Increased frequency of 6-thioguanine-resistant lymphocytes in peripheral blood of workers employed in cyclophosphamide production

Mutation Research, 243 (1990) 101-107

101

Elsevier MUTLET 0284

Increased frequency of 6-thioguanine-resistant lymphocytes in peripheral blood of workers employed in cyclophosphamide production E. Hiittner, U. Mergner, R. Braun and J. Sch6neich 1 Central Institute of Genetics and Research in Cultivated Plants, Academy of Sciences of the G.D.R., 4325 Gatersleben (G.D.R.) and 1Institute of Biology, School of Medicine, Martin-Luther- University Halle, 4020 Halle/Saale (G.D.R.)

(Accepted 17 August 1989)

Keywords." Cyclophosphamide,occupational exposure to; HPRT-deficientvariant frequency; Peripheral lymphocytes

Summary The frequency of 6-thioguanine-resistant peripheral blood lymphocytes has been determined by autoradiography in a control population and a population of cyclophosphamide-exposed individuals. The mean variant frequency in a non-exposed population was found to be 2.76 + 1.48 × 10- 5. Subpopulations of smokers and non-smokers revealed statistically significantt differences in the variant frequencies, i.e. 3.52 _ 1.55 × 10 -5 and 2.07 _+ 1.05 × 10 -5 respectively. In 20 out of a total of 23 individuals employed in cyclophosphamide synthesis and manufacturing, the variant frequency of 6-thioguanine-resistant lymphocytes was found to be higher than the m a x i m u m individual frequency found in the control population. The mean variant frequency in the cyclophosphamide-exposed population was 13.64 + 13.56 × 10 -5, a statistically significant increase as compared to the mean control frequency. There was no correlation between variant frequency and duration of employment suggesting that this test reflects the actual exposure and not a cumulative effect.

Somatic cell mutation which occurs in vivo can be monitored by study of the frequencies of 6-thioguanine-resistant (TG r) T-lymphocytes in peripheral blood (PBL) either by a direct cloning procedure (Strauss, 1982; Morley et al., 1983) or by autoradiography (Strauss and Albertini, 1977; Albertini, 1982). Resistance to T G denotes cells lacking the enzyme hypoxanthine-guanine phosCorrespondence: Dr. Edith Hiittner, Central Institute of Genetics and Research in Cultivated Plants, Academy of Sciences, 4325 Gatersleben (G.D.R.).

phoribosyl transferase (HPRT). This cellular defect is presumably a consequence of inactivation or loss of the hprt locus which is present normally at the end of the long arm of chromosome X (Chinault and Caskey, 1984). The mutant character of most 6-TG r PBL in human blood has been demonstrated by c D N A hybridization, DNAsequencing and enzyme analysis (Bradley et al:, 1987; Liber et al., 1987; Nicklas et al., 1987; Albertini et al., 1989). The procedure for detection of mutant or variant TG r PBL has been used to study the genetic effects of cytostatic and immunosup-

0165-7992/90/$ 03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)

102 pressive chemotherapy (Dempsey et al., 1985); Ammenheuser et al., 1988), radiotherapy (Dempsey et al., 1985; Messing and Bradley, 1985), diagnostic -y-radiation (Seifert et al., 1987) and radiation exposure in atomic bomb survivors (Hakoda et al., 1988a, b). We used the indirect autoradiographic procedure for detection of 6-TG r PBL in workers exposed to the cytostatic drug cyclophosphamide during production and processing.

Material and methods

Sample acquisition and populations studied Heparinized blood was obtained by venipuncture and the mononuclear cells containing T lymphocytes were recovered by Ficoll-Visotrast density gradient centrifugation (Boyum, 1968) of a mixture of human blood and Hanks' salt solution (1:1 v/v). A population of industrial workers employed in cyclophosphamide production and processing (E) was selected. This population derived from a pharmaceutical plant in the G.D.R. and included workers in the manufacturing process of raw cyclophosphamide (3), production of cyclophosphamide drug and packing (11), quality control (7) and technicians (2) and was described by the following criteria: age, sex, period of employment, tobacco smoking, alcohol drinking and medical therapy. A non-exposed control group (C) was selected from the same location and matched in order to have appropriate conditions for comparison and characterization of cyclophosphamide effects. Both groups comprised 23 persons with a mean age of 37.4 years (E) and 28.8 years (C), respectively. The period of employment in cyclophosphamide production ranged between 1 and 20 years.

