Biological Amplification Factor for Sunlight-induced ... · Biological Amplification Factor for Sunlight-induced Nonmelanoma Skin Cancer at High Latitudes1 Johan Moan,2 Arne Dahlback,

[CANCER RESEARCH 49. 5207-5212. September 15. 1989)

Biological Amplification Factor for Sunlight-induced Nonmelanoma Skin Cancerat High Latitudes1

Johan Moan,2 Arne Dahlback, Thormod Henriksen, and Knut Magnus

Institute for Cancer Research, Montebello [J. M.J, Institute of Physics, University of Oslo [A. D., T. H.J, and The Norwegian Cancer Registry, Montebello /K. M.J,0310 Oslo 3, Norway

ABSTRACT

Data for the incidence of basal cell carcinomas (BCCs) and squamouscell carcinomas (SCCs) of the skin, registered for six regions of Norwayduring 10 years (1976-1985), were used to evaluate the biological amplification factor Ab for induction of these cancers by sunlight. A,, is theratio of the increment in skin cancer production to the increment incausative sunlight exposure. Two different approximations were used forthe action spectrum for carcinogenesis: an erythema action spectrum; andan action spectrum for mutagenesis of cells in the basal layer of the skin.These two fundamentally different approaches yielded Ab values thatwere similar to within about 10%: 2.1-2.3 for BCCs; and 1.6-1.8 forSCCs. Using a radiation amplification factor for ozone depletion of 0.8-1.1, we find that the total amplification factor for BCCs is within therange 1.6-2.1 and that that for SCCs is within the range 1.3-1.7 atnorthern latitudes of 60-70 degrees. Thus, an ozone depletion of 1% willresult in an increase in the incidence of BCCs by 1.6-2.1% and of SCCsby 1.3-1.7%. There were no significant differences between the valuesfor men and women. Neither was there any significant difference betweenAh values found for skin commonly exposed to sunlight (face) and forskin sites normally covered by clothes and therefore receiving much lowerexposures, in spite of the fact that the tumor density per unit skin areawas a factor of 20 or more larger at the former sites. This observation,as well as the curves relating cancer incidence with annual exposure tocarcinogenic sunlight, supports a power law relationship between cancerincidence and annual sun exposure.

Sunlight appears to be the main cause of BCCs and SCCs even at thehigh latitudes of Northern Norway. All over, BCCs were found to beabout 6 times more frequent than SCCs. The ratio of the incidence ofBCCs to that of SCCs seemed to be independent of the latitude. Finally,BCCs were found to be equally frequent among men and women, whileSCCs were found to be about twice as frequent among men as amongwomen.

INTRODUCTION

The Antarctic "ozone hole" was recognized only four seasons

ago (1, 2). In 1988 the Ozone Trend Panel, an internationalgroup of atmospheric scientists, announced, based on Dobsonmeasurements, a small negative trend for the Northern Hemisphere in the period 1969-1986 (3). CFCs1 seem to play a role

as catalysts in the ozone depletion process (1, 4). The lifetimeof CFCs in the atmosphere is expected to be long. Thus, if theCFC theory is correct one may expect the depletion to go onfor decades, even if the production of CFCs is reduced fromnow on.

In order to estimate the consequences of an eventual ozonedepletion for the incidence of skin cancer it is important toknow the biological amplification factor, Ah. Ah is the ratio ofthe increment in skin cancer production to the increment in

Received 3/14/89; revised 6/2/89; accepted 6/15/89.The costs of publication of this article were defrayed in part by the payment

of page charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

' Supported by the Norwegian Research Council for Science and the Human

ities.2To whom requests for reprints should be addressed.' The abbreviations used are: CFC, chlorofluorocarbon; SCC, squamous cell

carcinoma; BCC. basal cell carcinoma; CÃ•E,Commission Internationale del'Ã‰clairÃ©;MUT, spectrum for mutation of cells in the basal layer of the skin.

causative sunlight exposure as defined by the equation

dR IdD

where R is the age-adjusted incidence of skin cancer and D is

the accumulated exposure to carcinogenic sunlight. In the calculation of the effective UV exposure (D) the action spectrumfor carcinogenesis is needed. Only approximations exist forsuch a spectrum. In the present work we have compared theresults of using two spectra based on fundamentally differentobservations: one is an approximation to the action spectrumfor erythema termed the CIE spectrum (5); and the other is aspectrum for mutation of cells in the basal layer of the skin,termed the MUT spectrum.

