referred to our institution for definitive cancertherapy. These cases were excluded from analysis becausethe second opinion request originated from a clinicianwithin the University of California at San Francisco(UCSF). The remaining 146 cases were submitted frompathologists outside of UCSF. Excluded from this cate-gory were 4 cases that were referred from an OMP. Theremaining 142 cases formed the basis of this study.

For each case, the submitted correspondence from thereferring PP, original pathology report, and matchedUCSF oral pathology report (second-opinion diagnosis)were retrieved and manually reviewed to extract predeter-mined demographic, clinical, and pathologic data. Eachdiagnosis was assigned to 1 of 11 disease categories(Table I) and each case subsequently reviewed to deter-mine if there was agreement, minor disagreement, ormajor disagreement between the submitting PP and thesecond opinion.1 Minor disagreements were defined asdifferences in diagnostic opinion that would not signifi-cantly alter the treatment and/or prognosis of the patient;major disagreements were defined as those that wouldsignificantly alter the evaluation plan, treatment, and/orprognosis of the patient. All minor and major disagree-ments were reviewed by at least 2 other OMPs at UCSF

for diagnostic confirmation.

RESULTSReferring pathologist demographics

A total of 81 separate pathologists submitted 142 casesfor second opinion over the study period. Seventy-eightPPs (96.3%) had MD degrees, 2 (2.5%) had DO degrees,and 1 (1.2%) had both MD and PhD degrees. Fifty-twocases (36.6%) were from the San Francisco Bay area.Additionally, there were 28 cases (19.7%) from other sitesin northern California, 11 (7.8%) from central California,and 2 (1.4%) from southern California; 49 cases (34.5%)came from 7 states outside of California. Pathology de-partments in community hospital settings submitted 129(90.8%) of the 142 cases, and 13 (9.2%) came frompathology departments within academic medical centers.Review of billing information showed that 73 patients(51.4%) used private insurance to pay for the second-opinion consultation, of which 25 (17.6%) were enrolledin the health maintenance organization Kaiser Perma-nente. Eight cases (5.6%) were billed to Medicare and 5(3.5%) to Medi-Cal (Medicaid), and 2 patients (1.4%) hadno health insurance. Fifty-four patients (38.0%) either hadtheir consultation fees billed directly to their referringpathologist or did not report their insurance type.

Biopsy sites, supplemental imaging studies, andspecial tests

For the 142 consultation requests, a total of 177 biopsyspecimens were examined. Biopsies were taken from severallocations in the head and neck, including mandible (26.6%),maxilla (19.2%), tongue (11.9%), gingivae (10.2%), buccalmucosa (9.0%), lip (5.6%), unspecified (4.5%), larynx(4.0%), palate (3.4%), floor of mouth (2.3%), parotid gland(1.1%), temporomandibular joint (0.6%), lymph node

Table I. Comparison of diagnostic categories for sub-mitted and second-opinion cases

Diagnostic category Submitted case Second opinion

Dysplasia/SCC 26 28Cyst 26 33Combination* 22† 6Odontogenic tumor 13 17Mucocutaneous disease 11 15BFOL 11 9Soft tissue tumor 10 17Salivary gland disease/tumor 8 8Bone 7 9No diagnosis given 7 0Skin 1 0Lymphoproliferative disorder 0 0

SCC, Squamous cell carcinoma; BFOL, benign fibro-osseous lesion.*Combination is defined as �1 diagnosis for �1 biopsies pertainingto the same patient from 1 consultation.†Combination diagnoses from referring pathologists: dysplasia/SCCvs. mucocutaneous/inflammatory disease � 8; cyst vs. odontogenictumor � 5; odontogenic tumor vs. bone � 2; other � 7.

(0.6%), nasal cavity (0.6%), and tonsil (0.6%) (Table II).

