International Journal of Gynecological Pathology 33:560–567, Lippincott Williams & Wilkins, Baltimore r 2014 International Society of Gynecological Pathologists
Original Article
Current Morphologic Criteria Perform Poorly in Identifying Hereditary Leiomyomatosis and Renal Cell Carcinoma Syndrome-associated Uterine Leiomyomas Sana Alsolami, M.D., Mona El-Bahrawy, MBBCh, Ph.D., FRCPath, Steve E. Kalloger, BSc., MSc., Nagla AlDaoud, M.D., Tilak B. Pathak, M.D., Catherine T. Cheung, BSc., M.D., Anna Marie Mulligan, M.D., FRCP(UK), Ian P. Tomlinson, B.A., M.A., Ph.D., Patrick J. Pollard, BSc., Ph.D., C. Blake Gilks, M.D., FRCPC, Ph.D., W. Glenn McCluggage, FRCPath, and Blaise A. Clarke, MBBCh, FRCPS Summary: The contemporary oncologic pathology report conveys diagnostic, prognostic, predictive, and hereditary predisposition information. Each component may be premised on a morphologic feature or a biomarker. Clinical validity and reproducibility are paramount as is standardization of reporting and clinical response to ensure individualization of patient care. Regarding hereditary predisposition, morphology-based genetic referral systems in some instances have eclipsed genealogy-based systems, for example, cell type in ovarian cancer and BRCA screening. In other instances such as Lynch syndrome, morphology-based schemas supplement clinical schemas and there is an emerging standard of care for reflex biomarker testing. Hereditary leiomyomatosis and renal cell carcinoma (HLRCC) syndrome predisposes patients to uterine and cutaneous leiomyomas (LMs) and renal cell carcinomas (RCCs). Several authors have emphasized the role pathologists may play in identifying this syndrome by recognizing the morphologic characteristics of syndromic uterine LMs and RCCs. Recently immunohistochemical overexpression of S-(2succinyl) cysteine (2SC) has been demonstrated as a robust biomarker of mutation status in tumors from HLRCC patients. In this blinded control-cohort study we demonstrate that the proposed morphologic criteria used to identify uterine LMs in HLRCC syndrome are largely irreproducible among pathologists and lack sufficient robustness to serve as a trigger to triage cases for 2SC immunohistochemistry or patients for further family/personal history inquiry. Although refinement of morphologic criteria can be considered, in view of the availability of a clinically robust biomarker, consideration should be given to reflex testing of uterine LMs with an appropriate age cut off or in the setting of a suspicious family history. Key Words: HLRCC—Hereditary leiomyomatosis and renal cell carcinoma syndrome—Morphologic criteria.
From the Department of Laboratory Medicine and Pathobiology (S.A.), Sunnybrook Health Sciences Centre; Department of Pediatric Laboratory Medicine (C.C.), The Hospital For Sick Children; Department of Laboratory Medicine and Pathobiology (A.M.M., B.A.C.), University Health Network, University of Toronto, Toronto, ON; Department of Pathology (S.K., C.B.G.), University of British Columbia, Vancouver, BC, Canada; Dammam Regional Laboratory (S.A.), Ministry of Health, Dammam, Saudi Arabia; Department of Histopathology (M.E.B.), Imperial College, Hammersmith Hospital, London; Molecular and Population Genetics Laboratory (I.P.T.), Wellcome Trust Centre for Human Genetics; Oxford Biomedical Research Centre (I.P.T.), Wellcome Trust Centre for Human Genetics; Henry Wellcome Building for Molecular Physiology (P.J.P.), University of Oxford, Oxford; Department of Pathology (W.G.M.), Royal Group of Hospitals Trust, Northern Ireland, UK; Department of Pathology and Microbiology (N.A.), Jordan University of Science and Technology, Ar Ramtha, Jordan; and Department of Pathology (T.B.P.), B.P.Koirala Memorial Cancer Hospital, Chitwan, Nepal. S.A. and M.E.B. contributed equally. This work was jointly supervised by W.G.M. and B.A.C. The authors declare no conflict of interest. Address correspondence and reprint requests to W.G. McCluggage, Department of Pathology, Royal Group of Hospitals Trust, Grosvenor Road, Northern Ireland BT12 6BL, UK. E-mail: [email protected].
DOI: 10.1097/PGP.0000000000000091
560
IDENTIFYING HEREDITARY LEIOMYOMATOSIS AND RENAL CELL CARCINOMA SYNDROME In addition to providing diagnostic, prognostic, and predictive information, pathologists are increasingly playing a critical role in the identification of possible germline/hereditary predisposition in cancer patients. Such identification may be predicated on: tumor morphology, for example, MsPath for Lynch syndrome (LS) in colorectal cancer (CRC) (1), highgrade serous morphology in ovarian/tubal carcinoma for BRCA (2); tumor topography, for example, rightsided colon cancer in LS; clinical data, for example young age of patient (3) or on reflex or clinician requested biomarker assessment, for example, mismatch repair immunohistochemistry (MMR-IHC) or microsatellite instability testing in LS (4). The College of American Pathologists (CAP) has stated that it behooves pathologists to recognize the histologic features of microsatellite unstable CRC (5). However, these remain only optional reporting elements in the CAP-CRC checklists and, acknowledging the inchoate nature of such features in endometrial carcinoma (EC), these are not included as optional checklist elements for the latter cancer type. Regarding recognition of possible hereditary predisposition in both cancer types, there is discrepant practice among pathologists (6). The hereditary leiomyomatosis and renal cell cancer (HLRCC) syndrome is characterized by mutation of the fumurate hydratase (FH, 1q42.3q43) gene which predisposes affected individuals to multiple leiomyomas (LMs) of the skin and uterus and to an aggressive form of renal cell carcinoma (RCC) (7–9). Recently, 3 of us (M.E.B., I.P.T., P.J.P.) identified aberrant succination of protein with overexpression of S-(2-succinyl) cysteine (2SC) as a robust biomarker of mutation status in HLRCC patients (10). Although data are extremely limited, it has been suggested that the uterine LMs in this syndrome have distinctive morphologic features characterized by increased cellularity, multinucleate cells, nuclear atypia, and tumor cell nuclei with large orangeophilic/eosinophilic nucleoli rimmed by a perinucleolar halo (11,12). The latter feature is also a histologic hallmark of the associated RCCs (13). Some authors have considered the morphologic features in LMs and RCCs in HLRCC to be distinctive and have suggested that recognition of these features in such tumors, especially in young patients, should serve as a tocsin for genetic counseling and work up for HLRCC syndrome (13). To date, no blinded control-cohort studies have been conducted of sporadic and HLRCC-associated
561
uterine LMs to assess the reproducibility and sensitivity and specificity of such proposed morphologic criteria. In this study, we report on the diagnostic accuracy of these features utilizing a retrospective cohort of uterine LMs, both sporadic and HLRCC-associated, where pathologists were blinded to the hereditary status of the subjects. MATERIALS AND METHODS The study set comprised 30 slides which included 19 slides of 17 separate uterine LMs from 12 patients with germline-confirmed HLRCC syndrome and 11 slides of sporadic LMs. For 9 patients with HLRCCassociated LMs, 1 slide from each was submitted. The additional 10 slides were from 3 patients, 7 from different LMs in 2 patients, and 3 from a single LM in 1. The control cohort was identified based on the absence of immunohistochemical expression of 2SC and the absence of family history. Slides were digitized and made available online. Six participants, including 3 specialist gynecologic pathologists (W.G.M., B.G., M.E.B.) and 3 nongynecologic pathologists (S.E.S., A.M.M., C.C.), were provided with login details. In the instruction sheet, the morphologic descriptions from previous studies of HLRCC-associated uterine LMs were listed and these included: prominent orangiophilic/eosinophilic nucleoli, perinucleolar halos, increased cellularity, presence of multinucleate cells, and nuclear atypia. In addition, 2 articles with images of HLRCC-associated LMs and RCCs were included for review by the participants (11,13). The participants reviewed whole-slide images and assessed and recorded the presence of prominent orangiophilic/eosinophilic nucleoli, perinucleolar halos, and multinucleate cells (0 = absent; 1 = present). Participants scored tumor cellularity as normal (0), cellular (1), or highly cellular (2). Finally, they recorded their ‘‘final morphologic appraisal’’ as to whether they considered each case a sporadic (0) or an HLRCCassociated (1) LM or as equivocal (2). Interobserver reproducibility for each diagnostic phenomenon was assessed with the Kappa statistic (k) computed for each reader by reader comparison within the gynecologic and nongynaecologic pathologist groups. To determine if specialist gynecologic patholologists had a superior ability to detect HLRCCassociated LMs based on morphologic assessment compared with their nonspecialist counterparts, sensitivity and specificity were calculated with 95% Int J Gynecol Pathol Vol. 33, No. 6, November 2014
562
ALSOLAMI ET AL.
