Document 5bzoq7j96o9be0rvQwznD80Ne

INTERIM REPORT API RESEARCH PROJECT MC-I U November 1, 1956 THE KETTERING LABORATORY) in the Department of Preventive Medicine and Industrial Health College of Medicine UNIVERSITY OF CINCINNATI, CINCINNATI, OHIO SC-API-2807 INTERIM REPORT - API RESEARCH PROJECT MC-1 November 1, 1956 Certain factors, which, were instrumental in bringing about a systematic investigation of potential hazards associated with carcinogenic materials encountered in various refining operations, were reviewed in some detail in an annual report dated October 1, 1955* This program has encompassed products from some 16 different cracking and lube refining processes. The carcinogenic potencies of the 120 samples, as gauged by the effects induced by their contact with the skin of mice, have varied over a wide range. As much as ten-fold differences have been observed between materials from similar processes, which were at least superficially comparable. These variations have demonstrated that the history of the sample, including the crude source and the operating conditions tinder which it was processed, were all-important in determining its final activity. The need for rapid methods for estimating potency became apparent. Many empirical correlations between physical or chemical properties and potencies have been tested. Some of these have found limited applicability to certain classes of oils but no generally applicable analytical technique has thus far been developed. By 1951 It was evident that the complexity of the problem was such that a 1 detailed understanding of the relationships between the chemical composition of the different oils and their relative potencies would be required before it would be possible to set up the desired analytical procedures to replace the lengthy biological test. At the same time, the rapid changes which had taken place in this dynamic industry, since the initiation of the project in 19M>> demonstrated that this type of fundamental understanding of the problem was essential for the recognition and control of the poten tial hazards associated with the handling of carcinogenic materials. . In the annual report for 1955 it was shown how the program of fractionation and biological testing of various components of selected carcinogenic oils had led to a useful hypothesis regarding the contributions of various classes of hydrocarbons to the effective potency of a given material. For the purpose of orienting the collection of experimental data these contributions may be expressed by the following equation: PMC = "a lC* + + 0>- The summation of C's represented the effective potency con tributed by typical polycyclic carcinogens, such as benzo pyrene. The contribution of the non-carcinogenic fraction of the oil, including the paraffins, cycloparaffins, and most of the aromatic constituents is represented by the multiplication factor, M . The considerable importance of these non-carcinogenic components is shown by the fact that M may reach values as high as 10. To determine in a syste matic manner the relative contributions of carcinogens and non-carcinogens to potency, a series of oils have been fractionated during 1958 by certain physical and chemical procedures in order to produce concentrates of the poly cyclic carcinogens as free as possible of other components of the oils. By chromatographic procedures all of the car cinogens may be collected into a concentrate suitable for biologic test to provide values for the summation (Cx + Cs + C3). A further, subdivision of the carcinogens has been possible through the process of extraction of the oils by maleic anhydride and subsequent chromatographic concentration of the components of the extract containing four or more rings. In Table 1 are shown the results of a series of