OOOOEVolume 109, Number 6 Jones and Jordan 867

Radiographic (e.g., dental x-ray, computerized tomog-raphy scan, magnetic resonance imaging [MRI], and soon) and/or photographic imaging studies accompanied 28cases (19.4%), for which additional imaging was recom-mended after review in 2 (7.1%) of these cases. For the114 cases sent without imaging, supplemental imagingwas recommended in 9 cases (6.3%). Further sections andhistochemical and immunohistochemical stains were per-formed by the OMP in 12 cases as part of the second-opinion assessment. Paraffin blocks sent by the PP accom-panied the microscopic sections in 52 (36.6%) of the 142consultation cases, and special stains accompanied 19cases (13.4%). For the majority of these cases (11 out of19, 57.9%) the interpretations of the special stains ren-dered in the original reports and consultations were inagreement. In 3 cases (15.8%) there was interpretativedisagreement between the PP and OMP. In 2 of thesecases the PP interpreted periodic acid–Schiff–stained sec-tions as positive for fungi, whereas the OMP interpretedthem as negative. In the third case the PP reported animmunohistochemical study as CD-68 positive whereasthe OMP interpreted it as negative. In 4 cases (21.1%) thereferring PP did not record an interpretation of the sub-mitted special stains. Based on the OMP consultationreport, follow-up was recommended in 28 cases (19.7%),which included recommendations for radiographic andclinical correlation (11 cases, 7.7%), radiographic assess-ment at a later date (6 cases, 4.2%), direct immunofluo-rescence studies (5 cases, 3.5%), repeat biopsy (2 cases,1.4%), supplemental immunohistochemistry (1 case,0.7%), antifungal therapy (1 case, 0.7%), and repeat MRI(1 case, 0.7%).

Diagnostic agreementOut of the 142 reviewed cases, there was diagnostic

agreement in 89 cases (62.7%). Agreement could not be

Table II. Summary of anatomic sites sent for secondopinionAnatomic site No. of biopsies (%)

Mandible 47 (26.6)Maxilla 34 (19.2)Tongue 21 (11.9)Gingiva 18 (10.2)Buccal mucosa 16 (9.0)Lip 10 (5.6)Unspecified 8 (4.5)Larynx 7 (4.0)Palate 6 (3.4)Floor of mouth 4 (2.3)Parotid 2 (1.1)TMJ 1 (0.6)Lymph node 1 (0.6)Nasal cavity 1 (0.6)Tonsil 1 (0.6)

assessed in 7 cases (4.9%), because no opinion was of-

fered by the contributing PP. There were a total of 46cases (34.1%) with differences in diagnostic opinion, 24(17.8%) of which were minor disagreements and 22(16.3%) major disagreements (Tables III and IV). For the24 minor disagreements, 15 were cases that dealt withodontogenic cysts or tumors, 5 soft tissue tumors, 3 mu-cosal lesions, and 1 fibro-osseous disease. Of the 22 majordisagreements, 5 were odontogenic cysts or tumors, 4were salivary gland diseases, 4 were bone or fibro-osseousdiseases, 4 were mucocutaneous/inflammatory diseases, 3were soft tissue tumors, and 2 were mucosal lesions. Forthe major disagreements, 9 cases (40.9%) initially diag-nosed as malignant by the PP were changed to benign inthe second-opinion report. Seven cases submitted with abenign diagnosis were changed to malignant. For theremainder of cases, the diagnostic category (benign ormalignant) did not change; however the second-opiniondiagnosis would result in a significantly different treat-

Table III. Minor diagnostic disagreements betweenoriginal and second-opinion diagnosesCase

# Submitted diagnosis Second-opinion diagnosis

8 Scar with atypical cells Peripheral giant cellgranuloma

9 Odontogenic fibroma Ossifying fibroma10 OKC Lateral periodontal cyst

(botryoid variant)15 Fibroma Pyogenic granuloma20 Dentigerous cyst vs. OKC Radicular cyst27 Peripheral giant cell granuloma Peripheral ossifying

fibroma28 Periapical cyst vs. dentigerous cyst Periapical granuloma31 Odontogenic fibroma vs. OKC Adenomatoid odontogenic

tumor37 Glandular odontogenic cyst Dentigerous cyst49 Black hairy tongue Focal fibrous hyperplasia50 Pyogenic granuloma Peripheral ossifying