confidence intervals. k values were categorized as poor (o0.4), fair to good (0.4–0.74), or excellent (Z0.75).
TABLE 1. Reproducibility of the morphologic features among the pathologist Comparisons
RESULTS The median age of patients with HLRCC syndrome uterine LMs was 39 yr (range, 32–70 yr). The mean age of patients with sporadic cases was 49 yr (range, 39–61 yr). For 2 of the slides, the quality of the scanned material was suboptimal and these were excluded (1 sporadic and 1 hereditary case), leaving 28 slides for analysis of which 18 were HLRCC-associated and 10 sporadic. Each morphologic feature (prominent nucleoli, perinucleolar halos, cellularity, and multinucleate cells) was assessed by the 6 participants stratified by gynecologic specialist or nonspecialist. The interobserver comparisons, quantified with the k statistic and SEs, for the 4 morphologic features revealed that most features were irreproducible (o0.4) among pathologists, regardless of subspecialty expertise with only the presence of prominent nucleoli showing borderline fair reproducibility (Table 1). Table 2 shows each participants’ percentage of correct final morphologic appraisals, the percentage of cases interpreted as equivocal, the percentage of hereditary and sporadic cases correctly identified, the percentage of equivocal cases which were actually hereditary or sporadic, and the percentage of hereditary cases which were identified correctly or as equivocal. Overall, the percentage of cases correctly identified ranged from 29% to 50% with the range being 11% to 39% for correct allocation of hereditary cases and 30% to 100% for sporadic cases. Rates of equivocation varied from 7% to 57%. The percentage of hereditary cases either correctly identified or regarded as equivocal ranged from 22% to 83%. Figures 1 and 2 show the spectrum of changes encountered in sporadic and hereditary LM cases. Examination of the sensitivity and specificity of the binarized final morphologic appraisal [hereditary (1) or equivocal (2) vs. sporadic (0)] as a screening tool to triage patients for further family/personal history inquiry or formal genetic counseling revealed that the sensitivity ranged from 22% to 83% and the specificity ranged from 30% to 100% (Table 3). The 95% confidence intervals revealed no evidence that either the gynecologic or nongynecologic pathologists had superior sensitivity or specificity for the detection of HLRCC patients by morphologic assessment of LMs. Int J Gynecol Pathol Vol. 33, No. 6, November 2014
k (SE)
Prominent nucleoli Gyn 1 vs. gyn 2 Gyn 1 vs. gyn 3 Gyn 2 vs. gyn 3 Gyn overall Nongyn 1 vs. nongyn Nongyn 1 vs. nongyn Nongyn 2 vs. nongyn Nongyn overall Cellularity Gyn 1 vs. gyn 2 Gyn 1 vs. gyn 3 Gyn 2 vs. gyn 3 Gyn overall Nongyn 1 vs. nongyn Nongyn 1 vs. nongyn Nongyn 2 vs. nongyn Nongyn overall Perinucleolar halos Gyn 1 vs. gyn 2 Gyn 1 vs. gyn 3 Gyn 2 vs. gyn 3 Gyn overall Nongyn 1 vs. nongyn Nongyn 1 vs. nongyn Nongyn 2 vs. nongyn Nongyn overall Multinucleate cells Gyn 1 vs. gyn 2 Gyn 1 vs. gyn 3 Gyn 2 vs. gyn 3 Gyn overall Nongyn 1 vs. nongyn Nongyn 1 vs. nongyn Nongyn 2 vs. nongyn Nongyn overall
2 3 3
2 3 3
2 3 3
2 3 3
0.5930 0.4011 0.4302 0.4643 0.5508 0.3600 0.5200 0.4734
(0.1507) (0.1769) (0.1677) (0.1091) (0.1616) (0.1765) (0.1630) (0.1091)
0.2272 – 0.0040 0.1630 0.0661 0.2104 0.3103 0.1034 0.1642
(0.1086) (0.1763) (0.1445) (0.0957) (0.1334) (0.1678) (0.1396) (0.1020)
0.2667 0.4842 0.3548 0.3318 0.4197 0.2804 0.2376 0.2841
(0.1401) (0.1675) (0.1250) (0.1091) (0.1710) (0.1393) (0.1584) (0.1091)
1.0000 0.3636 0.3636 0.4615 0.1429 0.6216 0.0000 0.1964
(0.0000) (0.2678) (0.2678) (0.1091) (0.1690) (0.1692) (0.15512) (0.1091)
DISCUSSION The HLRCC syndrome predisposes affected individuals to multiple LMs of the skin and uterus and to an aggressive form of RCC. Several authors have stressed the role of pathologists in identifying patients likely to have this syndrome by recognizing the hallmark features of such tumors, whether RCC or LMs, namely the presence of orangeophilic/eosinophilic nucleoli with clear perinucleolar halos in the majority of tumor cells (11,13). Recently, 3 of us (M.E.B., I.P.T., and P.J.P.) showed that immunohistochemical detection of aberrant succination of proteins is a robust biomarker of germline mutation of the fumurate hydratase gene (10). In that study we showed that the 2SC bioassay was equivalent or superior to conventional screening methods as a diagnostic test, and was sufficiently sensitive and specific for use as a routine clinical biomarker.
IDENTIFYING HEREDITARY LEIOMYOMATOSIS AND RENAL CELL CARCINOMA SYNDROME
563
TABLE 2. Final Morphologic Appraisal for Each Participant n (%) Gyn 1 Cases correctly identified Hereditary cases identified Sporadic cases identified Overall % equivocal cases Equivocal cases which are actually hereditary Equivocal cases which are actually sporadic Hereditary cases identified correctly or as equivocal
11/28 4/18 7/10 8/28 6/18 2/10 10/18
(39) (22) (70) (29) (33) (20) (56)
Increasingly, along with prognostic and predictive morphologic features and biomarkers, morphologic features and/or biomarkers suggesting genetic predilection are being incorporated into pathology reports, to allow individualized patient management. Although predictive biomarker assay and reporting protocols have rapidly been standardized and integrated to ensure an appropriate clinical response, reporting and integration of morphologic features and biomarkers of genetic predilection in patients with neoplasms is less well developed (14). Recognizing the possibility or likelihood of such syndromes is not only important for screening for metachronous or synchronous tumors and for identification of at risk and not-at-risk family members with subsequent tumor screening, but increasingly therapeutics may be predicated on the presence of these syndromes (15,16). Morphologybased genetic referral schemas have been developed to supplement the genealogy-based schemas which are often insensitive and implemented poorly in the clinical setting (17). However, as with clinical schemas, we need to be cautious and consider the reproducibility, sensitivity, and specificity of morphologic criteria as well as issues related to implementation as pathology reports grow in complexity (18). Some schemas are predicated on intrinsic basic morphology criteria such as tumor subtype in ovarian cancer, where it is recommended in some regions, for example, Ontario, Canada, that all patients with high-grade serous carcinoma of the ovary/tube/peritoneum undergo BRCA testing (2). This is only feasible now that the contemporary approach to ovarian carcinoma subtype allocation is highly reproducible and that reliable biomarker validation of cell type is available in difficult cases (19,20). With the diminishing cost of sequencing, one could argue for germline sequencing of a panel of hereditary ovarian cancer genes on all ovarian carcinomas. Identification of BRCA alteration may serve as a quality assurance tier for cell type allocation with any nonserous case in a BRCA 1/2 patient being reviewed.
