Observational prospective cohort study of patients ...

Downloaded from bjo.bmj.com on November 9, 2012 - Published by group.bmj.com

BJO Online First, published on October 31, 2012 as 10.1136/bjophthalmol-2012-302443 Clinical science

Observational prospective cohort study of patients with newly-diagnosed ocular sebaceous carcinoma Mahiul M K Muqit,1 Barny Foot,2 Stephen J Walters,3 Hardeep S Mudhar,4 Fiona Roberts,5 Ian G Rennie6 ▸ Additional supplementary files are published online only. To view these files please visit the journal online (http://dx.doi. org/10.1136/bjophthalmol2012-302443). 1 Oxford Eye Hospital, The John Radcliffe Hospital, Oxford University Hospitals NHS Trust, Oxford, UK 2 The British Ophthalmological Surveillance Unit (BOSU), The Royal College of Ophthalmologists, London, UK 3 Department of Medical Statistics, School of Health and Related Research, University of Sheffield, Sheffield, UK 4 National Specialist Ophthalmic Pathology Service (NSOPS), Department of Histopathology, The Royal Hallamshire Hospital NHS Trust, Sheffield, UK 5 Scottish Regional Ophthalmic Pathology Service, Department of Pathology, Western Infirmary, Glasgow, UK 6 Department of Ophthalmology and Orthoptics, The Royal Hallamshire Hospital NHS Trust, Sheffield, UK

Correspondence to Dr Mahiul MK Muqit, Oxford Eye Hospital, The John Radcliffe Hospital, Oxford University Hospitals NHS Trust, Oxford OX39DU, UK; [email protected] This work was presented in May 2011 at the Royal College of Ophthalmologists Annual Congress, Birmingham, UK Accepted 30 September 2012

ABSTRACT Purpose To investigate the epidemiology and clinicopathological management for ocular sebaceous carcinoma (OSC) in the UK. Methods Observational prospective cohort study of patients with newly-diagnosed OSC. The British Ophthalmological Surveillance Unit captured incident cases of OSC between 2008 and 2010. Incident and 6-month follow-up questionnaires from reporting ophthalmologists captured OSC demographic and clinical data. Results Data were available on 51 patients with unilateral OSC (response rate 85%). The UK estimated annual incidence was 0.41 cases per million population (95% CI 0.31 to 0.54). Median age was 70 years (SD 14, range 28–98) with 57% women. OSC location was upper lid (54%), lower lid (20%), multicentric (14%) and caruncle (12%). Most common misdiagnoses included chalazion (42%), basal cell carcinoma (30%) and blepharoconjunctivitis (16%), with median delay in diagnosis of 10 months (SD 9, range 0.5–36). Specialist ophthalmic pathologists performed diagnostics in 62%, with pagetoid/intraepithelial spread present in 39%. Misdiagnosis of chalazion ( p=0.019) and pagetoid tumour spread ( p=0.016) was associated with a significant diagnostic delay (one-way ANOVA/R2). Primary surgical management involved excision with reconstruction (49%), primary exenteration (10%) and Mohs surgery (8%). There were three deaths (out of 51) during the study period; one patient died of OSC-related disease and the other two due to other causes. Conclusions This population-based prospective study confirms OSC as a rare cancer in the UK. Masquerade syndromes result in significant diagnostic delays and increase the risk of pagetoid tumour spread. There is considerable UK variation in pathological and surgical management, and ocular reconstruction and radical surgery is often required for OSC due to delayed presentation.