fibroma61 Fibrohistiocytic giant cell

proliferationPeripheral giant cell

granuloma64 Atypical squamous proliferation Chronic mucositis77 OKC Calcifying odontogenic cyst81 Dental follicle vs. odontogenic

myxomaDesmoplastic fibroma

92 Aneurysmal bone cyst Ameloblastoma94 Pyogenic granuloma Peripheral ossifying

fibroma98 Periapical granuloma Radicular cyst

102 OKC vs. gingival cyst Lateral periodontal cyst(botryoid variant)

114 OKC OKC, atypical121 Lichen planus with atypia Epithelial dysplasia137 Calcifying epidermal

odontogenic cystAdenomatoid odontogenic

tumor144 Inflammatory process Peripheral ossifying

fibroma146 Ulceration Lichen planus

OKC, Odontogenic keratocyst.

ment plan for the patient.

OOOOE868 Jones and Jordan June 2010

DISCUSSIONThe primary objective of this study was to analyze

referral patterns and results of second diagnostic opinionsfor cases sourced from PPs. We based this study on anoral and maxillofacial pathology consultation practice lo-cated within a tertiary-care academic center in a largemetropolitan center. The majority of consultations origi-nated from PPs practicing in community hospital settings.Pathology practices located in academic medical centersprovided relatively few cases. The disproportionate num-ber of referrals originating from nonacademic centerscould be interpreted in a couple of ways. First, pathologylaboratories in community hospitals far outnumber thosewithin academic settings, and the referral patterns seenhere might simply mirror this demographic. Additionally,pathology practices within academic settings have accessto “in-house” pathology subspecialists, thus decreasingthe need for extramural consults.

In general, we found that the majority of second opin-ions were sought for the diagnosis and grading of oralepithelial dysplasia and carcinoma and the classificationof odontogenic cysts and tumors. There are several expla-

Table IV. Major diagnostic disagreements betweenoriginal and second-opinion diagnosesCase

# Submitted diagnosis Second-opinion diagnosis

17 Chronic mucositis Pemphigus vulgaris34 Pleomorphic adenoma Mucoepidermoid carcinoma,

intermediate grade36 SCC Hyperparakeratosis39 Benign fibro-osseous lesion Sclerosing osteitis48 Low-grade osteosarcoma Juvenile ossifying fibroma56 Paget disease Sclerotic bone67 Spindle cell tumor Dentigerous cyst75 Chronic sialadenitis Mucoepidermoid carcinoma,

low grade78 Odontogenic myxoma Dentigerous cyst80 Sialadenoma papilliferum vs.

intraductal papillomaPapillary

cystadenocarcinoma93 Calcifying epithelial

odontogenic tumorDentigerous cyst

97 Mucocele Granulomatous sialadenitis108 Pemphigus vs. lichen planus Nonspecific ulceration110 Unicystic ameloblastoma Calcifying odontogenic cyst111 Carcinoma in situ Mucin-rich salivary duct

carcinoma117 Basal cell carcinoma Ameloblastoma118 SCC Epithelial dysplasia, severe129 Schwannoma Malignant peripheral nerve

sheath tumor132 Ossifying fibroma Sclerosing osteitis145 Ameloblastoma Fibrous tissue, inflamed151 Ameloblastic fibroma Fibrosarcoma152 Benign angioleiomyomatous

neoplasmLow-grade myofibroblastic

sarcoma

SCC, squamous cell carcinoma.

nations for these results. First, medical school curricula

and pathology residencies often do not deal extensivelywith diseases of the oral cavity and maxillofacial com-plex.15 Second, there may be a general lack of exposureand experience with diseases of the oral and maxillofacialregion, because many of the diseases are quite rare. In ananalysis of cases received in an oral and maxillofacialpathology department in Sheffield, U.K., odontogenic tu-mors comprised only 1% of all specimens and 5% of jawlesions, giving an estimated incidence of fewer than 0.5cases per 100,000 per year.16 Third, the classification ofodontogenic cysts, tumors and salivary gland diseases iscomplex and not well known. For example, the histologicclassification of odontogenic tumors includes more than30 named entities, but may be somewhat overly complexowing to the inclusion of variants. Additionally, some ofthese variants are extremely rare and may not be encoun-tered in a lifetime of work. Finally, the grading of epithe-lial dysplasia can be somewhat subjective, particularly forlower-grade lesions that require more expert opinion forthis category of disease.17,18