biological tests on such fractions which are underway or have been completed during the current year. It will be noted that only about 15 per cent of the effective potency of the heavy catalytic cycle gas oil, API-89, was contributed by polycyclic carcinogens (API-89-3). In contrast, the entire potency of the thermal tar, API-65, can be ascribed to such polycyclic components. In both instances 25 to 30 per cent of the basic potency due to polycyclic aromatics could be ascribed to compounds ex tracted by maleic anhydride, such as derivatives of benz(a)anthracene. The contribution of the 5-ringed class, benzopyrenes, can be estimated from the content of this class as determined by direct chemical analysis. The third -k- major class of carcinogens, those of lowest molecular weight, are being investigated by a mor9 direct approach, which will be discussed in a subsequent section. Measurement of the multiplication factor, Ma,,, cannot be made by any direct method. This contribution due to accelerators may be estimated from the ratio of the effective potency of the oil to the base potency contributed by the polycyclic fraction. Thus, for oils API-89 and API-65, the estimated values of M would be 7 and 1, respectively. The potential difficulties which preclude a direct measurement of M. arise from the fact that a significant a part of the non-carcinogens of the oils have physical properties which are quite similar to those of the carcino gens. Hence, separation of these aromatic non-carcinogens from the carcinogens requires highly efficient application of the techniques of distillation and chromatography. Such separations are being carried out with a few selected materials, but require much more time than the simpler separations which are being applied to quite a variety of oils. It is possible, however, to obtain some useful information by direct tests of the effective accelerating properties of the non-carcinogenic components which are easily separated from the whole oil by chromatography. Thus, the fractions containing essentially all of the paraffins and cycloparaffins and monocycloaromatics, as well as most of the naphthalene derivatives, are under 1 test as the vehicles for 0.05 per cent of methylcholanthrene. In addition, tests are being started with solutions of methylcholanthrene in a medium devised to simulate as closely as possible the combination of the non-carcinogenic components of the original oils. These media are being blended from the non-carcinogenic chromatographic fractions together with certain liquid aromatic hydrocarbons to replace the polycyclics separated in the chromatography. These liquids, tetrahydronaphthalene and diphenylmethane, satisfy the theoretical requirement for introduction into the general medium of the aromatic system, with the potentially significant influence of it's -rf-orbitals on the physical characteristics of the liquid phase, without the complications of insolubility, on the one hand, or high volatility, on the other, encountered in initial efforts to employ phenanthrene or benzene for this purpose. It will be recalled from discussions in previous reports that, if the concentration of accelerators in an oil falls in the range 20 to 50 per cent (typical of eatalytically cracked residua), some indication of the magnitude of their influence may be estimated by tests of the whole oils at dosage levels varying from 20 mg. to 100 mg. per application. One is, of course, limited by the toxic nature of an oil in determining the maximum effect of such accelerators. Hence, a direct test using appropriate fractions obtained in the chemical laboratory may serve this purpose better. Since the results of testa on simple mixtures of relatively pure hydrocarbons normally lead to a more rapid understanding of the relative