Several estimates of the Ah can be found in the literature (6-12). It has been reported that Ah changes with the latitude, i.e.,with the annual exposure to sunlight (6). Therefore, and sincean eventual global ozone depletion most likely will manifestitself early at high northern and southern latitudes (1, 11), it isimportant to determine the biological amplification factor relevant for such regions.

Since the induction of melanoma is related to episodes ofsunburn, notably in the childhood, and therefore supposedlynot only to the total exposure but also to the fluence rate (13),we have included only SCCs and BCCs in in the present work.

MATERIALS AND METHODS

The fluence rate of carcinogenically effective solar radiation is defined by the expression Ec = Â¡E(t><d\,the integration being performedover the wavelength region of the solar spectrum. E is the solarirradiano.- at earth's surface, and o, is the action spectrum for carcino

genesis.Â£was determined by using a discrete ordinate algorithm to calculate

the propagation of light in vertically inhomogenous, plane parallelmedia (14). The model atmosphere used was the U. S. StandardAtmosphere 1976 which was divided in 39 homogenous layers with athickness of 2 km. We used the extraterrestrial solar radiation spectraas well as all orders of scattered light (Rayleigh scattering) from theatmosphere. The ground albedo was set equal to 0.2. The absorptionspectrum of ozone was taken from Ref. 15. The terrestrial spectracomputed in this way coincide almost completely (both with respect toshape and absolute values) with the spectra recorded in the same areaand at the same zenith angle and ozone concentration (16). o. forhuman carcinogenesis is not known. In the present work we have usedtwo different approximations for 0. : (a) .â€¢/,is assumed to be identicalwith the "reference spectrum", 0CIE,for erythema in humans proposed

by the CIE (5); (b) <t>eis approximated by T<t>â€ž,where T is the transmit-tance through the epidermis as given in Ref. 17, and <t>â€žis the actionspectrum for mutation of human cells. Thus, according to this procedure, the assumption is made that photocarcinogenesis is induced bymutation of cells in the basal layer of the skin. This is justified by theobservation that there seems to be a correlation between mutagenesisand malignant transformation (18).

Several action spectra for mutation of mammalian cells in culturecan be found in the literature. The spectrum used in the present work(Fig. I) is based on the data for human fibroblasts (19), for humanepithelial P3 cells (20), and for Chinese hamster ovary cells (21). With

5207

on June 14, 2020. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

SUNLIGHT-INDUCED NON-MELANOMA SKIN CANCER

a couple of exceptions all the action spectra are similar. Exceptionallylow values for mutation were found at 334 nm (20) and at 365 nm (21).This may be related to the fact that light in this wavelength region canmodify (6-4) photoproducts (22). Thus, the approximate average spectrum may give a contribution which is too high at certain wavelengthsin the UVA. However, a systematic error which contributes in theopposite direction is the fact that the integrations include radiationonly up to 405 nm. No data points exist to allow an extension to longerwavelengths. Furthermore, animal experiments indicate that UVA iscarcinogenic and, as indicated by Fig. 1, the spectrum T<)>â€žbears closeresemblance to the action spectrum for carcinogenesis in mice (23, 24).

The annual exposure to carcinogenic radiation from the sun Â¡sD =fE,dt, the integral being taken over 1 year. In our calculations theintegrals were approximated by the sums

and

=1.1

OCIE= I I E, 0c,EAAAf

with Ai = l h and AX = 1 nm. The seasonal average ozone levels atdifferent latitudes were used.

Epidemiological Data. Data for incidence of skin cancer were provided by the Norwegian Cancer Registry. All the incidences were ageadjusted to the European standard population (25). Standard errors aregiven as JÃ‘/N where N is the number of new cases during the observation period.

Norway was divided into six zones as shown in Fig. 2. The areaaround the capital Oslo is densely populated compared with the otherregions and we have reason to believe that different practice of reportingskin cancer may apply in this region. Therefore, it was excluded apriori. In all areas between 50 and 70% of the population live in ruralareas. Practically all inhabitants are Caucasian and we have no reasonto believe that there is any difference between the regions with respectto the distribution of persons with different skin types. A small fractionof the total exposure of the population to carcinogenic solar light maybe acquired during vacations to countries at low latitudes. Using information about the vacation habits of Norwegians, we have estimated anupper limit of the error due to this factor.

