Key Points
Question
Is there intrapatient concordance of mismatch repair protein (MMRP) immunohistochemical (IHC) staining pattern between different sebaceous neoplasms in patients with known Muir-Torre syndrome (MTS)?
Findings
In this case series of 11 patients with MTS, there was high intrapatient concordance of MMRP IHC staining between different sebaceous neoplasms of individual patients (range, 2-4 lesions per patient), with 36 of 38 lesions matching the confirmed germline mutation (sensitivity, 94.7%; 95% CI, 82.3%-99.4%). Extrafacial lesions (n = 16) corresponded to patients’ known germline mutation in 100% of cases.
Meaning
These findings may suggest that MMRP IHC has high reproducibility across sebaceous lesions within an individual patient with MTS, and that this test may be used to support a diagnosis of MTS when clinical suspicion is strong.
This case series assesses the concordance of mismatch repair protein immunohistochemical staining between different sebaceous neoplasms in patients with Muir-Torre syndrome.
Abstract
Importance
Appropriate use criteria for Muir-Torre syndrome (MTS) screening suggest that mismatch repair protein (MMRP) immunohistochemical (IHC) testing is usually appropriate in patients with 2 or more sebaceous neoplasms (SNs). While MTS is known to be caused by a germline mutation in mismatch repair genes, data are limited as to whether individual sebaceous tumors in these patients with multiple lesions show identical MMRP IHC staining patterns.
Objective
To determine concordance of MMRP IHC staining patterns in lesions of patients with MTS who have multiple SNs.
Design, Setting, and Participants
This retrospective single-center case series evaluated 38 SNs in 11 patients with MTS confirmed by genetic testing for MMRP IHC staining patterns. Tumor sites were classified as either facial or extrafacial. Data were collected between January 1, 2007, and January 1, 2018.
Main Outcomes and Measures
In each patient, MMRP IHC staining patterns for SNs were compared with one another to evaluate intrapatient concordance between lesions, and to the patient’s known germline mutation.
Results
A total of 11 patients (7 women and 4 men) with MTS, with a mean (SD) age of 59.3 (10.6) years at time of SN biopsy, were identified. There was high concordance between MMRP IHC staining results (2-4 lesions per patient) and the patient’s mutation status, with 36 of 38 total lesions (95%) matching (sensitivity, 94.7%; 95% CI, 82.3%-99.4%). Extrafacial site tumors represented 16 of 38 total lesions (42%) and demonstrated 100% concordance of IHC results to germline mutation. Only 1 of 11 patients (9%) demonstrated discordant results, with both lesions in this patient occurring on a facial site.
Conclusions and Relevance
In patients with known MTS, SNs present with highly concordant MMRP IHC staining profiles across multiple lesions. There is also a strong association with underlying germline mutations. A diagnosis of MTS might be supported by MMRP IHC when the pretest probability is high.
Introduction
Sebaceous neoplasms (SNs), including sebaceous adenoma, sebaceous epithelioma, and sebaceous carcinoma, are defining features of Muir-Torre syndrome (MTS), a variant of Lynch syndrome (LS).1 Universal screening of SNs with mismatch repair protein (MMRP) immunohistochemistry (IHC) has previously been recommended by some physicians to detect potential probands2 but has been debated given the high costs and variable sensitivities of detecting MTS, ranging from 25% to 85% depending on the IHC panel used.3,4,5 Sensitivity may increase with targeted screening, such as when a patient (of any age) presents with multiple SNs.6 Because MTS is caused by germline genetic mutations, it is presumed that individual SNs should harbor the same mutation. However, data evaluating intrapatient concordance of MMRP IHC staining patterns between different SNs are limited. This case series examined patients with MTS who had multiple SNs to determine the concordance of MMRP IHC staining patterns between lesions and with germline mutation.
Methods
This retrospective study was approved by the institutional review board of the University of Pennsylvania (protocol No. 828426). Need for informed consent was waived because the research was deemed to be of minimal risk to participants. Patients with MTS confirmed by germline genetic testing were identified at the Hospital of the University of Pennsylvania between January 1, 2007, and January 1, 2018. Patients were considered for inclusion in the study if they had more than 2, but ideally more than 3, SNs biopsied for histopathology documented in NewPath, the dermatopathology database of the Department of Dermatology at the Hospital of the University of Pennsylvania. Hematoxylin-eosin–stained sections of SNs from these patients were reviewed to confirm the diagnosis. Cases were excluded if the tissue block was inadequate for further assessment. Mismatch repair protein IHC staining, namely for MLH1, PMS2, MSH2, and MSH6 (Biocare Medical clones G168-15, A16-4, FE11, and 44, respectively, ready-to-use), was performed on a Bond-III IHC stainer model No. 3212367 (Leica Biosystems). Evaluation of nuclear staining was determined by a board-certified dermatopathologist (E.Y.C.) and a dermatopathology fellow (K.J.G.). Equivocal cases were reviewed by another board-certified dermatopathologist (J.T.S.). Immunohistochemistry staining results for individual lesions were compared with the patient’s known germline mutation to evaluate concordance rates, as well as with each other.