PBL short-time culture and autoradiographic assay The mononuclear cells were removed after centrifugation and washed 3 times in cell culture medium (Eagle's minimal essential medium, EMEM). Finally, the cell sediment was resuspended in prewarmed RPMI 1640 medium supplemented with 15% (v/v) fetal calf serum and ad-

ditives (Albertini, 1985) at a density of 106 cells/ml. The viability of these cell suspensions was in the range of 95%. The cell suspension from each person was divided into 4 subcultures prepared in cell culture tubes containing 1.8 ml lymphocyte suspension supplemented with phytohemagglutinin ( P H A ) (0.025 ml P H A per ml; Wellcome, Beckenham, England). 2 subcultures received 0.2 ml TG at a final concentration of 5 × 10- 5 M and 2 subcultures were prepared with RPMI 1640 and P H A alone. All cultures were incubated in an humidified 5% CO2/95% air atmosphere at 37°C for 27.5 h, labeled with 1.5 #Ci/ml [3H]TdR and incubated for an additional 16-h period. Cultures were terminated by centrifugation and adding 2 ml 0.075 M KC1. This 15-min hypotonic treatment is followed by fixation with 2 ml chilled (0°C) methanol/glacial acetic acid (3:1 v/v) 3 times. Finally, the nuclei were resuspended in 0.4 ml fresh fixative. The nuclei in each suspension were counted in a hemocytometer and 0.05 ml suspension from cultures made with TG were added to coverslips affixed to glass slides (4 coverslips/culture), dried and autoradiographed by standard procedures. Suspensions derived from TABLE 1 VARIANTFREQUENCIESIN EXPERIMENTSWITH DIFFERENT CONCENTRATIONSOF 6-THIOGUANINE Individualsa

Conc. 6-TG (moles/l)

Vf

1

5.0 × 10-5 2.5 × 10-5 1.25 × 10-5

2.08 × 10-5 2.22 × 10-S 3.81 x 10-5

5.0 x 10-5 2.5 × 10-5 1.25 × 10-5

2.34 × 10-5 5.96 x 10-5 9.65 × 10-5

5.0 × 10-5 2.5 × 10-5 1.25 × 10-5

2.95 x 10-5 3.20 × 10-5 7.39 × 10-5

5.0 × 10-5 2.5 × 10-5 1.25 × 10-5

5.84 × 10-5 7.53 × 10-5 8.23 × 10-5

aAll individuals are non-smokers and aged between 22 and 30 years.

103

cultures without T G were plated in a volume of 0.025 ml/coverslip and processed as above. For autoradiography photoemulsion O R W O K6, photodeveloper O R W O A71 and fixative O R W O A300 were used (ORWO, Wolfen, G.D.R.). The slides were exposed for 8 days at - 2 0 ° C . The counterstaining of nuclei was done in 2% acetoorcein. All slides were coded. Variants are characterized by heavy labeling, while normal nuclei do not show silver grains. Slide scoring and variant frequencies The principles given by Strauss and Albertini (1979) were followed. Labeling indices (L/) were

determined f r o m culture with and without the addition of 6-thioguanine: LI-TG

= number of labeled nuclei per 5000 5000

L I + T G = number of labeled nuclei on all slides total number of nuclei on all slides

