Download - By KARTIN, H. AND PETERS€¦ · obstruction of the common bile duct with stones. Since they werenot explored, the latter possibility cannot be denied. The course of their illnesses

THE LIPIDS OF SERUMANDLIVER IN PATIENTS WITHHEPATIC DISEASES1

By EVELYNB. MAN, BERNARDL. KARTIN, STANLEYH. DURLACHERANDJOHN P. PETERS

(From the Departments of Internal Medicine, Psychiatry and Pathology of the Yale UniversitySchool of Medicine and the Medical Service of the Nezw Haven Hospital, New Haven)

(Received for publication December 27, 1944)

The investigations of dietary fatty livers havethrown into prominent relief the function of theliver in the intermediary metabolism of lipids.The discovery that these disorders may be reme-died or prevented by lipotropic agents may haveimportant therapeutic implications since there issome evidence that the chemical and anatomicaldisturbances of the liver produced by some dis-eases and poisons resemble those found in dietarylivers and that they may be amenable to the actionof these lipotropic agents (1 to 3). These wereamong the reasons that led to an examination ofthe serum lipids in patients with diseases of theliver and pancreas and obstruction of the biliarysystem.. Such studies have been made before buthave usually been confined to the determinationof cholesterol, with or without fractionation, oftenby methods of doubtful accuracy. Because theinterplay of the various lipid components appearsto have significance apart from the concentrationof any single constituent, as often as possible thelipids were fractionated into their 4 major compo-nents: fatty acids, lipid phosphorus, total choles-terol, and free cholesterol. From the values thusobtained, the cholesterol esters and neutral fatwere estimated (4). In addition, samples oflivers, secured at autopsy from some of the pa-tients who had been studied, were analyzed forthe same compounds. Finally, the effects of lipo-tropic agents on the course of the disease in cer-tain of the patients are mentioned.

EXPERIMENTAL

Clinical material. Altogether 174 determina-tions of serum lipids were made on 70 patients.On some occasions, especially in the earlier stud-ies, the lipids were not completely fractionated.The numbers of analyses for the various lipid frac-

1Aided by a grant from the Fluid Research Fund ofthe Yale School of Medicine.

tions, therefore, vary and complete comparisonscannot be made in all instances. All fractionswere, however, measured in 125 observations on54 of the cases. Proteins and protein fractionsalso were determined on 146 occasions.

The cases can be roughly classified under thefollowing headings:

DiagnosisObstruction of the commonbile ductInfectious hepatitisToxic hepatitis

CertainQuestionable

Portal cirrhosisTypicalAtypical

Biliary cirrhosisTypicalAtypical

Acute hemorrhagic pancreatitis

Number of cases15116

33

28

8

1612

62

2

Some of these categories require explanation.Obstruction of the common bile duct includes pa-tients with common duct stones, carcinoma of thepancreas, and other tumors which occluded theduct. In most instances, the diagnosis was con-firmed by operation or post mortem examination.In all, there was evidence of acholia at some time.Studies were made before and after relief of ob-struction in many cases; in a few only after op-eration; in others, who had inoperable conditions,at intervals during their illness.

The group termed infectious hepatitis is some-what heterogeneous. All had febrile illnesses as-sociated with icterus, usually accompanied by en-largement and tenderness of the liver. No evi-dence of exposure to hepatic poisons could beelicited from any of them. One had an ante-cedent history suggesting cholelithiasis of thegall-bladder, and x-rays at that time had re-vealed imperfect filling of the gall-bladder. In 4instances, the liver remained somewhat enlargedafter the acute attacks had subsided. In 1 of