Gyn 2 8/28 5/18 3/10 16/28 9/18 7/10 14/18
(29) (28) (30) (57) (50) (70) (78)
Gyn 3 14/28 7/18 7/10 2/28 1/18 1/10 8/18
(50) (39) (70) (7) (6) (10) (44)
Nongyn 1
Nongyn 2
Nongyn 3
15/28 8/18 7/10 8/28 7/18 1/10 15/18
10/28 2/18 8/10 2/28 2/18 0/10 4/18
14/28 4/18 10/10 8/28 8/18 0/10 12/18
(54) (44) (70) (29) (39) (10) (83)
(36) (11) (80) (7) (11) (0) (22)
(50) (22) (100) (29) (44) (0) (67)
In relation to CRC, the CAP has stated that pathologists need to recognize the histologic and clinical features that should prompt at least a recommendation for MMR testing (5). Several groups recognizing low overall referral rates and cognizant of lack of standardization of such features have suggested reflex testing of CRC and EC with MMR-IHC (21,22). A reflex testing policy represents a marked change in practice with such testing traditionally being clinician initiated (6). In many ways, the use of morphologic features and the development of a biomarker to identify HLRCC is analogous to LS. As with LS, although autosomal dominant, HLRCC shows variable penetrance limiting the use of family history. Successful identification of this syndrome in patients presenting with uterine LMs (patients with HLRCC often present initially with uterine LMs) will allow screening for RCC, potentially reducing mortality. The proposed morphologic criteria for LS and HLRCC are not routinely reported in EC, LM, or RCC. Similar to microsatellite testing, the use of 2SC immunohistochemistry to identify this syndrome is not a direct test of the altered protein but an assay of the functional consequence of the genomic alteration. Unlike LS in which the MMR protein can be assayed using immunohistochemisty, anti-FH antibodies will detect both mutant and normal protein, rendering it ineffective as a screening tool. This is the first blinded case-control study to assess the reproducibility and diagnostic accuracy of proposed morphologic features to identify uterine LMs in patients with HLRCC. Despite previous studies describing the presence of various morphologic features in HLRCC-associated LMs, our study definitively demonstrates that the morphologic features we assessed are largely irreproducible among pathologists, regardless of whether they are specialist gynecologic pathologists or not. None of the morphologic features assessed were sufficiently robust to allow for the development of a predictive algorithm Int J Gynecol Pathol Vol. 33, No. 6, November 2014
564
ALSOLAMI ET AL.
FIG. 1. (A) A high-power view highlighting the characteristic large orangiophilic nucleoli and perinucleolar halos in an hereditary leiomyomatosis and renal cell carcinoma syndrome (HLRCC)-associated leiomyoma (40 ). (B) HLRCC-associated leiomyoma showing characteristic nuclear features and numerous pink globules. Five of 6 participants correctly assessed this as a hereditary case (40 ). (C) HLRCC-associated leiomyoma lacking the characteristic cytologic features. All participants assessed this incorrectly as a sporadic case (20 ). (D) In other areas of the hereditary case shown in (C), the tumor does show pericytomatous vasculature.
Int J Gynecol Pathol Vol. 33, No. 6, November 2014
IDENTIFYING HEREDITARY LEIOMYOMATOSIS AND RENAL CELL CARCINOMA SYNDROME
565
screening tool to triage patients for further family/ personal history inquiry or formal genetic counseling. The poor performance of the morphologic criteria is particularly concerning in view of the enrichment of this cohort for mutation-positive cases. We acknowledge some limitations to this study. For the majority of hereditary LMs, only 1 slide was viewed by the participants. The use of digitized images was new for some of the participants. The prestudy training offered to study participants was limited consisting of written description of such features and attached articles with images. However, in reality most pathologists would only encounter the description of these features through articles, with few having access to confirmed hereditary or sporadic cases. Further, refinement of morphologic criteria to identify HLRCC-associated tumors may be possible. In a recent abstract, Reyes and colleagues reviewed the morphologic features of 10 uterine LMs with 2SC-IHC overexpression or FH germline mutation. They highlighted additional syndrome-associated morphologic features, namely pericytomatous vasculature and highly fibrillary cytoplasm with pink globules, which were present in some of these neoplasms (23). However, to achieve such refinements, well-annotated cohort-control cases need to be available, both for analysis and teaching, and rigorous study design must be employed. As has been recently proposed, borrowing from the genomics community, publically available annotated digital slide repositories would be critical in such an endeavor (24). Alternatively, mirroring the development of reflex testing in CRC and EC with MMR-IHC, a similar strategy could be deployed with RCCs and LMs. Appropriate cost-effectiveness studies would be needed especially in view of the ubiquitous nature of uterine LMs and the relatively lower incidence of
FIG. 2. (A) A hereditary leiomyomatosis and renal cell carcinoma syndrome (HLRCC)-associated leiomyoma in which there are focal prominent nucleoli with halos (4), whereas other nuclei appear vesicular. Four of 6 participants equivocated on this case and 2 misdiagnosed it as a sporadic case (40 ). (B) A sporadic leiomyoma with vesicular nuclei and small nucleoli (4). Two participants assessed this case as being hereditary, 2 as sporadic and 2 equivocated (40 ).
to triage cases for 2SC immunohistochemistry. Predictably, the subjective final morphologic assessment (hereditary or equivocal vs. sporadic) lacked adequate sensitivity and specificity to be used as a
TABLE 3. Sensitivity and Specificity Achieved by Each Participant in the Final Morphologic Appraisal
Reader Gyn 1 Gyn 2 Gyn 3 Nongyn 1 Nongyn 2 Nongyn 3
Correct (HLRCC) Gold Standard N = 18
Correct (Sporadic) Gold Standard N = 10
False Positive
False Negative
10 14 8 15 4 12
7 3 7 7 8 10
3 7 3 3 14 0
8 4 10 3 2 6
Sensitivity (95% CI) 55.6 77.8 44.4 83.3 22.2 66.7
(30.8–78.4) (52.4–93.5) (21.6–69.2) (58.6–96.2) (6.6–47.6) (41.0–86.6)
Specificity (95% CI) 70.0 30.0 70 70.0 80.0 100
(34.8–93.0) (7.0–65.2) (34.8–93.0) (34.8–93.0) (44.4–96.9) (69.0–100)
Cases assessed as being hereditary or equivocal were combined and the binarized scores [hereditary (1) or equivocal (2) versus sporadic (0)] compared with the gold standard, either germline status or 2SC immunohistochemistry for the control group. CI indicates confidence interval; HLRCC, hereditary leiomyomatosis and renal cell carcinoma.
Int J Gynecol Pathol Vol. 33, No. 6, November 2014
566
ALSOLAMI ET AL.
