Exp. Pathol. 1991 ; 41: 169-174 Gustav Fischer Verlag Jena
1) Institute of Toxicology, Bayer AG, Wuppertal, F.R.G. 2) Institute of Experimental Pathology, Hannover Medical School, Hannover, F.R.G.
On the question of a carcinogenic action of cobaltcontaining compounds By D. STEINHOFF and U. MOHR Received: November 28, 1989; Accepted: December 5, 1989
Address for correspondence: Dr. D. STEINHOFF, Institut fUr Toxikologie, Bayer AG, W -5600 Wuppertal I, BRD Key words: Carcinogenesis; Cobalt-containing compounds; carcinogenic effect, cobalt oxide
Summary
After subcutaneous injections of doses of 5 x 2 and 1 x 10 mg cobalt (II) oxide/kg/week (total dose 1,000 mg/kg), respectively 5 out of 10 and 4 out of 10 rats in a lifetime study developed local malignant tumours (controls 0/10). After intraperitoneal injections of 3 x 200 mg cobalt(1I) oxide/kg 14 out of 20 rats developed malignant intraperitoneal tumours (controls 1120). With a Co/AI/Cr spinel the same treatment produced malignant intraperitoneal tumours in only 2 out of 20 rats. After intratracheal instillation of cobalt (II) oxide in single doses of 2 mg/kg (total dose 78 mg/kg) and 10 mg/kg (total dose 390 mg/kg) there were 2 benign pulmonary tumours among 100 rats in the low-dose goup and 2 benign and 4 malignant pulmonary tumours among 100 rats in the high-dose group. The Co/AIlCr spinel was tested intratracheally only in the high dose (total dose 390 mg/kg); among the 100 rats this produced 3 malignant pulmonary tumours. Alternating intratracheal instillation of cobalt (II) oxide (total dose 470 mg/kg) and benzo[a]pyrene (total dose 200 mg/kg) led to 9 malignant pulmonary tumours in 20 rats, whereas benzo[a]pyrene alone caused only 1 malignant pulmonary tumour in 20 rats. The results suggest that cobalt (II) oxide (76.7% Co) has a weakly carcinogenic effect. The Co/AI/Cr spinel investigated is still less active, and even in very sensitive tests (i.p. and i.t. administration) shows no statistically significant carcinogenic effect. This smaller effect may possibly be explained by its lower cobalt content (24.0%) or its much lower solubility (only less than 10% of the solubility of CoO in 0.1 n HCI).
Introduction To date there have been no reports of an increased incidence of tumours after occupational exposure to cobalt. On the other hand, the alloy constituent cobalt is ascribed particular importance in pulmonary fibrosis after chronic inhalation of hard metal dusts (I, 7). Chronic inhalational intake of cobalt dust can lead to diffuse-inflammatory reactions of the bronchial mucosa and chronic respiratory tract disorders (8). The cardiotoxic effects of cobalt are well known, as are the haematological changes it produces (polycythaemia) (6). Cobalt powder, cobalt oxide, and cobalt sulphide showed a distinct local carcinogenic activity in rats after single intramuscular injections of relatively high doses (I x 20- 30 mg/animal) 13
Exp. Pathol. 41 (1991) 4
169
(38 -93 % of the 10- 32 evaluable animals had local malignant tumours). A single intramuscular injection of 10 mg cobalt oxide/animal did not induce any tumours in 50 mice (2-4). By contrast, in a similar orienting test (25 x ~ 10 mg/rat i.m.), cobalt chloride seemed somewhat less active. This dose, which was roughly 10 times higher, induced local tumours in 73% of the rats (8/11) (10). 51 male golden hamsters were exposed 7 h per day, 5 times per week, over their lifetime to a concentration of 10 mg cobalt oxide/m3 respiratory air. This resulted in lung damage, but the survival times were as for the controls. No carcinogenic effect could be detected (13). After intratracheal instillation of cobalt(II, III) oxide (4 mg/animal/week over 30 weeks) in golden hamsters there were reports of bronchial epithelial reactions, but no reports of a carcinogenic effect, although among 50 treated golden hamsters there were 2 pulmonary tumours (5, 9). All in all, only rough orienting studies have been done on the question of a carcinogenic action of cobalt and cobalt compounds. As regards the present results, it should be noted that the injection tests are extremely sensitive to local carcinogenic effects, though on the other hand it is questionable whether the golden hamster is a suitable model for pulmonary tumour induction by inhalation (14). Given this situation it seemed necessary to investigate again, in a sensitive test, the carcinogenic activity of cobalt-containing inorganic compounds after their intake via the respiratory tract. Since intratracheal instillation in rats has shown itself to be a very suitable test method for, among others, chromium-containing compounds (12), cobalt-containing compounds should also be tested by intratracheal instillation. To this end the hitherto best investigated cobalt oxide and a cobalt-containing spinel were included in the investigations.