Ocular sebaceous carcinoma (OSC) is a rare tumour that accounts for one in 2000 of cutaneous malignancies.1 OSC is an important ocular malignancy as it is associated with significant morbidity and mortality. The tumour clinical appearances are seldom pathognomonic and OSC is commonly misdiagnosed, masquerading as other benign and malignant lesions that lead to a delay in referral and definitive treatment. Incorrect histopathological interpretations have been reported in 39–77% of cases, and this may be due to the tumour rarity, low clinical suspicion and lack of familiarity of pathologists with characteristic histological features of OSC.2–5 Br Copyright J Ophthalmol 2012;0:1–5. Articledoi:10.1136/bjophthalmol-2012-302443 author (or their employer)

In a large retrospective study conducted by a US National Cancer Registry, OSC was detected in 38.7% of sebaceous carcinoma cases.6 Curtis and co-workers estimated the annual incidence of OSC in Caucasians older than 20 years to be 0.5 per million in Florida.7 To date, there have been no prospective national studies to estimate the incidence of OSC, although there is an impression that the incidence of OSC may be increasing.8 In the UK, two tertiary referral centres have reported up to 1–2 OSC cases being managed annually over the last 30 years.4 9 The delay in diagnosis and treatment has a significant adverse effect on patient mortality. The 5-year mortality is 38% with greater than 6 months history, and 14% if the history is less than 6 months.3 Despite improvements in surgical techniques and other therapies, the mortality rates are reported to be high. More recently, improved survival rates have been reported in Scotland and the USA.4 10 There are no preferred standard surgical practices to manage OSC in UK eye units, and OSC cases are managed by pathologists and oculoplastic surgeons based in both non-tertiary and specialist centres. The British Ophthalmological Surveillance Unit (BOSU) is a well-recognised nationwide active surveillance system for the effective case ascertainment and epidemiological investigation of rare eye diseases.11 The BOSU was developed to assist in the investigation of uncommon ocular conditions that are of public health or scientific importance. There have been no previous prospective studies on the incidence of OSC, and previous estimates of incidence have been based on an estimate of the population served by single centres. The primary aim of this study is to provide contemporary epidemiological data on the UK incidence of OSC. The secondary objectives include describing the aetiology for delayed diagnosis and current diagnostic pathways of OSC in the pathology and ophthalmology sectors, and evaluation of surgical interventions and patient outcomes for OSC.

METHODS Study design, setting and population Patients with newly diagnosed OSC were identified prospectively through population-based surveillance by the BOSU during a 24-month period from March 2008 to February 2010 inclusive. BOSU operates a monthly active surveillance scheme throughout the UK.11 The surveillance scheme involves all permanently employed ophthalmologists in the UK with clinical autonomy (consultants and associate specialists), who

2012. Produced by BMJ Publishing Group Ltd under licence.

1


Clinical science form the reporting base. Before the initiation of a study, the BOSU informs all ophthalmologists about the new ocular condition under investigation, including the specific case definition. At the end of each month, a ‘yellow report card’ is sent to each ophthalmologist, who then returns it specifying any new case of interest or indicating that no new cases had been seen that month. After case notification, the investigators send incident and follow-up questionnaires to the reporting ophthalmologists. The study was approved by the Leeds Research Ethics Committee. The study adhered to the tenets of the Declaration of Helsinki and the STROBE guidelines.

Case report validation This was undertaken on different levels, and enabled capture re-capture analysis for a more stable estimate of the incidence to quantify the public health problem. The BOSU reports sent to the study group were used to identify and exclude duplicate case reporting at a national level. For example, an ophthalmologist may report the same case twice during initial diagnosis and at subsequent follow-up visits. In cases of tertiary referral to another hospital unit, we identified duplicate reports from within a single hospital using the questionnaires and by directly contacting reporting ophthalmologists. At 3-monthly intervals, the consultant pathologist at each reporting hospital was directly contacted to establish the incident BOSU case and whether further OSC cases were diagnosed during the study period. The cases reported through the study were externally validated and supplemented through the network of UK cancer registries. The register number was compared with the actual reported case pick-up from each region in the UK. The registry cases helped maximise detection of new OSC cases, and avoided duplication of new cases from hospital-BOSU reporting. All cases from both reporting systems were collated during the study annually.

Sebaceous tumour advisory group We formed a collaborative clinical advisory group known as the Sebaceous Tumour Advisory Group (STAG). This consisted of experts in the fields of ocular oncology, oculoplastic surgery and ophthalmic pathology actively involved in managing OSC in the UK. During the study conception, the study questionnaires were piloted through the STAG and outcome parameters were finalised prior to study commencement. During the study period, the STAG were consulted regarding study progress and OSC case notifications. In particular, members of the four National Specialist Ophthalmic Pathology Service (NSOPS) centres and the Scottish Regional Ophthalmic Pathology Service participated in the STAG. Additional external validation of incident OSC was performed through the five centres that form the core national reference pathology service and STAG that permitted additional tiers of capture re-capture analysis.

