Since Nitrogen Is at Least 5 Times More Soluble in Fat Than In

STUDIES ON DYSBARISM I. DEVELOPMENT OF DECOMPRESSION SYNDROME IN GENETICALLY OBESE MICE

WILLIAM ANTOPOL, M.D.; JOHN KALBERER, JR., M.S.; SAMUEL KOOPERSTEIN, M.D.; STEPHEN SUGAAR, M.D., AND CHRYSSANTHOS CHRYSSANTHOU, M.D. From the Joseph and Helen Yeamans Levy Laboratories, Beth Israel Hospital, and the Medical Department of the Port of New York Authority, New York, N.Y. Since nitrogen is at least 5 times more soluble in fat than in other tissues,l" the proportion of adipose tissue in the body influences the amount and rate of nitrogen released into the bloodstream after rapid decompression from high atmospheric pressure.5 It was the purpose of this investigation to find a modality in which the decompression syndrome could be produced regularly in small animals, so that a great number of them could be exposed simultaneously to pressure. In view of the influence of adipose tissue in the decompression syndrome, genetically obese mice were employed in these studies. Decompression illness ("bends") could be produced in the obese mice but not in their normal nonobese siblings or other strains of normal mice. These facts are especially significant in the light of recent reports cor- relating air crew obesity with fatal cases of dysbarism.8

MATERIAL AND METHODS Hereditary obese hyperglycemic mice of both sexes, 3 to 6 months of age were used. There were 2 weight ranges, 2I to 38 gm. (average 32 gm.) and 38 to 65 gm. (average 54 gm.), and, in addition, corresponding thin siblings weighing I7 to 27 gm. (average I9 gm.). The mice were obtained from the Roscoe B. Jackson Memorial Laboratory, Bar Harbor, Maine. In addition, thin yellow mice (YBR/W obtained from Dr. J. W. Wilson, Brown University, Providence, Rhode Island), C57 black, C58, and DBA mice were employed. The pressure chamber used for compression was 4 feet, 7 inches long and 2 feet, 4 inches inside diameter. The pressure was automatically controlled and could be maintained at any predetermined level up to go psi (pounds per square inch), ab- solute. Decompression to atmospheric pressure could be accomplished in less than i minute. Air entering the chamber was oil-free. An adjustable air-cooling-circulating system and a humidity control provided constant temperature (72 ± 20 F.) and relative humidity (50 per cent). All the animals were housed in metal cages in animal rooms with controlled temperature (71 + 20 F.) and relative humidity (50 per cent) and were fed Purina Laboratory Chow and water ad libitum. This work was sponsored by the Aerospace Medical Division, Air Force Systems Command, U.S.A.F. under Contract #AF 41(609)-1557, and aided by the United States Public Health Service, the Saul Singer Foundation and the Charles H. Silver Fund. Accepted for publication, February io, I964. lI5 I I6 ANTOPOL ET AL. VoI. 45, No. z

