Arch Dermatol Res (1992) 284:353-357

9 Springer-Verlag1992

UVA tanning devices interact with solar-simulated UV radiation in skin tumor development in hairless mice N. Bech-Thomsen 1, T. Poulsen 2, F. G. Christensen 2, K. Lundgren 1, and H. C. Wulf 1 1 Laboratory of Photobiology, Department of Dermatology, Copenhagen University Hospital, Rigshospitalet H-5192, Blegdamsvej 9, DK-2100 Kobenhavn O, Denmark 2 The Department of Pathology, Sonderborg Sygehus, Sonderborg, Denmark Received March 17, 1992

Summary. The carcinogenic effect of three UVA tanning sources was studied in lightly pigmented hairless mice. The three tanning sources (Bellarium-S SA-I-12, Philips T L 09R and Philips T L 10R) have different emission spectra, and emit different amounts of UVB. Radiation from the tanning sources was administered for 20 min/day, 5 day/week in daily doses equivalent to those used in suntan salons. The radiation was given alone or after 12 weeks of exposure to solar-simulated UV radiation ( S O L A R UV) (10min/day, 4 d a y / w e e k ; daily dose, 19.5 k J / m 2 UVA and 3.9 k J ] m 2 UVB). Irradiation with Bellarium-S SA-I-12 for 47 weeks and Philips T L 09R for 74 weeks induced skin tnmours in 20/20 and 13120 of the animals, respectively. When irradiation with Bellarium-S SA-I-12 and Philips T L 09R was administered after 12 weeks of S O L A R - U V exposure, a strong enhancement of SOLAR-UV-induced photocarcinogenesis was observed (p < 0.001). Irradiation with Philips T L 10R was only slightly carcinogenic, and during 85 weeks of irradiation only one skin tumor appeared in a group of 20 mice. However, when irradiation with Philips T L 10R was administered after 12 weeks of exposure to S O L A R UV, an enhancement of SOLAR-UV-induced carcinogenesis was observed (p < 0.001). Our results suggest that the hazards of exposure to commercial tanning devices are increased when they are used after a period of natural sun exposure. Even tanning sources with a low carcinogenic potential are able to increase S O L A R UV-induced carcinogenesis significantly.

Key words: UV-photocarcinogenesis - hairless mice U V A hazards

over 2 0 - 3 0 years will result in an increased skin t u m o r incidence of 1.3 [6]. U V A tubes used for commercial tanning purposes have different emission spectra. F r o m animal experiments and cell studies it is k n o w n that the biological effect of UVB (280-320 nm) is different f r o m that of U V A (320-400 nm) [13]. UVB radiation is approximately a thousand times more carcinogenic to hairless mice than U V A radiation [4]. Therefore, it is important not only to focus on the total energy output in the U V area, but also on the spectral emission characteristics o f the different tanning sources. In this study we c o m p a r e d the carcinogenic effect in hairless mice of irradiation with three c o m m o n tanning tubes with different spectral distributions. The three U V A tanning sources were administered in doses comparable to a tanning session, alone and after a period o f solar-simulated U V radiation ( S O L A R UV).

Materials and methods Animals

Lightly pigmented, hairless, female hr/hr mice with a C3H/Tif background (n = 240) were used. The animals were 29 weeks old at the start of the experiment. They were divided into eight groups (Table 1). Groups A1, B1, C1 and D comprised 40 mice, and the remaining groups comprised 20 mice each. Consecutive numbers were tattooed on the abdomen of the mice. Each group was housed in a separate box. The animals had free access to water and standard laboratory feed, and were kept under a light-dark cycle of 12-12 h. The room temperature was kept at 23-24 ~ The animals were irradiated in their boxes from above. Light sources

There is increasing concern a b o u t the long-term risks of exposure to commercial tanning devices. W h a t impact the use of such devices will have on the frequency of skin tumors in the populations of northern Europe and north America is unknown. Using a mathematical model, it has been estimated that weekly use of a U V A solarium Correspondence to: N. Bech-Thomsen