HLRCC in the LM and RCC population compared with LS and BRCA in the CRC, EC, and ovarian carcinoma population. The age range of HLRCC uterine LMs is 18 to 52 yr with a mean age 30 yr at presentation which contrasts with a mean age of 45 yr in the sporadic population (25–27). Reflex screening of all uterine LM specimens in patients 30 yr or less or in those who have a family history suggestive of HLRCC regardless of age is 1 strategy that could be critically assessed, for example. According to our data, using the currently proposed morphologic criteria to initiate 2SC-IHC testing lacks the requisite sensitivity and specificity. Germane to all hereditary predisposition biomarker or morphologic schema, it is imperative that the vicarious nature of the pathologist-patient relationship is addressed. We need to ensure that pathologist recommendations for screening and results of reflex-performed screening tests are appropriately followed up on by the clinician. Overall genetic referral rates for patients with serous carcinoma in Ontario still remain low and Sanchez et al (28) highlighted that recommendations to screen for LS rather than actual MMR-IHC results were less likely to result in genetic referral of the patient. Finally, we need to ensure appropriate feedback to the reporting pathologists as to the outcome of such recommendations, for clinical, quality assurance, and educational purposes. The burgeoning of hereditary predisposition morphology and/or biomarker schemas represents an exciting development in pathology and is yet another area in which laboratory physicians can midwife personalized medicine. As illustrated in this study, morphologic schema will need to be rigorously tested before implementation. Furthermore, as with predictive biomarkers, standardization of both technical and clinical aspects of hereditary predisposition biomarkers are needed. REFERENCES 1. Jenkins MA, Hayashi S, O’Shea AM, et al. Pathology features in Bethesda guidelines predict colorectal cancer microsatellite instability: a population-based study. Gastroenterology 2007;133:48–56. 2. Schrader KA, Hurlburt J, Kalloger SE, et al. Germline BRCA1 and BRCA2 mutations in ovarian cancer: utility of a histologybased referral strategy. Obstet Gynecol 2012;120:235–240. 3. Steinhagen E, Shia J, Markowitz AJ, et al. Systematic immunohistochemistry screening for Lynch syndrome in early age-of-onset colorectal cancer patients undergoing surgical resection. J Am Coll Surg 2012;214:61–67. 4. Beamer LC, Grant ML, Espenschied CR, et al. Reflex immunohistochemistry and microsatellite instability testing of colorectal tumors for Lynch syndrome among US cancer
Int J Gynecol Pathol Vol. 33, No. 6, November 2014
5. 6. 7.
8.
9.
10.
11.
12.
13.
14.
15. 16.
17.
18. 19. 20. 21. 22.
programs and follow-up of abnormal results. J Clin Oncol 2012;30:1058–1063. Turner JW, Frederick LB, Bloom KJ, et al. Prognostic Uses of MSI Testing. 2011. Available at: http://www.cap.org/apps/docs/ committees/technology/microsatellite_testing.pdf. Kalloger SE, Allo G, Mulligan AM, et al. Use of mismatch repair immunohistochemistry and microsatellite instability testing: exploring Canadian practices. Am J Surg Pathol 2012;36:560–569. Tomlinson IP, Alam NA, Rowan AJ, et al. Germline mutations in FH predispose to dominantly inherited uterine fibroids, skin leiomyomata and papillary renal cell carcinoma. Nat Genet 2002;30:406–410. Alam NA, Rowan AJ, Wortham NC, et al. Genetic and functional analyses of FH mutations in multiple cutaneous and uterine leiomyomatosis, hereditary leiomyomatosis and renal cancer, and fumarate hydratase deficiency. Hum Mol Genet 2003;12:1241–1252. Bayley JP, Launonen V, Tomlinson IP. The FH mutation database: an online database of fumarate hydratase mutations involved in the MCUL (HLRCC) tumor syndrome and congenital fumarase deficiency. BMC Med Genet 2008;9:20. Bardella C, El-Bahrawy M, Frizzell N, et al. Aberrant succination of proteins in fumarate hydratase-deficient mice and HLRCC patients is a robust biomarker of mutation status. J Pathol 2011;225:4–11. Garg K, Tickoo SK, Soslow RA, et al. Morphologic features of uterine leiomyomas associated with hereditary leiomyomatosis and renal cell carcinoma syndrome: a case report. Am J Surg Pathol 2011;35:1235–1237. Sanz J, Teller L, Vockes C, et al. Genetic alterations of uterine fibroids in hereditary leiomyomatosis and renal cancer (HLRCC) syndrome. Modern Pathology Annual Meetings Abstracts 223 A. Sanz-Ortega J, Vockes C, Stratton P, et al. Morphologic and molecular characteristic of uterine leiomyomas in hereditary leiomyomatosis and renal cancer (HLRCC) syndrome. Am J Surg Pathol 2013;37:74–80. Wolff AC, Hammond ME, Schwartz JN, et al. American Society of Clinical Oncology/College of American Pathologists guideline recommendations for human epidermal growth factor receptor 2 testing in breast cancer. J Clin Oncol 2007;25:118–145. Hewish M, Lord CJ, Martin SA, et al. Mismatch repair deficient colorectal cancer in the era of personalized treatment. Nat Rev Clin Oncol 2010;7:197–208. Martin SA, McCabe N, Mullarkey M, et al. DNA polymerases as potential therapeutic targets for cancers deficient in the DNA mismatch repair proteins MSH2 or MLH1. Cancer Cell 2010;17:235–248. Ryan P, Mulligan AM, Aronson M, et al. Comparison of clinical schemas and morphologic features in predicting Lynch syndrome in mutation-positive patients with endometrial cancer encountered in the context of familial gastrointestinal cancer registries. Cancer 2012;118:681–688. Clarke BA, Cooper K. Identifying Lynch syndrome in patients with endometrial carcinoma: shortcomings of morphologic and clinical schemas. Adv Anat Pathol 2012;19:231–238. Ko¨bel M, Kalloger SE, Baker PM, et al. Diagnosis of ovarian carcinoma cell type is highly reproducible: a transCanadian study. Am J Surg Pathol 2010;34:984–993. Kalloger SE, Ko¨bel M, Leung S, et al. Calculator for ovarian carcinoma subtype prediction. Mod Pathol 2011;24:512–521. Boland CR, Shike M. Report from the Jerusalem workshop on Lynch syndrome-hereditary nonpolyposis colorectal cancer. Gastroenterology 2010;138:2197. Palomaki GE, McClain MR, Melillo S, et al. EGAPP supplementary evidence review: DNA testing strategies aimed at reducing morbidity and mortality from Lynch syndrome. Genet Med 2009;11:42–65.
IDENTIFYING HEREDITARY LEIOMYOMATOSIS AND RENAL CELL CARCINOMA SYNDROME 23. Reyes C, Karamurzin Y, Frizzell N, et al. Uterine smooth muscle tumors with features suggesting hereditary leiomyomatosis and renal carcinoma (HLRCC) syndrome: detailed morphologic analysis and correlation with 2SC imnohistochemistry. Mod Pathol 2013;1224:293. 24. Hipp JD, Lucas DR, Emmert-Buck MR, et al. Digital slide repositories for publications: lessons learned from the microarray community. Am J Surg Pathol 2011;35:783–786. 25. Farquhar CM, Steiner CA. Hysterectomy rates in the United States 1990-1997. Obstet Gynecol 2002;99:229–234.
567
26. Toro JR, Nickerson ML, Wei MH, et al. Mutations in the fumarate hydratase gene cause hereditary leiomyomatosis and renal cell cancer in families in North America. Am J Hum Genet 2003;73:95–106. 27. Alam NA, Barclay E, Rowan AJ, et al. Clinical features of multiple cutaneous and uterine leiomyomatosis: an underdiagnosed tumor syndrome. Arch Dermatol 2005;141:199–206. 28. Sanchez JA, Vogel JD, Kalady MF, et al. Identifying Lynch syndrome: we are all responsible. Dis Colon Rectum 2008;51:1750–1756.
