Histological subtypes of periocular basal cell carcinoma

Authors

  • Albert Wu,

    Corresponding author
    1. South Australian Institute of Ophthalmology, University of Adelaide, Adelaide, South Australia, Australia
    • Correspondence: Mr Albert Wu, South Australian Institute of Ophthalmology, Level 8, East Wing, Royal Adelaide Hospital, Adelaide, SA 5000, Australia. Email: albert.wu3@gmail.com

    Search for more papers by this author
    • Albert Wu and Michelle T Sun contributed equally.
  • Michelle T Sun MBBS,

    1. South Australian Institute of Ophthalmology, University of Adelaide, Adelaide, South Australia, Australia
    Search for more papers by this author
    • Albert Wu and Michelle T Sun contributed equally.
  • Shyamala C Huilgol MBBS(Hons) FACD,

    1. Department of Dermatology, Royal Adelaide Hospital, University of Adelaide, Adelaide, South Australia, Australia
    Search for more papers by this author
  • Simon Madge FRCOphth,

    1. South Australian Institute of Ophthalmology, University of Adelaide, Adelaide, South Australia, Australia
    2. Department of Ophthalmology, Hereford County Hospital, Hereford, UK
    Search for more papers by this author
  • Dinesh Selva FRANZCO

    1. South Australian Institute of Ophthalmology, University of Adelaide, Adelaide, South Australia, Australia
    Search for more papers by this author

  • Competing/conflicts of interest: No stated conflict of interest.
  • Funding sources: No stated funding sources.

Abstract

Background

To determine the proportion of different subtypes of periocular BCC in South Australia.

Design

Retrospective review.

Participants

One thousand seven hundred thirteen consecutive periocular basal cell carcinoma (BCC) excision specimens.

Methods

Histological analysis of consecutive periocular BCC specimens.

Main Outcome Measures

Date of resection, patient age at resection, gender, tumour location, histological subtype and perineural invasion.

Results

From 2006 to 2012, a total of 1713 consecutive periocular BCC excision specimens were analysed. The mean age at resection was 68.8 years (median: 71, range: 21–101). Most specimens (56.4%) were removed from male patients. 52.7% involved the lower eyelid, 29.0% the medial canthus, 10.9% the lateral canthus and 7.5% the upper eyelid. The main histological subtypes identified were nodular (65.7%), infiltrative (17.5%), superficial (12.6%) and micronodular (4.2%). Of the specimens, 25.6% had more than one subtype. The most common subtype combinations were nodular with infiltrative (49.7%), and nodular with superficial (26.0%).

Conclusions

The majority of periocular BCC were located on the lower lid and classified histologically as nodular. Infiltrative BCC occurred more frequently than the superficial subtype. As the proportion of mixed BCC containing aggressive subtypes is high, surgical excision with margin control should be considered for periocular BCC.

Introduction

Basal cell carcinoma (BCC) is the most common eyelid cancer in Australia, accounting for up to 90% of all malignant eyelid tumours.[1, 2] The high incidence of BCC in Australia has been attributed to high levels of sunlight exposure and a fair-skinned population of increasing age.[1]

The classifications of histological subtypes advocated by the World Health Organisation and the Royal College of Pathologists have been widely used in the recent literature.[3, 4] The main growth patterns recognized are superficial, nodular, infiltrative and micronodular. The latter two are associated with significantly increased risk of local recurrence and increased morbidity.[3-5] It has been found that nodular BCC predominantly occur on the head and neck, and that infiltrative BCC predominate over superficial BCC on the face.[6-10] The higher proportion of infiltrative BCC on the face may prompt more aggressive treatment, especially in the periocular region where local invasion can result in significant morbidity. Few studies have examined the proportion of periocular BCC subtypes.[11, 12] To aid in the management of periocular BCC, we aimed to investigate the proportion of BCC subtypes occurring in the periocular region.

Methods

A retrospective histological review was conducted of all periocular (upper lid, lower lid, medial canthus or lateral canthus) excision specimens diagnosed as BCC at the Institute of Medical and Veterinary Science (IMVS) Main Laboratory, Adelaide in the 7-year period from 2006 to 2012. Approval was granted from the Ethics Committee of the Royal Adelaide Hospital (RAH). Reports were retrieved from the IMVS Histology database. All specimens were examined with paraffin section and reported by an IMVS pathologist. Tumour resection was performed at the RAH, The Queen Elizabeth Hospital, Port Pirie Hospital, Whyalla Hospital and other smaller medical centres in South Australia. The following data were analysed: date of resection, patient age at resection, gender, tumour location (laterality and periocular region), histological subtype and perineural invasion. The diagnostic criterion of perineural invasion, which was consistent with guidelines,[13] was the observation of cytologically malignant cells in the perineural space of nerves. In addition, total or near-total circumferential involvement, presence of perineural tracking in tangential sections and intraneural involvement all supported the diagnosis.