Broad-spectrum UV radiation, simulating the UV part of the solar spectrum (SOLAR UV) was obtained from one Philips TL 12 and five Bellarium-S SA-I-12 tubes (Philips, Eindhoven, The Netherlands). The emission spectrum is shown in Fig. 1. The emission spectra of the three commercially available UVA tubes (Bellarium-S SA-I-12 (Bellarium), Philips TL 09R (TL09) and Philips TL 10R (TL10)) are shown in Fig. 2. The emission spectra of the radiation sources were measured at 1 nm intervals using a Jobin Yvon H10

354 Table 1. U V treatment schedule and administered doses of UVA

SOLAR UV" (12 weeks)

UVA treatment (daily dose)

A

-

A1

+

B B1 C CI D E

+ + + -

Bellarium-S SA-1-12 Bellarium-S SA-I-12 Philips TL 09R Philips TL 09R Philips TL 10R Philips TL 10R None None

Group

Source

UVA (kJ/m 2)

UVB (kJ/m 2)

B-MED b

Duration c (weeks)

81.7

6.1

1.2

47

81.7

6.1

1.2

22

162.8 162.8 198.7 198.7 0 0

3.7 3.7 0.4 0.4 0 0

0.8 0.8 0.4 0.4 0 0

74 29 85 47

a Daily dose during SOLAR U V exposure: 19.5 kJ/m 2 U V A and 3.9 kJ/m 2 UVB, equivalent to 4.67 B-MED b B-MED, basic minimal erythema dose (as defined in the text) ~ Irradiation continued to death or killing of animals

double m o n o c h r o m a t o r (slit widths 0.5, 1.0 and 0.5 ram) and an E G & G 550 radiometer with a calibrated detector (EG & G 550-2 multiprobe). The irradiance of the radiation sources was measured with an International Light (IL) 1700 research radiometer. An IL SED 240 detector with a UVB filter an a W quartz diffuser was used for the Philips TL12 tube. An IL SED 400 detector with a WBS 320 filter and a W quartz diffuser was used for the U V A sources. Measurements were corrected on the basis of the spectral sensitivities of the detectors, and the emission spectra of the lamps. The Doses REL ~,TIVE E M I S S I O N 1.0l'

0.5

0

280

'

'"

320"

'

' 3(i0"

'

' 4bO . . . .

440 nm

administered, as kJ/m 2 and as h u m a n basic minimal erythema doses (B-MEDs), are shown in Table 1. One B-MED = 3 1 2 J / m z at 296 nm (24 h erythema) [9]. Doses in B - M E D were calculated from the CIE h u m a n erythema action spectrum, [8], which is very similar to the mouse oedema action spectrum [3]. The CIE h u m a n erythema action spectrum includes both the biological effect of UVA and UVB radiation [8].

Exposure schedule (Table 1) Groups A, B and C were irradiated with the Bellarium, TL09 and TL10 tubes, respectively. The groups were irradiated from week 13; they were not irradiated during the first 12 weeks of the experiment. The distance from the tubes to the level of exposure was 14 cm. The groups were irradiated for 20 min/day, 5 days/week until skin tumours (groups A and B) or age (group C) necessitated killing. G r o u p D was irradiated with SOLAR UV for 10 rain/day, 4 days/week for 12 weeks. The distance from the tubes to the level of exposure was 45 cm. During the 12 weeks a total dose of 187 kJ/m 2 UVB and 936 kJ/m 2 U V A was given. Groups AI, B1 and C1 were initially irradiated with SOLAR U V for 12 weeks as for G r o u p D. F r o m week 13 the groups were irradiated with the Bellarium, TL09 and TL10 tubes, respectively, as for groups A, B and C. G r o u p E was not irradiated.

Fig. 1. Spectral output of the SOLAR U V

Registration and statistics

REo•ATIVE EMISSION

9 '

~

~i

"~

0,5

~, "'~% 0 280

320

ii

ii ii

360 rim

Fig. 2. Spectral output of the three U V A tanning sources. - Bellarium-S; - - - - - , Philips TL 09R; . . . . . . . . Philips TL 10R

The first ten tumours with a diameter _> 1 m m were mapped separately for each animal. Changes in size, solour and ulcerations were noted weekly for every registered tumour. Tumours that regressed were not excluded from the analyses. The difference in time from the start of the experiment to the appearance of seven tumor end-points between the groups was analysed using the Log-rank test. The end-points were: the first, second third and fifth visible tumor with a diameter > 1 mm; the first visible tumor with a diameter _>2mm; total tumor area _>25 m m z and _> 100 mm2; and death. The time from the appearance of the first tumor with a diameter _> 1 m m to the appearance of the second, third and fifth visible tumor, and the growth of the tumours, were analysed using the M a n n Whitney test. The probability of survival without a certain tumour end-point was calculated using the actuarially adjusted K a p l a n - M e i e r method

[10].

