292 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS however, a large increase in the number of sudden deaths immediately fol- lowing inhalation in many of these cases, some exertion or stress had directly preceded the death. Reasoning that most of these people were oxygen-defi- cient as a result of confining their breathing to the contents of the bag, and realizing that, as has been known for many years, halogenated hydrocarbons will sensitize the heart to the effects of epinephrine, he hypothesized that the stressful activity had resulted in the endogenous release of epinephrine in the victim, and that sensitization by the fluorinated hydrocarbon propellants had resulted in a fatal cardiac arrhythmia. Also during the 1960's, an epidemic of sudden deaths among asthma-suffer- ers who used acrosolized bronchodilators was noted (34). This large increase in the death rate seemed to be confined, however, to England and Wales, Ire- land, Scotland, New Zealand, and Australia and reached a peak around 1967, after which the rate declined. Putting these two phenomena together, Taylor and Harris published a pa- per (35) in 1970 in which they presented experiments that purported to sup- port their theory that the asthma deaths had resulted from cardiac toxicity induced by the propellants in the victims' nebulizers. Since its publication, debate has continued on the errors in design and conclusions of this study, but unfortunately, the damage has been done to the confidence of the asthma pa- tient in his medication and to the general confidence of the consumer in aerosolized products. One major fault in these experiments was that the mice that were used were administered one puff from an antiasthma nebulizer, dis- charged directly into their nose and mouth. The normal volume of air inspired by a mouse varies from 0.20-0.25 ml, so that the 6.5 ml of propellant gas de- livered by the nebulizer assured that the animal would breathe 100% propel- lant for the three inspirations which it was allowed. The normal volume of air inspired by a human is about 450 ml, approximately 1800 times that of the mouse, so that the propellants would normally be inhaled by a human at a concentration of about 1% of the inspired air in the initial breath. In addition to this, the head of the animal was thrust into a tightly-fitting plastic bag after exposure to the nebulizer until cardiac abnormalities ap- peared on the electrocardiograph. The purpose of this bag was to simulate the lack of air which asthmatics may suffer before using the nebulizer. Taylor and Harris reported serious cardiac arrhythmias following propellant plus asphyx- ia, but none following asphyxia alone of up to 4 minutes. Investigations by 5 different laboratories (36-40) have been unable to duplicate these results and have found that the asphyxia component alone is sufficient to induce se- rious changes in cardiac rhythm. In a recent answer to his critics, Harris (41) now contends that the asphyxia originally described as being induced by "... a form-fitting plastic bag wrapped tightly around the nostrils and mouth ß . ." (35) was actually only partial asphyxia. It is for this reason, he states, that his asphyxia alone did not result in arrhythmias.

TESTING FOR INHALATION TOXICITY 293 In addition to all of this, it has been shown that the blood half-life of P-11 is between 0.3 and 1.5 rain, while the maximum blood concentration measured after two consecutive puffs from a nebulizer was •2.6 ptg/ml (4•2). It is known that at least 10 times this level is required to cause cardiac sensitization in nonanesthetized dogs when challenged with intravenous epinephrine (43, 44). We can safely conclude, then, that the propellants were not the culprits in the increase in asthma deaths. Just as a point of information, it has been re- ported (45) that this increase coincided with an increase in sales of one brand of nebulizer which contained 5 times the concentration of isoproterenol as was present in the other brands in addition, this more potent formula was not licensed for sale in this country, but was sold in the countries which experi- enced the epidemic of sudden asthma deaths. The myocardium of the dog has been shown to be particularly sensitive to isoproterenol when arterial oxygen tension is low (46), as would be the ease in a severe attack of asthma. Although the aerosol propellants were not implicated after all in these deaths, there is still a great deal of concern over their safety as a result of the findings of several investigations which may have been stimulated by the orig- inal, unfounded charges. One of the first such studies which supplied critical data on this problem was conducted by Reinhardt and his group (47). Briefly stated, they found that conscious dogs breathing a minimum of 0.35% P-11, or about 10 times as much p-19., experienced serious cardiac arrhythmias when challenged with an intravenous injection of epinephrine the concentra- tion of epinephrine approximated the amount thought to be released under conditions of stress. Additionally, they found that when the animal breathed 80% P-ll for 30 see, and was then frightened by a loud noise to stimulate en- dogenous epinephrine release, serious disturbances in heart rhythm could be produced in some of the dogs. Using barbiturate-anesthetized dogs, Flowers and Horan (48) have reported similar findings when the animals breathed the propellant gases from a plastic bag and were also given oxygen supple- mentation to prevent anoxia. A second study by Taylor et al. (49), better planned than the first, subject- ed monkeys to inhalation through an endotraeheal tube of an air mixture con- taining 30% P-19. and 9% P-114. They reported several different types of arrhythmias in response, without the use of injected ep'nephrine. Further work has also been done by Reinhardt's group (50). Using conscious beagle dogs which were running on a treadmill, they exposed the animals to atmos- pheres containing P-11, P-19., or P-114. Up to 1.0% P-11 produced no cardiac arrhythmias. At a level of 5.0% P-114, and 10.0% P-l•2, ventricular arrhythmias were produced in one of the dogs. For purposes of comparison, to return again to the calculation for propel- lant released by total discharge of a can of deodorant, we find that even the lowest level of P-19. which had an adverse effect on the heart in the above ex-