A total of 438 obese and thin mice of varying weight range, in a series of 32 separate experiments, was placed in the pressure chamber and exposed to 75 psi air pressure (gauge) for 6 hours and then were rapidly decompressed (in less than I minute). Tables I, II and IV show the type, weight range, and number of mice used in each experiment. In I 2 of the experiments, mice of different types or weight ranges were subjected to compression-decompression simultaneously to insure identical experimental conditions. These experiments are designated in the tables by the same experiment number. Following decompression, the mice were observed for at least 2 hours for clinical manifestations. Roentgenologic examination was performed in some mice before and at varying intervals after decompression, in order to observe the progressive accumulation of gas in various tissues and organs. When a mouse died, the survival period from the time of decompression was recorded. In several experiments some animals were sacrificed immediately after decompression. To eliminate the possibility of postmortem formation or fusion of bubbles, necropsy was performed in all cases immediately after death and with minimal manipulation. Sections of various organs were examined histologically. Several obese and thin mice which survived compression-decompression were sacrificed 4 to I2 months after decompression. In some of these, histologic examination of the femur or sternum was performed. In addition, one group of 20 obese mice ranging in weight from 52 to 6o gi. (average 55 gm.) was subjected to 75 psi for 3 hours instead of the usual 6 hours and then rapidly decompressed. Another group of 20 obese mice was subjected to 6o psi for 6 hours and then rapidly decompressed. RESULTS Obese Mice Shortly after decompression the obese mice exhibited difficulty in breathing, scratching, decreased motor activity, and most of them died within 30 minutes. In those animals which succumbed, severe respiratory distress, with panting, gasping and hiccough-like spells, and wobbling, twitching and erratic running about were observed several seconds pre- ceding death. The mortality of obese mice in the high weight range (39 to 65 gm.), based on a total of I75 mice in I9 separate experiments was 90.8 per cent (Table I). More than 8o per cent of these died less than 30 minutes after decompression. In the intermediate weight range (2I to 38 gm.) 6o obese mice in separate experiments exposed to identical conditions had a mortality of 53.3 per cent (Table II). Mice of the same strain but of different weight ranges, exposed to the same experimental conditions had a varied response (Table III). Obese mice of high weight range, exposed to 75 psi for 3 hours (or to 6o psi for 6 hours) and followed by rapid decompression, had a lower mortality (not exceeding 6o per cent) than animals of same weight range exposed to 75 psi for 6 hours. Gross examination of obese mice which succumbed to treatment revealed gas bubbles in the subcutaneous and intra-abdominal fat, the July, 1964 DECOMPRESSION SYNDROME II7 spleen and adrenals. The stomach and intestine were distended with great amounts of gas (Fig. i). The inferior vena cava was usually found filled with air bubbles, as were the right atrium and ventricle; this TABLE I MORTALITY OF OBESE MICE (39 TO 65 GM.) SUBJECTED TO COMPRESSION-DECOMPRESSION

Total Number of dead mice Exper. no. of Minutes after decompression no. mice 0-IO 10-30 30-60 over 6o Total I 8 4 3 I 0 8 6 4 0 2 0 0 2 7 4 2 2 0 0 4 8 4 2 2 0 0 4 9 4 2 2 0 0 4 10 5 I 3 0 0 4 II 24 7 5 6 0 18 12 14 3 9 0 0 I2 13 29 20 6 2 0 28 14 4 0 3 I 0 4 15 8 3 5 0 0 8 I6 4 0 4 0 0 4 I7 10 I 8 0 0 9 18 10 7 2 I 0 IO I9 6 o 6 o o 6 20 12 2 4 2 I 9 21 6 I 2 3 o 6 22 9 4 5 0 0 9 23 10 2 7 0 I 10

Total 175 6I 80 I6 2 I59 (9o.8)%

TABLE II MORTAITrY OF OBESE MICE (2I TO 38 GM.) SUBJECTED TO COMPRESSION-DECOMPRESSION

Total Number of dead mice Exper. no. of Minutes after decompression no. mice 0-10 10-30 3o-60 over 60 Total 2 I5 4 4 I 0 9 3 15 5 3 0 0 8 4 15 4 5 0 I 10 5 15 0 3 I I 5 Total 6o I3 I5 2 2 32 (53.3%)

