Document 6B9LqVr4Y7n2NoxZqv45RJB06

Meh'IN*e Copyright 1067 by The Williams & Wilkins Co. NOTICE THIS MATERIAL MAY Bfc PROTECTED BY COPYRIGHT I AW (TITLE 17 U.S. CODE) GGU-1 Vol. 46, Xo. 2 Printed in U.S.A. THE EARLY STAGE OF LIVER INJURY IN THE ALCOHOLIC HUGH A. EDMONDSON, M.D., ROBERT L. PETERS. M.D., HOWARD H. FRANKED, M.D., Ph.D., and STEPHEN BOROW8KY. M.D. Injury to liver cells and reactive fibrosis of many years and each had recently been drink the mesenchyme may occur after only a few ing more than usual. Most of them described rears of excessive ethanol intake or after a this episode as a "spree" or a "binge." Most much longer span, sometimes even decades. We of them had eaten little or nothing for vary have been particularly interested in the early ing periods, from one day to one month. : stages of liver injury and in the progression Of the symptoms on admission to the hos I or repetitive character of the pathologic le pital, those related to the gastrointestinal sions that lead to recognizable alcoholic liver tract were by far the most common: 40% com disease. Innumerable articles have been writ plained of vomiting, often of more than 24 ten on the pathology of liver injury in the hours' duration. Twenty-two patients com alcoholic, but these have been concerned with plained of abdominal pain, most often in the biopsies taken from patients with sympto epigastrium. Five of these were later proved matic liver disease or they have been retro to have acute pancreatitis. A history of bleed spective studies based upon autopsy findings. ing from the gastrointestinal tract, consisting materials and methods of melena or hematemesis, or both, was ob tained from 21 patients. The .origin of the The major portion of this study concerns bleeding was proved in only 12 patients; in 100 chronic alcoholic patients who had been these cases, peptic ulcer disease was found to drinking heavily just prior to hospitaliza be the cause. Symptoms related to the nervous tion. They were chosen at random from those system, including hallucinations, impending I on the admitting ward at the Loi Angeles delirium tremens, and neuromuscular disor ! County Hospital. Blood for numerous chem- ders, were noted in 29% of the patients. j ical tests was drawn before any patient re- Nearly all the patients recovered within a i ceived treatment, food or fluids. Biopsies were week and were able to leave the hospital. How I performed on 45 of the 100 patients within ever, 6 patients expired, one of hemorrhagic 12 to 72 hours after admission. After paraffin pancreatitis, one of myocardial infarction, , embedding of the sections, slides were stained one of cardiac arrest, one of mesenteric with hematoxylin and eosin and with Mas- thrombosis, and one of pneumonia and a fatty , son's trichrome stains. In addition, at time liver. The cause of death in the sixth patient ' of biopsy, a small portion of the specimen was not ascertained even after autopsy. was immediately fixed in osmium tetroxide Laboratory findings. Although many labora Jand prepared for electron microscopic study. tory tests were done, only the electrolyte dis No patient with histologic or other evidence turbances and serum bilirubin will be con I of cirrhosis was included in the final tabula-sidered briefly in this article. Abnormalities ' tion. were noted in the serum potassium, phos ! RESULTS phorus and calcium levels. Hypokalemia was the most common electrolyte deficiency, being ^ Clinical findings. The study group con found in 50% of the patients. Whereas in most sisted of 75 men and 25 women. Of the instances the hypokalemia was mild, in 10 pa group, 57% were between 40 and 60 years of tients the serum potassium was below 3 mEq/ age. All had been addicted to alcohol for liter. In nearly every instance serum potas sium levels below 3 mEq/liter were associated From the Departments of Pathology at the University of Southern California School of Medi- with vomiting, or lack of food, or diarrhea. j cine and the Los Angeles and John Wesley This was invariably true of those with levels / County Hospitals, Los Angeles, California. below 2.5 mEq/liter. In this latter group, only ) ! This work was supported bv grants of the National Institutes of Health No. AM 5801 and one patient had a mild alkalosis. GM 758. Hypocalcemia occurred in 13 patients and 119 24873002 .'