355 Table 2. Time in weeks to 0.5 level on the life-table for seven selected tumor end-points

Group

Tumor with a diameter _>1 mm

First tumor _>2 mm

Tumor area > 25 mm 2

> 100 mm 2

41 24

44 28

56 28

72 30

75 36

77 4l

Only one mouse developed one tumor 41 46 51 55 43

54

59

D

55

68

_a

E

No mice developed tumors

1st

2nd

3rd

5th

A A1

41 24

43 25

43 26

43 28

B B1

69 28

69 32

69 33

73 36

C C1

57

60

69

56

a Insufficient number of observations

Pathology

PROBABILITY OF SURVIVAL WITHOUT A TUMOR > 1 M M

1.0

The dorsal skin was removed from all killed animals and fixed in 4% buffered formaldehyde. Histological examination was performed by light microscopy after H & E staining on 20 of the mice, in which none of the skin tumors could fulfil the ctinical criteria for malignancy. The criteria were ulceration, progressive growth and a final diameter > 4 mm.

o,

Results

Acute effects 0,0

All S O L A R - U V - i r r a d i a t e d mice developed erythema, which persisted for approximately 1.5 weeks, at the beginning of the experiment. S O L A R - U V radiation induced pigmentation within 3 weeks. O f the three U V A tanning sources only the Bellarium tube was able to induce pigmentation in the mice.

'

0

30

I

50

'

I

70

I

90 WEEKS

Fig. 3. Kaplan-Meier plot showing time to first tumor with a diameter > 1 mm for the groups irradiated with SOLAR UV + Bellarium-S SA-I-12 (UVR-BELL, group A1), SOLAR UV + Philips TL 09R (UVR-TL09, group BI), SOLAR UV + Philips TL 10R (UVR-TLIO, group C1) and SOLAR UV alone (UVR, group D)

Chronic effects S O L A R UV versus S O L A R UV + UVA tanning sources (Fig. 3). Time to all tested t u m o r end-points was significantly longer in group D ( S O L A R U V alone), c o m p a r e d with the groups that were first irradiated with S O L A R U V and were subsequently irradiated with the Betlarium (A1), TL09 (B1) and TL10 (C1) tubes, respectively (p < 0.001). pairwise comparisons between the groups A1, B1 and C1 also showed a significant difference in time to all tested t u m o r end-points (p < 0.001). Table 2 shows the time in weeks to the 0.5 level on the K a p l a n - M e i e r plot for the seven t u m o r end-points. UVA tanning sources alone (Fig. 4). Irradiation with the Bellarium tubes (group A) for 47 weeks at a daily dose of 1.2 M E D , and with the TL09 tubes (group B) for 74 weeks at a daily dose of 0.8 M E D resulted in skin t u m o r development in 20/20 and 13/20 of the animals, respectively. The time to all tested end-points was significantly shorter in the Bellarium group (A) c o m p a r e d with the TL09 group (B) (p < 0.0001). One tumor-bearing mouse was observed (week 60) in group C, irradiated with the TL10 tubes at a daily dose

of 0.4 M E D for 85 weeks. The group is not shown in Fig. 4.

S O L A R UV + UVA tanning sources versus UVA tanning sources alone (Fig. 4). A pairwise comparison between the groups A1 vs. A, and B1 vs. B showed that t u m o r development was significantly enhanced in the groups that were first irradiated for 12 weeks with S O L A R U V and subsequently with the U V A tanning sources, compared with the groups irradiated only with the U V A tanning sources (p < 0.001). The results of the M a n n - W h i t n e y tests showed that the appearance of additional tumors was highly dependent on the daily M E D dose during exposure to the U V A sources. In contrast, t u m o r growth was independent of the daily dose. T u m o r growth f r o m 1 m m to 3 ram, and from 1 m m to 4 m m was not significantly different in group A1 c o m p a r e d with group D, and group C1 c o m p a r e d with group D (p _> 0.25; sample sizes from 7 to 68). The histological examination showed that 18 out of the 20 selected mice had squamous cell carcinomas (SCCs). Two mice did not have malignant tumors: one,

356 PROBABILITY OF SURVIVAL WITHOUT A TUMOR ~ 1 MM 1.o

..........