Results
We identified 20 patients with 2 or more biopsy-proven SNs (Figure 1) who otherwise met clinical criteria for MTS. A total of 11 patients (7 women and 4 men; mean [SD] age at time of SN biopsy, 59.3 [10.6] years) had confirmation of MTS via genetic testing (Table). Mutations of MSH2 type were most common (8 of 11; 73%), followed by MSH6 (2 of 11; 18%) and MLH1 (1 of 11; 9%). Reports of the specific type of mutation (eg, deletion, substitution) were available in 7 of the 11 total patients. A total of 38 specimens (28 sebaceous adenoma, 4 sebaceous endothelioma, and 6 unclassifiable SN) were obtained (median [range] specimens per patient, 3 [2-5]). At the time of diagnosis of first SN, only 4 patients were older than 60 years (median [range] at time of diagnosis, 60 [33-78] years). A total of 7 of 11 patients (64%) presented with SNs prior to diagnosis of MTS. A total of 3 of 11 patients (27%) had 1 keratoacanthoma, and 4 of 11 (36%) had multiple keratoacanthomas. Colorectal cancer developed in 4 of 11 (36%) patients. Of 7 female patients, 4 had endometrial cancer and 2 had ovarian cancer. One patient had bladder cancer. A total of 2 patients had a family history of multiple LS-associated internal malignant neoplasms.
Figure 1. Biopsy Specimen of Sebaceous Adenoma from Patient 5.
Histopathologic image of a yellowish papule demonstrates lobules of bland adipocytes arising from the epidermis (hematoxylin-eosin, original magnification ×45).
Table. Patient Demographic Characteristics and Mismatch Repair Protein Immunohistochemistry Staining Results.
| Patient; mutation | Sebaceous neoplasm | Site | Staining resulta | Concordance with germline mutation | |||
|---|---|---|---|---|---|---|---|
| MLH1 | MSH2 | MSH6 | PMS2 | ||||
| 1; MSH2 (338delCA) | Atypical SN | EF | + | − | − | + | Yes |
| SE | EF | + | − | − | + | Yes | |
| Cystic SN | EF | + | − | − | + | Yes | |
| 2; MSH2 (2222del4) | Benign SN | EF | + | − | − | + | Yes |
| SA | EF | + | − | − | + | Yes | |
| SA | EF | + | − | − | + | Yes | |
| 3; MSH6 | SA | F | + | + | − | + | Yes |
| SA | F | + | + | − | + | Yes | |
| SE | F | + | + | − | + | Yes | |
| 4; MSH2 (IVS4 + 2T>G) | SA | EF | + | − | − | + | Yes |
| SA | F | + | − | − | + | Yes | |
| SA | F | + | − | − | + | Yes | |
| 5; MLH1 (R226L) | SA | EF | − | + | + | − | Yes |
| SA | F | − | + | + | − | Yes | |
| SA | F | − | + | + | − | Yes | |
| SA | F | + | + | + | + | No | |
| SA | F | + | + | + | + | No | |
| 6; MSH2 (IVTS5 + 3A>T) | Surface of SN | F | + | − | − | + | Yes |
| SA | EF | + | − | − | + | Yes | |
| Surface of SN | F | + | − | − | + | Yes | |
| Cystic SA | EF | + | − | − | + | Yes | |
| 7; MSH2 | SN | EF | + | − | − | + | Yes |
| SA | EF | + | − | − | + | Yes | |
| SA | F | + | − | − | + | Yes | |
| Surface of SN | F | + | − | − | + | Yes | |
| 8; MSH2 (c.39 48del10) | SA | F | + | − | − | + | Yes |
| SA | F | + | − | − | + | Yes | |
| 9; MSH2 | SA | F | + | − | − | + | Yes |
| SA | F | + | − | − | + | Yes | |
| SA | EF | + | − | − | + | Yes | |
| 10; MSH6 (p.R1035X) | SA | F | + | + | − | + | Yes |
| SE | F | + | + | − | + | Yes | |
| SE | F | + | + | − | + | Yes | |
| SA | F | + | + | − | + | Yes | |
| 11; MSH2 | SA | F | + | − | − | + | Yes |
| SA | EF | + | − | − | + | Yes | |
| SA | EF | + | − | − | + | Yes | |
| SA | EF | + | − | − | + | Yes | |
Abbreviations: EF, extrafacial; F, facial; SA, sebaceous adenoma; SE, sebaceous epithelioma; SN, sebaceous neoplasm.