Sample size Ocular sebaceous carcinoma is a rare tumour and the reported incidence varies between 0.38 and 0.5 per million in published studies from the UK.4 9 Extrapolation from the literature indicates that there may be approximately 30 cases of OSC newly diagnosed per year in the UK. With an estimated 30 new cases of OSC per year and 2 years of observation, this leads to an estimated UK annual incidence rate of 0.5 per million, and we anticipated that we would be able to estimate this incidence rate within +/−0.12 (95% CI 0.38 to 0.62 per million population per year).

Statistical analysis The analysis will be descriptive, and the rate of incidence reported per 1 000 000 of the population in the UK with CIs. The diagnosis and management outcomes will be expressed as percentages of the total number of reported cases. Risk factors were analysed and reported as mean, SD with 95% CIs. The one-way ANOVA test and t-test was used to compare tumour location, differential diagnosis, and delay in diagnosis for OSC. The comparison of multiple risk factors was undertaken using Fisher’s least significant difference and the Bonferroni method. A significance level of p5 mm Data unavailable

Cases, n (%)

21 17 10 6 5 1

(42%) (33%) (20%) (12%) (10%) (2%)

25 6 5 4 1 2 4

(49%) (12%) (10%) (8%) (2%) (4%) (8%)

1 1 2 9 1 1 27

*Total number of cases (n=61) exceeds the study patient number (n=51) as some patients underwent more than one diagnostic technique in some centres. †Data unavailable for eight cases. ‡Data unavailable for nine cases.

DISCUSSION This study is the first prospective study of OSC in the UK. We report a contemporary incidence rate for OSC in the UK, and have identified the misdiagnosis of chalazion to be a significant factor in diagnostic delay from the community to the ophthalmologist. We estimate the UK incidence of OSC to be approximately 0.41 cases per million UK population. General practitioners (GPs) and ophthalmologists should have a high index of clinical suspicion for OSC if patients present with a non-resolving, persistent chalazion. Clinical signs of pagetoid tumour spread are easily missed on ophthalmoscopy, and incomplete tumour removal will occur with excisional techniques without map biopsy. We found significant delays of up to 6 months occurred before finally diagnosing OSC, and this increased the risk of intraepithelial neoplasia. Despite the increasing ethnic populations within the UK, OSC is rare in the UK with mainly OSC found in white Caucasians.7 Tryggyason and coworkers found a 633 cases of eyelid sebaceous carcinoma (34.5%) in a series of 1836 cases of head and neck sebaceous carcinoma in the USA.14 In the Far East, OSC comprised 33% of all malignant eyelid tumours.15 However, the incidence of OSC in Taiwan at 0.2–0.3 per million is half the UK rate.16 Previous studies reported that females are more frequently affected with OSC, and we found a slightly higher rate of OSC among women.17 18 In elderly individuals, the eyelid skin laxity increases that may mask an OSC mass masquerading as a chalazion, and this patient group is less likely to contact a doctor for an apparent benign chalazion-like eyelid swelling. Our study confirms the importance of prompt diagnostic biopsy and referral for a specialist ophthalmic opinion in cases of recurrent, unilateral chalazion or in cases where the chalazion shows atypical features. 4