Materials and Methods The following were tested for a carcinogenic action in rats: cobalt (II) oxide and Col AI/Cr spinel. Cobalt (II) oxide (CoO), a black powder, is a chemically pure substance made by decomposition of cobalt (II) nitrate AR at fairly high temperatures, followed by ignition at 1,000 dc. Cobalt(II) oxide is poorly dispersible in an ultrasonic field. Electron micrographs (TEM) show coarse, angular particles. Fine particulate material measuring approximately 0.1 flm is frequently present in the form of small agglomerates. In some cases these particles are attached to the surface of coarser particles. Because of the lack of differentiation as to whether the substance is made up of individual particles or of agglomerates, an evaluation of the particle size distribution can only be very tentative. Approximately 80% of the particles are in the 5-40 flm range. The spinel, a blue powder, is a chemically pure substance with the empirical formula Co 110.66 (Al 070 Cr IIIo.30h 03.66 , made of a mixture (ignited at 1,250°C) of CoO, AI(OHh, and Cr203; the ignited product was ground. The spinel is used as a pigment in plastics and paints. Electron micrographs (TEM) show coarse, mostly angular particles as well as fine material. The fine fraction (measuring < 0.1 flm) is mostly agglomerated. Here too, because of the irregularity of the particle size spectrum, the evaluation of the particle size distribution cannot be relied upon. Approximately 80% of the particles are in the range of < 1.5 flm. The animals used in the study were SPF-bred lO-week-old Sprague-Dawley rats from MUS Rattus GmbH, Brunnthal. The animals were kept individually under conventional conditions in type II Makrolon® cages, on dust-free wood granulate bedding. The feed was Altromin® 1324standard diet (pellets, Altromin GmbH, Lage, F.R.G.) and was provided ad libitium. The room temperature was 23 ± 2 DC, the air was changed 10 times per h; the relative humidity was between 30 and 70%. Preliminary experiments had shown that in the case of a single intratracheal instillation in Sprague-Dawley rats it was not the toxicity but the administrability that limited the size of the doses used in the chronic investigations. For both compounds the highest dose was set at 1 X 10 mg/kg b. wt. every 14 days. Cobalt (II) oxide was also investigated at a dosage of 2 mg/kg every 14 days. The intratracheal treatment was given over a period of 2 years. The study was carried out over the lifetime of the animals, the treatment period being followed by an observation period. All the animals either died of a natural death or were sacrificed in a clearly moribund state (C0 2), 170
Exp. Pathol. 41 (1991) 4
All the animals were subjected to a thorough autopsy; the macroscopic findings were documented in writing and the fresh weights of the following organs were determined: brain, heart, testes, liver, lungs (+ trachea), spleen, kidneys, adrenals, and ovaries. All tumours, all organs and tissues suspected of having tumours, and the following organs and tissues of all rats were stored for histological investigation in 6 % neutral buffered formalin: adrenals, bladder, brain, epididymis, oesophagus, eyes, femur, heart, intestines, kidneys, larynx, liver, lungs, mammary glands, mesentery, nasopharynx, omentum, ovaries, pancreas, pituitary, prostate, salivary glands, seminal vesicles, skeletal muscle, skin, spinal cord with marrow, spleen, sternum, stomach, submaxillary gland with lymph nodes, testes, thymus (if possible), thyroids, tongue, trachea, and uterus. In the following groups all preserved organs and tissues were subjected to a histological examination: cobalt(Il) oxide high-dose, spinel, and NaCI control. In the cobalt(II) oxide lowdose and untreated control groups, only organs and tissues suspected of having tumours and the respiratory tract were subjected to histological examination. The sections were prepared from paraffin blocks