The variant frequency (Vf) is defined as: V f - L I + TG L I - TG

TABLE 2 VARIANT FREQUENCIES

FOUND IN A POPULATION

Individual

Age

Sex

OF CHEMICALLY

Tobacco

Number of

smoke

cells a n a l y z e d

NON-EXPOSED

Variants

LI- TG

INDIVIDUALS

V f × 10- 5

x 10 ~ 1

f

23

-

390

3

0.37

2.08

2

f

24

+

338

3

0.38

2.34

3

f

35

+

332

5

0.51

2.95

4

f

35

+

276

5

0.31

5.84

5

f

33

+

248

3

0.54

2.24

6

f

29

+

328

5

0.42

3.63

7

f

30

+

190

6

0.52

6.07

8 9

f m

30 25

+ -

356 200

5 3

0.60 0.46

2.28 3.26

10

f

29

+

372

2

0.37

1.45

11

f

31

+

252

2

0.28

2.84

12

f

16

+

166

3

0.36

5.02

13

f

33

+

160

2

0.31

4.03

14

f

22

-

322

3

0.31

3.01

15

f

24

-

196

2

0.47

2.17

16

f

32

-

284

1

0.42

0.84

17

f

33

-

204

2

0.33

2.97

18

f

34

-

292

3

0.38

2.70

19

f

20

-

384

1

0.43

0.61

20 21

f m

29 40

-

298 312

1 2

0.42 0.54

0.80 1.19

22

m

20

-

228

1

0.27

1.63

23

m

37

-

260

4

0.42

3.66

23

m/f

28.8

+ / -

0.41

2 . 7 6 + 1.48

12

m/f

28.2

-

0.44

2 . 0 7 +_ 1.05

11

f

29.5

+

0.41

3.52 + 1.55 a

aSignificantly different from V f i n non-smoking individuals at p<0.05.

104

The Vfof each person from the exposed group and control group was determined individually. For comparison between groups of exposed and nonexposed individuals, ~ and standard deviation have been calculated. Differences between groups were proven by M a n n - W h i t n e y test.

were cultivated with different amounts of 6-TG (Table 1). There is a reverse correlation between concentrations of T G and the observed variant frequencies. We used 5.0 × 10- 5 M 6-thioguanine for selection in all experiments done in this study. Variant frequencies for 23 non-exposed individuals were studied by this technique and a mean of 2.76 (SD 1.48) × 10- 5 was found. The range for all 23 single data was between 0.61 × 10 -5 and 6.07 × 10 -5 (Table 2). While no age- or sex-related differences of variant frequencies could be found in the total

Results

Variant frequencies in non-exposed individuals For selection of appropriate 6-TG concentration blood samples from 4 non-exposed individuals

TABLE 3 V A R I A N T F R E Q U E N C I E S FOUND .IN A P O P U L A T I O N PHOSPHAMIDE PRODUCTION AND PROCESSING

OF

INDUSTRIAL

WO R K ER S

EMPLOYED

IN

CYCLO-

Individual

Sex

Age

Tobacco smoke

Number of Variants cells analyzed × 10 3

Years of employment

LI-TG

Vf × 10-5

Working place a

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

f f f f f f m f f m f f f f m f f f

31 31 37 36 41 53 34 40 31 50 40 48 39 23 30 58 52 43

+ + + + + + + + + +

268 202 62 98 156 180 170 244 172 246 156 221 157 183 216 294 102 222

18 11 5 10 7 6 14 15 1 9 6 63 5 3 8 8 6 8

12 2 1 6 1 11 3 11 6 20 12 14 13 1 1 14 19 14

0.59 0.44 0.46 0.41 0.37 0.47 0.53 0.34 0.44 0.45 0.45 0.41 0.29 0.35 0.42 0.26 0.38 0.18

11.38 12.38 17.53 24.89 12.13 7.09 15.54 18.08 1.32 8.13 8.55 69.53 10.98 4.68 8.82 10.47 15.48 20.02

Q Q Q Q v P T P V S Q TS P V TS P P Q

19 20 21 22 23

m m f f f

27 30 29 29 29

+ + + -

154 283 263 118 152

1 4 5 2 3

4 4 10 13 5

0.39 0.21 0.30 0.10 0.39

1.67 6.73 6.34 16.95 5.06

S S v P Q

23

m/f

37.4

+/-

8.5

0.37

13.64 + 13.56

10

m/f

37.5

-

8.1

0.34

11.82 +

13

m/f

37.3

+

8.9

0.40

15.04 + 17.75

4.73

aWorking place: Q, quality control; V, vial-filling apparatus; P, processing of pharmaceutical drug; T, tablet mass production; S, synthesis; TS, technical seryice.

105

population, an increased mean variant frequency is obvious for cigarette smokers. This frequency (3.52 × 10-5) is significantly higher as compared to that in non-smoking individuals (2.07 × 10-5). All individual Vfs higher than 4 × 10-5 occur in that particular subgroup.

mean variant frequencies found in the respective control subgroups revealed significance o f p < 0.01 in all cases. Thus, it can be concluded that occupational cyclophosphamide exposure causes genotoxic effects which can be measured as T G r variants in peripheral blood lymphocytes.