+ indicates preserved staining; − indicates absent/loss of staining.
As seen in the Table, concordance of MMRP staining pattern to known germline mutation was high, in 36 of 38 cases (sensitivity, 94.7%; 95% CI, 82.3%-99.4%). Discordance was identified in only 1 of 11 patients (9%), with 2 lesions in this patient showing discordance with germline mutation occurring at facial sites. Extrafacial SNs comprised 16 of 38 (42%) total cases and demonstrated 100% concordance with germline mutation. Immunohistochemistry demonstrated loss of both MSH2 and MSH6 staining in patients with MSH2 germline mutation, but only loss of MSH6 with retained MSH2 staining in patients with MSH6 germline mutation (Table). In the patient with MLH1 mutation, loss of both MLH1 and PMS2 staining was identified (Figure 2).
Figure 2. Mismatch Repair Protein Immunohistochemistry Staining in a Sebaceous Adenoma Biopsy Specimen from Patient 5.
There is loss of MLH1 (A, B) and PMS2 (C, D) nuclear staining in the basaloid cells of the sebaceous adenoma. Note the retained nuclear staining in the epidermis compared with loss of MLH1 and PMS2 nuclear staining in the sebaceous lobules. There is positive, retained nuclear staining in MSH2 (E, F) and MSH6 (G, H).
Discussion
In this study of 11 patients, the median (range) age at time of SN biopsy was 60 (33-78) years. Based on appropriate use criteria for MTS screening by IHC in patients 60 years or younger, it is deemed usually appropriate (the strongest consensus agreement category) to screen patients with multiple SNs (>2) or those with 1 SN on an extrafacial site.6 Everett et al7 demonstrated that patients with MTS are more likely to present with multiple SNs and at a younger age. However, data were limited as to whether lesions in the same individual show concordant MMRP IHC staining patterns. Retrospective studies looking at the utility of universal MTS screening, which included patients with multiple SNs, did not specify the MMRP IHC staining pattern for each lesion.2,8 A study of synchronous/metachronous LS-associated neoplasms (internal and cutaneous) examined 4 of 13 patients who had 2 SNs evaluated by MMRP IHC.9 Of those 4 patients, 3 showed concordant staining patterns while 1 patient with germline MLH1 showed intact MMRP in one lesion and loss of MLH1/PMS2 in another.9 Examination of internal (intestinal, uterine) LS-associated neoplasms in the same study showed discordant MMRP IHC with germline mutations in 2 of 9 cases.
The present study demonstrated that most SNs in patients with MTS show MMRP IHC staining that reflects their germline genetic mutations. This high concordance (94.7%) of MMRP IHC to germline mutation provides further support for screening patients who have multiple SNs with MMRP IHC. Only 1 in 11 patients (9%) demonstrated discordance between IHC staining results and germline mutation. This discordance arose in facial lesions in a patient with an MLH1 intron-adjacent substitution mutation. This mutation may have resulted in a biologically inactive protein, which retained the moiety critical for detection by MMRP IHC.
In a prior study of 216 SNs, those located on nonfacial sites had a high specificity for detecting MMRP abnormalities (96%), but a low sensitivity (41%).2 Although the present study’s sample size is small, the 100% concordance between IHC results and patient’s germline mutation for extrafacial tumors provides further evidence for IHC testing any tumor at a nonfacial site for MTS screening.
Similar to data previously published in the gastroenterology literature, the data in this study may also support the use of a 2-antibody panel (MSH6 and PMS2).10 Antibodies MSH6 and PMS2 are obligate partners of MSH2 and MLH1, respectively.11 Loss of either MSH6 or PMS2 should allow detection of loss of MSH2 or MLH1, respectively.
Limitations
This study had limitations, because it was a small retrospective study of patients with known MTS from 1 academic center. Larger prospective and/or retrospective studies in all patients with multiple SNs, regardless of known MMRP mutation status, are needed to evaluate the concordance of MMRP IHC staining patterns and thus validate the findings of this study. The cause of the discordant IHC results in 1 patient is unclear. While this may be due to the specific mutation seen in this patient, additional studies including patients with similar substitutions are needed to assess if it is an issue with MMRP IHC reliability in general, or with the inability of IHC to detect certain mutations. Regarding the latter possibility, it should be noted that a small study of 13 patients suggested that a subset of patients with MTS do not appear to have detectable microsatellite instability, as assessed by MLH1 and MSH2 IHC.12
Conclusions
In patients with MTS, there is high intrapatient concordance of MMRP IHC between different SNs and with the patient’s underlying germline mutation, particularly in lesions on extrafacial sites. The findings of this study may suggest that MMRP IHC has high reproducibility across SNs within individual patients with MTS and, moreover, that this test may be used to support a diagnosis of MTS when clinical suspicion is high.