RESULTS AND DISCUSSION

An action spectrum for mutagenesis of cell in the basal celllayer of the skin can be estimated by using the transmissionspectrum T for epidermis ( 17) and the action spectrum <>,â€žformutagenesis of mammalian cells (Fig. 1). The latter spectrumis probably quite reliable up to about 310 nm since in this

300 350 400

- MUT spectrumCIE spectrumCarcinoganasis in mice

60.9"-

59.0Â°- -

Population

M : 115100109800

Â®M: 123700F: 120400

M: 302600F: 303000

M: 147 500F: 149000

M: 184 200F: 183000

M: 344 145F: 348270

Fig. 2. Map of Norway indicating the six regions considered in the presentwork.

20 80

Fig. 1. Action spectra: ( ) 7"0m(see the text); ( ) CIE action spectrum

for erythema in humans (5); ( ) carcinogenesis in mice (24).

5208

Â¿0 60

Latitude (degrees)

Fig. 3. Annual exposure to carcinogenic light at sea level at different latitudescalculated by use of the CIE and the MUT action spectra. Insel, data for theactual latitudes for maximum sun exposure (24 h) and for 4 h (10 a.m.-2 p.m.)sun exposure every day in a year.

wavelength region it follows the absorption spectrum of DNA,the most likely chromophore for photocarcinogenesis in wavelength regions where it absorbs. At longer wavelengths, however, DNA has a very low absorbance and carcinogenesis mustbe related to light absorption in other, yet unknown, chromo-phores. The action spectrum T<t>m(Fig. 1) seems to be realisticfor photocarcinogenesis since it agrees surprisingly well withthe action spectrum for carcinogenesis in mice (24) as well as




Table 1 Relative exposures to carcinogenic sunlight at different latitudes

Exposure (CIE)Latitude Sunshine/yr Av. sky

Region (degrees)" (h)4 coverc Uncorrected'' Corrected'

12345669.666.963.160.960.959.01200120012701200150015000.720.710.660.730.630.62.00.10.27.39.39.50.00.10.31.38.46.59

" Latitude of the center of gravity for the population.4 Averaged over the time period 1956-1980. Data from the Norwegian Mete

orological Institute.f Averaged over 1 yr and over several (>3) stations of observation in each

region. Data from the Norwegian Meteorological Institute.d Total exposure to carcinogenic sunlight calculated by using the CIE action

spectrum. The data are normalized to 1 in region 1.' Corrected exposures to carcinogenic sunlight. The corrections were carried

out using data from Fig. 4.3 in Ref. 16.

with the CIE erythema spectrum (5). Mutations of cells in thebasal layer are likely to play important roles in skin photocar-cinogenesis. For comparison, all further calculations were carried out with both the T<t>mspectrum (termed the MUT spectrum) and the CIE spectrum.

As expected, the variation of the annual exposure to carcinogenic sunlight with the latitude is slightly dependent onwhether the CIE spectrum or the MUT spectrum is used in thecalculations (Fig. 3). In these calculations we have taken intoaccount the zonal monthly averages of the ozone level throughout the year. For the latitudes of the present epidemiolÃ³gica!investigation we have also calculated the annual exposure between 10 a.m. and 2 p.m. (Fig. 3). This was done in order tosee how critically the further calculations would be dependenton the habits of sun exposure that are assumed for the population.

Sun exposure during vacations to southern latitudes may

present a systematic error in the present work. We have triedto estimate an upper limit of this error. According to information from Norwegian traveling agencies there is no significantdifference in the frequency of vacations to southern latitudesper inhabitant in Northern and Southern Norway. Furthermore,the frequencies for men and women are similar to within 2%.Finally, the annual average stay at southern latitudes per inhabitant in the time period 1976-1985 is about 1.5 days (data fromthe Norwegian Transporteconomical Institute), which wouldadd less than 5% to the annual exposure to carcinogenic lighteven for people from the northern part of the country. Thismight, at worst, lead to an underestimation of/!,, of about 10%.