Int J Gynecol Pathol Vol. 33, No. 6, November 2014
Original Article
Current Morphologic Criteria Perform Poorly in Identifying Hereditary Leiomyomatosis and Renal Cell Carcinoma Syndrome-associated Uterine Leiomyomas Sana Alsolami, M.D., Mona El-Bahrawy, MBBCh, Ph.D., FRCPath, Steve E. Kalloger, BSc., MSc., Nagla AlDaoud, M.D., Tilak B. Pathak, M.D., Catherine T. Cheung, BSc., M.D., Anna Marie Mulligan, M.D., FRCP(UK), Ian P. Tomlinson, B.A., M.A., Ph.D., Patrick J. Pollard, BSc., Ph.D., C. Blake Gilks, M.D., FRCPC, Ph.D., W. Glenn McCluggage, FRCPath, and Blaise A. Clarke, MBBCh, FRCPS Summary: The contemporary oncologic pathology report conveys diagnostic, prognostic, predictive, and hereditary predisposition information. Each component may be premised on a morphologic feature or a biomarker. Clinical validity and reproducibility are paramount as is standardization of reporting and clinical response to ensure individualization of patient care. Regarding hereditary predisposition, morphology-based genetic referral systems in some instances have eclipsed genealogy-based systems, for example, cell type in ovarian cancer and BRCA screening. In other instances such as Lynch syndrome, morphology-based schemas supplement clinical schemas and there is an emerging standard of care for reflex biomarker testing. Hereditary leiomyomatosis and renal cell carcinoma (HLRCC) syndrome predisposes patients to uterine and cutaneous leiomyomas (LMs) and renal cell carcinomas (RCCs). Several authors have emphasized the role pathologists may play in identifying this syndrome by recognizing the morphologic characteristics of syndromic uterine LMs and RCCs. Recently immunohistochemical overexpression of S-(2succinyl) cysteine (2SC) has been demonstrated as a robust biomarker of mutation status in tumors from HLRCC patients. In this blinded control-cohort study we demonstrate that the proposed morphologic criteria used to identify uterine LMs in HLRCC syndrome are largely irreproducible among pathologists and lack sufficient robustness to serve as a trigger to triage cases for 2SC immunohistochemistry or patients for further family/personal history inquiry. Although refinement of morphologic criteria can be considered, in view of the availability of a clinically robust biomarker, consideration should be given to reflex testing of uterine LMs with an appropriate age cut off or in the setting of a suspicious family history. Key Words: HLRCC—Hereditary leiomyomatosis and renal cell carcinoma syndrome—Morphologic criteria.
From the Department of Laboratory Medicine and Pathobiology (S.A.), Sunnybrook Health Sciences Centre; Department of Pediatric Laboratory Medicine (C.C.), The Hospital For Sick Children; Department of Laboratory Medicine and Pathobiology (A.M.M., B.A.C.), University Health Network, University of Toronto, Toronto, ON; Department of Pathology (S.K., C.B.G.), University of British Columbia, Vancouver, BC, Canada; Dammam Regional Laboratory (S.A.), Ministry of Health, Dammam, Saudi Arabia; Department of Histopathology (M.E.B.), Imperial College, Hammersmith Hospital, London; Molecular and Population Genetics Laboratory (I.P.T.), Wellcome Trust Centre for Human Genetics; Oxford Biomedical Research Centre (I.P.T.), Wellcome Trust Centre for Human Genetics; Henry Wellcome Building for Molecular Physiology (P.J.P.), University of Oxford, Oxford; Department of Pathology (W.G.M.), Royal Group of Hospitals Trust, Northern Ireland, UK; Department of Pathology and Microbiology (N.A.), Jordan University of Science and Technology, Ar Ramtha, Jordan; and Department of Pathology (T.B.P.), B.P.Koirala Memorial Cancer Hospital, Chitwan, Nepal. S.A. and M.E.B. contributed equally. This work was jointly supervised by W.G.M. and B.A.C. The authors declare no conflict of interest. Address correspondence and reprint requests to W.G. McCluggage, Department of Pathology, Royal Group of Hospitals Trust, Grosvenor Road, Northern Ireland BT12 6BL, UK. E-mail: [email protected].
DOI: 10.1097/PGP.0000000000000091
560
IDENTIFYING HEREDITARY LEIOMYOMATOSIS AND RENAL CELL CARCINOMA SYNDROME In addition to providing diagnostic, prognostic, and predictive information, pathologists are increasingly playing a critical role in the identification of possible germline/hereditary predisposition in cancer patients. Such identification may be predicated on: tumor morphology, for example, MsPath for Lynch syndrome (LS) in colorectal cancer (CRC) (1), highgrade serous morphology in ovarian/tubal carcinoma for BRCA (2); tumor topography, for example, rightsided colon cancer in LS; clinical data, for example young age of patient (3) or on reflex or clinician requested biomarker assessment, for example, mismatch repair immunohistochemistry (MMR-IHC) or microsatellite instability testing in LS (4). The College of American Pathologists (CAP) has stated that it behooves pathologists to recognize the histologic features of microsatellite unstable CRC (5). However, these remain only optional reporting elements in the CAP-CRC checklists and, acknowledging the inchoate nature of such features in endometrial carcinoma (EC), these are not included as optional checklist elements for the latter cancer type. Regarding recognition of possible hereditary predisposition in both cancer types, there is discrepant practice among pathologists (6). The hereditary leiomyomatosis and renal cell cancer (HLRCC) syndrome is characterized by mutation of the fumurate hydratase (FH, 1q42.3q43) gene which predisposes affected individuals to multiple leiomyomas (LMs) of the skin and uterus and to an aggressive form of renal cell carcinoma (RCC) (7–9). Recently, 3 of us (M.E.B., I.P.T., P.J.P.) identified aberrant succination of protein with overexpression of S-(2-succinyl) cysteine (2SC) as a robust biomarker of mutation status in HLRCC patients (10). Although data are extremely limited, it has been suggested that the uterine LMs in this syndrome have distinctive morphologic features characterized by increased cellularity, multinucleate cells, nuclear atypia, and tumor cell nuclei with large orangeophilic/eosinophilic nucleoli rimmed by a perinucleolar halo (11,12). The latter feature is also a histologic hallmark of the associated RCCs (13). Some authors have considered the morphologic features in LMs and RCCs in HLRCC to be distinctive and have suggested that recognition of these features in such tumors, especially in young patients, should serve as a tocsin for genetic counseling and work up for HLRCC syndrome (13). To date, no blinded control-cohort studies have been conducted of sporadic and HLRCC-associated
561
uterine LMs to assess the reproducibility and sensitivity and specificity of such proposed morphologic criteria. In this study, we report on the diagnostic accuracy of these features utilizing a retrospective cohort of uterine LMs, both sporadic and HLRCC-associated, where pathologists were blinded to the hereditary status of the subjects. MATERIALS AND METHODS The study set comprised 30 slides which included 19 slides of 17 separate uterine LMs from 12 patients with germline-confirmed HLRCC syndrome and 11 slides of sporadic LMs. For 9 patients with HLRCCassociated LMs, 1 slide from each was submitted. The additional 10 slides were from 3 patients, 7 from different LMs in 2 patients, and 3 from a single LM in 1. The control cohort was identified based on the absence of immunohistochemical expression of 2SC and the absence of family history. Slides were digitized and made available online. Six participants, including 3 specialist gynecologic pathologists (W.G.M., B.G., M.E.B.) and 3 nongynecologic pathologists (S.E.S., A.M.M., C.C.), were provided with login details. In the instruction sheet, the morphologic descriptions from previous studies of HLRCC-associated uterine LMs were listed and these included: prominent orangiophilic/eosinophilic nucleoli, perinucleolar halos, increased cellularity, presence of multinucleate cells, and nuclear atypia. In addition, 2 articles with images of HLRCC-associated LMs and RCCs were included for review by the participants (11,13). The participants reviewed whole-slide images and assessed and recorded the presence of prominent orangiophilic/eosinophilic nucleoli, perinucleolar halos, and multinucleate cells (0 = absent; 1 = present). Participants scored tumor cellularity as normal (0), cellular (1), or highly cellular (2). Finally, they recorded their ‘‘final morphologic appraisal’’ as to whether they considered each case a sporadic (0) or an HLRCCassociated (1) LM or as equivocal (2). Interobserver reproducibility for each diagnostic phenomenon was assessed with the Kappa statistic (k) computed for each reader by reader comparison within the gynecologic and nongynaecologic pathologist groups. To determine if specialist gynecologic patholologists had a superior ability to detect HLRCCassociated LMs based on morphologic assessment compared with their nonspecialist counterparts, sensitivity and specificity were calculated with 95% Int J Gynecol Pathol Vol. 33, No. 6, November 2014
562
ALSOLAMI ET AL.