. * ** -if 120 EDMONDSON ET AL hypophosphatemia in 20. Although low serum phosphorus or calcium levels might occur alone, in a majority of patients they were associated with hypokalemia. It is not known how a low serum level of phosphorus or calcium might affect the liver cells, but it is of interest that one patient with a calcium level of 3.3 mEq/liter and potassium of 2.2 mEq/litcr died of cardiac arrest soon after ad mission to the hospital. This case brings to mind the many eases of sudden and unex plained deaths among alcoholics that are noted in coroners' offices throughout this country. Serum bilirubin levels above 2.0 mg/100 ml of blood were noted in 19 patients, but in only 12 of these was jaundice observed on physical examination. Biopsies were obtained on 3 of these patients; each showed a fatty liver but cholestasis was not observed. Microscopic findings. The most common microscopic abnormality was fatty change: this was noted in 41 of the 45 biopsy speci mens, or 91.1%. Included among the livers with fatty change were 5 in which the changes were slight, less than 1% of the liver cells showing any fat vacuoles. The fatty change in 36 patients was classified on the basis of 14to 44-, using the classification of 1+ for fatty change that involved a maximum of 25% of the cells; 2+ for 25 to 50%; 3+ for 50 to 75%; and 4+ for more than 75% of the cells. In 22 patients, the fatty change was 14-; in 10 it was 2+; in 4 it was 3+; none was graded 4+. There was good correlation between the lack of food and the severity of fatty change as all patients who gave a history of not eat ing for two or more weeks had 2+ or 3+ fat in the liver. In contrast, those who said they had been eating at least a small amount of food daily until 24 hours before entry had either no fat or only 1+ change. One patient who claimed he had not eaten for one week had 14fat in the liver, and another who had not eaten for three days had 24- fatty change. In one patient with severe fatty change, a repeat biopsy performed 33 days later showed no stigmata of disease. Hydropic change was the second most prev alent type of injury to the liver cells; this was seen in 18 patients, or 40%. The exces sive, watery cytoplasm either had a pale granular appearance or it was vesiculated (both types of change are shown in Fig. 1). The vesicles were most often multiple and filled the cytoplasm. Sometimes larger, poorly outlined vacuoles were present that were diffi cult or impossible to distinguish from fat vacuoles. The lack of a sharp margin around the larger vacuoles indicated that they might be hydropic in origin, but, on examination with the electron microscope, all large vacuoles appeared to be composed of fat. Hydropic swelling was most often centrolobular; rarely were cells at the periphery affected. There was no constant association between the hy dropic swelling and either fatty change or early fibrosis. With the electron microscope, the hydropic cytoplasm seen with the light microscope appeared as ill-defined radiohieent areas devoid of any organelles or other struc tures (Fig. 2). In addition, in some of the specimens the cisternae were greatly dilated (Fig. 3), although the ribosomes shower! no significant change. It was considered that the vacuolar type of hydropic change in the liver cells might be related to the hypokalemia in a manner similar to vacuolar nephropathy, but the presence of hydropic cells in 5 patients with normal serum potassium indicates that this is not necessarily true. Nevertheless, hy pokalemia was observed in 10 of 17 patients in whom hydropic change was noted on biopsy,, whereas in 28 patients without such change only 10 had a low serum potassium. Intracellular potassium levels were not de termined. The mitochondria in some instances ap peared normal, and occasionally they had an increased number of cristae. The most promi nent change was