In light of the varying diagnostic services available (specialist ophthalmic vs general pathologist) for GPs, we identified a greater number of patients requiring confirmatory diagnosis of OSC in specialist pathology centres. Intraepithelial OSC spread is not visible on clinical biomicroscopy, and clinical signs such as conjunctival inflammation is non-localising and fairly nonspecific sign for neoplastic disease in the anterior segment. Diffuse OSC spread within the eyelid and conjunctiva thereby can masquerade as a chronic blepharoconjunctivitis. In a large retrospective series from Shields and co-workers, the authors recommend a diagnostic biopsy of OSC with 4 mm surgical margins and intraoperative frozen sections.19 The majority of UK centres perform a full thickness eyelid biopsy, using incisional or excisional techniques. Our study highlights the high incidence of pagetoid eyelid epidermal and conjunctival epithelial tumour spread. Perhaps conjunctival map biopsy should be undertaken as a primary procedure in all suspected cases of OSC.20 In addition, all cases in our study that required primary exenteration underwent diagnostic map biopsies presumably to delineate the extent of pagetoid tumour spread. A delay of longer than 6 months between onset of symptoms and establishing the correct diagnosis has been noted for invasive OSC, and this has a significant adverse effect on patient mortality.21 Our study corroborates published reports of the mean delay varying between 1 and 3 years.22 Specialist pathologists from five national centres undertook the majority of OSC diagnostics in the UK OSC study. The study follow-up questionnaire reported a history of previous curettage or biopsies, with misdiagnosis at non-tertiary referral centres by general non-ophthalmic pathologists. It is important to clearly demonstrate sebaceous differentiation within the group of eyelid neoplasms, as the differential includes basal cell carcinoma with adnexal differentiation and squamous cell carcinoma.23 24 In 50% of the OSC cases in this study, careful attention to the cytomorphology on the H&E stained material and an additional panel of immunohistochemical stains was required. The primary management of OSC is surgical intervention, and this can be challenging in the presence of the ill-defined nature of pagetoid invasion.5 21 Once the patient reaches the referral centre, ophthalmologists may have had little clinical experience of dealing with OSC and this is reflected by the proportion of patients undergoing the controversial technique of primary curettage and the variation in diagnostics by ophthalmologist. This inexperience at the primary and secondary care levels is known to result in a low clinical suspicion for this tumour. The differences in surgical practice such as variations in surgical tumour margins reflect this. Currently in the UK, patients may endure a delay of up to 11 months between the onset of symptoms and establishing the correct diagnosis. The delays in hospital referral lead to most patients requiring complex ocular reconstruction surgery following excision of the tumour. One in 10 patients underwent radical primary exenteration, with pagetoid spread detected in these cases. This figure is similar to previous large series of OSC.19 The use of topical mitomycin-C in OSC with pagetoid conjunctival spread is increasingly reported, and in the UK, this treatment modality is available at a number of tertiary referral centres.25 Mohs surgery was not frequently undertaken in the ophthalmic/ oculoplastic services, however, there are well-recognised combined dermatology/ophthalmology services active in the UK. Muir-Torre syndrome is an autosomal dominant condition and is characterised by the association of a visceral neoplasm (often, colorectal and genitourinary) and sebaceous tumours.26 Br J Ophthalmol 2012;0:1–5. doi:10.1136/bjophthalmol-2012-302443