TABLE III RELATIONSHIP OF WEIGHT TO MORTALITY IN DECOMPRESSION SICKNESS Weight Average No. range weight of MIortality Strain (gm.) (gm.) mice (%) Obese hyperglycemic 39-65 54 175 90.8 Obese hyperglycemic 2I-38 32 6o 53.3 Thin siblings 17-27 19 125 0 I I8 ANTOPOL ET AL. Vol. 45, No. z could be demonstrated in roentgenograms (Fig. 2). In only rare instances were bubbles found in the left chambers of the heart. Bubbles were not grossly visible in the remaining organs. In the bone marrow, however, the presence of gross bubble formation could not be verified because of the trauma inherent in fracturing the cortex to expose the marrow. For this reason, intact bones were decalcified and examined histologically. Microscopic examination revealed the bubbles to be round, ovoid, or irregularly-shaped clear spaces. On superficial examination the accumulations of gas appeared lodged in tissue spaces, probably because of the relative size of the vacuoles in comparison with the small blood channels. However, serial sections usually exhibited a continuity between the bubble space and the vessel lumen. Widely separated nuclei of flattened endothelial cells were occasionally noted about the bubble. In the spleen, gas bubbles rendered the organ sponge-like in appearance (Fig. 3); the very small bubbles were usually perifollicular in location in the region of the terminal opening of the sheathed arteriole. Rouleaux formation was evident here; larger bubbles appeared in the pulp and sinuses, often lim- ited and molded by the trabeculae. In the adrenal glands they occurred predominantly in the cortex (Fig. 4), at times forcing apart and disrup- ting cortical cell columns. In the bone marrow, pronounced hyperemia was evident in both diaphysis and epiphysis, and occasionally hemor- rhagic and necrotic foci were observed, predominantly in the diaphysis. In addition an occasional large ovoid space was filled with coagulum. Bubbles were present in both the epiphysis and diaphysis and, when large, were distorted and molded by bony trabeculae (Fig. 5). The cytoplasm of the liver and adrenal cells occasionally contained small vacuoles which failed to stain with Sudan III. Similar observations were reported by Gersch, Hawkinson and Rathbun in guinea pigs after decompression.7 The islands of Langerhans, which in obese mice are ordinarly enlarged, were hyperemic. Occasionally the perivascular spaces in the lungs were dilated and contained mononuclear cells and an eosinophilic coagulum. Gas bubbles were present in the pulmonary arteries but could not be detected in pulmonary veins. Pulmonary hemorrhage was infrequent. Numerous scattered petechial hemorrhages were frequent in the brain. In addition, in some cases the perivascular space was widened, but it could not be determined whether this was a pathologic change or an arti- fact. No gas bubbles were observed. Bubbles were apparent in the adeno- hypophysis; they were occasionally present in the skin but were rare in the kidney, liver or intestinal wall. In obese mice sacrificed immediately after removal from the chamber July, I964 DECOMPRESSION SYNDROME II9 (before any decompression symptoms developed) no gas bubbles were found in the subcutaneous or intra-abdominal fat or in any of the viscera. This indicated that an element of time was essential for the formation of gas accumulations and for the fusion of minute bubbles to form larger ones. Roentgenograms of an obese mouse surviving i hour after exposure to 75 psi gauge pressure for only 3 hours were taken before pressure, immediately after, and i hour following rapid decompression. The films taken immediately after decompression showed no discernible gas in the subcutaneous fat but those taken i hour later exhibited numerous gas pockets (Fig. 6). This further documented the time factor required in bubble formation. In surviving obese mice, sacrificed 4 to I2 months after decompression, random histologic examination of one femur occasionally revealed fibrosis and fibrochondromatous structures in the marrow. In the diaphysis, these lesions measured up to one third the diameter of the medul- lary canal. In the epiphysis, they occupied many spaces of the can- cellous bone, replacing the bone marrow. Occasionally the marrow contained cystic areas completely or partially filled with a coagulum or fibrous tissue. In some instances these lesions involved more than 50 per cent of the epiphysis. Thin Mice None of the 203 nonobese mice exposed to 75 psi for 6 hours died (Table IV). In I2 of the 2I experiments obese mice and their thin siblings were subjected to pressure simultaneously so that the effects of the treatment could be compared under identical conditions. Shortly after removal from the pressure chamber, thin mice exhibited slight panting and, in some cases, scratching, but none of the other symptoms encountered in the obese animals. Mice sacrificed I5 minutes to several hours after rapid decompression showed no bubble formation on gross examination. Occasionally, the intestine was slightly distended, but the configuration was not nearly so striking as in the obese mice. In thin mice sacrificed 24 to 48 hours after rapid decompression, bone marrow sinuses were markedly hyperemic, as in the obese mice, but no gas bubbles were present. In some of these animals sacrificed several months after decompression, the femur revealed alterations similar to those in the obese mice. DISCUSSION The obese mice used in these experiments offered many advantages as subjects for the study of decompression illness. They were regularly I20 ANTOPOL ET AL. Vol. 45, No. z susceptible to this syndrome and many, together with their thin siblings, could be placed in the pressure chamber simultaneously, assuring exposure of all animals to identical conditions. There are at least 6 known types of obesity8; one type, observed in mice,9 is determined and controlled by a single Mendelian recessive gene.