Fortunately, the average sky cover is not dramatically different for the different regions of the country (Table 1). Thedependence of the exposure to carcinogenic sunlight upon theaverage sky cover has been determined by Josefson (16). Usingthis information we have corrected the exposures for differencesin sky cover between the regions (Table 1). The reliability ofthis correction method was tested on data in Ref. 26 where theRobertson-Berger exposures were measured and averaged overthe time period 1974-1985 are given, together with the averagesky cover at eight different locations in the United States. Someof these locations (i.e., Oakland and Philadelphia) are at asimilar latitude and elevation but have a different sky cover,i.e., 0.48 versus 0.62. If we had used the correction factors forsky cover that can be estimated from Ref. 26, we would haveobtained Ah values that are less than 10% lower than the oneswe have calculated by use of the data in Ref. 16. Taking intoaccount that pollution may play a larger role in attenuating theUV light in large American cities than in Scandinavian ruralareas, we chose to use the correlation factors from Ref. 16 inthe further calculations.

Data for the relative tumor density at different anatomical

Table 2 Relative tumor density of basal cell carcinomas of the skin in different regions in Norway

MalesSite-Skin

area(f)4Region0123456A0.00520"220200180300440B0.001250018002500290021003100C0.059227258397444439632D0.28182138433468E0.6553.72.96.46.15.511.8A-C0.065246278414462458655A-E1.023.526.041.446.342.669.3(A-O/E669665768356A0.00540208060120120B0.00117009002000220021002500C0.059192264325359430568FemalesD0.28142131283153E0.6552.34.04.46.04.08.4A-C0.065203255332365432563A-E1.018.625.133.335.056.839.5(A-O/E8864756110867

Â°The capital letters correspond to the following sites on the body: A, outer ear; B, eyelid plus corner of the eye; C. face and head except (A+B); D, neck plus trunk;

E, all other sites.4 f, fraction of total skin area (27).' Regions of the country according to Fig 1.Â¿Values are R/f, where R is the annual age-adjusted incidence of cancer per 100,000 averaged over the 10 years 1976-1985. The standard error for R/f is

â€”¿�â€”-,being the population in the region divided by 100,000 (see Fig. 1).lOn

Table 3 Relative tumor density ofsquamous cell carcinomas of the skin in different regions in Norway

MalesSite"Skin

area(f)4Region'123456A0.005240*520520520680720B0.001600300100200300200C0.059375671697386D0.280.71.42.93.22.94.3E0.6551.82.42.62.32.75.5A-C0.0656195106106126134A-E1.05.48.29.39.210.813.4(A-O/E344041464724A0.005202020204040B0.001300100100100100200C0.059353146617175FemalesD0.280.71.81.11.82.51.8E0.6550.31.11.11.51.42.3A-C0.065383145586974A-E1.02.93.34.05.55.97.1(A-O/E(28)(41)394932

a'a See Table 2, Footnotes a-d.

5209




Table 4 Ratio of density of basal cell carcinomas to squamous cell carcinomas

Males Females

Site" B C D E A-C A-E A B C D E A-C A-E

Region1234560.080.420.380.350.440.614.16.025157.0166.1264.6155.6136.4136.0127.3

162.11.22.52.72.02.14.02.93.94.43.64.94.43.24.45.03.95.22

5.719.04203223213

125.58.57.15.96.17.620

7.7193.6284.0164.0122.929

3.75.38.27.46.36.37.66.47.68.36.49.65.6

Â°For explanation of symbols, see Table 2, Footnotes a-d.

Fig. 4. Data for tumor density A//of BCCs and for SCCs for men and womenin six regions of Norway (Fig. 2). The total annual exposures are given relativeto the exposure in region 1. Upper four curves, data for skin areas A-C [face andhead (Table 2)| receiving high exposures. Lower four curves, data for skin area E(see Table 2, Footnote a). Bars, magnitude of the standard errors at the differentlevels.

Table 5 At values obtained by using the CIE action spectrum

CancerBCCsecSexMFMFMFMFSite"A-CA-CEEA-CA-CEE4hAt2.022.092.422.131.501.781.121.86r'0.970.970.840.870.930.870.811.81a"0.180.150.700.650.280.420.330.70

â€¢¿�See Table 2, Footnote a.* Values obtained by use of the total annual exposure to sunlight.' Statistical significance.d Standard errors in /(Â».

sites of basal cell carcinomas and squamous cell carcinomas ofthe skin observed in different regions in Norway are given inTables 2 and 3. The relative tumor density is here defined asR/f, i.e., the annual incidence per unit area of skin. The annualincidence is averaged over 10 years (1976-1985) and given in

5210

age-adjusted numbers of cases per 100,000 (see "Materials andMethods"). The standard errors of the numbers R/f given in

the tables can be estimated as described in the legend of Table2. For instance, standard errors for Columns A-C for males inTable 2 range between 8% (region 1) and 3% (region 6) of thenumbers given in the table. The standard errors for the corresponding column in Table 3 range between 6 and 16% of thenumbers given. The standard errors for the numbers in columnsA and B are large (in certain cases >40% of the numbers given).Therefore, these numbers are only rough indications of thetumor density at these locations and cannot be used for adetailed comparison of the incidence in different regions of thecountry.