Clinical science We captured a single case of Muir-Torre syndrome, however, perhaps due to the rarity of this syndrome, ophthalmologists in our study had not questioned patients about the potential systemic associations in a third of patients. In all cases of OSC, we recommend a multidisciplinary team approach, with systemic enquiry for any colorectal, uterine, pancreatic, and ureteric symptoms that may be indicate neoplastic disease elsewhere. A limitation of this study could be the under-reporting of OSC from referring UK dermatologists and plastic surgeons, hence a potential for underestimating the incidence of OSC in the UK. However, the main strengths of this the study was the methods used to maximise case capture of new, incident OSC cases. A variety of external mechanisms were used; prospective case ascertainment directly from treating ophthalmologists through the BOSU; external validation over 2 years using the STAG and national ophthalmic pathology networks; and referencing with the UK cancer registry network over the study period. The mortality data from this study should be viewed with caution. In the three cases that died, one case was attributed to metastatic disease associated with OSC. The other two patients died of unrelated causes. Although the rate was 2.6%, the average follow-up was only 1 year for this cohort of patients. The survival rates for OSC in this cohort of UK patients needs evaluation at 3- and 5-year intervals to provide a true reflection of tumour-related mortality. Our study highlights the current NHS diagnostic pathway that involves the primary care referral followed by either local ophthalmology and general pathology services or investigation by a tertiary referral ophthalmic centre with specialist ophthalmic pathology. The significant delay in OSC diagnosis of up to 11 months needs to be addressed at a primary care level. Ophthalmic practitioners and GPs in the community should have a high index of suspicion of OSC in cases of a nonresolving chalazion, diffuse blepharoconjunctivitis, and atypical unilateral blepharitis. In order to plan an effective diagnostic and treatment pathway for patients with OSC, the authors would support the recommendation that diagnostic and ophthalmic clinical management of this rare, malignant tumour within a dedicated eyelid and adnexal tumour clinic by a multidisciplinary team that includes oculoplastic surgery, ophthalmic pathology, and dermatopathology support. Acknowledgements The authors thank the British Ophthalmological Surveillance Unit (BOSU) for support in the study. The authors thank the Sebaceous Tumour Advisory Group: P Atkinson, S Verity, M Beaconsfield, R Bonshek, R Caesar, B Chang, S Coupland, Prof B Damato, J Dickenson, M Hayward, Prof Hiscott, A Foss, K Hakin, M Halliwell, J Hungerford, C Hutchinson, E Kemp, P Kearns, C Inkster, B Leatherbarrow, Prof Luthert, S Livesey, H McLean, R Manners, R Malhotra, G Meligones, H Mudhar, J Olver, B Parkin, R Radford, R Redmond, F Roberts, G Rose, A Saadiq, R Sampath, S Desai, J Tan, V Thaller, A Tyers, J Uddin, S Vardy, C Vize, M Wright, M Burton, S Woodruff, D Hughes, T Reuser, D Cheung. Contributors Substantial contributions to the intellectual content of the paper were made as follows: Design and hypothesis of study: MMK, IGR; data acquisition: MMK, BF; data analysis and interpretation: MMK, BF, SW, FR, HM, IGR; patient recruitment and data acquisition: MMK, BF. Substantial contributions to the written content of the paper as follows: Drafting of the manuscript: MMK, BF, SW, FR, HM, IGR; critical revision of the manuscript for important intellectual content: MMK, BF, SW, FR, HM, IGR; statistical expertise: MMK, SW.

Br J Ophthalmol 2012;0:1–5. doi:10.1136/bjophthalmol-2012-302443

Funding The UK ocular sebaceous carcinoma (OSC) Study was funded by a grant provided by Cancer Research UK. The British Ophthalmological Surveillance Unit (BOSU) is funded by The Guide Dogs for the Blind Association. Competing interests None. Ethics approval Leeds Research Ethics Committee. Provenance and peer review Not commissioned; externally peer reviewed.

REFERENCES 1.

2. 3. 4. 5. 6. 7.

8.

9. 10.

11.

12.

13. 14.

15. 16. 17.

18. 19. 20. 21. 22. 23. 24. 25.

26.

