(CANCER RESEARCH 51, 979-984, February 1, 1991] Development of Skin Tumors in Hairless Mice after Discontinuation of Ultraviolet Irradiation1 Frank R. de GruijI2 and Jan C. van der Leun Institute of Dermatology IAZV, State University of Utrecht, PB. 85500, NL-3508 GA Utrecht, The Netherlands ABSTRACT The development of skin tumors (mainly squamous cell carcinomas) in hairless Skh-HRl mice after discontinuation of a course of daily UV irradiations (wavelengths, 280-370 nm) is compared to that when the daily irradiations are continued. Under conditions of continued daily exposures 50% of 22 animals contracted tumors with diameters of at least 4 mm in 135 days. With exposures stopped after 35 or 19 days (2 groups with 24 and 23 mice) this time interval increased to 280 and 645 days, respectively; the rate at which multiple tumors developed on the mice was correspondingly lower. A mathematical model, derived from a larger experiment (223 mice) with different levels of chronic UV expo sure, successfully predicts the tumor development after discontinuation of UV exposure. This model is similar to those used in risk assessments for skin cancers in human populations, e.g., in relation to stratospheric ozone depletion, sunbeds, etc. The model separates UV-driven processes from purely time-dependent processes. These stochastic processes, de scribed by Weibull statistics, form stages in the tumorigenesis. This interpretation of the data indicates that a late, UV-independent stage occurs between the smallest observable tumors and larger ones with diameters of over 4 mm. This could be a simple growth stage, but histopathology suggests that it may also entail a transition from actinic keratosis to squamous cell carcinoma. INTRODUCTION The induction of skin cancer by UV radiation is an important public health concern. Epidemiológica! data form the basis of risk assessments (1-3), but these data have severe restrictions. It is, for example, impossible to determine from epidemiológica! data which part of the solar spectrum causes skin cancer. One needs a priori knowledge to interpret the data in terms of exposure to UV radiation. This knowledge is most directly derived from animal experiments: e.g., Roffo (4) and Blum's group (5), who established that wavelengths below 315 nm (so- called UVB) are the most effective in sunlight. Recently, more detailed spectral data on UV tumorigenesis have become avail able (6, 7). The combination of epidemiológica! and experimental data has proved to be useful in risk assessments for a depletion of stratospheric ozone (2, 8-10). These assessments mainly dealt with stationary situations with various levels of annual UV exposure. How changes in annual UV load during a man's life affect his ultimate risk has been a matter of conjecture. Based on so-called multihit, multistage, or multievent theories, one can factorize the risk into a purely dose-related component and a purely time-related component (11-13). Risk assessments along these lines have been applied to situations with changes in annual UV loads during a lifetime (14, 15). The experiment presented in this paper provides information on the tenability of such risk assessments. Received 6/21/90; accepted 11/20/90. 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 in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1Financed by the Dutch Cancer Foundation, Grant UUKC D77-6 and by a fellowship of the Royal Dutch Academy of Sciences (KNAW) for F. R. G. 2 To whom requests for reprints should be addressed. The present experiment was designed to investigate how discontinuation of a course of daily exposures would affect the ultimate tumor formation in comparison to the situation with continued exposures. A larger experiment on tumorigenesis by chronic daily exposure (16) served as a frame of reference from which model predictions could be made for the present experi ment. MATERIALS AND METHODS Mice. Albino hairless mice, Skh-HRl (not athymic), entered the experiment at 6-8 weeks of age; both males and females were used. The animals came from a random-bred colony in our animal laboratory; the breeding stock originated from the Skin and Cancer Hospital in Philadelphia, PA. These animals do not spontaneously develop squa mous cell carcinomas of the skin; an animal may occasionally get a papilloma late in life (16). Maintenance. The animals were housed separately in cages that were subdivided into 12 compartments. The mice had free access to mouse chow and tap water. Irradiation. The dorsal surfaces of the mice were irradiated daily with Westinghouse FS40 T12 sunIamps (wavelengths in the UV between 280 and 370 nm with a maximum output around 310 nm) that were mounted above the cages. These lamps were switched on automatically from 12 noon to 1:15 p.m.; the irradiance was adjustable. The daily surface exposure applied to the top of the cages was 1.9 kJ/m2 (meas ured with a Kipp El l thermopile). At the dorsal level of the animals (inside the cage) the exposure was reduced by about 20%. The room was illuminated from 7:00 a.m. to 7:00 p.m. by yellow fluorescent lamps (Philips, TL40W16), which emitted no UV radiation. The Experiment. In 2 groups, of 23 and 24 mice, the daily irradiations were stopped after 19 and 35 days, respectively. In one group of 22 mice the irradiations were continued. This latter group was also the group exposed to the highest daily dose in the larger experiment on chronic UV exposure (13,16) referred to in "Introduction." The present experiment ran in parallel with this larger experiment. Mice with a large total tumor mass (0.5 to 1 cm3) and/or with ill health were sacrificed. A group was taken out of the experiment when about 30% of the mice were lost, or when all of the surviving mice had contracted tumors with diameters of over 4 mm. Animal Observation. Initially, the animals under chronic, daily ex posure were checked for tumors weekly. The animals in which irradia tions were discontinued after 35 days were checked every 2 weeks until a first tumor was spotted. Then this group was also checked every week, and the group in which irradiations were stopped after day 19 was checked every 4 weeks. After spotting a first tumor in the latter group, it was checked every 2 weeks. Tumors less than 1 mm in diameter could be spotted by a trained observer. If it was the first tumor on an animal, the observation needed to be confirmed in the next checkup or else the observation was disregarded. This was done to reduce the error that a small transient lesion could introduce in recording the number of tumor-bearing ani mals. For more details on the experimental setup the reader is referred to our earlier publications (13, 16). Definitions and Data Analysis. The tumor development in a group in the course of time is measured in two ways: (a) prevalence, i.e., the percentage of tumor-bearing mice, and (b) the yield, i.e., the average number of tumors per survivor. 979 on July 15, 2015. © 1991 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from