Differences in subtype proportion according to tumour location were analysed with chi-square tests. Differences in the laterality of the lesion according to gender were also analysed with chi-square tests. One-way analysis of variance was employed to compare mean age at resection according to subtype with post-hoc Bonferroni testing.

Results

From 2006 to 2012, a total of 1713 consecutive periocular BCC excision specimens were analysed. The mean age at resection was 68.8 ± 14.3 years (median: 71, range: 21–101). Of the specimens, 966 (56.4%) were removed from male patients and 747 (43.6%) from female patients. BCC occurred on the right side in 838 (51.1%) cases and on the left in 801 (48.9%). There was no significant interaction between gender and the side of the lesion (P = ns). Of the specimens, 742 (52.7%) involved the lower eyelid, 409 (29.0%) the medial canthus, 153 (10.9%) the lateral canthus and 105 (7.5%) the upper eyelid.

The most common histological subtype was nodular, accounting for 1278 (65.7%) cases. Infiltrative BCC was identified in 341 (17.5%) cases, superficial in 245 (12.6%) cases and micronodular in 81 (4.2%) cases. Subtype was not reported for 195 specimens. Squamous differentiation was identified in 81 (4.7%) specimens. Perineural invasion was identified in 16 (0.9%) cases. Table 1 shows the relationships between subtype and mean age at resection, gender and tumour location. Mean age at resection varied by subtype (P = 0.03), with infiltrative subtypes diagnosed at older age compared with nodular subtypes (P < 0.05). There were significant differences in the proportion of subtypes by tumour location (P = 0.004). Compared with lower lid and medial canthus, lateral canthus had higher proportions of infiltrative and superficial subtypes and upper lid had lower proportions of these subtypes.

Table 1. Relationships between basal cell carcinoma subtype and mean age at resection, gender and tumour location
 NodularInfiltrativeSuperficialMicronodular
Mean age at resection (years)68.5 ± 14.471.0 ± 14.169.8 ± 14.770.2 ± 14.1
Gender, % (n)    
Male65.3 (740)18.2 (206)12.0 (136)4.6 (52)
Female66.3 (538)16.6 (135)13.4 (109)3.6 (29)
Periocular region, % (n)    
Lower lid64.7 (548)18.5 (157)13.0 (110)3.8 (32)
Medial canthus68.1 (322)15.9 (75)11.6 (55)4.4 (21)
Lateral canthus56.1 (115)25.4 (52)16.6 (34)2.0 (4)
Upper lid76.1 (83)11.0 (12)7.3 (8)5.5 (6)

There were 388 (25.6%) specimens containing more than one subtype. The most common subtype combinations were nodular with infiltrative (193 specimens), and nodular with superficial (101 specimens), accounting for 49.7% and 26.0% of mixed BCC, respectively. Table 2 lists the frequency of occurrence of all the combinations.

Table 2. Combinations of subtypes in mixed basal cell carcinoma
SubtypesNumber of specimens, n (%)
  1. I, infiltrative; M, micronodular; N, nodular; S, superficial.
N + I193 (49.7)
N + S101 (26.0)
N + M29 (7.5)
N + I + S28 (7.2)
I + S21 (5.4)
S + M4 (1.0)
N + I + M4 (1.0)
N + S + M3 (0.8)
I + M2 (0.5)
I + S + M2 (0.5)
N + I + S + M1 (0.3)

Discussion

The majority of periocular BCC were located on the lower lid and classified histologically as nodular. Infiltrative BCC occurred more frequently than the superficial subtype, consistent with previous studies of facial and periocular BCC. Additionally, we found that 25.6% of all BCC had mixed histology, of which the combination of nodular with infiltrative accounted for close to half of these specimens.

The proportion of subtypes in our study was compared with that in five previous studies shown in Table 3. These studies were conducted at different latitudes, with the latitude of Adelaide greater than Queensland but less than Melbourne, the United Kingdom and France. Our study demonstrated higher rates of infiltrative BCC compared with superficial subtype, consistent with previous reports from Townsville, Brisbane and the United Kingdom of facial and periocular BCC.[7, 11, 12] These findings may be attributable to the theory which suggests that superficial BCC results from acute, intense sun exposure to skin that is less exposed to sunlight and thus less tanned.[14] As the face and periocular region are chronically exposed to sunlight and more tanned, fewer superficial BCC may develop at these sites.[6-8] Another theory suggests that BCC subtype progresses from superficial to nodular to infiltrative, and therefore, if chronic sun exposure triggers progression to infiltrative BCC, more infiltrative BCC may develop in the periocular region.[15]

Table 3. Comparison of basal cell carcinoma subtypes in the literature
 Nodular (%)Infiltrative (%)Superficial (%)Micronodular (%)
  1. n, number of specimens examined.