TABLE IV STRAINS OF NONOBESE MICE SUBJECTED TO COMPRESSION-DECOMPRESSION Exper. No. of no. Strain mice Mortality I 4 0 2 I5 0 3 I5 0 4 15 0 5 Thin siblings to 14 0 6 2 0 8 hereditary obese mice 4 0 I2 6 0 14 4 0 15 7 0 17 10 0 20 12 0 24 I7 0 25 C7Bk15 0 26 C57/Blk 15 0

28 YBR/W 4 0

29 10 0 30 I0 0 31 DBA 10 0 32 10 0 Total 203 0

This type of obesity, as in the strain used in these experiments, has been referred to as the "hereditary obese-hyperglycemic syndrome" by Mayer, Bates and Dickie.Y The obese-hyperglycemic mouse is charac- terized by extreme overweight; Mayer, Russel, Bates and Dickie 11 reported weights of 50 to I I5 gm. as compared to 25 to 30 gm. in thin lit- termates. The excess weight is due almost entirely to an increase in body fat. It has been reported that the release of nitrogen from fat plays a significant role in the development of decompression illness.5"2 Boycott and Damant'2 chemically estimated the quantity of fat in rats and guinea pigs, and correlated this with the survival and mortality rates after rapid decompression. A direct relation was found between the amount of fat and the susceptibility to decompression illness. Gersh and July, I964 DECOMPRESSION SYNDROME I21 co-workers,7 working with guinea pigs, found bubbles in both obese and lean animals but observed that gas collections occurred more readily and to a greater extent in the former. The results obtained in our experiments are in agreement with the above findings. In our own mice subjected to compression-decompression, mortality could be correlated with the degree of obesity (Table III). It is usually stated that bone pain in individuals working in compressed air is due to accumulated gas bubbles in the yellow marrow, 70 to 90 per cent of which is fat.7 In the obese-hyperglycemic mice the marrow of the femur and tibia was highly cellular and contained very little fat. Despite this, however, numerous large gas bubbles were present, indicating that factors other than fat may also play a role in de- termining localization and accumulation of gas. It is possible that the circulation in the marrow is slow when the blood sinuses are distended, and this retardation of flow favors gas bubble aggregation.5 This prin- ciple is also applicable to the spleen with its open circulation, to the adrenal cortex with its irregularly anastomosing sinusoids, to the pars anterior of the pituitary with its rich blood supply, and to the liver with its venous sinuses and portal circulation. 13,14 In keeping with these observations, Harvey and colleagues15"6 and Behnke and associates"5 concluded that formation and growth of vacuoles is dependent upon factors other than the concentration of gas in the tissue itself. This is particularly so with regard to bubble formation in the spleen, since the spleen contains no fat and receives its blood supply from an arterial circulation in which very few, if any, bubbles appear. Moreover, few bubbles are found in the portal vein or liver on gross or microscopic examination, despite the fact that these drain blood from the spleen, which may exhibit massive gas accumulation. All of these observations are compatible with active (in addition to passive) extraction of consider- able amounts of gas from the blood in the spleen. Because of these ex- tractive and accumulative features, the spleen may act as a safety reservoir, sequestering gas and holding it in check. A study of the effects of decompression in splenectomized animals is now being planned. Because of the constant marked gaseous distention of the intestine in obese animals after decompression, a feature which is minimal in thin mice, the capacity of the intestine to function as a gaseous exchange organ should be investigated. Elimination of nitrogen by the intestine may be an additional factor responsible for the infrequent presence of gas in the portal vein and liver. The presence of bubbles in the left ventricle in an occasional mouse may have stemmed from a "nonfunctioning" patent foramen ovale which would not permit the passage of blood but could permit the passage I2 2 ANTOPOL ET AL. Vol. 45, No. z of gas from the right to the left atrium. Air bubbles could also gain en- trance to the left cardiac chamber by way of thebesian vessels and venae parvae which may form a communication between the right and left chambers. In addition, one should consider those instances in which the nitrogen reaching the lung cannot be blown off in sufficient amounts dur- ing expiration; this might then traverse the capillary bed and pulmonary veins and reach the left side of the heart. The late-appearing bone lesions were observed in both obese and thin mice. Similar bone changes have been observed in human cases of caisson disease.'7 The pathogenesis of latent bone alterations is obscure. They may represent changes secondary to hyperemia, hemorrhage, infarction or necrosis occurring early after decompression. Gas bubbles obstructing nutrient vessels of bone could be responsible in obese mice. However, in thin mice, similar changes were observed even though these animals revealed no bubble formation and did not develop decompression sickness. It is possible, therefore, that compression-decompression initiates changes which are independent