contribution of different types of components, a series of parallel tests are being carried out on solutions of benzopyrene in blended media, including a standard accelerator such as n-dodecane, together with cycloparaffins of high molecular weight (from white mineral oils), on the one hand, or with more volatile hydrocarbons of relatively low molecular weight, on the other, fhe results are shown in Table 2 and in Figure 1. It will be noted that, at the $0 mg. level of dosage, dilution of dodeoane with $0 per cent of either of the non-accelerating solvents, decahydronaphthalene or beta-methylnaphthalene, did not reduce the accelerating activity of the n-paraffin On the other hand, dilution with only 2$ per cent of white mineral oil (viscosity, 3^0 S.S.tJ, at 100F.) reduced the activity considerably. These results suggest that some major component of the white oils has the property of counteracting accelerating influences to a significant extent. It has been shown previously that some of these white oils contain long-chain accelerators. The falltire to obtain from them any significant adduct with urea indicates that these are not n-paraffins, but presumably isoparaffins and long-chain alkylcyclohexane s or cyclo pentanes, the ring being near the end of the chain (1-cyclohexyldeeane is an active accelerator, whereas 7-cyclohexyltridecane is almost inactive). It has been found that, to a useful extent, the monocyeloparaffIns can be extracted from the white oils by their formation of adducts with thiourea, the polycycloparaffins concentrating in the non-adduct* Such fractionations are being carried out on API-80 to obtain material with which to determine whether the polycycloparaffins are responsible for the inhibitory effect observed in the blends discussed above. Since it la suspected that the physiological as well as the physical properties of the liquid media are also influenced signif icantly by the presence of non-carcinogenic aromatics, suitable blends will also be tested containing such components together with the polycycloparaffins and the accelerators. The preceding paragraphs Indicate the necessity for consideration of the non-carcinogenic medium as a whole in the approach to the final goal of this phase of the project, a quantitative measure of the magnitude of the factor Mad under any given set of conditions of exposure* In addition, a continuing effort has been made to define the general characteristics of the components of this medium which are primarily responsible for its catalytic (accelerating) influence on carcinogenesis. The earlier results of tests of a variety of relatively pure hydro carbons for their ability in this respect have been summarized in a paper entitled "Carcinogenesis of the Skin. II. The Catalytic Effects of Aliphatic and Related Hydrocarbons", -8- A fairly general relationship has been noted between the molecular length (in the extended configuration) of the solvent and its ability to accelerate the Induction of tumors by polycyclic carcinogens. As shown in Pig. 2, activity seems to reach a maximum level at a length equivalent to that of the n-paraffins containing 12 to lip earbon atoms. Further, the maximum effect for alkyl derivatives of benzene and naphthalene seems to be somewhat greater than that for saturated hydrocarbons. The fairly active accelerating ability of triiso propylbenzene seems to be an exception to this relationship of molecular length and activity. As discussed previously, the initial sample used showed in its mass spectrum evidence of an impurity having a lower molecular weight. A pure sample of 1,3,5-triisopropylbenzene, kindly supplied by Dr. E. W. Adams, has now bean tested. It has shown the same accelerating activity as the original sample (Ma~ 4)* The shape of the molecule has also been