The following conclusions can be drawn from Tables 2, 3,and 4 but will not be discussed in detail, since they are notclosely related to the evaluation of Ah.

1. Data for all six regions and for men and women combinedindicate that BCCs are about 6 times more frequent than SCCs(Table 4, columns A-E). This is a slightly higher ratio thanfound by others (9, 28).

2. The ratio of the frequency of BCCs to SCCs is large forsite D (neck and trunk) and low for sites A and B (ear andeyelid plus corner of the eye, respectively) (Table 4). This doesnot necessarily mean that important induction mechanismsother than sun exposure exist.

3. For males the ratio of the frequency of BCCs to SCCs atsite A (ears) is surprisingly small. Furthermore, from Tables 2and 3 one can evaluate that the average ratio between theincidence of BCCs at site A for men and women is 3.9 Â±1.5while the corresponding ratio for SCCs is much larger, at 21 Â±2. None of these ratios shows any decreasing or increasing trendwith decreasing latitude. Different hair fashions, giving rise toa much higher exposure of men's ears than women's ears to

sunlight, explain why the ratio is larger than unity. The strikingdifference between the ratio for BCCs and SCCs may be explained on the basis of protective and/or potentiating factorsat this specific site for one or both of the sexes and deservesfurther attention.

4. With the possible exception of site A for men, the ratioof the frequency of BCCs to SCCs is not dependent on thelatitude (Table 4).

5. Overall, BCCs are about as frequent among women asamong men while SCCs are about twice as frequent among menthan among women (Tables 2 and 3). Thus, the ratio of BCCsto SCCs is higher for women (7.3 Â±1.0) than for men 4.3 Â±0.6) (Table 4).

6. The relative tumor density of BCCs is highest at site A(eyelid + corner of the eye, Table 2).

The present work indicates that sunlight is the major causeof both BCCs and SCCs of the skin. Thus, in both cases theratio of the tumor density at light-exposed sites, such as theface, to that at sites which are normally shielded from light byclothes is large. This is demonstrated by columns (A-Q/E inTables 2 and 3.

On the average this ratio is 75 Â±10 for BCCs and 38 Â±7 forSCCs. The fact that the ratio is only one-half as large for SCCsas it is for BCCs may indicate that factors other than sunlightplay a slight role in the case of SCCs. The ratio does not appearto be dependent on the region (Tables 2 and 3), an observationthat indicates that Ah is similar for sites which receive highexposures of UV light (face) and for sites that receive muchlower exposures.

Fig. 4 shows the relationship between R/f and annual expo-




Table 6 At values

CancerBCCsecSexM

FMFTotal'2.022.091.50

1.78CIÃ‰"8

a.m. -8p.m.*1.84

1.911.37

1.6410

a.m.-2p.m.'1.54

1.601.15

1.63Total'2.29

2.371.70

2.02MUT*8

a.m. -8p.m.*2.02

2.101.50

1.78IOa.m.-2

p.m.'1.63

1.691.21

1.43' Calculated using the CIE action spectrum.*Calculated using the MUT action spectrum.

Exposures calculated by integration over 1 yr.*â€¢'Exposures calculated by integration over the time periods 8 a.m.-8 p.m. and 10 a.m.-2 p.m. for each day in a yr.

sure to carcinogenic sunlight, /). Our data are in agreementwith a simple power law:

R/f= const Dp

(6). Thus, to a first approximation Ahâ€”¿�p and can be determinedfrom the slopes of the curves in Fig. 4. The results of a linearregression analysis of the data in Fig. 4 are given in Table 5.We have also calculated the Ah values using the MUT actionspectrum and integrating over different times of the day (8a.m.-8 p.m. or 10 a.m.-2 p.m., respectively) (Table 6).

From Tables 5 and 6 we can conclude the following.1. Ah values obtained by use of the CIE and the MUT action

spectrum, respectively, are similar to within about 10% (Table6).