confidence intervals. k values were categorized as poor (o0.4), fair to good (0.4–0.74), or excellent (Z0.75).
TABLE 1. Reproducibility of the morphologic features among the pathologist Comparisons
RESULTS The median age of patients with HLRCC syndrome uterine LMs was 39 yr (range, 32–70 yr). The mean age of patients with sporadic cases was 49 yr (range, 39–61 yr). For 2 of the slides, the quality of the scanned material was suboptimal and these were excluded (1 sporadic and 1 hereditary case), leaving 28 slides for analysis of which 18 were HLRCC-associated and 10 sporadic. Each morphologic feature (prominent nucleoli, perinucleolar halos, cellularity, and multinucleate cells) was assessed by the 6 participants stratified by gynecologic specialist or nonspecialist. The interobserver comparisons, quantified with the k statistic and SEs, for the 4 morphologic features revealed that most features were irreproducible (o0.4) among pathologists, regardless of subspecialty expertise with only the presence of prominent nucleoli showing borderline fair reproducibility (Table 1). Table 2 shows each participants’ percentage of correct final morphologic appraisals, the percentage of cases interpreted as equivocal, the percentage of hereditary and sporadic cases correctly identified, the percentage of equivocal cases which were actually hereditary or sporadic, and the percentage of hereditary cases which were identified correctly or as equivocal. Overall, the percentage of cases correctly identified ranged from 29% to 50% with the range being 11% to 39% for correct allocation of hereditary cases and 30% to 100% for sporadic cases. Rates of equivocation varied from 7% to 57%. The percentage of hereditary cases either correctly identified or regarded as equivocal ranged from 22% to 83%. Figures 1 and 2 show the spectrum of changes encountered in sporadic and hereditary LM cases. Examination of the sensitivity and specificity of the binarized final morphologic appraisal [hereditary (1) or equivocal (2) vs. sporadic (0)] as a screening tool to triage patients for further family/personal history inquiry or formal genetic counseling revealed that the sensitivity ranged from 22% to 83% and the specificity ranged from 30% to 100% (Table 3). The 95% confidence intervals revealed no evidence that either the gynecologic or nongynecologic pathologists had superior sensitivity or specificity for the detection of HLRCC patients by morphologic assessment of LMs. Int J Gynecol Pathol Vol. 33, No. 6, November 2014
k (SE)
Prominent nucleoli Gyn 1 vs. gyn 2 Gyn 1 vs. gyn 3 Gyn 2 vs. gyn 3 Gyn overall Nongyn 1 vs. nongyn Nongyn 1 vs. nongyn Nongyn 2 vs. nongyn Nongyn overall Cellularity Gyn 1 vs. gyn 2 Gyn 1 vs. gyn 3 Gyn 2 vs. gyn 3 Gyn overall Nongyn 1 vs. nongyn Nongyn 1 vs. nongyn Nongyn 2 vs. nongyn Nongyn overall Perinucleolar halos Gyn 1 vs. gyn 2 Gyn 1 vs. gyn 3 Gyn 2 vs. gyn 3 Gyn overall Nongyn 1 vs. nongyn Nongyn 1 vs. nongyn Nongyn 2 vs. nongyn Nongyn overall Multinucleate cells Gyn 1 vs. gyn 2 Gyn 1 vs. gyn 3 Gyn 2 vs. gyn 3 Gyn overall Nongyn 1 vs. nongyn Nongyn 1 vs. nongyn Nongyn 2 vs. nongyn Nongyn overall
2 3 3
2 3 3
2 3 3
2 3 3
0.5930 0.4011 0.4302 0.4643 0.5508 0.3600 0.5200 0.4734
(0.1507) (0.1769) (0.1677) (0.1091) (0.1616) (0.1765) (0.1630) (0.1091)
0.2272 – 0.0040 0.1630 0.0661 0.2104 0.3103 0.1034 0.1642
(0.1086) (0.1763) (0.1445) (0.0957) (0.1334) (0.1678) (0.1396) (0.1020)
0.2667 0.4842 0.3548 0.3318 0.4197 0.2804 0.2376 0.2841
(0.1401) (0.1675) (0.1250) (0.1091) (0.1710) (0.1393) (0.1584) (0.1091)
1.0000 0.3636 0.3636 0.4615 0.1429 0.6216 0.0000 0.1964
(0.0000) (0.2678) (0.2678) (0.1091) (0.1690) (0.1692) (0.15512) (0.1091)
DISCUSSION The HLRCC syndrome predisposes affected individuals to multiple LMs of the skin and uterus and to an aggressive form of RCC. Several authors have stressed the role of pathologists in identifying patients likely to have this syndrome by recognizing the hallmark features of such tumors, whether RCC or LMs, namely the presence of orangeophilic/eosinophilic nucleoli with clear perinucleolar halos in the majority of tumor cells (11,13). Recently, 3 of us (M.E.B., I.P.T., and P.J.P.) showed that immunohistochemical detection of aberrant succination of proteins is a robust biomarker of germline mutation of the fumurate hydratase gene (10). In that study we showed that the 2SC bioassay was equivalent or superior to conventional screening methods as a diagnostic test, and was sufficiently sensitive and specific for use as a routine clinical biomarker.
IDENTIFYING HEREDITARY LEIOMYOMATOSIS AND RENAL CELL CARCINOMA SYNDROME
563
TABLE 2. Final Morphologic Appraisal for Each Participant n (%) Gyn 1 Cases correctly identified Hereditary cases identified Sporadic cases identified Overall % equivocal cases Equivocal cases which are actually hereditary Equivocal cases which are actually sporadic Hereditary cases identified correctly or as equivocal
11/28 4/18 7/10 8/28 6/18 2/10 10/18
(39) (22) (70) (29) (33) (20) (56)
Increasingly, along with prognostic and predictive morphologic features and biomarkers, morphologic features and/or biomarkers suggesting genetic predilection are being incorporated into pathology reports, to allow individualized patient management. Although predictive biomarker assay and reporting protocols have rapidly been standardized and integrated to ensure an appropriate clinical response, reporting and integration of morphologic features and biomarkers of genetic predilection in patients with neoplasms is less well developed (14). Recognizing the possibility or likelihood of such syndromes is not only important for screening for metachronous or synchronous tumors and for identification of at risk and not-at-risk family members with subsequent tumor screening, but increasingly therapeutics may be predicated on the presence of these syndromes (15,16). Morphologybased genetic referral schemas have been developed to supplement the genealogy-based schemas which are often insensitive and implemented poorly in the clinical setting (17). However, as with clinical schemas, we need to be cautious and consider the reproducibility, sensitivity, and specificity of morphologic criteria as well as issues related to implementation as pathology reports grow in complexity (18). Some schemas are predicated on intrinsic basic morphology criteria such as tumor subtype in ovarian cancer, where it is recommended in some regions, for example, Ontario, Canada, that all patients with high-grade serous carcinoma of the ovary/tube/peritoneum undergo BRCA testing (2). This is only feasible now that the contemporary approach to ovarian carcinoma subtype allocation is highly reproducible and that reliable biomarker validation of cell type is available in difficult cases (19,20). With the diminishing cost of sequencing, one could argue for germline sequencing of a panel of hereditary ovarian cancer genes on all ovarian carcinomas. Identification of BRCA alteration may serve as a quality assurance tier for cell type allocation with any nonserous case in a BRCA 1/2 patient being reviewed.