McLean IW, Jackobiec FA, Zimmerman LE, et al. Tumours of the eye and ocular adnexa. Vol 3. Washington: Maryland Armed Forces Institute of Pathology, 1993:28–35. Nelson BR, Hamlet KR, Gillard M, et al. Sebaceous carcinoma. J Am Acad Dermatol 1995;33:1–15. Ni C, Kuo PK. Meibomian gland carcinoma: a clinicopathologic study of 156 cases with long-period follow-up of 100 cases. Jpn J Ophthalmol 1979;23:388–401. Muqit MMK, Roberts F, Lee WR, et al. Improved survival rates in sebaceous carcinoma of the eyelid. Eye (Lond) 2004;18:49–53. Shields JA, Demirci H, Marr BP, et al. Sebaceous carcinoma of the ocular region: a review. Surv Ophthalmol 2005;50:103–22. Dasgupta T, Wilson LD, Yu JB. A retrospective review of 1349 cases of sebaceous carcinoma. Cancer 2009;115:158–65. Margo CE, Mulla ZD. Malignant tumors of the eyelid: a population-based study of non-basal cell and non-squamous cell malignant neoplasms. Arch Ophthalmol 1998;116:195–8. Khan JA, Doane JF, Grove AS. Sebaceous and meibomian carcinomas of the eyelid. Recognition, diagnosis, and management. Ophthal Plast Reconstr Surg 1991;7:61–6. Zurcher M, Hintschich CR, Garner A, et al. Sebaceous carcinoma of the eyelid: a clinicopathological study. Br J Ophthalmol 1998;82:1049–55. Esmaeli B, Nasser QJ, Cruz-Carillo H, et al. American Joint Committee on Cancer T category for eyelid sebaceous carcinoma correlates with nodal metastasis and survival. Ophthalmology 2012;119:1078–82. Foot B, Stanford M, Rahi J, et al. British Ophthalmological Surveillance Unit steering committee. The British Ophthalmological Surveillance Unit: an evaluation of the first 3 years. Eye (Lond) 2003;17:9–15. Mahmood S, von Lany H, Cole MD, et al. Displacement of nuclear fragments into the vitreous complicating phacoemulsification surgery in the UK: incidence and risk factors. Br J Ophthalmol 2008;92:488–92. United Kingdom National Statistics. http://www.statistics.gov.uk/hub/population/ index.html Tryggvason G, Bayon R, Pagedar NA. Epidemiology of sebaceous carcinoma of the head and neck: Implications for lymph node management. Head Neck 2012. doi :10.1002/hed.22009. Ni C, Searl SS, Kuo PK, et al. Sebaceous cell carcinomas of the ocular adnexa. Int Ophthalmol Clin 1982;22:23–61. Lin HY, Cheng CY, Hsu WM, et al. Incidence of eyelid cancers in Taiwan. A 21-year review. Ophthalmology 2006;113:2101–7. Rao NA, Hidayat AA, McLean IW, et al. Sebaceous carcinomas of the ocular adnexa: A clinicopathologic study of 104 cases, with five-year follow-up data. Hum Pathol 1982;13:113–22. Kass LG, Hornblass A. Sebaceous carcinoma of the ocular adnexa. Surv Ophthalmol 1989;33:477–90. Shields JA, Demirci H, Marr BP, et al. Sebaceous carcinoma of the eyelids: personal experience with 60 cases. Ophthalmology 2004;111:2151–7. Putterman AM. Conjunctival map biopsy to determine pagetoid spread. Am J Ophthalmol 1986;102:87–90. Doxanas MT, Green WR. Sebaceous gland carcinoma: review of 40 cases. Arch Ophthalmol 1984;102:245–9. Boniuk M, Zimmerman LE. Sebaceous carcinoma of the eyelid, eyebrow, caruncle and orbit. Trans Am Acad Ophthalmol Otolaryngol 1968;72:619–41. Muthusamy K, Halbert G, Roberts F. Immunohistochemical staining for adipophilin, perilipin and TIP47. J Clin Pathol 2006;59:1166–70. Johnson JS, Lee JA, Cotton DW, et al. Dimorphic immunohistochemical staining in ocular sebaceous neoplasms: a useful diagnostic aid. Eye (Lond) 1999;13:104–8. Shields CL, Naseripour M, Shields JA, et al. Topical mitomycin-C for pagetoid invasion of the conjunctiva by eyelid sebaceous gland carcinoma. Ophthalmology 2002;109:2129–33. Ponti G, Ponz de Leon M. Muir-Torre syndrome. Lancet Oncol 2005;6:980–7.

5


Observational prospective cohort study of patients with newly-diagnosed ocular sebaceous carcinoma Mahiul M K Muqit, Barny Foot, Stephen J Walters, et al. Br J Ophthalmol published online October 31, 2012

doi: 10.1136/bjophthalmol-2012-302443

Updated information and services can be found at: http://bjo.bmj.com/content/early/2012/10/30/bjophthalmol-2012-302443.full.html

These include:

References

This article cites 24 articles, 3 of which can be accessed free at: http://bjo.bmj.com/content/early/2012/10/30/bjophthalmol-2012-302443.full.html#ref-list-1

P