Variant frequencies in cyclophosphamide-exposed individuals

Discussion

The variant frequencies of T G ~ lymphocytes in peripheral blood of individuals employed in cyclophosphamide production and processing were found to be increased (Table 3). The individual variability ranged from 1.32 × 10- 5 up to 69.53 × 10 -5 with a mean of 13.64 × 10 -5 . For 20 individuals in this group the actual Vf exceeded the highest value found in the non-exposed group and only 3 persons were found to have frequencies comparable with the control group. The mean variant frequency was increased by 5-fold as compared to the control group and this difference is statistically significant (p < 0.001). There is no correlation between time of employment (duration of exposure) and the actual variant frequency. According to these data it can be expected that the actual variant frequency is predominantly a consequence of exposure just before blood sampling. It is important to point out that the mutant nature of the variants observed cannot be confirmed by this technique. The clear-cut effect of cyclophosphamide exposure underlines the usefulness of this test as a monitor for genotoxic action regardless of the origin of the variants induced. While individuals exposed during chemical synthesis and preparation o f drug vials have only moderately increased variant frequencies, others employed in tablet mass production, packing, quality control and technical service showed considerably higher frequencies of T G r lymphocytes. Although the mean variant frequency for cyclophosphamide-exposed smokers is increased as compared with the mean variant frequency for the exposed but not smoking subpopulation, this difference is not statistically significant (p>0.2). A comparison of the mean variant frequencies in smoking and non-smoking subpopulations exposed to cyclophosphamide with

The autoradiographic assay for T G r lymphocytes arising in vivo is a rapid and sensitive test for genotoxins and recommended for human specific-locus monitoring (Albertini, 1985). It is generally accepted that cryoconservation of PBL is useful and results in the elimination of phenocopies (Albertini, 1985; Albertini et al., 1988). Without such cryoelimination the variant frequencies in normal PBL of adults were found to be in the range o f 5.5 × 10 -5 (Lange and Pranter, 1982) and 1.1 × 10 - 4 (Albertini, 1980), while cryoconservation will decrease this frequency to values between 1.9 X 10 - 6 and 8.7 × 10 - 6 (Ammenheuser et al., 1988; Albertini et al., 1988). Although a cryoconservation step was omitted in our protocol and the selective concentration of T G was reduced to 25°70 of that used by Albertini et al. (1988), we found a low variant frequency of about 2 × 10- 5 This frequency might include phenocopies too. The low but statistically significant increase in the mean variant frequency of T G r lymphocytes in smokers demonstrates quite well the sensitivity of our protocol. In a more extensive study using the T-cellcloning approach Albertini et al. (1988) found a 2.5-fold increase in the variant frequency for smokers as compared with non-smokers. After elimination of one subject with extraordinary high variant frequencies, there was only a 1.5-fold increase in the mean Vfof smokers as compared with non-smokers, exactly the same as the increase we found in our study. We used this test for the monitoring of genotoxic effects due to occupational exposure to cyclophosphamide just before technical reconstruction of the production facility. T L C analysis of air samples revealed a significant amount of this drug in the environment of some working places. Cyclophosphamide is a well

106 known mutagenic and carcinogenic compound (IARC, 1981; M o h n et al., 1976) and was shown to exert genotoxic effects not only in cancer patients under chemotherapy but also in laboratory staff and nurses handling this drug as well as other cytostatics. The occupationally exposed individuals in this report were not involved in any medical therapy at the time of blood sampling, so that the effect of cyclophosphamide can be studied without interference from other cytostatics. Our data demonstrate the induction of 6-TG r variants in PBL due to occupational exposure to cyclophosphamide. This conclusion was drawn after tobacco smoking, drug use and other environmental factors had been ruled out. These results are in good concordance with those published by Ammenheuser et al. (1988) for multiple sclerosis patients undergoing cyclophosphamide therapy with a monthly dose of 750 m g / m 2. Furthermore, our finding that the individual variant frequency is not dependent on cumulative exposure due to employment time, but describes the actual exposure, is in agreement with data presented by Ammenheuser et al. (1988). The high variability of the variant frequencies found in our exposed group might reflect differences in individual exposure and sensitivity as was also found in a population of cancer patients undergoing multiple chemotherapy (Dempsey et al., 1985) or radiotherapy (Seifert et al., 1987). Individual differences in pharmacokinetic behavior of alkylating metabolites of cyclophosphamide in man (Torkelson et al., 1974) m a y contribute as well. Occupational handling of anticancer drugs in hospitals has been shown to be associated with genotoxic effects as demonstrated by cytogenetic surveillance methods and mutagenicity studies of excreta (Kolmodin-Hedman et al., 1983; N o r p p a et al., 1980; Sorsa et al., 1982; Pohlova et al., 1986; Falck et al., 1979). The same effects can be demonstrated in our population of exposed workers, where both urinary excretion of mutagenic compounds and frequency of c h r o m o s o m a l aberrations in blood lymphocytes is significantly higher as compared to the control population (Grummt et al., in preparation).