considered as a potentially important characteristic in addition to its length. The saturate, 7-cyclohexyltrids cane, demon strated only a relatively low activity (M_ ** 2.5), even though its molecular length should be optimum (see structures below). In contrast, however, the corresponding 7-phenyl and 7-n-hexyl derivatives have shown a significantly more active accelerating influence (these samples were kindly supplied by API Research Project 42 at Pennsylvania State College). CH3 (CB2 ) ^CH (CH2) ^CH3 6 CH3 (CH2) 5CH (012)5^3 (CH2p)'5 7-cyclohexyltridecane Ma ~ 2.5 7-phenyltridecane Ma=* 9 7-n-hexyltridecane H INVESTIGATION DIRECTED TOWARDS THE IDENTIFICATION OF THE CARCINOGENS OF FRACTIONS, DISTILLING IN THE RANGE 720 TO 760F., FROM CRACKED AND STRAIGHT RUN DISTILLATES OR RESIDUA These are the carcinogens responsible for the major part of the potency of heavy catalytic cycle gas oils, coker gas oils, and light lubricating stocks which contain little or no benzopyrene (occasionally the cracked gas oils have high enough end points to contain signif icant amounts of benzopyrene, but this seems to' be the exception). In a concentrated attack on this problem of the nature of the carcinogen of lowest molecular weight in petroleum oils, the polycyclic aromatics obtained by chroma tography from a highly active residual product of Fluid Catalytic cracking, API-8, was fractionated through an 80-plate distillation column in 195jE> The biological tests on narrow fractions distilling in the range, 60 to 760, were completed In 196. The results are shown in Table 3. It has been demonstrated that all of these fractions contain accelerating components, although the activity in this respect apparently drops to a relatively low level by the time the concentration of carcinogen (in fraction 8-26-k or -1) has increased to a point capable of inducing tumors in 9 per cent of the C3H mice. 10 - The carcinogenic fractions, 8-26-1, -m, and -n, (730 to 760F.) were selected for further fractionation. In order to reduce the complexity of the system the three fractions were first extracted with aqueous acid and base and then with maleic anhydride. The concentration of the major components of the maleic anhydride adduct from 8-26-n was estimated from its mass spectrum (ionization potential, 7 volts) as follows: n n2n-l8 1? 0.7$ 18 5.1 19 6.1+ 20 3.1 21 3- a 1 CnH2n-20 o.a% 11.1 11.8 5.6 1.8 Suggested structures: %H2n-22 0.3% 3.8 ' 6.1+ 2.3 0.8 cn^2n-2l+ _ 13.3* 2.2 0.8 0.3 Cycloalkylanthracenes Alkylbenzofluorenes Benz(&)anthracene On the basis of the mas3 spectrum at high voltage, the cyclopentano substitution is considered to be the most likely structure for the CnH2n_20 hydrocarbons. A six-membered 1 >rr 11 cycloalkyl ring would be expected (on the basis of the spectrum of tetralin) to yield a strong fragment peak at m/e of 204 by loss of -CHb-CHb-. The intensity of the 204 ion in the spectrum of the mixture was negligible. The components extracted by maleic anhydride apparently contributed about 1/6 of the potency of the original distillate, 8-26-n. About l/ of the accelerated potency of 0.10 shown by this fraction in the test 8-26-n-l can be ascribed to benzanthracene. It seems likely that the major contribution came from the 4-**inged compounds having one partially saturated ring, but this cannot be stated with certainty until additional experimental evidence is available. The combined material from 8-26-1, -m, and -n, which had been freed from the components soluble in aqueous acid and base and from the anthracenic compounds removed by maleic anhydride, was next subjected to chromatographic fractionation on activated alumina. Mass and absorption spectrometry were used to determine the distribution of the various classes of hydrocarbons through the chromatographic fractions (see Table 4). The biological test on the fraction which contained almost all of the 4-ringed chrysene and triphenylene (8-26-14) mixed with the more strongly adsorbed portion of the benzofluorenes, indicated the presence of only a trace of the carcinogens of the parent distillate. The test on the most strongly adsorbed fraction (8-26-1^), that which includes nitrogen compounds such as derivatives 12 of carbazole, has shown mainly that this fraction is exceed ingly toxic to the mice. No tumors have been produced . in 34- weeks. One of the intermediate fractions (8-26-11), that containing cyclopentanophenanthrenes, fluoranthenes, and benzofluorenes, but essentially free of compounds having ij. fully aromatic rings, has shown some carcinogenic activity, but was not responsible for more than about 10 per cent of the potency of the original distillates. The foregoing tests on fractions derived from 8-26-1, -m, and -n were accelerated by the use of dodecane as a solvent. This permitted a relatively rapid elimination of a possible 4-ringecl aromatic carcinogen from consideration. Tests of the remaining fractions are being carried out at their original concentration (or at a doubled concentration) in benzene as a solvent in order to obtain a better measure of their relative potencies, Pmcj and for a more direct comparison with the potencies of the parent distillates', which were tested as %0 per cent solutions in benzene. These new tests, 8-26-7, -8, -9, -10, and -12, have been under way about four months. The results of these tests should indicate the particular classes of polycyclic hydro carbons responsible for the major part of the potency in this distillation range, 720 to 760P. According to the physical and chemical behavior of the classes indicated, a further concentration of the carcinogens will be carried out until their certain identification has been accomplished. Analytical techniques for the direct determination of this class in complex oils may then be developed. - 13 - API EXPERI MENT NUMBER DESCRIPTION FRACTION DISTIL LING BENZO GRAVITY <75of. PYRENE (API) (per cent) (per cent) RELATIVE CARCINO GENIC POTENCY, *MC (at 100 mg.) 89 Catalytically 2?.0 cracked sidestream (F.C.C. heavy cycle gas oil) 100 0,001 fill 33 Coker gas oil 16.8 60 0.008 Q.4 65 Thermal tar 6.6 43.8 0.04 O.lix 120 Straight run dis 19.5 tillate {naphthenic ) 52 5.003 0.07 (50 mg.) 105 Straight run dis tillate (waxy) 28.9 72.9 115 Straight run distil - 23.2 -^50 late (naphthenic) 0.0003 .002 0.11 O.O3 - 13 TABLE I RELATIVE CONTRIBUTIONS OP CARCINOGENS AND NON-CARCINOGENS TO THE EFFECTIVE POTENCY OF VARIOUS REFINERY STREAMS CHROMATOGRAPHIC FRACTIONATION FRACTION EXTRACTED BY MALEIC ANHYDRIDE Saturates Thru Alkyl- Poly BASIC naphthalenes cyclic CONTRI ESTI Yields Contri Aromatics BUTION MATED bution API *MC and OF ACCELER of EXPERI MENT with Non-hydro POLY- ATING 0.05^ carbons CYCLICS FACTOR 2-3 4+ 4+ Ring Fractlor NUMBER Yield MC (yield) TO Pjkj1 Ma rings rings ` ^c1 89-1 89-2 89-3 89-4 89-6 8l2^ 81 ^0.11 33-3 33-4 33-5 67 inc. 65-1 65-2 65-3 120-1 40 Inc. 120-2 120-3 105-1 . 67 814. inc. inc. 19^ 33 60 33 0.04 1 inc. 0.12 1 inc. 1.64^ 0.2% 0.00 0.012 (2.1) 2.2 inc. (0.9) 1.1 O.03 (1.2) 0.4 A- O.OOlj. 115-1 91 inc. 1. Determined by biological test on solution of polycyclic fraction in benzene at appropriate concentration, three 20 mg. appli cations each week. 2. A direct test of this fraction has produced 1 benign papilloma in 36 weeks. * TABLE 2 INFLUENCE OF VARIOUS DILUENTS ON THE ACCELERATING ACTIVITY OF n-DODECANE API EXPERI MENT NUMBER ACCELERATOR DILUENT CONCEN TRATION SCHEDULE OF OF BENZOPYRENE APPLICATIONS ($ by RELATIVE (no./week - weight) SPREAD1 mg.) MEAN TIME OF . APPEARANCE OF TUMORS (weeks 663 100$ Dodecane None 0.04 ' 2.9 2-30 23.7 RELATIVE CARCINO GENIC POTENCY, P^ rMC 0.14 666 100$ Dodecane None 0.08 2.9 2-30 20.4 0,18 293 100$ Dodecane None 0.20 2.9 3-30 14.7 0.21 638 None 100$ Decahydronaphthalene 0.08 1.8 e s t, 2-30 0) 43 a^0.03 639 30$ Dodecane 30$ Decahydronaphthalene 0.08 2.4 est . 