In about two-thirds of the cases in the present study,the PP opinion and second opinion were in agreement.Thus, in about one-third of the cases there was either aminor (24 cases) or a major (22 cases) disagreement indiagnostic opinion. Although there are limited reportson the major differences in diagnostic pathology opin-ion between initial and second diagnoses in the headand neck, the 16.3% identified here is somewhat higherthan the published range of the sample sets of 7% inWestra et al.6 and 11.4% in Manion et al.1

Minor differences in opinion did not result in changesto the likely management of patients’ disease and reflectminor differences in classification. For example, the dif-ferences between a peripheral ossifying fibroma and aperipheral giant cell granuloma (Table III, case #27) arethe presence of bone and giant cells; however, both enti-ties are benign, etiologically similar, and managed bylocal excision. Similar to Westra et al.,6 differences ingrading of epithelial dysplasia (mild, moderate, severe)were recorded but not reported as minor disagreements,because grading is often subjective and can lack repro-ducibility.17,18

In contrast, the impact of major disagreements is moreimportant because the revised diagnosis would have asignificant impact on preoperative evaluation, alterationsin nature and extent of surgery, chemotherapy, or radia-tion therapy, and long-term follow-up. For example, alow-grade osteosarcoma (Table IV, case #48) would bemanaged by wide local excision supplemented with radi-ation and chemotherapy, whereas a juvenile ossifyingfibroma is managed by local excision. Sialadenoma pap-illiferum (Table IV, case #80) is managed by local exci-sion, but papillary cystadenocarcinoma would require an

oncologic work-up, a larger resection, and long-term fol-

OOOOEVolume 109, Number 6 Jones and Jordan 869

low-up for recurrence and metastasis. These differenceshighlight the importance of recognizing the many differ-ent diseases in the oral and maxillofacial region. They alsoillustrate the effect of subspecialty pathology practice,because many diseases of the oral and maxillofacial com-plex are uncommon or rare, and accurate classification ofthese diseases can have a significant effect on patientmanagement and outcome. Interestingly, the percentageof cases submitted by academic and community pathol-ogy labs with which the OMPs had major disagreementswas similar, 23.1% (3 out of 13 cases) and 14.7% (19 outof 129 cases), respectively. Minor disagreement percent-ages were also similar: 23.1% (3 out of 13 cases) foracademic labs and 16.3% (21 out of 129 cases) for com-munity labs. Thus it appears that PPs in academic andcommunity settings had similar concerns in the diagnosisof some maxillofacial lesions.

Only 25 of the 142 consultations were submitted withsome type of supplemental radiographic imaging of thelesion in question. Although academic pathology centersrepresent only �9% of the referring institutions, they senta disproportionate one-third (8 out of 25 cases) of theconsultations with some type of radiographic imagingstudy. It is difficult, however, to determine how oftenthese images were actually useful in achieving the diag-nosis. With the exception of mucosal diseases, manyproblematic lesions in the oral and maxillofacial complexoccur in bone, and radiographic imaging of these casesusually provides information vital to determining a finaldiagnosis. This is especially pertinent because many of thePP cases dealt with odontogenic cysts and tumors of thejaws but were not sent with corresponding radiographicwork-ups. In general, radiographic images would beenrelevant in the 11 out of 142 cases (7.7%) that requiredadditional imaging to determine definitive diagnoses sub-sequent to the second opinion.

One of the limitations of this study is the sample size.Because the sample is relatively small and the populationof referring pathologists dominated by those in California,wider conclusions about PP referral patterns to OMPscannot be made. However, despite these limitations, sev-eral conclusions can be drawn. First, the relatively fre-quent occurrence of major disagreements in this studysuggests the important role OMPs can play in the diag-nosis of diseases in the oral and maxillofacial region of thehead and neck. Second, several diagnostic categories ofdisease may pose challenges for the PP, including thediagnosis and grading of squamour cell carcinoma ordysplasia and the classification of odontogenic cysts andodontogenic tumors. These results support the frequentneed for supplemental imaging for consultation cases inthe oral and maxillofacial complex. Taken together, theresults of this study support the interaction between PP

and OMP in determining definitive diagnoses for specific

lesions of the oral and maxillofacial complex, includingthe oral cavity.