623

MAN, KARTIN, DURLACHER,AND PETERS

them, slight enlargement of the liver had beennoted earlier. It may be argued that these sub-jects had cirrhosis or, since there was evidence ofinitial acholia, that they had painless transitoryobstruction of the common bile duct with stones.Since they were not explored, the latter possibilitycannot be denied. The course of their illnessesand of the icterus, however, was not that usuallyseen with common duct obstruction but was char-acteristic of hepatitis. All but 2 of the cases canbe divided into 2 groups: (1) to which those justmentioned belong, in which icterus was an earlyand striking symptom, with evidence of initialacholia and variable febrile reaction; (2) cases inwhich icterus was a less prominent or relativelylate phenomenon, while signs of hepatic insuffi-ciency dominated the picture. All the fatalitiesoccurred in this last group; but the distinctionbetween the 2 types was not altogether one ofseverity, because 1 case belonging to the secondgroup recovered. One of the 2 exceptional casesdeveloped jaundice and signs of hepatic failureafter a criminal abortion. E. coli was culturedfrom the blood of this patient. Another developedjaundice and a fatal acute nephritis after delivery.Both these patients died. They probably do notbelong in the same category as the other patientswith infectious hepatitis.

The diagnosis of toxic hepatitis also requiresexplanation. In 2, the hepatitis could be defi-nitely attributed to an organic halide compound.In another, it arose from arsenical treatment ofsyphilis. A similar origin was suspected in afourth. The fifth ran an extremely rapid fatalcourse resembling that of a toxic hepatitis. Theonly drug to which she is known to have beenexposed is veronal, but it was impossible to securean adequate history. An infectious etiology can-not be excluded.

The term portal cirrhosis signifies nothing butscarring of the liver that assumes a particulartopographical distribution. Its causes are un-known. It has no more real standing as a dis-ease than did the old renal scrapbasket, chronicinterstitial nephritis, which has been abandoned.It may be produced by infectious or toxic agentsand it may arise from dietary or nutritional de-ficiencies. The differentiation of typical fromatypical cirrhosis, therefore, is somewhat specious.For this discussion, however, it may serve a use-

ful purpose. The term typical is applied here tocases in which icterus was a comparatively in-significant, usually late, manifestation; in whichthe onset and development of the disease wasinsidious; and in which hepatomegaly, spleno-megaly and ascites were prominent relativelyearly. The diagnosis was established at autopsyin 6 of the 16 cases. Of the atypical cases, 3had histories of early jaundice with acholic stools.One of these and 2 additional patients had beensubjected to cholecystectomy. Two of the patientshad histories of intermittent or chronic diarrheafor long periods. In 1, x-ray revealed a gastricdefect suggesting carcinoma. Two had chronicheart failure, 1 associated with cystic bronchiec-tasis. One had adhesive calcific pericarditis forwhich she was subjected to 2 unsuccessful opera-tions. In the last, fibrosis of the pancreas anddilation of the pancreatic ducts were discoveredat autopsy.

The diagnosis of typical biliary cirrhosis wasestablished by autopsy 3 times. The other 3 pa-tients had suffered from recurrent obstructivejaundice for periods of years. In every instancestudies were made in the terminal stages of thedisease. Protocols of 2 of these patients, 44700and B43742, are presented below.

Of the 2 cases of acute hemorrhagic pancreatitis,1 had a fulminating fatal course. The other wasill for 3 weeks. Autopsy revealed, in addition tofat necroses and other evidences of the pancreaticdisease, a fistula -to the gall-bladder, purulentbronchitis, and necrotizing cystitis and pyelitis.

Abstracts of the protocols of selected cases, in-cluding those from which liver specimens weretaken, are given at the end of the paper.

ANALYTICAL METHODS

Blood was drawn when patients were in the postabsorp-tive state. Serum, separated with anaerobic precautions,was analyzed for lipids by methods previously described(4). Total protein, albumin, and globulin were also meas-ured in most sera. Fractions were precipitated accordingto Howe and nitrogen was determined by macro Kjeldahlprocedure. Van den Bergh tests, cephalin flocculation,prothrombin time, and bromsulfalein tests were done inmany instances, icterus indices quite frequently.

Liver samples, after they had been weighed on an ana-lytical balance and minced under alcohol and ether, weresubjected to similar treatment. Attempts were made tosecure samples from the most representative portions ofthe liver, usually the middle of the right lobe. Specimens

624

LIPIDS OF SERUMAND LIVER IN HEPATIC DISEASES

for microscopic analysis were taken from the same re-gion. The middle of the right lobe was chosen because itis usually possible in this region to find larger homogene-ous areas of liver parenchyma. Although it has been sug-gested that there are inherent differences in the composi-tion of the 2 lobes, depending upon the sources from whichthey derive their blood supply (5), the right lobe com-prises the great mass of the liver.