2. Integration between 10 a.m. and 2 p.m. yields Ah valueswhich are about 25% lower than those obtained by integrationover the whole day. The most likely situation is probably thatthe population gets its sunlight between 8 a.m. and 8 p.m. Inthat case, AI,values are obtained which are only 10% lower thanthose obtained by integration over the whole day.

3. Ab values for SCCs seem to be lower than those for BCCs.This is in contrast to what others have found (9). A possibleexplanation might be that BCCs are less frequently reported tothe Cancer Registry by doctors in the northern part of thecountry. This is not very likely, however, since the ratio of therates of BCCs to that of SCCs is high compared to data inother reports (9, 28) and not dependent on the latitude of theregion (Table 4).

4. Ab values for highly exposed sites of the body (columnsA-C) are not significantly different from those for less exposedsites (column E) (Fig. 4; Table 5). This observation supportsthe power law relationship between incidence and exposure tocarcinogenic sunlight. Data for lower latitudes seem to indicatethat Ah values increase with increasing annual UV exposures(6). Consistent with this, our A,, values are slightly lower thanthose found from epidemiological data at lower latitudes.

Some of the possible error sources in the present investigationhave already been mentioned. Combined they might make upat most 20% of the values given for Ah. An additional errorsource might be that people in the northern part of the countrytend to spend more time indoors than people in southern partsof the country, simply because the weather is colder. A correction for this would lead to lower A/, values. To correct for thiserror one would need detailed data about peoples' habits with

respect to indoor/outdoor life at working hours as well as later.No such information is available.

If the ozone amount (O3) in the atmosphere is reduced by</(Oi), the annual exposure D to carcinogenic sunlight willincrease by dD. The radiation amplification factor is defined by

, dD'

[0,]

Using the CIE and the MUT action spectra, we have foundthat A, equals 1.1 and 0.8, respectively, at latitudes between 60and 70 degrees (29). The overall amplification factor for skincancer induction related to ozone depletion is equal to theproduct of A, and Ah. Using data from columns D in Table 6the total amplification factor for BCCs is 1.6-2.1, dependingon whether the MUT or the CIE action spectrum is used.Correspondingly, the value for SCCs is within the range of 1.3-1.7. Thus, according to the present work, an ozone depletionby 1% will result in an increase in the incidence of BCCs of1.6-2.1% and in the incidence of SCCs of 1.3-1.7% at northernlatitudes of 60-70 degrees.

REFERENCES

1. Karman, J. ( '.. Gardiner. II.. and Shanklin, J. D. Large losses of total ozone

in Antarctica reveal seasonal CIOâ€ž/NOXinteraction. Nature (I mid.). 315:207-210, 1985.

2. Stolarski, R. S. The antarctic ozone hole. Sei. Am., 25Â«:20-26, 1988.3. Lindley, D. CFCs cause part of global ozone decline. Nature (Lond.), 323:

293, 1988.4. Barrett, J. W., Solomon, P. M., de Zafra, R. L., Jaramillo, M., Emmons, L.,

and Parrish, A. Formation of the Antarctic ozone hole by the CIO dimermechanism. Nature (Lond.), 336:455-458, 1988.

5. McKinlay, A. F., and Diffey, B. L. A reference action spectrum for ultravioletinduced erythema in human skin. CIE J., 6: 17-22, 1987.

6. Green, E. E. S., Findley, G. B., Klenk, K, F., Wilson, W. M., and Mo, T.The ultraviolet dose dependence of non-melanoma skin cancer incidence.Photochem. Photobiol., 24: 353-362, 1976.

7. Rundel, R. D. and Nachtwey, D. S. Skin cancer and ultraviolet radiation.Photochem. Photobiol., 28: 345-356, 1987.

8. De Gruijl, F. R., and van der Leun, J. C. A dose-response model for skincancer induction by chronic UV exposure of a human population. J. I hemBiol., 83:487-504, 1980.

9. Rundel, R. D. Promotional effects of ultraviolet radiation on human basaland squamous cell carcinoma. Photochem. Photobiol., 38, 569-575, 1983.

10. Setlow, R. B. The wavelengths in sunlight effective in producing skin cancer:a theoretical analysis. Proc. Nati. Acad. Sci. USA, 71: 3363-3366, 1974.

11. Kerr, R. A. Stratospheric ozone is decreasing. Science (Wash. DC), 239:1489-1491, 1988.

12. Fears, T. R., and Scotto, J. Estimating increases in skin cancer morbiditydue to increases in ultraviolet radiation exposure. Cancer Invest.. /: 119-126, 1983.