and stained with haematoxylin and eosin. Two groups of Sprague-Dawley rats served as controls, one group untreated and the other treated with physiological saline in the same manner as the study animals. The administration volume was in all cases 1 ml/kg per treatment. For the intratracheal instillation all the actively treated and control groups consisted of 50 male and 50 female Sprague-Dawley rats. The doses were checked several times at random by gravimetric determinations. In addition to being used in the experiment with intratracheal instillation, cobalt (II) oxide and the cobalt spinel were also investigated, for the purposes of orientation, in the intraperitoneal injection test, which is highly sensitive for locally carcinogenic substances (II). Both these substances (with corresponding NaCl controls) were given in doses of 3 x 200 mg/kg b. wt. (at intervals of 2 months in each case). Each group comprised 10 male and 10 female SpragueDawley rats. Cobalt(Il) oxide was also administered by subcutaneous injection to groups of 10 male Sprague-Dawley rats in doses of 5 x 2 mg/kg/week or 1 x 10 mg/kg/week over a period of 2 years. In each case a control group was treated correspondingly with physiological saline. A further group of 20 female Sprague-Dawley rats was treated intratracheally with cobalt(II) oxide (10 mg/kg, once a week; from the 8th treatment 20 mg/kg every 14 days; total of 27 treatments, total dose 470 mg/kg) and benzo [aJ pyrene (20 mg/kg, once a week; from the 8th treatment dose given every 14 days; total of 10 treatments, total dose 200 mg/kg). The period between treatment with cobalt (II) oxide and benzo [aJ pyrene was in each case 4 days. A control group was treated correspondingly with benzo [aJ pyrene and physiological saline. These small-scale orienting studies were also carried out over the lifetime of the animals, i.e. the rats died naturally or were sacrificed in a clearly moribund state. All organs and tissues suspected of having tumours, all tumours in the injection region (s.c. and i.p.), and in the case of the intratracheal instillation of cobalt (II) oxide and benzo [aJ pyrene the entire respiratory tract, were subjected to histological examination. Tables 1 and 2 give an overview of all the carcinogenesis studies carried out by us with cobalt(II) oxide and cobalt-containing spinel and also show the total doses administered in each case.
Results Overall, intratracheal treatment was tolerated relatively well by the animals. However, because of a few isolated treatment-dependent deaths, the interval between the individual doses had to be increased form 2 to 4 weeks from the 19th to the 30th treatment (total doses: table 1). In body weights and the survival times there were no appreciable differences between the intratracheally treated animals and the controls treated correspondingly with physiological saline. In the group receiving the high cobalt(II) oxide dose the relative and absolute lung weights of both sexes were elevated compared to the NaCl controls (p <0.01). The same holds for the spinel, 13*
Exp. Pathol. 41 (1991) 4
171
~
~
-.c
~
-
9-
:;.
0:
"C
"?
><
rr.
tv
-.J
50 50
50 50
Controls-untreated Controls-untreated
NaCl controls NaCl controls M F
M F
M F
F
M F M
Sex
10 mg/kg 10 mg/kg
10 mg/kg
2 mg/kg 2 mg/kg 10 mg/kg
Individual doses
390 mg/kg 390 mg/kg
390 mg/kg
78 mg/kg 78 mg/kg 390 mg/kg
Total doses
No. of animals
20
20
20 10 [0 10 10 20
20
Substances
Cobalt (II) oxide
Col AI/Cr spinel
NaCl controls Cobalt(II) oxide Cobalt(II) oxide N aCl controls NaCI controls Cobalt(II) oxide + Benzo [a] pyrene
Benzo [a] pyrene + NaCI
F
10M; 10F M M M M F
1OM;1OF
10M; IOF
Sex
20mg/kg
5 X 2 mg/kg/week 1 x 10 mg/kg/week 5x Ix 10-20 mg/kg 20mg/kg
200 mg/kg
200 mg/kg
Individual doses
200 mg/kg
470 mg/kg 200 mg/kg
1000 mg/kg 1000 mg/kg
600 mg/kg
600 mg/kg
Total doses
Table 2. Overview of the results of the small-scale orienting carcinogenesis studies in rats.