Calculations. Concentrations of neutral fat were esti-mated by factors employed in previous studies (4). Thisinvolves the assumption that the proportions of the vari-ous phospholipids are the same in disease as they are inhealth, a hazardous, but inescapable assumption. Whenlipid phosphorus is high, any departure from these pro-portions will have a significant effect on the estimatedvalue for neutral fat. In normal subjects and in manydiseases the fatty acid combined with cholesterol can beestimated with accuracy from total cholesterol alone be-cause the proportion of esters is so constant. In liver dis-ease, this proportion is disturbed with great frequency.Unless the fractions are measured separately, therefore,estimations of fatty acids combined with cholesterol, andconsequently of neutral fat, are unreliable. This also be-comes an important source of error when total cholesterolis high. For this reason neutral fat has been estimatedonly when cholesterol esters were separately measured.

The normal standards used for the evaluation of dataare those given in the first 2 papers of this series (4, 6).

RESULTS

The concentration of lipids in the serumThe concentrations of the 3 chief lipid fractions,

cholesterol, lipid phosphorus and neutral fat, areshown in Figure 1, together with the diseases ofthe patients from which the sera were obtained.Certain rough correlations are at once evident.

Cholesterol is usually high in patients with ob-structive jaundice, falling when the obstructionis relieved. This has been frequently noted be-fore (7 to 11). The concentration of cholesterolcontinues to rise for some time after obstructionhas been established, reaching extreme heights ifobstruction persists for a considerable period.Examples of the course of the lipids in illustrativecases are presented in Figure 4. The degree ofhypercholesterolemia in this and other series can-not, however, be correlated with the duration ofthe jaundice. This has led certain observers (8)to conclude that ultimately, if the obstruction isnot relieved, cholesterol will again diminish, some-times to subnormal concentrations. The featureswhich determine the course of the hypercholes-terolemia do not seem to be quite so simple asthis. In this series, concentrations greater than

400 mgm. per cent were seen within 4 to 12 daysafter the onset of obstruction; while values below300 mgm. per cent were found from 4 to 9 weeksfrom the onset. In 1 case 862 mgm. per cent wasfound in a patient whose common duct had beenpresumably completely obstructed by a pancreaticcarcinoma for 4 months. Concentrations below300 mgm. per cent in the face of obstruction wereobserved under the following conditions: 1. A pa-tient with carcinoma of the ampulla of Vater de-veloped jaundice with chills in November of 1942.His serum cholesterol, which was 243 mgm. percent on May 28, 1943, rose during a series ofoperations for relief of the condition to 296 mgm.per cent on June 24, to fall off again when theobstruction had been relieved. He appears tohave had remissions of jaundice during this longcourse. 2. In another case, values of 245 and 258mgm. per cent were obtained from a patient witha metastatic carcinoma at a time when obstructionwas probably incomplete; on the day of his death,16 days later, when the icterus index had risento 150, his serum cholesterol was 530 mgm. percent (see Figure 4, Case B36866). 3. A concen-tration of 222 mgm. per cent was found 3 daysbefore death in a patient with primary carcinomaof the liver. The cholesterol was evidently fallingsince it had been 252 mgm. per cent a week ear-lier. 4. The serum cholesterol was only 207 mgm.per cent in a woman of 62 who had icterus for3 months. The stools had been clay-colored untilabout a week before the determination of lipids.Although chemical tests for bile were reportednegative in the hospital, the stools were describedas brown or greenish. The icterus index alsodiminished. The obstruction had, therefore, prob-ably been relieved. On Figure 1 appears a cho-lesterol of 485 mgm. per cent after operative reliefof obstruction. In the 15 days after operation,this had fallen from 657 mgm. per cent (still ear-lier it had been 862 mgm. per cent). The patientdid not remain long enough to permit further ob-observations, but it may be presumed that the cho-lesterol continued to fall (see Figure 4, CaseB49208). Biliary obstruction, therefore, causescholesterol to rise strikingly and progressively.In the few instances in which hypercholesterolemiawas not observed, obstruction appears to havebeen incomplete or else had been associated with

625


biliary obstruction

iofectious hepatitistoxic bepatil isporlal cirrbosisbiliary cirrhosisacute pancreatitis

El active El relieved0 active 0 recoyeredE active mrecoveredU typical Uaty pical0 typical 1 atypical

300 400 500 600 700 800 900Cholesterol, mg.per cent.