REFERENCES1. Manion E, Cohen MB, Weydert J. Mandatory second opinion in

surgical pathology referral material: clinical consequences ofmajor disagreements. Am J Surg Pathol 2008;32:732-7.

2. Lueck N, Jensen C, Cohen MB, Weydert JA. Mandatory secondopinion in cytopathology. Cancer Cytopathol 2009;117:82-91.

3. Coblentz TR, Mills SE, Theodorescu D. Impact of second opin-ion pathology in the definitive management of patients withbladder carcinoma. Cancer 2001;91:1284-90.

4. Kronz JD, Westra WH, Epstein JI. Mandatory second opinion surgicalpathology at a large referral hospital. Cancer 1999;86:2426-35.

5. Kronz JD, Westra WH. The role of second opinion pathology inthe management of lesions of the head and neck. Curr OpinOtolaryngol Head Neck Surg 2005;13:81-4.

6. Westra WH, Kronz JD, Eisele DW. The impact of second opinionsurgical pathology on the practice of head and neck surgery: a decadeexperience at a large referral hospital. Head Neck 2002;24:684-93.

7. Chafe S, Honore L, Pearcey R, Capstick V. An analysis of theimpact of pathology review in gynecologic cancer. Int J RadiatOncol Biol Phys 2000;48:1433-8.

8. Hahm GK, Niemann TH, Lucas JG, Frankel WL. The value ofsecond opinion in gastrointestinal and liver pathology. ArchPathol Lab Med 2001;125:736-9.

9. Hamady ZZ, Mather N, Lansdown MR, Davidson L, Maclennan KA.Surgical pathological second opinion in thyroid malignancy: impact onpatients’ management and prognosis. Eur J Surg Oncol 2005;31:74-7.

10. Staradub VL, Messenger KA, Hao N, Wiley EL, Morrow M.Changes in breast cancer therapy because of pathology secondopinions. Ann Surg Oncol 2002;9:982-7.

11. Wurzer JC, Al-Saleem TI, Hanlon AL, Freedman GM, Patchef-sky A, Hanks GE. Histopathologic review of prostate biopsiesfrom patients referred to a comprehensive cancer center: corre-lation of pathologic findings, analysis of cost, and impact ontreatment. Cancer 1998;83:753-9.

12. Cook IS, McCormick D, Poller DN. Referrals for second opinionin surgical pathology: implications for management of cancerpatients in the UK. Eur J Surg Oncol 2001;27:589-94.

13. Vivino FB, Gala I, Hermann GA. Change in final diagnosis onsecond evaluation of labial minor salivary gland biopsies.J Rheumatol 2002;29:938-44.

14. Barret AW, Speight PM. Use of oral pathology services bygeneral histopathologists and their attitudes to training of oralpathologists. J Clin Pathol 1996;49:565-9.

15. Lewis CW, Grossman DC, Domoto PK, Deyo RA. The role ofthe pediatrician in the oral health of children: a national survey.Pediatrics 2000;106:E84.

16. Jones AV, Franklin CD. An analysis of oral and maxillofacialpathology found in adults over a 30-year period. J Oral PatholMed 2006;35:392-401.

17. Pindborg JJ, Reibel J, Holmstrup P. Subjectivity in evaluatingoral epithelial dysplasia, carcinoma in situ and initial carcinoma.J Oral Pathol 1985;14:698-708.

18. Brothwell DJ, Lewis DW, Bradley G, Leong I, Jordan RC, Mock D,et al. Observer agreement in the grading of oral epithelial dysplasia.Community Dent Oral Epidemiol 2003;31:300-5.

Reprint requests:

Richard C. K. JordanS-512, 513 Parnassus AvenueUniversity of California San FranciscoSan Francisco, CA 94143-0424

richard.jordan@ucsf.edu