50 50
50
50 50 50
No. of animals
ColAVCr spinel ColAIICr spinel
Cobalt (II) oxide
Cobait(II) oxide Cobalt (II) oxide Cobalt (II) oxide
Substances
benign squamous epithelial turnoUT bronchioal veolar adenoma bronchioalveolar adenomas bronchioalveolar adenocarcinoma adenocarcinomas (p = 0.02) bronchioalveolar carcinoma
i. t.
i.p . s.c. s.c. s.c. s.c. i. t. i. t.
i.p.
i.p.
Route of admin.
9 pulmonary tumours (8 squamous cell carcinomas, I adenocarcinoma) I pulmonary tumour (squamous cell carcinoma)
14 i.p. (10 mal. histiocytomas, 3 sarcomas, [mal. mesothelioma) 2 i.p. (l mal. histiocytoma, I sarcoma) I i .p. (mal. histiocytoma) 5 s.c. } (mal. histiocytomas 4 s.c. + sarcomas)
Local tumours
I squamous cell carcinoma 2 squamous cell carcinomas
I I 2 I 2 I
Pulmonary tumours (encountered only as individual tumours)
Table 1. Overview of the total doses and the pulmonary tumour frequencies observed after intratracheal instillation of the test substances in rats.
though only with p < 0.05. As regards histologically detectable non-neoplastic changes, one can only instance those in the lungs in the form of bronchio-alveolar proliferations in more than half of the treated rats (N aCI controls 0, cobalt (II) oxide low dose 51, cobalt (II) oxide high dose 70, spinel 61). In comparison to the control groups (untreated and NaCI controls), there was evidence of a carcinogenic activity due to the test substances cobalt(II) oxide and spinel only in the lungs. The frequencies of pulmonary tumours in each group of 100 rats were as follows: untreated controls 0, NaCI controls 0, cobaIt(II) oxide low dose 2 (benign), cobalt(II) oxide high dose 6 (4 malignant, 2 benign) spinel 3 (malignant). All the lung tumours were so small that they were only discovered on histological examination; indeed, the first pulmonary tumour (a squamous cell carcinoma) occurred in a rat that died 617 days after the start of intratracheal instillation of the spinel. Table 1 gives an overview of the total doses, the tumour types, and their distribution among the study groups after intratracheal instillation. Table 2 presents the results of the small-scale orienting carcinogenesis experiments, which show a locally carcinogenic action for cobalt(II) oxide, but not for the spinel.
Discussion Inflammatory reactions of the bronchial mucosa after chronic cobalt dust inhalation have been described in man (8). The pulmonary fibrosis observed after chronic inhalation of hard metal dusts is attributed in particular to the cobalt content of such dusts (1, 7). No evidence of a carcinogenic activity due to cobalt in man has so far come to light. After intramuscular injection of relatively high doses, cobalt powder, cobalt oxide, and cobalt sulphide exert a local carcinogenic effect in rats (2-4). In a similar orienting study cobalt chloride proved to be less effective (10). Injection experiments of this kind are, as is generally known, very sensitive to local carcinogenic effects. In the inhalation experiment, however, it was just as impossible to demonstrate a carcinogenic effect for cobalt oxide in golden hamsters (13) as after intratracheal instillation, although in the latter case [cobalt(I1, III) oxide] 2 pulmonary tumours were encountered among 50 treated animals (5, 9). In respect of the latter negative experiments it should be noted that the respiratory tract of golden hamsters seems to be less sensitive to the local carcinogenic action of metals than the respiratory tract of rats (14). From other examples, too, it can be deduced that in this respect the rat can be regarded as a suitable model for the situation in man (12). In our own orienting experiments with subcutaneous and intraperitoneal injections cobalt(II) oxide showed a not very pronounced local carcinogenic activity [very sensitive methods, high doses, 40-70% of the animals with local malignant tumours (controls 0-5%)]. After intratracheal instillation of cobalt (II) oxide there was a dose-dependent carcinogenic effect in the lungs. In the low-dose group (total dose 78 mg/kg) 2 benign pulmonary tumours were observed among a total of 100 rats; in the high-dose group (total dose 390 mg/kg) there were 4 malignant and 2 benign pulmonary tumours among 100 rats (controls 0/100). This finding is underlined by the experiment with alternating intratracheal instillation of benzo [a] pyrene and cobalt(II) oxide. 