ARTICLE IN PRESS. Increased prevalence of left-sided skin cancers. Susan T. Butler, MD, a and Scott W. Fosko, MD a,b,c,d,e Saint Louis, Missouri

Transcription

1 ORIGINAL ARTICLE Increase prevalence of left-sie skin cancers Susan T. Butler, MD, a an Scott W. Fosko, MD a,b,c,,e Saint Louis, Missouri Backgroun: Previous research has shown an increase in photoamage an precancers on the left sie of the face. Objective: We sought to etermine whether there is a higher frequency of skin cancer evelopment on the left sie of the boy than the right. Methos: The stuy was a retrospective review of patients with skin cancer referre to our Mohs micrographic surgery an cutaneous oncology unit in Results: When incluing all types of skin cancers an both sexes, more cancers occurre on the left (52.6%) than the right (47.4%) (P =.059), with a stronger tren in men (P =.042). There were significantly more malignant melanoma in situ on the left (31/42, 74%) than the right (11/42, 26%) (P =.002). Limitations: Population was comprise of patients referre to an acaemic meical center an often for Mohs micrographic surgery. Conclusions: There were significantly more skin cancers on the left than the right sie in men. This iscrepancy was even more profoun in malignant melanoma in situ. ( J Am Aca Dermatol /j.jaa ) Key wors: left sie; photoamage; skin cancer; ultraviolet A. Currently, between 2 an 3 million nonmelanoma skin cancers an 132,000 melanoma skin cancers occur globally each year. 1 Ultraviolet (UV) exposure is a major causative factor in melanoma an nonmelanoma skin cancer. UV raiation is comprise of UVA an UVB rays. UVA rays are able to penetrate glass winows whereas UVB rays cannot. Automobile rivers are expose to a significant amount of UVA rays that penetrate From the Departments of Dermatology, a Otolaryngology, b Ophthalmology, c an Internal Meicine, Saint Louis University School of Meicine; an the Saint Louis University Cancer Center. e Funing sources: None. Conflicts of interest: None eclare. Portions of these ata were inclue in a poster presentation entitle The Association of Asymmetric Skin Cancers with Time Spent in an Automobile at the American Acaemy of Dermatology Annual Meeting, February 1-5, 2007, Washington, DC. Accepte for publication November 9, Reprint requests: Susan T. Butler, MD, Department of Dermatology, Saint Louis University, 1402 S Gran Blv, Saint Louis, MO suzybutler1@gmail.com. Publishe online March 10, /$36.00 ª 2009 by the American Acaemy of Dermatology, Inc. oi: /j.jaa Abbreviations use: LM: lentigo maligna MMIS: malignant melanoma in situ UV: ultraviolet river s sie winow glass or pass easily through open winows, sunroofs, or convertible tops. In our own an others personal experience in the Unite States, it has been observe that significant photoamage, precancers, an possibly skin cancers were occurring on the river s sie or left sie of the face. A previous stuy by Singer et al 2 showe an asymmetry of photoamage on the face, an a correlation between left-sie photoamage an percentage of time spent as the river. The photoaging effects of UVA can also be emonstrate through a striking case stuy of an office worker with extensive Favre-Racouchot isease on her left cheek, which face her office winow for 15 years, but virtually no photoamage on the contralateral sie. Because UVB oes not pass through winow glass, this observation ientifies the role of UVA in photoaging. 3 A stuy performe in Australia foun an increase in incience of precancers on the sie of 1

2 2 Butler an Fosko JAM ACAD DERMATOL the boy correlating to the sie of the car subjects most frequently sat on. Men forme more actinic keratoses on the right sie, whereas women forme more on the left sie (river s sie is on right in Australia). 4 This supports the association of UV exposure an subsequent evience of clinical amage to the occupant s expose sie of the face. No stuies to ate have shown that more skin cancers form on the left sie of CAPSULE SUMMARY the boy than the right for countries where riving habits are similar to those in the Unite States. The goal of this stuy was to etermine whether there is a higher frequency of skin cancer evelopment on the left sie of the boy than on the right. METHODS This stuy was approve by our institutional review boar (No ). The esign was a retrospective review of all patients referre to the Mohs micrographic surgery an cutaneous oncology unit at our university epartment of ermatology between January 2004 an December Information was collecte regaring the type of skin cancer, location of the skin cancer on the boy, an age an sex of the