2-30 20.0 0.19 660 30$ Dodecane 30$ Beta-methylnaphthalene 0.08 2-30 19.1 0.20 613 None 83$ Beta-methylnaphthalene, 13$ Benzene 0.20 2-30 30.6X = 0.10 614 73$ Dodecane 23$ white oil (340 ssu/1000) 0.20 2.63 est . 3-30 20.4 0.12 297 30$ Dodecane 30$ white oil 0.20 2.4 est 3-30 20.2 0.12 296 None 100$ white oil 0.20 1.8 3-30 27.9 0.07 1. Based upon relative spread of 1.0 for solutions of benzopyrene in benzene. 2. Experience indicates an experimental variation of 13$. 3. In these tests, essentially all of the C3H mice develop squamous cell papillomas and ' carcinomas if they survive. ', - - 15 - TABLE 3 TUMOR-IKDTJCING ACTIVITY OP PRODUCTS OP FRACTIONAL DISTILLATION OP THE POLYCYCLIC AROMATICS OP A CATALYTICALLY CRACKED OIL API EXPERI MENT NUMBER 8-26-a^ FRACTION DISTILLED 28.5 - 30.8$ t5OfpTMA fTrRTi BOILING RANGE (in 0 P. at 760 mm.) 641 _ 646P. INCIDENCE OP TUMORS {$ - weeks) 0$- 80 RELATIVE CARCINO GENIC POTENCY, PMC - 8-26-b 30.8 - 32.0 646 - 651 0 - 80 - 8-26-c 32 _ 34.1 651 - 656 0 - 80 - 8-26-d 8-26-e 34.1 - 36 36 - 38 656 - 662 662 *N> 668 0 _ 80 0 - 80 -- - 8-26-f 38 - 40 668 - 6?4 7 - 80 - 8-26-g 40 - 43.9 674 - 686 28 - 80 - 8-26-h 43-9 - 46 686 - 693 12 _ 80 - 8-26-i 46 - 50.3 693 - 709 36 - 80 ''v'O.Ol 8-26-j 8-26-k*1 2 8-26-1 8-26-m1 50.3 - 52.5 52.5 - 54*8 54.8 - 57.3 57.3 - 60.4 709 - 717 717 - 727 727 - 736 736 - 747 71 - 73 83 - 28 90 - 58 87 - 45 0.03.01 S0.211 a*^90-^,,*0052 0.291 8-26-n 60,4 - 62.4 747 - 759 100 - 46' 8-26-2 8-26-4 2 Solution of Benzopyrene in fraction of 8-26-g, -h, and -I; plus 20$ benzene Solution of Benzopyrene in fraction2 of 8-26-1, ~m, and -n; plus 30$ benzene Acceler ating -JLagjLQE-- 5 2 1. All fractions tested as 50$ solutions in benzene except "B-26-k" and -m, which were accelerated with $0% dodecylbenzene. 2. Ghromatographie fractions including saturates through naphth alenes, 2?$ of 8-26-g, -h, and -1 ; 2- 3$- of 8~ -2'6-1,'-m, and -n, 1 TABLE 4 RELATIVE POTENCIES OP CHROMATOGRAPHIC FRACTIONS OP CARCINOGENIC FRACTIONS OP LOWEST MOLECULAR WEIGHT PROM CATALYTICALLY CRACKED RESIDUUM, API-8 CHROMATOGRAPHY BIOLOGICAL TESTS PERCENT ELUTED APPARENT COMPOSITION API EXPEBI- MENT. NUMBER PERCENT ELUTED. CONCENTRATION . IN. SOLUTION TESTED SOLVENT SCHEDULE OP.APPLI- INCIDENCE CATIONS . OF. (no./week TUMORS - mg.) (/-weeks) 0 - 22/ Saturates 21 - 27 Naphthalenes 25 - 90 Phenanthrenes 50 - 87 Pyrenes 80 - 91 Fluoranthenes 84 - 94 Benzofluorenes 91 - 94 Chrysene, etc. 94 - loo Non-hydrocarbons 8-26-7 0-26-8 8-26-9 8-26-10 8-26-11 8-26-12 8-26-14 8-26-15 26 - 64/ 64 - 80 80 - 85 85 - 88 88 - 91 91 - 93 93 ' 94 94 - 100 33. 4/ 32.6 9.6 6.2 4.1 3.8 0.9 10.4 Benzene Benzene Benzene Benzene Dodecane Benzene Dodecane Dodecane 3 - 15 0/ - 19 3-15 18 - 19 3-15 0-19 3 - 15 o - 19 3-50 70 - 30 3-15 0-19 3-50 io - 35 3-50 0-35 MEAN TIME OP APPEARANCE OP PAPILLOMAS (w e e ks) FIGURE 1 - INFLUENCE OF CYCLOPARAFFINIC DILUENTS ON THE ACCELERATING ACTIVITY OF n-DODECANE -HJ FIGURE 2 - MOLECULAR LENGTH OF SOLVENT vs. ACCELERATING ACTIVITY ..r - 19 - Prom The Kettering Laboratory In the Department of Preventive Medicine and Industrial Health, College of Medicine, University of Cincinnati, Cincinnati, Ohio. Investigative Team: Bernard H. Braun, Ph. D. Prank P. Cleveland, M. D. Ralph T. Denham, B. S. Mary Jane Graf, B, S. Francis P. Heyroth, Ph. D., M. D. A. Wesley Horton, Ph. D. John J. Phair, M. D Pred Shaffer, M. S. Klaus L. Stammer, M. D. Russell Tye, M. S. Waldo J. Younker, M. S. Patricia Clapsaddle Lenore Hull Irvin Rapien Efl'ie West Report A. Wesley Horton, Ph. D, Date: November 1, 19^6 Director J J