15 20 25 30 40 50 60Lipid pbosphorus, MSg.per cent

LLzK. 0.PE3

10 15 2.b 25 30Neutral tat, roEq.per liter

FIG. 1. CHOLESTEROL, LIPID PHOSPHORUSAND NEUTRAL FAT IN THE SERUM OF PATIENTS WITHVARous TYPES OF LivER DISEASE

20

I0

0 _

25

a 200

._

> 15L0)

oO.0

I Il1

05

i.

I.

I.

L0

L0

-0E (Z 20

15

10

5

F

0'

626


extreme cachexia or with extensive destructionor degeneration of the parenchyma of the liver.

In general, cholesterol was higher in the pa-tients with infectious hepatitis than it was in thosewith toxic hepatitis. There were, however, dis-tinct exceptions to this rule. In 2 of the caseswith infectious hepatitis, values below 100 mgm.per cent were observed. One of these accom-panied an acute postpartum nephritis. True hy-percholesterolemia, values greater than 300 mgm.per cent, was encountered only twice, in conjunc-tion with acholic stools. It might have been notedmore frequently if patients had been seen earlierin their illnesses. Illustrative cases are shownin Figure 5. In 3 cases with somewhat similarcourses, studied only after icterus had begun toimprove, cholesterols lay within normal limits. Atleast 1 of these had acholic stools earlier. Of the8 patients with normal or high cholesterols, only

5c

4c0

0~

_ 3v£;3

.0R2.

6

4D 40

.4_0_9-CL

3..00

03

0

0 * ' Ao

00 olite

1 died; her icterus was a terminal feature accom-panying pneumonia. Of the 7 with hypocholes-terolemia (cholesterol less than 150 mgm. percent), on the other hand, only 1 survived. It wasnoted that this patient had icterus when he was ad-mitted to the hospital immediately after he hadbeen scalded. How long the jaundice had persistedis as uncertain as its origin. As it subsided theserum cholesterol rose from 66 to 173 mgm. percent. This is the only uncomplicated infectiouscase in which a cholesterol lower than 100 mgm.per cent was observed. On the other hand, con-centrations below 100 mgm. were encountered in4 patients in the terminal stages of hepatitis causedby drugs and chemicals. In these, jaundice was alate phenomenon overshadowed by symptoms ofhepatic insufficiency. A diagnosis of arsphenaminehepatitis and encephalitis was made on a negresswho, after antisyphilitic treatment, was admitted

xc~~~x x . xO * x * x

0

0+ 0

000 o

oh~~~~~~~yO.. C#* x

*m A 0* ** #4-ISA-iVo A+*h A* X

A

xx

0* *

KK

'Cx IC

*

100 200 300 400 500 600 710 800Cbolesterol, f,a per cent

Biliary ob.strcactionactive *,relieved 0; iofectioas bepatits,adive &,recovered &;toxic hopatitis,active m, recovered Q; portal cirrbosis,typical., atypical, o;biliary cirrhosis, typical x, atypical +.

FIG. 2. THE RELATION OF CHOLESTEROLTO ALBUMIN AND TO GLOBULIN INTHE SERUMOF PATIENTS WITH VARIous TYPES OF LIvR DISEASE

627

MAN, KARTIN, DURLACHER, AND PETERS

to the hospital with a temperature of 104.40 F. ina semistuporous condition, with a number ofanomalous features, including x-ray evidence ofa tumor in the sella turcica. Serum cholesterolwhen she was improving was 318 mgm. per cent,the icterus index was 50. Despite a distortion ofthe interrelationships between the lipid fractionsindicating a disorder of the liver, the cephalin floc-culation test was negative. The diagnosis in thiscase, therefore, seexms doubtful. The data on thewhole suggest that hypercholesterolemia in hepa-titis is an indication that the inflammatory condi-tion has involved the bile ducts, causing obstruc-tion. Whether this is a distinctive feature of cer-tain types or stages of hepatitis is a suitable sub-ject for inquiry.