patients. Patients were exclue if their skin cancer was locate along the miline of the boy, or if their meical recors were unavailable or incomplete. RESULTS The charts of 1047 patients were reviewe. Of the 890 patients who ha non-miline skin cancers, 557 (63%) were male an 333 (37%) were female (P =.001). The mean age was 68 years. Basal cell carcinoma was the most common type of skin cancer with a total of 608 (68.3%) (Table I). The next most common cancer was squamous cell carcinoma occurring in 178 (20.0%) patients, followe by malignant melanoma in situ (MMIS) in 42 (4.7%) patients, squamous cell carcinoma in situ in 38 (4.3%) patients, an invasive malignant melanoma in 23 (2.6%) patients. The MMIS group was compose of both lentigo maligna (LM) an superficial spreaing subtypes. Of note, the melanomas an MMIS were not treate by the Mohs micrographic technique, rather by wie local excision an permanent section confirmation of margins. The most common location for skin cancer was the hea/neck with 731 (82.1%) occurrences. The next most common location was the arm with There are more skin cancers occurring on the left sie of the boy than the right. This tren is more pronounce in men an in certain cancers such as melanoma in situ. This fining may be explaine by ultraviolet exposure to the left sie of the boy while riving. 59 (6.6%) occurrences, followe by the leg with 30 (3.4%), an the han with 27 (3.0%) cancers (Table II). When incluing all types of skin cancers an both sexes, more cancers i occur on the left (52.6%) than the right (47.4%) (P =.059). When iviing by sex, there were significantly more skin cancers on the left (301) than on the right (256) in male patients only (P =.042). There was no significant ifference note in female patients (Fig 1). The cancers occurring on the hea an neck were analyze separately as these areas are most irectly expose to UV rays while riving. The left-sie preominance was even more pronounce when looking at only these sites in both sexes, with the percentage of skin cancers on the left increasing from 52.6% (468/890) to 54.0% (395/731). This preominance was only statistically significant in male patients, who showe a greater increase from 54.0% (301/557) to 55.8% (255/457) (P =.017). When incluing only MMIS, there were significantly more cancers on the left (31/42, 74%) than the right (11/42, 26%) (P =.002) (Fig 2). This tren was consistent in both sexes, although stronger in men (17/22, 77%) compare with women (14/20, 70%). DISCUSSION For all cancers in both men an women, a leftsie istribution was seen. There were significantly more skin cancers forme on the left sie of the boy than the right in men an especially of the expose areas of the hea an neck. There was no significant ifference observe in women. Several explanations for these finings may exist. Increase exposure to UV rays uring riving may lea to the evelopment of more skin cancers on the left sie of the boy. Previous research has shown that the left sie of the hea, neck, arm, an han receive up to 6 times the ose of UV raiation as the right sie in those sitting on the left sie of the car. 5 Our finings further support this hypothesis, as there was a slightly larger iscrepancy between left- an right-sie skin cancers in those cancers occurring on the hea an neck. There were 54% left-sie cancers when just incluing hea an neck locations versus 52.6% leftsie cancers for all locations. It is also interesting to note that when examining the istribution of skin cancer location for both sies, 82% (731/890)

3 J AM ACAD DERMATOL Butler an Fosko 3 Table I. Skin cancer types by sex an sie of boy Male Female Left Right Total Percent BCC SCC MMIS SCCIS MM Total 557 (63%) 333 (37%) 468 (53%) 422 (47%) BCC, Basal cell carcinoma; MM, malignant melanoma; MMIS, malignant melanoma in situ; SCC, squamous cell carcinoma; SCCIS, squamous cell carcinoma in situ. Table II. Location of skin cancer by sex Male Anatomic location Left Right Left Right Total Percent of population Hea an neck Arm Han Foot Back Chest Leg Total Female Fig 1. Skin cancers on left versus right sie of boy by sex. occurre on the sun-expose areas of the hea an neck an this increase to 92% (817/890) when incluing the arm an han locations (Table II). Most of the UV exposure while riving affects the hea an neck locations, so this fining supports a potential role for the sun exposure gaine while riving. The sie winows of automobiles are mae of nonlaminate clear glass, which blocks UVB but allows 63% transmission of UVA. 6 Though UVB has traitionally been recognize as a contributing factor in skin cancer formation, there have been no conclusive stuies regaring the role of UVA. In vivo stuies have shown that UVA alone can inuce mutations in human DNA, 7,8 an observational stuies have recognize an increase risk of melanoma in patients receiving large oses of UVA from tanning bes or in combination with psoralen. 