In a recent study of Sorsa et al. (1988) it could be demonstrated that high safety standards including protective clothing, gloves, respirators, and, in special circumstances, gas-protective suits, will yield negative data for chromosomal damage and SCE in peripheral blood lymphocytes and absence of mutagenic urine in workers employed in cyclophosphamide production despite significant air concentrations of this compound. The plant under study is being reconstructed to provide a safety area for processing of the drug without direct exposure to the working environment. This reconstruction includes the use o f safety cabins with the air exhausted to the outside of the working room. Data shown by Albertini et al. (1988) for the induction of T G r variants and by Sorsa et al. (1988) for the cytogenetic and mutagenic effects demonstrate that with maximal personal protection, use of safety instructions and a high standard of technology, the safe handling of cytostatics is possible both in hospitals and in industry. This study will be repeated after the start of cyclophosphamide production in the new facilities.

References Albertini, R.J. (1980) Drug-resistant lymphocytesin man as indicators of somatic cell mutation, Teratogen. Carcinogen. Mutagen., 1, 25-48. Albertini, R.J. (1982) Studies with T-lymphocytes:an approach to human mutagenicity monitoring, in: B.A. Bridges, B.E. Butterworth and I.B. Weinstein (Eds.), Banbury Report 13, Indicators of Genotoxic Exposure, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp. 393-412. Albertini, R.J. (1985) Somatic gene mutations in vivo as indicated by the 6-thioguanine-resistant T-lymphocytes in human blood, Mutation Res., 150, 411-422. Albertini, R.J., L.M. Sullivan, J.K. Berman, C.J. Greene, J.A. Stewart, J.M. Silveira and J.P. O'Neill (1988) Mutagenicity monitoring in humans by autoradiographic assay for mutant T lymphocytes, Mutation Res., 204, 481-492. Albertini, R.J., I.N. Gennett, B. Lambert, W.G. Thilly and H. Vrieling (1989) Meeting report, mutation at the hprt locus, Mutation Res., 216, 65-88. Ammenheuser, M.M., J.B. Ward Jr., E.B. Whorton Jr., J.M. Killian and M.S. Legator (1988) Elevated frequencies of 6-thioguanine-resistant lymphocytes in multiple sclerosis patients treated with cyclophosphamide: a prospective study, Mutation Res., 204, 509-520.