9 out of 20 rats treated in this way showed malignant pulmonary tumours; this compares with only 1 out of 20 rats treated with benzo [a] pyrene alone. The spinel was less effective than cobalt(II) oxide. Under the same intratracheal treatment (total dose 390 mg/kg) 3 malignant pulmonary tumours developed with the spinel and 4 malignant and 2 benign pulmonary tumours developed with cobalt(II) oxide. The difference in effect between the spinel and cobalt(lI) oxide was even clearer in the sensitive intraperitoneal injection test. Unter the same dosage (3 x 200 mg/kg) malignant local tumours occurred in only 2 out of 20 rats receiving the spinel, but in 14 out of 20 rats receiving cobalt(II) oxide (NaCl controls 1120). If the orienting injection studies indicate a local carcinogenic action due to cobalt(II) oxide, the more informative intratracheal instillation experiment shows that this substance is only weakly carcinogenic. The spinel (24.0% Co) is even less effective than cobalt(II) oxide (76.7 % Co), even though the number of lung tumours are not statistically different from control. This lower incidence is possibly related to the lower cobalt content of the spinel or of the much lower solubility. Exp. Pathol. 41 (1991) 4
173
References I. BECH, A. 0., KIPLING, M. D., H~ATHER, J. C.: Hard metal disease. Br. J. Ind. Med. 1962; 19: 239-252 2. GILMAN , J. P. W. , RUCKERBAUER , G. M.: Metal carcinogenesis. I. Observations on the carcinogenicity of a refinery dust, cobalt oxide, and colloidal thorium dioxide. Cancer Res . 1962a; 22: 152-157. 3. - Metal carcinogenesis. II. A study on the carcinogenic activity of cobalt, copper, iron , and nickel compounds. Cancer Res. 1962b; 22: 158-165. 4. HEATH, J. C.: The production of malignant tumours by cobalt in the rat. Br. 1. Cancer 1956; 10: 668-673. 5. NIOSH: Criteria for controlling occupational exposure to cobalt. DHHS (NIOSH) Publication No. 82-107, 1981. 6. PAYNE, L. R.: The hazards of cobalt. J. Soc. Occupat. Med. 1977; 27: 20-25. 7. REINL, W., SCHNELLBACHER, F., RAHM, G.: Lungenfibrosen und entziindliche Lungenerkrankungen nach Einwirkung von Kobaltkontaktmasse. Zbl. Arbeitsmed. 1979; U: 318-325. 8. RoTO, P.: Asthma symptoms of chronic bronchitis and ventilatory capacity among cobalt and zinc production workers. Scand. J. Work Environ. & Health 1980; 6 (Suppl.): 1-49. 9. SAFFIOTTI, U. , CEfIS, F., KOLB, L. H., GROTE , M. I.: Intratracheal injection of particulate carcinogens into hamster lungs. Proc. Am. Ass. Cancer Res. 1963; 4: 59. 10. SHABAAN, A. A., MARKS, Y. , LANCASTER, M. ~., DUFEU, G. N.: Fibrosarcomas induced by cobalt chloride (CoCI 2 ) in rats. Lab. Anim. 1977 ; 11: 43-46. II. STEINHOFF, D.: Possibilities for an adequate stepwise carcinogenicity testing procedure. In: NORPOTH, K. H., GARNER, R. C. (eds.): Short-term test systems for detecting carcinogens. Springer-Verlag, Berlin-Heide\bergNew York 1980; pp. 127-138. 12. - GAD, SH. C., HATFIELD, G. K., MOHR, U.: Carcinogenicity study with sodium dichromate in rats. Exp. Pat hoI. 1986; 30: 129-141. 13. WEHNER, A. P., BUSCH, R. H., OLSON, R. J., CRAIG, D. K.: Chronic inhalation of cobalt oxide and cigarette smoke by hamsters. Am. Ind. Hyg. Assoc . J. 1977; 38: 338-346. 14. - STUART, B. 0., SANDERS, C. L.: Inhalation studies with Syrian golden hamsters. Prog. Exp. Tumor Res. 1979; 24: 177-198.
174
Exp. Pathol. 41 (1991) 4