9 Our observations that significantly more MMIS occur on the sie of the boy expose consistently to UVA penetrating car winow glass suggest that cumulative UVA exposure over years may inee play an important role in the evelopment of this cutaneous malignancy. The significant left-sie preominance of MMIS in this stuy was striking an has prompte further investigation into this fining (Tables III an IV). Future stuies are planne with increase numbers in this subgroup to confirm this observation an further stuy the potential photocarcinogenic role of UVA in MMIS. Interestingly, our invasive melanomas emonstrate a right-sie preominance (70%, 16/23). Although the overall number of invasive melanoma is low, this raises the possibility that both the type of UV exposure, chronic versus intermittent, an boy location, sun expose versus sun protecte, may play a role in the subtype of melanoma evelopment an its site of occurrence. Curiously, this may also suggest that chronic sun exposure may have a protective effect in the evelopment of invasive melanoma. Our stuy was not esigne to aress this possibility an our overall patient numbers for melanoma are small an a significant limitation to any interpretation. This observation is striking in our cohort an eserves further stuy. Chang et al 10 an others have reporte

4 4 Butler an Fosko JAM ACAD DERMATOL Table IV. Anatomic locations for melanoma in situ an invasive melanoma MMIS Melanoma Total Anatomic location Left Right Left Right MMIS MM Hea an neck Arm Han Back Chest Leg Total Fig 2. Melanoma in situ on left versus right sie of boy by sex. MM, Malignant melanoma; MMIS, malignant melanoma in situ. Table III. Melanoma in situ an invasive melanoma by sex an sie of boy Male Female Left Right Total MMIS MM Total 36 (55%) 29 (45%) 38 (58%) 27 (42%) 65 MM, Malignant melanoma; MMIS, malignant melanoma in situ. a relationship of melanoma boy site istribution to patterns an amounts of sun exposure, with recreational intermittent sun exposure associate with melanomas of the trunk an occupational continuous sun exposure with melanomas of the hea an neck. This meta-analysis i not inclue melanomas in situ, but i inclue LM melanomas. Others have reporte that the pattern of sun exposure is relate to the risk of invasive melanoma evelopment, with intermittent, intense perios of exposure associate with an overall increase risk, an more chronic, continuous exposures with a ecrease risk of invasive melanoma evelopment The sex ifference observe in our ata may be explaine by ifferences in riving behaviors between men an women. In this stuy, women i not show a significant left-sie preominance of skin cancers. However, when iviing the ata by ecae, it was note that younger women (\50 years) i in fact show a left-sie preominance (Fig 3). At age 50 years, this tren reverses as rightsie skin cancers become more common (with the exception of those age[ 80 years). In the subgroup of women younger than 50 years, 31 of 54 or 57% of cancers were on the left sie. In women oler than 50 years, 141 of 286 or 49% of cancers were on the left sie. It is important to note that these numbers are relatively small an that analysis of the subgroups is not statistically significant. This tren may represent Fig 3. Location of skin cancer in female patients by age. a shift in riving behavior by women. Perhaps the oler generations have spent more time in the passenger sie of the car, receiving more UV raiation to the right sie of the boy. Although we can only speculate about the riving habits in women, the shift in the tren seen aroun age 50 years may inicate that women born in 1954 or later have spent more time as the river, receiving more exposure on the left sie of their boy. The number of vehicles per househol has nearly ouble in the last 25 years, which may affect the amount of time women spen on the river s sie of the car. 14 It was particularly interesting to fin such a significant asymmetry with regar to MMIS. LM, one subtype of MMIS in particular, is known to evelop on chronically sun-expose skin, so it is possible that the higher frequency of left-sie LM is a result of the cumulative exposure to UV rays while riving a car. The existence of a left-sie tren in both nonmelanoma skin cancer an noninvasive melanomas unerlines the importance of minimizing sun exposure while in the car. This can be achieve by laminating or tinting winow glass. Clear films can be applie to winow glass that block 99.9% of UVA an UVB. One stuy irectly measure cell eath upon exposure to UV light, an note a 93% reuction in cell eath when UV exposure was filtere through laminate auto glass. 15 Our ata also unerline the

