Senator HART. Our next witness, as the Surgeon General has already indicated, is Dr. Jacqueline Verrett.

STATEMENT OF DR. JACQUELINE VERRETT, FOOD AND DRUG ADMINISTRATION, DEPARTMENT OF HEW

Dr. VERRETT. Thank you, Mr. Chairman, for this opportunity to discuss our investigations of the relationships between chlorinated phenoxy herbicides, chlorinated dibenzo-p-dioxins, and the chick edema factors.

Chick edema disease was first recognized in 1957, when large numbers of broiler flocks in the United States suffered what appeared to be an epidemic disease. The affected birds appeared droopy, with ruffled feathers, and had difficulty breathing. In many flocks, more than 50 percent of the birds died as a result of the disease. Of the millions of birds affected, those autopsied consistently displayed hydropericardium (accumulation of fluid in the pericardial sac), accumulation of fluid in the abdominal cavity, subcutaneous edema, and additionally liver and kidney damage.

In 1958 the investigations of a number of laboratories indicated that the causal agent was contained in fats, and specifically in the unsaponifiable fraction of fats in the commercial poultry rations. In laying hens the toxic fat caused a rapid drop in egg production. Pullets receiving toxic fat during the full growing period did not come into production, and mortality was very high. Hydropericardium, the most common lesion found in young birds, was not found in birds of laying age.

The chick edema factor found in the toxic fat in the 1957 outbreak was presumed to have arisen as a byproduct of industrial production of stearic and oleic acids, since the unsaponifiable materials from this process were the components of fat in the poultry ration. Subsequently, the toxic substance was found to be present in several different types of fats. It was demonstrated to be present in samples of commercially produced oleic acids and triolein, in acidulated vegetable oils, and in inedible animal tallows. The demonstration of the presence of the chick edema factor in commercial fats led to the ruling by the Food and Drug Administration in 1961 that higher fatty acids intended for food additive use must be free of the chick edema factors. The presence of the factor was to be ascertained by a chick bioassay based on the volume of pericardial fluid in birds fed the fat under investigation.

Beginning in 1958, fat that had first been proved to be toxic to chicks was used by various investigators in experiments with other species, and was demonstrated also to produce deleterious effects in rats, mice, turkeys, pigeons, guinea pigs, swine, dogs, and monkeys. Early investigations of feeding toxic fats to rats indicated that they are more resistant than chicks in short term feedings, but when fed in sufficient dosage, extracts of the toxic fat produced definite deleterious effects as shown by growth depression, enlarged and fatty livers, marked involution of the thymus, and enlarged adrenals.

Guinea pigs fed 219 percent toxic fat stopped growing at 6 weeks, and death losses occurred at 8 weeks. At a level of 4 percent the

weight losses occurred after 3 weeks, and death at 4 weeks. The pathology observed was congestion of the lungs and mottled livers. Dogs fed 10 percent toxic fat in their rations lost hair on their backs and shoulders (alopecia), and there was poor reproduction and lactation performance. Whelped pups were either dead or weak, and the mothers seemed to have an insufficient milk supply. Pups removed before weaning and fed a normal ration showed an immediate and dramatic increase in growth. Other litters maintained on the toxic fat ration post weaning demonstrated inferior growth performance.

Monkeys have demonstrated considerable sensitivity to toxic fat materials. In one study nine monkeys received a toxic triolein at a level of 25 percent in their diets. One monkey died at 1 month, and four died at 3 months. At the 3-month period, corn oil was substituted for the toxic triolein, but the other four monkeys died from 3 weeks to 5 weeks later, in spite of this substitution. Of the nine monkeys fed the toxic triolein, eight were autopsied and showed signs of jaundice, pancreatic atrophy, and fibrosis, hemosiderosis, fatty liver with necrosis, bile duct proliferation, and gross hemorrhage in the intestinal tract. No such pathology was seen in the control monkeys in this study.

A second study with 36 monkeys given a toxic fat at levels from 0.125 to 10 percent of their diet, demonstrated an inverse relationship between the concentration of the toxic fat in the diet and their mean survival time. Those given the highest level (10 percent) had a mean survival time of only 91 days, while those given the lowest level (0.125 percent) had a mean survival time of 445 days. It has been estimated that the highest level provided approximately 728 micrograms total chick edema factors, while the lowest level diet provided approximately 100 micrograms total intake. The toxic fat was lethal at all levels studied, and the animals were sacrificed when possible just before death. During the last 30 days of life, all monkeys developed alopecia, generalized subcutaneous edema, accumulation of fluid in the abdominal and thoracic cavities, and hydropericardium. There were decreases in red and white blood cell counts, total serum protein values, and altered serum-protein ratios. There was also cardiac dilatation and myocardial hypertrophy and edema. Finally, the experimental monkeys had reduced hematopoiesis and spermatogenesis, degeneration of the blood vessels, focal necrosis. of the liver and gastric ulcers.