13. Pearl, D. K., and Scott, E. L. The anatomical distribution of skin cancer. Int.J. Epidemici., 15: 502-506, 1986.

14. Stamnes, K., Tsay, S., Wiscombe, W., and Jayaweera, K. Numerically stablealgorithm for discrete-ordinate-method radiative transfer in multiple scattering and emitting layered media. Appi. Optics, 32: 2502-2509, 1988.

15. World Meteorological Organization. Global ozone research and monitoringproject. Ref. No. 16, Atmospheric ozone 1985, Vol. 1, pp. 355-358, 1985.

16. Josefson, W. Solar ultraviolet radiation in Sweden. SMHI Reports, Meteorology and Climatology, S-60176 Norrkoping, Sweden, pp. 34-39, 1986.

17. Bruls, W. A. G., and van der Leun, J. C. The use of diffusers in themeasurement of transmission of human epidermal layers. Photochem. Photobiol., 36: 709-713, 1982.

18. Moustacci, E. Aspects of radiation-induced mutagenesis and malignant transformation. In: C. HÃ©lÃ¨ne,M. Charlier, T. Montenay-Garestrier, and G.Laustriat (eds.). Trends in Photobiology, pp. 217-228. New York: PlenumPublishing Corp.. 1986.

19. Enninga, H. C., Groenendijk. R. T. L.. Filon, A. R.. van Zeeland, A. A., andSimons, J.W.I. M. The wavelength dependence of U.V.-induced pyrimidinedimer formation, cell killing and mutation induced in human diploid skinfibroblasts. Carcinogenesis (Lond.), 7: 1829-1836, 1986.

20. Jones, C. A.. Huberman, B., Cunningham, M. L., and Peak, M. J. Mutagenesis and cytotoxicity in human epithelial cells by far- and near-ultravioletradiation: action spectra. Radial Res., 110: 244-254, 1987.

5211




21. Wells, R. L., and Hahn, A. Action spectra for killing and mutation of Chinese Ltd., 1971.hamster cells exposed to mid- and mear-ultraviolet monochromatic light. 26. Scotto, J., Cotton, G., Urbach, F., Berger, D., and Fears, T. BiologicallyMutÃ¢t.Res., 129: 751-758, 1984. effective ultraviolet radiation: surface measurements in the United States,

22. Mitchell, D. L., and Rosenstein, B. S. Use of specific radio-immune assays 1974 to 1985. Science (Wash. DC), 239: 762-763, 1988.to determine action spectrum for the photolysis of (6-4) photoproducts. 27. Lund, C. C., and Browder, N. C. The estimation of area of burns. Surg.Photochem. Photobiol., 45: 781-786, 1987. Gynecol. Obstet., 79: 352-361, 1944.

23. van Weelden, II.. de Gruijl, F. R., and van der I curi. J. C. Tumors induced 28. Scotto, J., Kopf, A. W., and Urbach, F. Non-melanoma skin cancer amongby UVA in mice. Pholochcm. Photobiol., 37: 579, 1983. Caucasians in four areas of the United States. Cancer (Phila.), 34: 1333-

24. Sterenborg, H. J. C. M. Investigations on the action spectrum of tumorigen- 1338, 1974.esis by ultraviolet radiation. Thesis, Chap. 7. Institute of Dermatology, State 29. Moan, J., Dablback, A., Larsen, S., Henriksen, T. and Stamnes, K. OzoneUniversity of Utrecht. The Netherlands, 1987. depletion and its consequences for the fluence of carcinogenic sunlight.

25. Hill, A. B. Principles of Medical Statistics, pp. 204-210. London: The Lancet Cancer Res., 49:4247-4250, 1989.

5212



1989;49:5207-5212. Cancer Res Johan Moan, Arne Dahlback, Thormod Henriksen, et al. Nonmelanoma Skin Cancer at High LatitudesBiological Amplification Factor for Sunlight-induced

Updated version

http://cancerres.aacrjournals.org/content/49/18/5207

Access the most recent version of this article at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/49/18/5207To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



Documents

Biological Amplification Factor for Sunlight-induced ... · Biological Amplification Factor for Sunlight-induced Nonmelanoma Skin Cancer at High Latitudes1 Johan Moan,2 Arne Dahlback,