References
- 1.Schwartz RA, Torre DP. The Muir-Torre syndrome: a 25-year retrospect. J Am Acad Dermatol. 1995;33(1):90-104. doi: 10.1016/0190-9622(95)90017-9 [DOI] [PubMed] [Google Scholar]
- 2.Jessup CJ, Redston M, Tilton E, Reimann JD. Importance of universal mismatch repair protein immunohistochemistry in patients with sebaceous neoplasia as an initial screening tool for Muir-Torre syndrome. Hum Pathol. 2016;49:1-9. doi: 10.1016/j.humpath.2015.10.005 [DOI] [PubMed] [Google Scholar]
- 3.Cesinaro AM, Ubiali A, Sighinolfi P, Trentini GP, Gentili F, Facchetti F. Mismatch repair proteins expression and microsatellite instability in skin lesions with sebaceous differentiation: a study in different clinical subgroups with and without extracutaneous cancer. Am J Dermatopathol. 2007;29(4):351-358. doi: 10.1097/DAD.0b013e318057713c [DOI] [PubMed] [Google Scholar]
- 4.Erten MZ, Fernandez LP, Ng HK, et al. Universal versus targeted screening for Lynch syndrome: comparing ascertainment and costs based on clinical experience. Dig Dis Sci. 2016;61(10):2887-2895. doi: 10.1007/s10620-016-4218-y [DOI] [PubMed] [Google Scholar]
- 5.Roberts ME, Riegert-Johnson DL, Thomas BC, et al. Screening for Muir-Torre syndrome using mismatch repair protein immunohistochemistry of sebaceous neoplasms. J Genet Couns. 2013;22(3):393-405. doi: 10.1007/s10897-012-9552-4 [DOI] [PubMed] [Google Scholar]
- 6.Vidal CI, Sutton A, Armbrect EA, et al. ; Task Force/Committee Members; Rating Panel . Muir-Torre syndrome appropriate use criteria: effect of patient age on appropriate use scores. J Cutan Pathol. 2019;46(7):484-489. doi: 10.1111/cup.13459 [DOI] [PubMed] [Google Scholar]
- 7.Everett JN, Raymond VM, Dandapani M, et al. Screening for germline mismatch repair mutations following diagnosis of sebaceous neoplasm. JAMA Dermatol. 2014;150(12):1315-1321. doi: 10.1001/jamadermatol.2014.1217 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Schon K, Rytina E, Drummond J, et al. Evaluation of universal immunohistochemical screening of sebaceous neoplasms in a service setting. Clin Exp Dermatol. 2018;43(4):410-415. doi: 10.1111/ced.13359 [DOI] [PubMed] [Google Scholar]
- 9.Roth RM, Haraldsdottir S, Hampel H, Arnold CA, Frankel WL. Discordant mismatch repair protein immunoreactivity in Lynch syndrome-associated neoplasms: a recommendation for screening synchronous/metachronous neoplasms. Am J Clin Pathol. 2016;146(1):50-56. doi: 10.1093/ajcp/aqw067 [DOI] [PubMed] [Google Scholar]
- 10.O’Regan T, Chau K, Tatton M, Smith T, Parry S, Bissett I. Immunochemistry screening for Lynch syndrome in colorectal adenocarcinoma using an initial two antibody panel can replace a four antibody panel. N Z Med J. 2013;126(1382):70-77. [PubMed] [Google Scholar]
- 11.Lee JB, Litzner BR, Vidal CI. Review of the current medical literature and assessment of current utilization patterns regarding mismatch repair protein immunohistochemistry in cutaneous Muir-Torre syndrome-associated neoplasms. J Cutan Pathol. 2017;44(11):931-937. doi: 10.1111/cup.13010 [DOI] [PubMed] [Google Scholar]
- 12.Entius MM, Keller JJ, Drillenburg P, Kuypers KC, Giardiello FM, Offerhaus GJ. Microsatellite instability and expression of hMLH-1 and hMSH-2 in sebaceous gland carcinomas as markers for Muir-Torre syndrome. Clin Cancer Res. 2000;6(5):1784-1789. [PubMed] [Google Scholar]