the presence of swollen mito chondria that contained few or no cristae (Fig. 3). In many of these, dissolution of the mitochondrial membrane was observed (Figs. 3 and 4). In many of the liver cells the mito chondria were concentrated both along the plasma membrane and around the nucleus, the more or less clear space between being devoid of organelles (Fig. 5). The striking appearance of the mitochondria along the cell membrane; is easily recognized with the electron mi croscope. Hyaline necrosis accompanied by a few neutrophils and moderate sclerosis of the ves- vi .ated in Fig. 1). multiple and ! rger, poorly t were diffi1 from fat rgin around they might j examination ge vacuoles . Hydropicilar; rarely ted. There Dn the hychange or microscope, the light radiolucent , ther structne of the :ly dilated showed no l that the i the liver lemia in a lathy, but 5 patients 1 :ate- at leles-. .,y- I ' patients | aoted on , tout such totassium. ' not de- nces ap had an t promien m. itoi cristae n of the d (Figs. he mitoang the leus, the ' devoid )earance nbranes on mi- y < j ' I j a few he ves- S | Fig. 1. A vesicular type of hydropic change is seen in many of the enlarged parenchymal cells. Other cells are swollen but are still slightly granular. The patient had a blood alcohol of 274 mg/100 ml on admission to the hospital. He had been on a 10-day "binge" and had not eaten for 3 days. Delirium tremens was the clinical diagnosis. (Hematoxylin and eosin staFinig; . X2.6I5n0) this cross section of an hepatic sinusoid, the surrounding parenchymal cells show a breakdown of the endoplasmic reticulum in some areas but not in others. The endoplasmic reticulum breakdown may correlate with granular hydropic (HY) swelling. Red blood cells (RBC); lipid droplets (LI); nucleus of liver cell (N). (Lead citrate stain; X 9,000) 24873004 ! Sfc-S'JiwoiiSSCffi : ig^gsfigKSSS&^sS?Sv` ggsg ?s 122 EDMONDSON ET AL SsfcM gWsPV&>.,sAt.e* gatses rjwk 5^*5,-^ *Mi . '3-5#S?-**? "Ml fsai &JJK?*?**'' ** `RER. mm1 >#a ,,sT.*Sw-Sa !W* F fcSu EfW '-* iSt-> LArf'? :V> M+mz ^SKr.S'iaarc'i Jgesss--? ,-m N jS ar m S(ftSiT3y5v? J?JJ? ?&Si *M [li Fie. 3. Vesiculation and swelling of endoplasmic reticular cisternae. Ribosomes still line cisternae but the mitochondria show early membrane disruption. Degenerating mitochondria (M); dilated cisternae of rough endoplasmic reticulum (RER). (Lead citrate stain; X 32,500) Fig. 4. Further vesiculation of endoplasmic reticulum with loss of some ribosomes. An hourglass-shaped mitochondrion (M) shows disrupted membranes. Markedly dilated cis tern,a of rough endoplasmic reticulum (RER); cisterna of smooth endoplasmic reticulum (SER). (Lead citrate stain; X 60,000) 2 a i> i I ne Fig. 5. Low-power electron photomicrograph shows the displacement of cell organelles ia to the cell membranes and to the nuclear membrane. The endoplasmic reticulum is almost X completely destroyed. A few strands of collagen between 2 liver cells can be recognized. Nucleus (N); mitochondria (M); hydropic change (HY); collagen (CO). (Uranyl acetate in staFinig;. X6. 1H1y,TaOliOn)e necrosis and neutrophilic exudate is seen at the lower left. Intense m sclerosis of sinusoids, diffuse hydropic swelling and moderate fatty change are also present. The patient, a 55-year-old woman, had been an alcoholic for 10 years. (Hematoxvlin and eosin stain; X 350) 123 i 124 EDMONDSON ET AL sels was present in only one patient (Fig. 6). A type of hyaline change involving only oc casional cells and characterized by small round hyaline bodies within the cytoplasm was seen in 3 biopsy specimens (Fig. 7). These bodies are similar to those reported by Porta and associates (S) in rats that had re ceived 3671 of their total calorics as ethanol for 2 to 12 weeks. The hyaline bodies were believed to be megamitochondria when examinecl with the electron microscope. To date we have been unable to identify the small round hyaline bodiesin our material with the electron microscope. Most interesting were the changes in the mesenchymal tissue. Although 20 patients, or 44.49c, had no centrolobular mesenchymal changes (Fig. S), the