. . . . . . .

',

I

I

I

I

i_ I 'BELL

0.5

h ~I

q

0,0

10

30

50

~

,

70

90 WEEKS

d

Fig. 4. Kaplan-Meier plot showing time to first tumor with a diameter > 1 mm for the groups irradiatedwith SOLARUV + Bellarium-S SA-I-12 (UVR-BELL, group A1), SOLARUV + Philips TL09R (UVR-TL09, group B1), Bellarium-S SA-I-12 alone (BELL, group A) and Philips TL 09R alone (TL09, group B) in group C1, did not show any sign of malignancy, and the other, in group D, had hyperplasia and slight dysplasia.

Discussion

Several studies have been conducted concerning the effect of UVA radiation applied after exposure to UVB or broad-spectrum UV radiation [7, 14-16]. In a study using Skh hairless albino mice, UVA from F-40 T12BL-O black lights filtered through 6-mm glass was applied after exposure to different periods of irradiation with FS-40 T12 Westinghouse sunlamps [7]. The tumor yield was lower in all groups irradiated with UVA compared with the groups that were only irradiated with the sunlamps. In another study two groups of Skh albino mice were irradiated either with UVA or UVB, while a third group alternated between the two regimes, changing every week [14]. The interchanging exposures to UVA and UVB did not result in an elevated tumor rate. A study from our laboratory using lightly pigmented hr/hr hairless mice has shown that mice initially exposed to artificial sunlight (Westinghouse 40 W FS-40 sunlamp and two Philips TL-40 W/09 tubes) and subsequent UVA (three Philips TL-40 W/09 tubes) for 2, 4 or 6 months had an increased tumor development [15]. A similar result was seen when the UVA radiation was filtered through a 2-mm thick plain glass plate [16]. The promotive effect seen in this study of the two UVA radiation sources with the largest output in the UVB area (Belarium 6.9% and TL09 2.2%) on SOLAR-UV carcinogenesis is not surprising. Both UVA tanning tubes were by themselves carcinogenic. The tumor-enhancing effect of the two UVA sources on SOLAR-UV-induced tumours may be explained by additive carcinogenic responses. It is more interesting that irradiation with the TL10 tubes (UVB content 0.2%) was able to enhance SOLARUV carcinogenesis. In group C1, which was initially

irradiated with SOLAR UV and subsequently with TL 10, the median time to the first tumour was 41 weeks. This means that tumors appeared 29 weeks after cessation of the initial carcinogenic stimulation with SOLAR UV. Two possible mechanisms can be envisaged. First, even though the TL10 is a very weak carcinogen, the concept of additivity may well account for our observations. A dose that is too small to produce tumors in a lifetime can significantly add to the effectiveness of a higher dose [7]. Second, irradiation with the TL 10 tubes might induce alterations in the immune response of the mice, leading to an impaired rejection of SOLAR-UV-induced skin tumors [11, 12]. Our study shows that irradiation from TL10 tubes at a dose sufficient to induce immunological alterations in humans [11], is also able to enhance UV-induced skin tumor development in mice. The tumorenhancing effect is present, even though the tumourinitiating ability of TL 10 alone, for all practical purposes, is absent. In a previous study in which we exposed a group of 20 hr/hr C3H/Tif mice to TL10 radiation for 98 weeks we observed that 6 of 20 mice developed tumors [2], which is a higher incidence than observed in the TL10irradiated group C in this study. This is probably due to a lower daily exposure dose in this study compared with the previous study (0.4 B-MED versus 0.6 B-MED). The Mann-Whitney tests showed that the growth of individual tumors > 1 mm was independent of UV exposure. This is surprising, because some mice were unirradiated during tumor growth, while other mice were exposed to 1.2 B-MED of Bellarium radiation during tumor growth. In contrast the appearance of additional skin tumors was highly influenced by the daily dose of the UVA sources. Our results are in accordance with the findings of earlier studies [5, 17]. Previous studies, in which we examined histologically all skin tumours, demonstrated that all larger tumors were SCCs [11, 17]. In the present study the mice were killed when necessitated by tumor development. This allowed the tumors to grow excessively. Histological examination was therefore restricted to 20 mice, with skin tumors that were not obviously malignant. Of the 20 examined mice, 18 had SCCs. We found that two out of the three UVA sources were able to induce skin tumors. The strongest carcinogen was Bellarium tube radiation. We observed that only the mice irradiated with the Bellarium tubes developed a visible pigmentation. If all the UVA tanning sources had been given in doses capable of inducing the same degree of pigmentation, then the carcinogenic effect of the Bellarium tubes might have been similar to the other UVA sources. All tested UVA tanning sources had the ability to increase significantly the carcinogenic effect of previously administered solar simulated UV radiation. This effect can be exerted even by irradiation with UVA tanning sources with a low output of UVB radiation that by itself is only minimally carcinogenic. Our data suggest that the hazards of exposure to commercial tanning devices is largely increased when the equipment is used after a period of natural sunlight exposure.