Periocular region

Adelaide (34.9°S)

n = 1681

2006–2012

65.717.512.64.2

Wong et al.[12]

Periocular region

Brisbane (27.5°S)

n = 596

1992–2001

61.221.510.96.4

Ho et al.[11]

Periocular region

United Kingdom (52.6°N)

n = 338

2000–2006

82.011.55.90.6

Raasch et al.[7]

Face

Townsville (19.3°S)

n = 3245

1997–1999

56.919.312.910.9

McCormack et al.[6]

Head and neck

Melbourne (37.8°S)

n = 1982

1991

81.710.47.9Not reported

Scrivener et al.[8]

Head and neck

France (48.5°N)

n = 9941

1967–1996

85.67.17.3Not reported

In general, infiltrative and superficial subtypes have been shown to occur more frequently in the periocular region and at lower latitudes compared with on the head and neck and at higher latitudes.[6-8, 11, 12] Thus, these subtypes may be associated with greater sun exposure which occurs at lower latitudes. Additionally, latitude has been shown to influence BCC subtype more than body site, with significantly higher proportions of infiltrative and superficial BCC found on the head and neck in the Townsville study compared with in the periocular region in the United Kingdom study.[7, 11] Our findings are in keeping with this, as we reported higher proportions of infiltrative and superficial BCC compared with studies of head and neck BCC conducted at higher latitudes.

We found that there was mixed histology in 25.6% of our specimens and Ho et al. identified mixed histology in 17.9% of their periocular BCC.[11] Our most common subtype combinations of nodular with infiltrative and nodular with superficial were consistent with previous studies.[6] Mixed histology has been reported to occur in up to 38.5% of all BCC and can result in an initial biopsy that misses one or more additional aggressive subtypes.[16-23] If an aggressive subtype is missed, initial treatment may be inadequate and BCC may recur. As the proportion of mixed BCC is high and aggressive subtypes require modified treatment, surgical excision with margin control may be favoured over nonsurgical destructive modalities for periocular BCC.[1, 2, 24]

Although we found that infiltrative subtypes were diagnosed at older age compared with nodular subtypes, previous studies of BCC from the whole body did not find such a relationship.[6-9] Patients with superficial BCC have been found to be markedly younger than those with other subtypes, suggesting that carcinogenesis of superficial BCC requires less sunlight exposure than nodular BCC.[5-7, 9] It has also been suggested that superficial BCC results from intermittent, intense sun exposure which occurs more frequently at younger ages and decreases with increasing age.[6-9] However, we found that the mean age for superficial BCC was not significantly different to that for nodular BCC. Thus, our findings differ from previously proposed pathways of BCC subtype progression of superficial to nodular to micronodular and superficial to nodular to infiltrative.[15] The periocular region, which is chronically exposed to sunlight and more tanned, may be less likely to sunburn and develop superficial BCC at a young age when acute, intense sun exposure usually occurs, resulting in the mean age for superficial BCC being similar to nodular BCC. Our findings suggest there may be differences in the pattern of BCC development in the periocular region compared with the rest of the body.[5-9] Further studies analysing the relationship between mean age and histological subtype of BCC in the periocular region and face are required to better characterize this pattern of development.

Previous studies have investigated whether greater ultraviolet exposure to one side of the body when driving would result in more BCC occurring on that side.[12, 25-29] In keeping with this theory, our study demonstrated a slight predilection for BCC to occur on the right side which is exposed to more sunlight when driving in Australia. Additionally, it has previously been proposed that as males drive more than females, more BCC would occur in males than females on the side of the face exposed to more sunlight when driving.[27-29] However, with increasing numbers of female drivers over the past few decades, the proportion of male to female drivers in Australia is now very similar.[30] In keeping with this, we found that there were no statistically significant differences in laterality of tumours between the two genders.

Our study has a number of limitations. The retrospective nature of this study precluded the collection of additional clinical information which may have provided further insights. Additionally, our study is not population based as we included resection specimens examined at a single laboratory in South Australia. Thus, bias could potentially result from tendency to send specimens of certain subtypes to the laboratory and use of destructive treatment modalities which preclude histological examination. Bias could also result because subtype was not reported for a number of specimens. Although IMVS is the main pathology laboratory servicing the public health sector in South Australia, studies including specimens from all pathology laboratories in the region would add to our findings.

In conclusion, we found that the most common BCC histological subtype occurring in the periocular region was nodular followed by infiltrative and superficial. Additionally, a quarter of all specimens had mixed histology, with nodular and infiltrative being the most frequently occurring combination. Given the high proportion of mixed BCC containing aggressive histological subtypes, surgical excision with margin control should be considered for periocular BCC given the morbidity associated with recurrence in the eyelids.

Ancillary