Prospective Consensus Reporting by Gynecologic Pathology and Dermatopathology Improves Diagnosis of Vulvar Biopsies Open Access

This study was performed (with approval by the institutional review board) in an academic tertiary care hospital where the anatomic pathology department follows a subspecialty sign-out format. For 6 months, consecutive vulvar biopsies were reviewed prospectively by 1 of 3 surgical pathologists and 1 of 3 dermatopathologists. In keeping with the routine daily workflow of the anatomic pathology division, biopsies from obstetrics and gynecology were reviewed by surgical pathologists while biopsies from dermatology were reviewed by dermatopathologists. All surgical pathologists had subspecialty sign-out expertise in gynecologic pathology (mean, 7 years; range, 4–11 years); 1 pathologist was fellowship-trained in gynecologic pathology. All dermatopathologists were fellowship-trained and practiced dermatopathology in an academic setting (mean, 4 years; range, 1–5 years).

For each case, preliminary diagnoses were independently generated and recorded and then subsequently shared in consensus review between 1 surgical pathologist and 1 dermatopathologist. A third pathologist from either the surgical pathology (SP) or dermatopathology (DP) cohort adjudicated cases without consensus after dual review. The following data elements were collected for each case: division from which the biopsy originated (SP or DP), patient age, site, provided clinical history, whether additional history was retrieved from the electronic medical record, diagnostic category (neoplastic, inflammatory, both, other), preliminary diagnoses (for both SP and DP), final (consensus) diagnosis, adjudication needed (yes/no), ancillary tests, and diagnostic discrepancy (yes/no; major or minor). A major discrepancy was defined as a diagnosis affecting patient care (eg, increased surveillance, different treatment). A minor discrepancy was defined as diagnoses or differences in terminology/wording that did not impact patient care.

Eighty-four biopsies (48 SP, 36 DP) from 70 patients (mean age, 53 years; age range, 17–85 years) were prospectively reviewed during the 6-month study period. Forty-two of 84 cases (50%) were categorized as neoplastic, 38 of 84 (45%) as reactive/inflammatory/infectious (“inflammatory”) with the remaining (5%) showing both or other features (Table 1). Clinical history was provided in 71 of 84 cases (84%) (SP: 36 of 48 cases [75%]; DP: 35 of 36 cases [97%]); detailed clinical information (description of lesion, clinical differential diagnosis) was provided in 27 of 48 SP (56%) and 33 of 36 DP (92%) cases. Retrieval of additional clinical information from the electronic medical record was required in 10 of 84 cases (12%) (SP: 9 of 10 cases [90%]; DP: 1 of 10 cases [10%]).

Independently generated preliminary diagnoses were in agreement between SP and DP in 62 of the 84 biopsies (74%). Of these vulvar biopsies, 34 (55%) were neoplastic, 27 (43%) were inflammatory, and 1 (2%) was categorized as “other.” In the neoplastic agreement category, diagnostic concordance was seen in all malignancies (5 of 5), pigmented lesions (12 of 12), and adnexal tumors (1 of 1). High agreement was also seen with squamous neoplasia (21 of 27, 78%), with near perfect agreement for the high-grade squamous intraepithelial lesion/vulvar intraepithelial neoplasia 3 (HSIL/VIN 3) cohort (11 of 12, 92%). Of note, none of the vascular neoplasms demonstrated independent diagnostic SP/DP agreement. Of the inflammatory biopsies, the granulomatous (1 of 1, 100%) and lichenoid reaction patterns (11 of 13, 85%) showed the highest SP/DP agreement.