357

References 1. Bech-Thomsen N, Wulf HC, Poulsen T, Lundgren K (1988) Pretreatment with long-wave ultraviolet light inhibits ultraviolet-induced skin tumor development in hairless mice. Arch Dermatol 124:1215-1218 2. Bech-Thomsen N, Wulf HC, Poulsen T, Christensen FG, Lundgren K (1991) Photocarcinogenesis in hairless mice induced by ultraviolet A tanning devices with or without subsequent solar-simulated ultraviolet irradiation. Photodermatol Photoimmunol Photomed 8:139-145 3. Cole CA, Davies RE, Forbes PD, Aloisio LD (1983) Comparison of action spectra for acute cutaneous responses to ultraviolet radiation: man vs albino mice. Photochem Photobiol 37:623-631 4. Cole CA, Forbes D, Davies RE (1986) An action spectrum for UV photocarcinogenesis. Photochem Photobiol 43:275-284 5. De Gruijl FR, van der Meer JB, van der Leun JC '(1983) Dose-time dependency of tumor formation by chronic UV exposure. Photochem Photobiol 3 7 : 5 3 - 6 2 6. Diffey BL (1987) Analysis of the risk of skin cancer from sunlight and solaria in subjects living in northern Europe. Photodermatol 4:118-126 7. Forbes PD, Davies RE, Urbach F (1978) Experimental ultraviolet photocarcinogenesis: Wavelength interactions and time dose relationships. NCI Monogr 50:31-38 8. McKinlay AF, Diffey BL (1987) A reference action spectrum for ultraviolet induced eryhtema in human skin. CIE J 6: 17-22

9. Parrish JA, Jaenicke KF, Anderson RR (1982) Erythema and melanogenesis action spectra of normal human skin. Photochem Photobiol 36:187-191 10. Peto R, Pike MC, Day NE et al. (1980) Annex to international agency for research on cancer supplement 2, IARC Monographs on the evaluation of the carcinogenic risk of chemicals to humans. IARC, Lyon, pp 311-426 11. Rivers JK, Norris PG, Murphy GM et al. (1989) UVA sunbeds: tanning, photoprotection, acute adverse effects and immunological changes. Br J Dermatol 120:767-777 12. Romerdahl CA, Kriepke ML (1988) Role of helper T-lymphocytes in rejection of UV-induced murine skin cancers. Cancer Res 48:2325-2328 13. Roza L, Baan RA, Leun JC van der, Kligman L, Young AR (1989) UVA hazards in skin associated with the use of tanning equipment. J Photochem Photobiol/[B] 3:281-287 14. SlaperH (1987) Skincancer and UV exposure : investigationson the estimations of risks. PhD thesis, Utrecht, pp 117-141 15. Staberg B, WulfHC, Poulsen T, Klemp P, Brodthagen H (1983) Carcinogenic effect of sequential artificial sunlight and UVA irradiation in hairless mice. Arch Dermatoi 119:641-643 16. Staberg B, WulfHC, Klemp P, Poulsen T, Brodthagen H (1983) The carcinogenic effect of UVA irradiation. J Invest Dermatol 81:517-519 17. Wulf HC, Poulsen T, Brodthagen H, Hou-Jensen K (1982) Sunscreens for delay of ultraviolet induction of skin tumors. J Am Acad Dermatol 7:194-202