Gyn 2 8/28 5/18 3/10 16/28 9/18 7/10 14/18
(29) (28) (30) (57) (50) (70) (78)
Gyn 3 14/28 7/18 7/10 2/28 1/18 1/10 8/18
(50) (39) (70) (7) (6) (10) (44)
Nongyn 1
Nongyn 2
Nongyn 3
15/28 8/18 7/10 8/28 7/18 1/10 15/18
10/28 2/18 8/10 2/28 2/18 0/10 4/18
14/28 4/18 10/10 8/28 8/18 0/10 12/18
(54) (44) (70) (29) (39) (10) (83)
(36) (11) (80) (7) (11) (0) (22)
(50) (22) (100) (29) (44) (0) (67)
In relation to CRC, the CAP has stated that pathologists need to recognize the histologic and clinical features that should prompt at least a recommendation for MMR testing (5). Several groups recognizing low overall referral rates and cognizant of lack of standardization of such features have suggested reflex testing of CRC and EC with MMR-IHC (21,22). A reflex testing policy represents a marked change in practice with such testing traditionally being clinician initiated (6). In many ways, the use of morphologic features and the development of a biomarker to identify HLRCC is analogous to LS. As with LS, although autosomal dominant, HLRCC shows variable penetrance limiting the use of family history. Successful identification of this syndrome in patients presenting with uterine LMs (patients with HLRCC often present initially with uterine LMs) will allow screening for RCC, potentially reducing mortality. The proposed morphologic criteria for LS and HLRCC are not routinely reported in EC, LM, or RCC. Similar to microsatellite testing, the use of 2SC immunohistochemistry to identify this syndrome is not a direct test of the altered protein but an assay of the functional consequence of the genomic alteration. Unlike LS in which the MMR protein can be assayed using immunohistochemisty, anti-FH antibodies will detect both mutant and normal protein, rendering it ineffective as a screening tool. This is the first blinded case-control study to assess the reproducibility and diagnostic accuracy of proposed morphologic features to identify uterine LMs in patients with HLRCC. Despite previous studies describing the presence of various morphologic features in HLRCC-associated LMs, our study definitively demonstrates that the morphologic features we assessed are largely irreproducible among pathologists, regardless of whether they are specialist gynecologic pathologists or not. None of the morphologic features assessed were sufficiently robust to allow for the development of a predictive algorithm Int J Gynecol Pathol Vol. 33, No. 6, November 2014
564
ALSOLAMI ET AL.
FIG. 1. (A) A high-power view highlighting the characteristic large orangiophilic nucleoli and perinucleolar halos in an hereditary leiomyomatosis and renal cell carcinoma syndrome (HLRCC)-associated leiomyoma (40 ). (B) HLRCC-associated leiomyoma showing characteristic nuclear features and numerous pink globules. Five of 6 participants correctly assessed this as a hereditary case (40 ). (C) HLRCC-associated leiomyoma lacking the characteristic cytologic features. All participants assessed this incorrectly as a sporadic case (20 ). (D) In other areas of the hereditary case shown in (C), the tumor does show pericytomatous vasculature.
Int J Gynecol Pathol Vol. 33, No. 6, November 2014
IDENTIFYING HEREDITARY LEIOMYOMATOSIS AND RENAL CELL CARCINOMA SYNDROME
565
screening tool to triage patients for further family/ personal history inquiry or formal genetic counseling. The poor performance of the morphologic criteria is particularly concerning in view of the enrichment of this cohort for mutation-positive cases. We acknowledge some limitations to this study. For the majority of hereditary LMs, only 1 slide was viewed by the participants. The use of digitized images was new for some of the participants. The prestudy training offered to study participants was limited consisting of written description of such features and attached articles with images. However, in reality most pathologists would only encounter the description of these features through articles, with few having access to confirmed hereditary or sporadic cases. Further, refinement of morphologic criteria to identify HLRCC-associated tumors may be possible. In a recent abstract, Reyes and colleagues reviewed the morphologic features of 10 uterine LMs with 2SC-IHC overexpression or FH germline mutation. They highlighted additional syndrome-associated morphologic features, namely pericytomatous vasculature and highly fibrillary cytoplasm with pink globules, which were present in some of these neoplasms (23). However, to achieve such refinements, well-annotated cohort-control cases need to be available, both for analysis and teaching, and rigorous study design must be employed. As has been recently proposed, borrowing from the genomics community, publically available annotated digital slide repositories would be critical in such an endeavor (24). Alternatively, mirroring the development of reflex testing in CRC and EC with MMR-IHC, a similar strategy could be deployed with RCCs and LMs. Appropriate cost-effectiveness studies would be needed especially in view of the ubiquitous nature of uterine LMs and the relatively lower incidence of
FIG. 2. (A) A hereditary leiomyomatosis and renal cell carcinoma syndrome (HLRCC)-associated leiomyoma in which there are focal prominent nucleoli with halos (4), whereas other nuclei appear vesicular. Four of 6 participants equivocated on this case and 2 misdiagnosed it as a sporadic case (40 ). (B) A sporadic leiomyoma with vesicular nuclei and small nucleoli (4). Two participants assessed this case as being hereditary, 2 as sporadic and 2 equivocated (40 ).
to triage cases for 2SC immunohistochemistry. Predictably, the subjective final morphologic assessment (hereditary or equivocal vs. sporadic) lacked adequate sensitivity and specificity to be used as a
TABLE 3. Sensitivity and Specificity Achieved by Each Participant in the Final Morphologic Appraisal
Reader Gyn 1 Gyn 2 Gyn 3 Nongyn 1 Nongyn 2 Nongyn 3
Correct (HLRCC) Gold Standard N = 18
Correct (Sporadic) Gold Standard N = 10
False Positive
False Negative
10 14 8 15 4 12
7 3 7 7 8 10
3 7 3 3 14 0
8 4 10 3 2 6
Sensitivity (95% CI) 55.6 77.8 44.4 83.3 22.2 66.7
(30.8–78.4) (52.4–93.5) (21.6–69.2) (58.6–96.2) (6.6–47.6) (41.0–86.6)
Specificity (95% CI) 70.0 30.0 70 70.0 80.0 100
(34.8–93.0) (7.0–65.2) (34.8–93.0) (34.8–93.0) (44.4–96.9) (69.0–100)
Cases assessed as being hereditary or equivocal were combined and the binarized scores [hereditary (1) or equivocal (2) versus sporadic (0)] compared with the gold standard, either germline status or 2SC immunohistochemistry for the control group. CI indicates confidence interval; HLRCC, hereditary leiomyomatosis and renal cell carcinoma.
Int J Gynecol Pathol Vol. 33, No. 6, November 2014
566
ALSOLAMI ET AL.