107 Boyum, A. (1968) Separation of leukocytes from blood and bone marrow, Scand. J. Clin. Invest., 21 (Suppl. 97), 51-76. Bradley, W.E.C., J.L.P. Garean, A.M. Seifert and K. Messing (1987) Molecular characterization of 15 rearrangements among 90 human in vivo somatic mutants shows that deletions predominate, Mol. Cell. Biol., 7,956-960. Chinault, A.C., and C.T. Caskey (1984) The hypoxanthine phosphoribosyltransferase gene: a model for the study of mutations in mammalian cells, in: W.E. Cohn and K. Moldave (Eds.), Progr. Nucleic Acid Res. Mol. Biol., 31, 295-313. Dempsey, J.L., R.S. Seshadri and A.A. Morley (1985) Increased mutation frequency following treatment with cancer chemotherapy, Cancer Res., 45, 2873-2877. Falck, K., P. Gr6hn, M. Sorsa, H. Vainio, E. Heinonen and L.R. Holsti (1979) Mutagenicity in urine of nurses handling cytostatic drugs, Lancet, i, 1250-1251. Hakoda, M., M. Akiyama, S. Kyoizumi, K. Kobuke, A.A. Awa and M. Yamakido (1988a) Measurement of in vivo HGPRTdeficient mutant cell frequency using a modified method for cloning human peripheral blood T-lymphocytes, Mutation Res., 197, 161-169. Hakoda, M., M. Akiyama, S. Kyoizumi, A.A. Awa, M. Yamakido and M. Otake (1988b) Increased somatic cell mutant frequency in atomic bomb survivors, Mutation Res., 201, 39-48. 1ARC (1981) Monographs on the Evaluation of Carcinogenic Risk of Chemicals to Humans, Vol. 26, Some Antineoplastic and Immunosuppressive Agents, International Agency for Research on Cancer, Lyon, p. 441. Kolmodin-Hedman, B., P. Hartwig, M. Sorsa and K. Falck (1983) Occupational handling of cytostatic drugs, Arch. Toxicol., 54, 25-33. Lange, B.J., and J.E. Pranter (1982) The emergence of 6-thioguanine resistant lymphocytes in pediatric cancer patients, Mutation Res., 94, 487-499. Liber, H.L., K.M. Call and J.B. Little (1987) Molecular and biochemical analysis of spontaneous and X-ray-induced mutants in human lymphoblastoid cells, Mutation Res., 178, 143-153. Messing, K., and W.E.C. Bradley (1985) In vivo mutant frequency rises among breast cancer patients after exposure to high doses of 3,-radiation, Mutation Res., 152, 107-112. Mohn, G., and J. Ellenberger (1976) Genetic effects of cyclophosphamide, ifosfamide and trofosfamide, Mutation Res., 32, 331-360. Morley, A.A., K.J. Trainor and R. Seshadri (1983) Cloning of human lymphocytes using limiting dilution, Exp. Hematol.,

11,418-424.

Nicklas, J.A., T.C. Hunter, L.M. Sullivan, J.K. Berman, J.P. O'Neill and R.J. Albertini (1987) Molecular analyses of in vivo hprt mutations in human T-lymphocytes, I. Studies of low frequency spontaneous mutants by Southern blots, Mutagenesis, 2, 341-347. Norppa, H., M. Sorsa, H. Vainio, P. Gr6hn, E. Heinonen, L. Holsti and E. Nordman (1980) Increased sister chromatid exchange frequencies in lymphocytes of nurses handling cytostatic drugs, Scand. J. Work Environ. Health, 67, 229-301. Pohlova, H., M. Cerna and P. R6ssner (1986) Chromosomal aberrations, SCE and urine mutagenicity in workers occupationally exposed to cytostatic drugs, Mutation Res., 174, 213-217. Seifert, A.M., W.E.C. Bradley and K. Messing (1987) Exposure of nuclear medicine patients to ionizing radiation is associated with rises in H P R T - mutant frequency in peripheral T-lymphocytes, Mutation Res., 191, 57-63. Sorsa, M., H. Norppa and H. Vainio (1982) Induction of sister chromatid exchanges among nurses handling cytostatic drugs, in: B.A. Bridges, B.E. Butterworth and I.B. Weinstein (Eds.), Banbury Report 13, Indicators of Genotoxic Exposure, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, p. 341-350. Sorsa, M., L. Pyy, S. Salomaa, L. Nylund and J.W. Yager (1988) Biological and environmental monitoring of occupational exposure to cyclophosphamide in industry and hospitals, Mutation Res., 204, 465-479. Strauss, G.H.S. (1982) Direct mutagenicity testing: the development of a clonal assay to detect and quantitate mutant lymphocytes arising in vivo, in: B.A. Bridges, B.E. Butterworth and I.B. Weinstein (Eds.), Banbury Report 13, Indicators of Genotoxic Exposure, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp. 423-441. Strauss, G.H.S., and R.J. Albertini (1977) 6-Thioguanine resistant lymphocytes in human blood, in: D. Sott, B.A. Bridges and F.H. Sobels (Eds.), Progress in Genetic Toxicology, Elsevier, Amsterdam, pp. 327-334. Strauss, G.H.S., and R.J. Albertini (1979) Enumeration of 6-thioguanine-resistant peripheral blood lymphocytes in man as a potential test for somatic cell mutations arising in vivo, Mutation Res., 61, 353-379. Torkelson, A.R., J.A.C. Budde and J.H. Weikel (1974) The metabolic fate of cyclophosphamide, Drug Metab. Rev., 3, 131-165. Communicated by B. Lambert