Independently generated preliminary diagnoses were discrepant in 22 of 84 cases (26%), but consensus review resulted in an agreed-upon diagnosis in all cases, with adjudication from a third pathologist being required in 6 of 22 cases (27%). Major and minor initial diagnostic discrepancies were recorded for 9 cases (11%) (Table 2) and 13 cases (15%) (Table 3) among the 84 biopsies, respectively. As might be anticipated, the most common category for a discrepancy (major and/or minor) for surgical pathologists was in the cohort of “inflammatory vulvar biopsies” (SP: n = 9 versus DP: n = 1), while the most common category for dermatopathologists was the “neoplastic vulvar biopsy” (DP: n = 5 versus SP: n = 3). The cohort of major initial diagnostic discrepancies consisted of 4 neoplastic and 3 inflammatory cases; 1 biopsy was categorized as “both” and 1 as “other.” Discrepancies in the diagnosis of vulvar neoplasia generally involved accurate identification of low-grade squamous intraepithelial lesions (LSILs) or distinction between low- and high-grade squamous intraepithelial lesions (Table 2). Discrepancies in the diagnosis of vulvar inflammatory dermatoses generally involved rare variants of common diseases (eg, hypertrophic lichen sclerosus or chronic vegetative herpetic infection; Table 2). The cohort of minor initial diagnostic discrepancies consisted of 3 neoplastic and 8 inflammatory cases; 2 biopsies were categorized as “other.” Minor diagnostic discrepancies were largely related to the differential reporting of lichen simplex chronicus versus chronic spongiotic dermatitis (Table 3). Consensus review of vulvar biopsies in which these 2 diagnoses were being considered led to a shared understanding that chronic spongiotic dermatitis would be a preferable diagnosis if there was evident epidermal edema (spongiosis) and any significant (more than sparse) dermal infiltrate. Lichen simplex chronicus was reserved for cases in which secondary changes of rubbing/scratching were evident but epidermal edema and dermal inflammation were minimal or lacking. Clinically, spongiotic dermatitis (which may often have secondary changes of lichen simplex chronicus) may respond to topical therapy, whereas treatment of primary lichen simplex chronicus relies more on cessation of the scratching cycle. In addition, case 1 in Table 3 illustrates a minor initial diagnostic disagreement for both SP and DP (SP: lichen simplex chronicus; DP: chronic spongiotic dermatitis); DP suggested a periodic acid–Schiff with diastase (PAS-D) stain during consensus review. The subsequently positive PAS-D staining led to the final diagnosis of a fungal infection. Select cases of major and minor discrepancies are illustrated in Figures 1 through 4. Importantly, diagnostic agreement of the preliminary diagnoses increased over time with a reduction in major (n = 6 to 3) and minor (n = 9 to 4) discrepancies between the first and second half of the study period. Ancillary stains were ordered in 32 of 84 biopsies (38%), with PAS-D being the most common.

Collaboration with DP resulted in better delineation of nonneoplastic vulvar biopsies but also helped identify uncommon (vascular) neoplasms. Specifically, collaboration with DP resulted in a different diagnosis in 9 cases (eg, condyloma to angiokeratoma), a more specific diagnostic line and/or more detailed and improved commentary in 7 cases (eg, “reactive squamous mucosa with hyperkeratosis and parakeratosis” to “lichen sclerosus, hypertrophic variant” or “benign squamous mucosa with dilated and inflamed adnexal structure” to “hybrid cyst with features of follicular cyst [infundibular type] and verrucous cyst”), and helpful ancillary stains in 3 cases (eg, identification of herpes simplex virus and fungal infections). Collaboration with SP resulted in better discrimination in the evaluation of vulvar squamous intraepithelial lesions in 5 cases (HSIL versus LSIL versus no dysplasia) and establishing a different inflammatory reaction pattern in 1 case.

Vulvar skin biopsies may encompass both cutaneous inflammatory processes and neoplasia encountered in the gynecologic tract. Diagnostic acumen in the interpretation and reporting of vulvar biopsies, particularly nonneoplastic biopsies, is not well studied in the literature. We hypothesized that collaboration between surgical pathologists and dermatopathologists may improve overall reporting through improved diagnostic concordance. Our study shows that nearly three-fourths of all biopsies were in complete diagnostic agreement between SP and DP with a slightly better overall agreement for neoplastic entities. Our data further suggest that in our practice environment, there is a high degree of shared competence between surgical pathologists and dermatopathologists in the accurate diagnosis of vulvar malignancies, the spectrum of pigmented vulvar lesions, and vulvar inflammatory lesions with a lichenoid reaction pattern. This diagnostic competence is likely related to the relative frequency with which these entities are encountered in our daily practice. The remaining one-fourth of cases demonstrated an initial diagnostic disagreement (9 major and 13 minor discrepancies). Consensus review resulted in an agreed-upon final diagnosis between SP and DP in all cases. Collaboration with DP predominantly improved the reporting of nonneoplastic lesions by replacing descriptive diagnoses with more precise terminology, using ancillary stains to identify infections, or adding detailed commentary to optimize the report. Collaboration with SP helped the dermatopathologist fine-tune thresholds for the diagnosis of squamous intraepithelial lesions of the vulva. These trends suggest that intradepartmental consult may be of use for DP when considering whether a biopsy shows low- or high-grade dysplasia and for SP when biopsies do not show classic inflammatory histopathologic features. We anticipate that more detailed and sophisticated reporting would potentially result in (1) improved clarity of reports for clinicians and (2) more effective patient care. A follow-up study with our clinical colleagues represents a future avenue to confirm these impressions.

A pathologist's comfort level with vulvar biopsies likely varies by exposure in residency or fellowship training as well as experience through one's daily practice setting. Anecdotally, surgical pathologists (in comparison to dermatopathologists) tend to approach vulvar biopsies, particularly nonneoplastic ones, with a sense of “dread” and more descriptive sign-out styles. This scenario is likely augmented by the often limited clinical history provided by gynecologists (in comparison to dermatologists), who in turn are forced to navigate the meaning of pathology reports with long descriptive diagnostic jargon. Our study supports this assumption; surgical pathologists struggled more with inflammatory vulvar biopsies than neoplastic biopsies, while dermatopathologists showed the opposite diagnostic proclivity. Detailed clinical history (description of the lesion with differential diagnosis) was almost routinely provided by dermatologists (92%) for the dermatopathologist but much less often provided by the gynecologist (52%) for the surgical pathologist. Our study also suggests that performance can improve over time, as collaboration between subspecialties reduced the overall number of major and minor discrepancies by at least 50% in the second half of the study period. Future investigations looking into expanding collaborative approaches to gynecologists and dermatologists in addition to pathologists could allow for identification of more ways to improve delivery of patient care through interprofessional education.