other 25 patients, or 55.6%, did show some degree of connective tissue proliferation. The milder degrees of proliferative changes were unexpected because biopsies on alcoholics without symptomatic liver disease had not been done on more than an occasional patient. An increase of connective tissue that stained positively for collagen often thickened the walls of the cen tral veins and sinusoids (Figs. 9 atid 10). The lumen of the vessels did not seem to be compromised in any way and the sclerotic process was not in every case associated with fatty, hyaline, or hydropic change in the parenchymal cells. In more advanced lesions, the collagen surrounded individual liver cells or groups of liver cells and at the same time obliterated the sinusoidal lumens (Fig. 11). The presence of collagen in the space of Disse was observed on examination with the electron microscope (Fig. 12). The amounts of collagen varied; in some specimens it was an outstanding feature and penetrated the spaces between individual liver cells (Fig. 5). Al though the centrolobular sclerosis might occur alone, in most instances there was an increase of periportal connective tissue of a lesser ot an equivalent amount; rarely' was it greater. A summary of the microscopic findings is given in Table 1. by necrosis was rarely observed; it occurred in only 3 of 45 biopsy specimens and in one patient at autopsy. In one of the specimens, die was DISCUSSION In this study, biopsies were performed on 45 of 100 patients, none of whom had cirrhosis, so there was an opportunity to study' the early k. Fig. 7. Small There is mild fr claimed to have VsJ X 500) Fig. 8. This the liver cells. T o out any food in Q irv uijrl LIVER INJURY IN THE ALCOHOLIC 125 3J}I3 P'3S'!) '(IV) ;qabaJB,j 'aarqaid uuoj<i zi 'MX (008 X iui^s > g joj ua}Ba }OU bj +Z 13 s? 9J3lIX 33 aiLL H 0IX p9A\ ajjtbj 8ut}B3 uaaq pBq aq }nq sjbsX 9} joj oqoqoo[B (00S X db tiaaq p' uWibqj}suaapuBiodjqoaqi.ix} usoajsosnbajB^A) }bj A\aj b Xpio sniB}noo }Bq} JSAq b m spB-ii jBptosnnts jo Suraaqoiq} XjjBg oi 'DiJ auiojqoi.p uossbjv) 'I 'Stj in sb }napsd auiBg -spiosnuE Stnpnnojjns aq} 3(uooo[eb xpjB:maqBn}o5 puajxa 'anssp aApaanuoo 'jo sinds puB- panaqoiq; st uiaA jBJ}nao siq} }0 {[Bav aqj, '6 'oij 931 LIVER INJURY IN THE ALCOHOLIC 127 TV* ' '4 Sjfcv. \&3 -sv vrv HS S7s| p i-g P s#ra ltsli M& 24873010 K?K,< t^s *jr - '%] : crfSifrr#*1 7a Fxc. 11. The coDnective tissue has encircled individual liver cells and obliterated sinusoids. There is a 2+ fatty change. The patient had been drinking heavily for 1 month and had not eaten for 3 days previous to hospitalization for delirium tremens. (Masson trichrome Fig. 12. Prominent collagen deposition in space of Disse (SD). Sinusoid (S) is at top of picture. Parenchymal cell microvilli extend around collagen (CO) fibrils. Mitochondrion (M). (Lead citrate stain, X 46,000) 128 EDMONDSON ET AL TABLE 1 Microscopic Findings in Ifi Patients Fat Hydropic Hyaline Centrolobular Periportal cnange necrosis sclerosis fibrosis 0-1+ in 9 1+ in 22 2+ in 11 3+ in 3 4 8 6 i 1+ in 5 1+ in 2 2 1+ in 1 1+ in 2 2+ in 3 2+ in 2 1-b in 1 potassium levels did not have any hydropic change. The possibility of widespread damage to the cell membranes cannot be ruled out, even though the only direct evidence of this was seen in the disruption of the mitochon drial membranes. At a recent symposium on the "Forms of Water in Biological Systems," Hechter (3) discussed the intracellular water structure and mechanism of cellular trans port. He presented a schematic representation of the "resting membrane," showing the pro tein layers interlocked by water layers in an icelike arrangement. Taking into considera tion both the light and the electron micro scopic findings in our biopsy sections, it seems inevitable that the normal relationship of water to the membrane systems is disturbed. Further investigation of the plasma mem brane as well as the membranes of the or ganelles is needed. Another possible cause for the hydropic change in the liver cells is suggested by the experiments of Kalant and co-workers (4) who showed that ethanol affects the transport