of bubble formation and the decompression syndrome. In this respect it is of interest that bone and joint alterations have been found in human subjects exposed to high pressures with no history of an acute or subacute attack of caisson disease.'8 Leriche 19 has described marrow changes similar to those observed in our experiments but due to vasomotor reactions. Reactions of this nature may also be con- sidered possible factors in the pathogenesis of the bone lesions described herein. A long-term experiment in search for latent and delayed effects, particularly in bone, should be undertaken. It will also be necessary to devise methods for distinguishing alterations due to changes in atmospheric pressure per se from those due to compression followed by decompression. SUMMARY The decompression syndrome was produced regularly in genetically obese mice after exposure to 75 psi (gauge) air pressure for 6 hours, followed by rapid decompression. Normal intact thin siblings or other nonobese mice exposed to the same conditions did not develop this syndrome. Gross and microscopic examination of mice succumbing to decompression sickness revealed striking bubble accumulation in the spleen, bone marrow, adipose tissue, adrenal cortex and pituitary. The intestines were markedly distended with gas; this was minimal in thin mice. It was postulated that the intestine might be an organ for gaseous exchange. The spleen seemed to act as a reservoir for bubbles, thus preventing the gas from entering the general circulation. July, I964 DECOMPRESSION SYNDROME I23 There was a direct correlation between the degree of obesity and susceptibility to decompression illness. Bone alterations (fibrochondromatous lesions and cystic areas) observed in both obese and nonobese mice, long after exposure to compression-decompression, are discussed. It appeared advantageous to use the genetically obese mouse as a subject in the study of decompression illness. REFERENCES I. BEHNKE, A. R.; THOMSON, R. M., and SHAW, L. A. The rate of elimination of dissolved nitrogen in man in relation to the fat and water content of the body. Am. J. Physiol., I935, I14, 137-146. 2. CAMPBELL, J. A., and HILL, L. Concerning the amount of nitrogen gas in the tissues and its removal by breathing almost pure oxygen. J. Physiol., 1931, 71, 309-322. 3. CAMPBELL, J. A. Studies in saturation of the tissues with gaseous nitrogen. II. The gaseous nitrogen content of certain fats, fatty acids, etc., saturated with air, as estimated by a new apparatus. Quart. J. Exper. Physiol., I933, 23, 2II-2I8. 4. VERNON, H. M. The solubility of air in fats and its relation to caisson disease. Proc. Roy. Soc. London, s.B, 1907, 79, 366-371. 5. BEHNKE, A. R., JR. Physiologic studies pertaining to deep sea diving and aviation, especially in relation to the fat content and composition of the body. Harvey Lect., I94I-I942, 37, I98-226. 6. POWELL, T. J.; CARRIGAN, E. P., and STANFIELD, M. J. Obesity in aircrew. Aerospace Med., I963, 34, 21-25. 7. GERSH, I.; HAWKINSON, G. E., and RATHBUN, E. N. Tissue and vascular bubbles after decompression from high pressure atmospheres-correlation of specific gravity with morphological changes. J. Cell. & Comp. Physiol., 1944, 24, 35-70. 8. MAYER, J. Genetic, traumatic and environmental factors in the etiology of obesity. Physiol. Rev., I953, 33, 472-508. 9. INGALLS, A. M.; DICKIE, M. M., and SNELL, G. D. Obesity, a new mutation in the house mouse. J. Hered., I950, 41, 3I7-3I8. IO. MAYER, J.; BATES, M. W., and DICKE:, M. M. Hereditary diabetes in genetically obese mice. Science, I95I, 113, 746-747. II. MAYER, J.; RUSSEL, R. E.; BATES, M. W., and DICKIE, M. M. Metabolic, nutritional and endocrine studies of the hereditary obesity-diabetes syndrome of mice and mechanism of its development. Metabolism, 1953, 2, 9-2I. I2. BOYCOTT, A. E., and DAMANT, G. C. Experiments on the influence of fatness on susceptibility to caisson disease. J. Hyg., Cambridge, I908, 8, 445-456. 13. POPA, G. T., and FIELDING, U. A portal circulation from the pituitary to the hypothalamic region. J. Anat., I930, 65, 88-9i. 14. WORTHINGTON, W. C., JR. Some observations on the hypophyseal portal system in the living mouse. Bull. Johns Hopkins Hosp., 1955, 97, 343-357. I5. HARVEY, E. N.; BARNES, D. K.; MCELROY, W. D.; WHITELEY, A. H.; PEASE, D. C., and COOPER, K. W. Bubble formation in animals. I. Physical factors. J. Cell. & Comp. Physiol., I944, 24, I-22. I24 ANTOPOL ET AL. Vol. 45, No. I i6. HARVEY, E. N.; WHITELEY, A. H.; MCELROY, W. D.; PEASE, D. C., and BARNES, D. K. Bubble formation in animals. II. Gas nuclei and their dis- tribution in blood and tissues. J. Cell. & Comp. Physiol., I944, 24, 23-34. I7. KAHLSTROM, S. C.; BURTON, C. C., and PHEMISTER, D. B. Aseptic necrosis of bone. I. Infarction of bones in caisson disease resulting in encapsulated and calcified areas in diaphyses and in arthritis deformans. Surg. Gynec. & Obst., I939, 68, I29-I46. I8. TAYLOR, H. K. Aseptic necrosis and bone infarcts in caisson and noncaisson workers. New York State J. Med., I943, 43, 2390-2398. I9. LERICHE, R. The problem of osteo-articular diseases of vasomotor origin. Hydroarthrosis and traumatic arthritis: genesis and treatment. J. Bone & Joint Surg., I928, io, 492-500.