A review of the literature demonstrates several publications on the diagnosis and reporting of vulvar biopsies. In a study of nondiagnostic vulvar biopsies, Day et al¹  state that “fundamental knowledge gaps are exacerbated by the use of imprecise terminology” by pathologists and that interpretation of a biopsy can differ based on clinical context. The authors encouraged a multidisciplinary review of biopsy results with clinicopathologic discrepancies and concluded that best correlation was achieved when clinicians provided a detailed description, location, and differential diagnosis for a lesion while working with an expert (vulvar) pathologist. Our study suggests that liberal use of interdepartmental consultation can close these knowledge gaps and decrease the use of imprecise terminology. In a review of 183 cases of vulvar dermatoses, Chan and Zimarowski²  reported reserving a descriptive sign-out diagnosis with a limited differential diagnosis in approximately one-tenth of cases. Contributing factors that resulted in descriptive reports included inadequate clinical history and diagnostic difficulty due to shared histopathologic features. Niamh et al³  corroborated that vulvar inflammatory dermatoses are a challenging area of pathology and concluded that a small subset of patients with vulvar inflammatory dermatoses do not fit a specific diagnostic category, resulting in a nonspecific, descriptive diagnosis with a comment for continued clinical surveillance. While we agree that there are certainly times in which a descriptive diagnosis is necessary (and indeed a descriptive diagnosis was ultimately rendered in 7 of our cases), consensus reporting seems to decrease the frequency with which such an approach is required. With regard to vulvar neoplasia, publications have predominantly focused on interobserver variability in measuring depth of invasion in squamous cell carcinoma,^4,5  reproducibility of grading vulvar dysplasia,^6–8  and accurate classification of challenging preneoplastic lesions such as differentiated vulvar intraepithelial neoplasia.^9,10  Abdel-Mesih and colleagues⁴  studied the interobserver agreement for stage 1A squamous cell carcinoma of the vulva and noted that agreement among gynecologic pathologists is only fair in the determination of in situ or superficially invasive disease. A recent study by Pouwer et al⁵  corroborated these findings. Of note, we did not encounter diagnostically challenging cases of superficially invasive squamous cell carcinomas in our cohort. In a multicenter European collaborative study on the interobserver variation in the diagnosis and grading of VIN, Preti et al⁶  reported a good agreement for grading VIN when current terminology was used by expert pathologists. Preti et al⁶  further noted good agreement for VIN 2/3 but poor agreement for VIN 1. In a study about differentiated vulvar intraepithelial neoplasia (dVIN), van den Einden et al⁹  showed that the agreement among pathologists-in-training, general surgical pathologists, and gynecologic pathologists ranged from poor to moderate with increased agreement following an educational session of the diagnostic criteria for dVIN. van de Nieuwenhof et al¹⁰  reviewed cases diagnosed as lichen sclerosus in women with and without progression to invasive squamous cell carcinoma; in women with progression, 42% of lichen sclerosus cases were retrospectively reclassified as dVIN. This particular type of vulvar squamous dysplasia is known to be fraught by interobserver variability and of note, no cases of dVIN were encountered in our study cohort.

The prospective nature of our study is a major strength. Cases were reviewed independently by the originating pathologist (SP or DP) on service and preliminary diagnoses were shared and discussed for consensus opinion before sign-out. This approach not only fostered a sense of collaboration and education but also provided a final report that we believe provided more meaningful information to the clinician. Another inherent strength of our study was the relatively even distribution of neoplastic versus inflammatory vulvar biopsies in our study cohort. Of note, owing to the proximity of both SP and DP offices, there was no significant delay in turnaround time during the study period. The weaknesses in our study include the small number of discrepant cases, which precludes a meaningful statistical analysis, and the lack of cases at high risk for interobserver variability (atypical nevi, differentiated VIN). In addition, all surgical pathologists had subspecialty expertise in the sign-out of gynecologic pathology; no general surgical pathologists or trainees participated in this study. However, we anticipate that shared consensus reporting would be valuable in these groups as well.

In summary, prospective review of vulvar biopsies by both surgical pathologists and dermatopathologists improves diagnostic concordance over time. Such interdisciplinary collaboration allows for a more sophisticated and detailed pathology report that we believe provides more meaningful information for the clinician. Consensus review also allows pathologists to gain more diagnostic confidence in interpretation of inflammatory (for SP) and neoplastic (for DP) vulvar biopsies. Therefore, intradepartmental consultation is of value in select cases.

The authors would like to thank Soheila Korourian, MD, for her contributions to this study.