5 J AM ACAD DERMATOL Butler an Fosko 5 importance of wearing a aily sunscreen that blocks both UVA an UVB on all expose boy parts. One limitation of this stuy was that our population was limite to patients who were referre to an acaemic meical center an often for Mohs micrographic surgery. Although this may have le to a higher percentage of skin cancers locate on the hea an neck, this selecte population, however, shoul not have skewe the frequency of cancers to the left versus the right. Another factor that may have influence our ata is the preominance of men in our population. The percentage of men in our stuy is consistent with previously reporte incience rates showing that men have nearly ouble the number of nonmelanoma skin cancer as women. 16 The tenency of men to sit on the river s sie of the car may have contribute to the number of left-sie skin cancers in our stuy. There were certainly ifferent trens in skin cancer formation observe when examining the sexes separately. In summary, there were significantly more skin cancers foun on the left than the right sie of the boy in men. For women, a tren of left-sie preominance was seen an was stronger for those younger than 50 years. This iscrepancy was even more profoun in MMIS an in all skin cancers occurring on expose parts of the boy, especially the hea an neck locations. This left-sie preominance of skin cancers may be in part a result of UV exposure uring time spent in an automobile. Future stuies nee to be conucte to etermine if riving behavior is relate to skin cancer formation. The authors are grateful to Evan Jones, MD, an Summer Youker, MD, for early stuy esign; to Elizabeth Duvall, MD, for ata collection; to Chris Kling, MD, for preliminary ata analysis; to Dana Oliver, MPH, MT (ASCP), for statistical analysis; an to Paula Turo an Carrie Angus for aministrative support. REFERENCES 1. Worl Health Organization. Ultraviolet raiation an the INTERSUN Programme Available from: URL: who.int/uv/faq/skincancer/en/inex1.html. Accesse January 17, Singer R, Hamilton T, Voorhees J, Griffiths C. Association of asymmetrical facial photoamage with automobile riving. Arch Dermatol 1994;130: Moulin G, Thomas L, Vigneau M, Fiere A. Un cas unilateral elastose avec kystes et comeons e Favre et Racouchot. Ann Dermatol Venereol 1994;121: Foley P, Lanzer D, Marks R. Are solar keratoses more common on the river s sie? Br Me J Clin Res 1986;293: Moehrle M, Soballa M, Korn M. UV exposure in cars. Photoermatol Photoimmunol Photome 2003;19: Hampton PJ, Farr PM, Diffy BL, Lloy JJ. Implication for photosensitive patients of ultraviolet A exposure in vehicles. Br J Dermatol 2004;151: Burren R, Scaletta C, Frenk E, Panizzon RG, Applegate LA. Sunlight an carcinogenesis: expression of p53 an pyrimiine imers in human skin following UVA I, UVA I 1 II an solar simulating raiations. Int J Cancer 1998;76: International Agency for Research on Cancer (IARC). A review of human carcinogensepart D: raiation. Vol 10. Available from: URL: Accesse August 28, Stern RS, Nichols KT, Vakeva LH. Malignant melanoma in patients treate for psoriasis with methoxsalen (psoralen) an ultraviolet A raiation (PUVA): the PUVA follow-up stuy [see comments]. N Engl J Me 1997;336: Chang Y, Barrett JH, Bishop DT, Armstrong BK, Bataille V, Bergman W, et al. Sun exposure an melanoma risk at ifferent latitues: a poole analysis of 5700 cases an 7216 controls. Int J Epiemiol 2009;38: Elwoo JM, Gallagher RP, Hill GB, Pearson JG. Cutaneous melanoma in relation to intermittent an constant sun exposureethe western Canaa melanoma stuy. Int J Cancer 1985;35: Green A, Siskin V, Bain C, Alexaner J. Sunburn an malignant melanoma. Br J Cancer 1985;51: Osterlin A. Malignant melanoma in Denmark: occurrence an risk factors. Acta Oncol 1990;29: McGuckin N, Liss S. Aging cars, aging rivers: important finings from the national househol travel survey. Institute Transportation Engineers 2005;75: Bernstein EF, Schwartz M, Viehmeyer R, Arocena MS, Sambuco CP, Ksenzenko SM. Measurement of protection affore by ultraviolet-absorbing winow film using an in vitro moel of photoamage. Lasers Surg Me 2006;38: American Cancer Society. Nonmelanoma skin cancer; Atlanta (GA); Available from: org/118.00/ pf. Accesse March 24, 2007.