LEGENDS FOR FIGURES Photomicrographs were prepared from sections stained with hematoxylin and eosin. FIG. I. Obese mouse exposed to 75 pounds per square inch (psi; gauge) for 6 hours and rapidly decompressed. The stomach and intestine exhibit marked gaseous distention. FIG. 2. Obese mouse exposed to 75 psi (gauge) for 6 hours and rapidly decompressed. A roentgenogram demonstrates the presence of air in the vena cava (arrow) and heart. (The film shown has approximately X 2 magnification; it was obtained by using 0.3 mm. fine focus x-ray tube and increasing the object- film distance.) July, I964 DECOMPRESSION SYNDROME I25

1or!.

i1S11 i..n~~~~~ 126 ANTOPOL ET AL. Vol. 45, No. z

;f;>" i.0....:: 77 - :f:fa:d: St02X..... FIG. 3. Mouse exposed to 75 psi (gauge) for 6 hours and rapidly decompressed. Died in I 7 minutes. The spleen is filled with air bubbles, rendering the organ sponge-like in appearance. X 2.5. FIG. 4. Adrenal from the same mouse shown in Figure 3. Abundant air bubbles are associated with tearing of adrenal tissue. X 50. FIG. S. Femur in an obese mouse exposed to 75 psi for 6 hours and rapidly decompressed. Air bubbles in marrow spaces are distorted by bony trabeculae. X So. FIG. 6. Obese mouse exposed to 75 psi for 3 hours and rapidly decompressed. Died in 65 minutes. Progressive bubble formation is evident in the subcutaneous tissue, and there is increasing gas content (arrows) in the intestine. A. Before exposure to pressure. B. Immediately after decompression. C. One hour after decompression. July, r964 DECOMPRESSION SYNDROME 127

_i LiE,;EE;StE ti--A ;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~1s-