Limited experimentation with mice, pigeons, and turkeys, indicated that toxic fat in the diet led to reduction in growth without hydropericardium or accumulation of abdominal fluid. Similarly, swine, fed toxic fat at a level of 9 percent of their ration, showed poor weight gain, but one pig sacrificed 6 weeks after the start of the study showed no gross or microscopic lesions attributable to the ration.

One important finding in the studies with chickens, is the apparent storage of the chick edema factors in chick tissues. The unsaponifiable fraction of carcasses (exclusive of intestines, head, and feet) of chickens fed the toxic fat was very potent in producing hydropericardium in other birds when incorporated in their rations. Other investigations of the distribution of the chick edema factors in the

chick tissues indicated significant levels in bone, heart, intestine, kidney, liver, and skin. The liver contained more than 80 percent of the total detected. A similar determination of the distribution in rats indicated the presence of chick edema factors only in liver and in the feces.

During the years that the previously described toxicity investigations were taking place, the toxic fats were undergoing intensive chemical analyses to concentrate, purify, and finally determine the nature of the compounds responsible for chick edema disease. At all steps of these procedures, the path of the toxic material was confirmed by assay in young chicks. This proved to be a time-consuming and difficult job because of the complexity of the fatty materials. A major breakthrough in this effort came when it was found that a highly purified crystalline material possessing the properties of chick edema factor contained chlorine. This indicated that it was not a natural component of the fat in which it occurred.

Work in several laboratories obtained similar results, and examination of the purified material by a variety of analytical techniques suggested that chick edema factors could be highly substituted (chlorinated) derivatives of naphthalene, biphenyl, anthracene, or even structures common to the chlorinated pesticides of the DDT family. These latter compounds were ruled out when tested in the chick feeding assay, but some derivatives of the former classes of compounds were tested and found in some instances to be toxic, and indeed produce similar lesions to those observed with authentic toxic fat. However, none of these compounds demonstrated the high order of toxicity, or the complete chick edema syndrome when so tested. Finally, by means of single crystal X-ray crystallography, it was demonstrated that a pure compound isolated from a toxic fat was a hexachloro-dibenzo-p-dioxin. This structure was verified by infrared, ultraviolet, and mass spectrometry data. Final confirmation came when this particular compound was synthesized and found to produce the same lesions in chicks as the compound isolated from the toxic fat.

The finding that a chlorinated dibenzo-p-dioxin was a chick edema factor explained why different investigators had isolated materials similar in their capacity to elicit chick edema disease, but yet in their purest forms, had slightly different chemical properties. The large number of isomers possible (more than 60) in this family ranging from mono- to octa-chloro-dibenzo-p-dioxins illustrates the complexity of the problem. It then became a problem of determining whether some or all of these compounds are in fact chick edema factors, and what their relative capacities in this regard might be.

The chlorinated dibenzo-p-dioxin structures have been known in organic chemistry many years, and became particularly noteworthy, when in manufacturing processes with chlorophenols, their formation as byproducts posed serious occupational hazards. The most potent in this regard seems to have been the symmetrical tetrachlorop-dioxin which was formed in the manufacture of 2,4,5-trichlorophenol. These chlorinated compounds were found to cause a serious and persistent disease referred to as chloracne. This disease was first described in 1899. Associations of this disease with chlorinated dienzo-p-dioxins were made by the Germans, who had several out

breaks of this disease in their factories. There have also been similar occurrences in the Netherlands and in this country, in factories manufacturing chlorophenol compounds. It should also be pointed out, that other compounds, such as the chlorinated naphthalenes, anthracenes, biphenyls and dibenzofurans, are known to be acnegenic, but as in the case of the toxic response in chicks, these materials are less potent than the chlorinated dibenzo-p-dioxins. In the case of the chloracne associated with dioxin, the human symptomatology extends to other mucous membrane irritation, porphyria cutanea tarda, hirsutism, hyperpigmentation, increased skin fragility, severe damage to the internal organs, particularly hepatotoxicity, and central nervous system disorders, as indicated by neuromuscular symptoms and psychologic alterations, and other systemic symptoms. Most of these occupational exposures in Germany occurred in the 1950's, and followup examination of these affected workers in recent years indicate that the recovery period is lengthy, with many workers still having demonstrable adverse effects from prior exposure. Similar observations have been made on exposed workers in the United States of America. The tetrachlorodibenzo-p-dioxin was demonstrated to produce the chloracne in humans after the application of only 20 micrograms. The rabbit ear is especially sensitive, with concentration of 0.001 percent to 0.005 percent producing severe reactions after local application. This assay using the rabbit ear is apparently used as an indicator in some plants of the content of this particular dioxin in the manufacturing process. Hence, the serious health significance of these compounds for humans has, inadvertently, been clearly documented.