of cations across cell membranes, leading to an increase in intracellular water and sodium and a possible fall in intracellular potassium. Severe mitochondrial changes in the livers of chronic alcoholic patients have been re ported (2, 7, 9, 10). Enlargement of mito chondria, with abnormality of shape, changes in the cristae, myelin figures and degeneration were reported by Porta and associates (7). Svoboda and Manning (10) noted crystal like inclusions, often with mitochondrial en largement. Abnormally large, dense bodies in the mitochondrial matrix, cavitation of mito chondria and rearrangement of cristae were also observed. Lane and Lieber (5) used hu man subjects who, after a period of normal diet, were given an analogous diet in which the carbohydrate was replaced isocalorically by ethanol to the extent that 46% of the total caloric intake was supplied by ethanol. The diets were continued for 16 to 18 days. The authors noted on biopsy that a minority of the mitochondria exhibited some degree of damage. The small number of biopsy specimens showing hyaline necrosis indicates that this type of injury is infrequent in patients who do not have symptoms of liver disease. The small, round hyaline bodies which are mor phologically dissimilar to "alcoholic hyalin" seen in otherwise normal appearing cells may or may not have any significance. It is prob able that liver cell injury, short of necrosis, in the alcoholic is reversible, even after repeated "binges" over a period of many years. The question of whether or not hydropic change and small hyaline bodies affecting the liver cells may, if the patient continues to drink and abstains from food, progress to scle rosing hyaline necrosis i left unanswered. It seems to us that patients who enter the hos pital with this disorder most likely go through a stage in which their liver shows only the milder changes. Perhaps the onset of either delirium tremens or nausea and vomiting, followed by hospitalization, prevents the more serious liver disorder. However, there may be unknown factors that lead to the onset of aeuje sclerosing hyaline necrosis, as we have occasionally seen the disease become clinically manifest several days or weeks after the pa tient has been hospitalized. The presence of increased connective tissue around the central veins and along the sinusoidal walls in the absence of any other significant lesion was unexpected. Although it cannot be definitely stated that this excess connective tissue forms without any con comitant liver cell injury, we were unable, in some biopsy specimens, to observe any such injury with the light microscope. We know that more advanced fibrosis is often associated with hyaline necrosis of parenchymal cells. The exact relationship of the two types of in jury is not clear. Some of the material in the walls of the sinusoids that stains with Mas son's trichrome stain may be material other than collagen; but, on electron microscopic studies, collagen fibers were easily demon strated Iterative tion win it is li . more or 1 tients, i chronic ble for / along t markabl process ; difficult I Altho \ made oi * this stu ment. coholic _ and wli I entry t ' We hoj tients i ' hospital > fact til: plained findingthe fre | the in ) arnylag' j The ' ' lytes a such p cies sh . lowed | enough Last r. not m fat. F ; liver 1 ( larged f the p "'l alway: One withoi the t: \ v si & o LIVER INJURY IN THE ALCOHOLIC 129 i it orically 5% of the total ' ethanol. The IS days. The a minority of me degree of j psy specimens ates that this patients who r disease. The tiich are moroholic hyalin" ring cells may :e. It is probof necrosis, in after repeated ears. not hydropic i affecting the continues to ogress to sclemanswered. It snter the hosily go through lows only the aset of either md niting, ents ..ie more there may be the onset of , as we have ome clinically after the pa- .nective tissue i along the of any other . Although it t this excess it any conire unable, in we any such . We know ten associated chymal cells. types of inaterial in the is with Maslaterial other microscopic isily demon- < , | ' I t ' j | | 1 : I l / I. | [ j 1 I V ( ' J / strated in our material. Although the pro liferative change may undergo slow resolu tion when the patient abstains from ethanol, it is likely that the sclerosing process is more or less progressive in many alcoholic pa tients, leading eventually to symptoms of chronic liver disease. The stimuli responsi ble for the proliferation of cells and collagen along the sinusoids are unknown. The re markable variation in the intensity of the process from one patient to another is most difficult to explain. Although a more detailed report will be made of the clinical and laboratory aspects of this study, a few findings are worthy of com ment. The mortality rate of 67c among al coholic patients who had been on a "spree" and who did not appear to be seriously ill on entry to the hospital was rather surprising. We hope to study further this group of pa tients if a significant number of them are hospitalized again for acute alcoholism. The fact that nearly one-fourth of those who com plained of abdominal pain had laboratory findings suggesting pancreatitis emphasizes the frequency of this complication and shows the importance of determining the serum amylase level on all such patients. The marked disturbances in serum electro lytes and death due to cardiac arrest in one such patient implies that electrolyte deficien cies should be promptly treated and not al lowed to remain until the patient can eat enough to regain electrolyte balance. Lastly, the absence of a palpable liver does not mean the organ may not contain excess fat. Five patients had 1+ to 2+ in their liver biopsy study, but the liver was not en larged on physical examination. However, if the patient is jaundiced, the liver is nearly always enlarged. SUMMARY One hundred chronic alcoholic patients without evidence of cirrhosis were studied at the time of an episode of acute alcoholism. Six patients died, each of a different com plication that was not related to alcoholic liver disease. The frequency of serum electrolyte deficiency was high and should be considered as a possible factor in sudden death. Biopsy specimens of the liver were taken from 45 patients. Microscopically, although fatty change was present in over 907, there was a lack of severe hepatic alterations. More than one-third of the patients had swollen hydropic liver cells that were characterized by organelle disorgani zation when viewed with the electron micro scope. About one-half of the patients had some eentrolobular intrasinusoidal collagen which seemed to have its origin in the spaces of Dissc but often could be found in the intercellular spaces by electron microscopy. REFERENCES 1. Aterman, K.: Observations on the nature of "watery vacuolation"; the response of the liver cell to the intravenous injection of hypertonic saline, Evans blue, dextran, and heparin. Lab. Invest., 7 : 577,1958. 2. Biava, C.: Mallory alcoholic'hynlin: a hereto fore unique lesion of hepatocellular ergastoplasm. Lab. Invest., IS: 301, 1964. 3. Hechter, O.: Intracellular water structure and mechanisms of cellular transport. Ann. N. Y. Acad. Sei., 125 : 625, 1965. 4. Kalant, H.,. Mons, W. and Mahon, M. A.: Acute effects of ethanol on tissue electrolytes in the rat. Canad. J. Physiol. Pharmacol., 44: 1,1966. 5. Lane, B. P. and Lieber, C. S.: Ultrastructural alterations in human hepatocytes following ingestion of ethanol with adequate diets. Amer. J. Path., 4S: 593,1966. 6. Lieber, C. S.: Hepatic and metabolic effects of alcohol. Gastroenterology, 50: 119, 1966. 7. Porta, E. A., Bergman, B. J. and Stein, A. A.: Acute alcoholic hepatitis. Amer. J. Path., 46 : 657,1965. 8. Porta, E. A., Hartroft, W. S. and De La Iglesia, F. A.: Hepatic changes associated with chronic alcoholism in rats. Lab. Invest., 14: 1437, 1965. 9. Schaffner, F., Loebel, A., Weiner, H. A. and Barka, T.: Hepatocellular cytoplasmic changes in acute alcoholic hepatitis. J.A.M.A., 183 : 343, 1963. 10. Svoboda, >. J. and Manning, R. T.: Chronic alcoholism with fatty metamorphosis of the liver; mitochondrial alterations in hepatic cells. Amer. J. Path., 44 645,1964. 24873012