HLRCC in the LM and RCC population compared with LS and BRCA in the CRC, EC, and ovarian carcinoma population. The age range of HLRCC uterine LMs is 18 to 52 yr with a mean age 30 yr at presentation which contrasts with a mean age of 45 yr in the sporadic population (25–27). Reflex screening of all uterine LM specimens in patients 30 yr or less or in those who have a family history suggestive of HLRCC regardless of age is 1 strategy that could be critically assessed, for example. According to our data, using the currently proposed morphologic criteria to initiate 2SC-IHC testing lacks the requisite sensitivity and specificity. Germane to all hereditary predisposition biomarker or morphologic schema, it is imperative that the vicarious nature of the pathologist-patient relationship is addressed. We need to ensure that pathologist recommendations for screening and results of reflex-performed screening tests are appropriately followed up on by the clinician. Overall genetic referral rates for patients with serous carcinoma in Ontario still remain low and Sanchez et al (28) highlighted that recommendations to screen for LS rather than actual MMR-IHC results were less likely to result in genetic referral of the patient. Finally, we need to ensure appropriate feedback to the reporting pathologists as to the outcome of such recommendations, for clinical, quality assurance, and educational purposes. The burgeoning of hereditary predisposition morphology and/or biomarker schemas represents an exciting development in pathology and is yet another area in which laboratory physicians can midwife personalized medicine. As illustrated in this study, morphologic schema will need to be rigorously tested before implementation. Furthermore, as with predictive biomarkers, standardization of both technical and clinical aspects of hereditary predisposition biomarkers are needed. REFERENCES 1. Jenkins MA, Hayashi S, O’Shea AM, et al. Pathology features in Bethesda guidelines predict colorectal cancer microsatellite instability: a population-based study. Gastroenterology 2007;133:48–56. 2. Schrader KA, Hurlburt J, Kalloger SE, et al. Germline BRCA1 and BRCA2 mutations in ovarian cancer: utility of a histologybased referral strategy. Obstet Gynecol 2012;120:235–240. 3. Steinhagen E, Shia J, Markowitz AJ, et al. Systematic immunohistochemistry screening for Lynch syndrome in early age-of-onset colorectal cancer patients undergoing surgical resection. J Am Coll Surg 2012;214:61–67. 4. Beamer LC, Grant ML, Espenschied CR, et al. Reflex immunohistochemistry and microsatellite instability testing of colorectal tumors for Lynch syndrome among US cancer
Int J Gynecol Pathol Vol. 33, No. 6, November 2014
5. 6. 7.
8.
9.
10.
11.
12.
13.
14.
15. 16.
17.
18. 19. 20. 21. 22.
programs and follow-up of abnormal results. J Clin Oncol 2012;30:1058–1063. Turner JW, Frederick LB, Bloom KJ, et al. Prognostic Uses of MSI Testing. 2011. Available at: http://www.cap.org/apps/docs/ committees/technology/microsatellite_testing.pdf. Kalloger SE, Allo G, Mulligan AM, et al. Use of mismatch repair immunohistochemistry and microsatellite instability testing: exploring Canadian practices. Am J Surg Pathol 2012;36:560–569. Tomlinson IP, Alam NA, Rowan AJ, et al. Germline mutations in FH predispose to dominantly inherited uterine fibroids, skin leiomyomata and papillary renal cell carcinoma. Nat Genet 2002;30:406–410. Alam NA, Rowan AJ, Wortham NC, et al. Genetic and functional analyses of FH mutations in multiple cutaneous and uterine leiomyomatosis, hereditary leiomyomatosis and renal cancer, and fumarate hydratase deficiency. Hum Mol Genet 2003;12:1241–1252. Bayley JP, Launonen V, Tomlinson IP. The FH mutation database: an online database of fumarate hydratase mutations involved in the MCUL (HLRCC) tumor syndrome and congenital fumarase deficiency. BMC Med Genet 2008;9:20. Bardella C, El-Bahrawy M, Frizzell N, et al. Aberrant succination of proteins in fumarate hydratase-deficient mice and HLRCC patients is a robust biomarker of mutation status. J Pathol 2011;225:4–11. Garg K, Tickoo SK, Soslow RA, et al. Morphologic features of uterine leiomyomas associated with hereditary leiomyomatosis and renal cell carcinoma syndrome: a case report. Am J Surg Pathol 2011;35:1235–1237. Sanz J, Teller L, Vockes C, et al. Genetic alterations of uterine fibroids in hereditary leiomyomatosis and renal cancer (HLRCC) syndrome. Modern Pathology Annual Meetings Abstracts 223 A. Sanz-Ortega J, Vockes C, Stratton P, et al. Morphologic and molecular characteristic of uterine leiomyomas in hereditary leiomyomatosis and renal cancer (HLRCC) syndrome. Am J Surg Pathol 2013;37:74–80. Wolff AC, Hammond ME, Schwartz JN, et al. American Society of Clinical Oncology/College of American Pathologists guideline recommendations for human epidermal growth factor receptor 2 testing in breast cancer. J Clin Oncol 2007;25:118–145. Hewish M, Lord CJ, Martin SA, et al. Mismatch repair deficient colorectal cancer in the era of personalized treatment. Nat Rev Clin Oncol 2010;7:197–208. Martin SA, McCabe N, Mullarkey M, et al. DNA polymerases as potential therapeutic targets for cancers deficient in the DNA mismatch repair proteins MSH2 or MLH1. Cancer Cell 2010;17:235–248. Ryan P, Mulligan AM, Aronson M, et al. Comparison of clinical schemas and morphologic features in predicting Lynch syndrome in mutation-positive patients with endometrial cancer encountered in the context of familial gastrointestinal cancer registries. Cancer 2012;118:681–688. Clarke BA, Cooper K. Identifying Lynch syndrome in patients with endometrial carcinoma: shortcomings of morphologic and clinical schemas. Adv Anat Pathol 2012;19:231–238. Ko¨bel M, Kalloger SE, Baker PM, et al. Diagnosis of ovarian carcinoma cell type is highly reproducible: a transCanadian study. Am J Surg Pathol 2010;34:984–993. Kalloger SE, Ko¨bel M, Leung S, et al. Calculator for ovarian carcinoma subtype prediction. Mod Pathol 2011;24:512–521. Boland CR, Shike M. Report from the Jerusalem workshop on Lynch syndrome-hereditary nonpolyposis colorectal cancer. Gastroenterology 2010;138:2197. Palomaki GE, McClain MR, Melillo S, et al. EGAPP supplementary evidence review: DNA testing strategies aimed at reducing morbidity and mortality from Lynch syndrome. Genet Med 2009;11:42–65.
IDENTIFYING HEREDITARY LEIOMYOMATOSIS AND RENAL CELL CARCINOMA SYNDROME 23. Reyes C, Karamurzin Y, Frizzell N, et al. Uterine smooth muscle tumors with features suggesting hereditary leiomyomatosis and renal carcinoma (HLRCC) syndrome: detailed morphologic analysis and correlation with 2SC imnohistochemistry. Mod Pathol 2013;1224:293. 24. Hipp JD, Lucas DR, Emmert-Buck MR, et al. Digital slide repositories for publications: lessons learned from the microarray community. Am J Surg Pathol 2011;35:783–786. 25. Farquhar CM, Steiner CA. Hysterectomy rates in the United States 1990-1997. Obstet Gynecol 2002;99:229–234.
567
26. Toro JR, Nickerson ML, Wei MH, et al. Mutations in the fumarate hydratase gene cause hereditary leiomyomatosis and renal cell cancer in families in North America. Am J Hum Genet 2003;73:95–106. 27. Alam NA, Barclay E, Rowan AJ, et al. Clinical features of multiple cutaneous and uterine leiomyomatosis: an underdiagnosed tumor syndrome. Arch Dermatol 2005;141:199–206. 28. Sanchez JA, Vogel JD, Kalady MF, et al. Identifying Lynch syndrome: we are all responsible. Dis Colon Rectum 2008;51:1750–1756.
Int J Gynecol Pathol Vol. 33, No. 6, November 2014