Research in Germany and Japan indicated that the magnitude of this problem was indeed large, since the formation of the chlorinated di-benzo-p-dioxins would be facilitated in the saponification procedures used in various processes involving chlorophenols. A further complication is that a given chlorophenol preparation is generally contaminated with other isomers, increasing the possibility of formation of a wide spectrum of chlorinated di-benzo-p-dioxins beyond those to be expected from the predominant component. Evidence that this does occur will be discussed shortly in connection with the chicken embryo studies of these materials.

During the time the previously described investigations of the chick edema factors were underway, many of which were carried out by FDA investigators, methodology was developed for detecting the chick edema factors using sensitive gas-liquid-chromatographic (GLC) techniques. It became apparent that authentic toxic fats consistently gave peaks with specific retention times, and these peaks were used as an indication of chick edema factor in a suspect sample. Confirmation of this was obtained using the chick feeding assay. In the light of recent knowledge of the chlorodioxins as chick edema factors, it has been possible to establish that the materials being detected were hexa-, hepta-, and octa-chlorodibenzo-p-dioxins. Although toxic fat samples did indeed contain peaks corresponding to dioxins of lesser chlorine content, i.e., di-, tri-, tetra-, penta-, these are not detectable with this particular analytical procedure because their particular peaks are obscured by other components,

including pesticide residues present in the samples. Other GLC procedures are being developed to detect these latter dioxins.

In the early 1960's the chicken embryo was being used in toxicological evaluations of a wide variety of materials. It was hoped to develop a rapid and sensitive screening system to pinpoint compounds with significant toxic and teratogenic effects for further study. In view of the demonstrated sensitivity of the chicken to chick edema factors, the chicken embryo was used to assay toxic fat samples, and found to present the same syndrome as observed in the chick feeding assay. A high mortality was observed with toxic fat extracts, and additionally, hydropericardium, generalized and massive edema, eye, beak, and leg defects, and necrotic livers were apparent on gross observation. No microscopic studies have been conducted on embryos or hatched chicks in these investigations.

In parallel with other investigations, the chicken embryo was used to test the toxicity of the chick edema factors isolated from toxic fats. It was also found that the chlorinated biphenyls, naphthalenes, anthracenes, and other compounds did indeed elicit a toxic response, and in some instances, the chick edema syndrome was present. But in no case were any of these materials as potent as the toxic components isolated from toxic fats, and were generally less potent by a few orders of magnitude.

After the identification in 1966 of a hexachloro dibenzo-p-dioxin as a chick edema factor, studies were initiated in which various isomers of chlorinated dibenzo-p-dioxin were prepared and tested in the embryos. Although the investigation was not extensive or complete, it illustrated that isomers prepared by pyrolyzing selected chlorophenols did give chloro-dibenzo-p-dioxins with GLC retention times duplicating those in the authentic toxic fats, and likewise, produced the chick edema syndrome in the treated embryos. It was also apparent from this study that the various isomers (that is, those with different chlorine content, and those with identical chlorine content, but with chlorine atoms positioned differently on the molecule) varied in their toxicity, although in all cases only microgram or less quantities were required to elicit the toxic response. It is not possible to give exact figures for the toxicities obtained in this study, since most of the individual dioxins were contaminated with traces of others. Nevertheless, it was apparent that the symmetrical chlorodioxin prepared from 2,4,5-trichlorophenol (2,3,6,7-tetrachloro dibenzo-p-dioxin) was more potent than any of the others tested, even recognizing its lack of purity.

During this investigation, samples of the chlorophenols, both technical and reagent grades, were examined by GLC to determine if preformed chloro-dioxins were present. The presence of chlorodioxins was demonstrated by GLC, and these materials, which can be removed by appropriate techniques, were then tested in the chicken embryo system and did indeed produce chick edema. A current study of similarly contaminated chlorophenols, containing from 18 ppb to 95 ppm of chloro-dioxins with six or more chlorine atoms are currently under test.

This investigation was not pursued further in view of the fact that there had been no known occurrences of chick edema disease since the late 1950s and, hence, such research had a low order of priority.