In patients with typical portal cirrhosis, choles-

terol was usually normal or low. In the cases withatypical portal cirrhosis, it lay in the same generalrange, although the average concentrations weresomewhat higher. This difference depends largelyon the inclusion in this group of certain cases.The patient with calcified adhesive pericarditis,despite intractable ascites, hepatomegaly andsplenomegaly, had a normal serum lipid patternthroughout. This case accounts for 7 of the cho-lesterols above 160 mgm. per cent. Two othersare contributed by another patient with heart fail-ure. Three are from the patient who had dilatedpancreatic ducts, and 1 from 1 of the patients withprevious cholecystectomy. In 1 instance, the cho-lesterol rose from 129 to 179 mgm. per cent undertherapy. Among the typical cases, 3 of the cho-lesterols above 160 mgm. per cent are from a pa-

*A

A

A

A

AOa0AVOS vaO *a

0 .t1XA4 x

x xA

x

'00 200 300 4b0 500 600Cholesterol,m g. per cent

tx

A

A

o AX40FS -\

x x.

ox a

5 '10 15 20 25 30 40 60Lipid phosphoras, ms. per cent

FIG. 3. THE RELATION OF NEUTRALFAT TO CHOLESTEROLAND TO LIDPHOSPHORUSIN THE SERUM OF PATIENTS WITH VAsous TYPES OFLIvER DISEASE

For interpretation.of the symbols, see Figure 2.

30.

L20La,

.-

Lo. lo

w2

;0 0-30L

4-

z20~

10

* *x xx

K%

700 800

xXXX X

x

628

x


tient who had successive values of 189, 185, 161,and 125 mgm. per cent. This probably illustratesthe usual course of the disease. In 3 cases, how-ever, the cholesterol rose in the course of the dis-ease. In 1 instance a rise from 143 to 225 mgm.per cent marked a period of improvement in whichascites, previously requiring frequent paracentesis,was controlled by diet and diuretics. A rise from154 to 208'mgm. per cent in another case attendedsymptomatic improvement. The third, when firstseen, extremely malnourished, had a cholesterol ofonly 64 mgm. per cent. This rose to 112 and thento 229 mgm. per cent. Despite the fact that thiswoman died of a ruptured esophageal varix andsuffered, besides, from heart failure, she neverdeveloped any considerable ascites. At autopsy,the liver was not characteristic of portal cirrhosis.Instead, there was a focal fibrosis, while the cellsand architecture of most of the organ were wellpreserved.

The figures in biliary cirrhosis above 300 mgm.per cent are all from 1 patient who had also dia-betes and necrobiosis (see Protocol 44700). Hercourse will be discussed in detail elsewhere. Thepatient with a cholesterol of 285 mgm. per centalso had a long-standing history of biliary stasis.Another patient with a similar condition had cho-lesterols of 167 and 177 mgm. per cent. It isknown that cholesterol was 230 and 254 mgm. percent 11/2 years and 1 year earlier, respectively.Another with an acute history had a cholesterolof 207 mgm. per cent 2 weeks before death. The2 cases with cholesterol below 150 mgm. per centdied after prolonged courses, in cachetic states, 1with extreme recurrent ascites. The condition ofthe atypical cases can be judged from the sketchesof their records given in the protocols. It maybe inferred that recurrent or partial biliary ob-struction gives rise to hypercholesterolemia whichdiminishes as hepatic destruction progresses, ulti-mately giving way to terminal hypocholestero-lemia, either as a result -of the extreme malnutri-tion or the loss of liver parenchyma.

In this connection, although serum proteins mayseem irrelevant to a discussion of lipids, they maythrow some light on the effect of malnutrition. Itmust be recognized, however, that they are cri-teria of nutrition in a special sense only. Muchemphasis has been placed upon hyperglobulinemiain cirrhosis. In the present series (see Figure

2A), serum globulin was below 3 per cent 7 timesin typical portal cirrhosis, 4 per cent or above 14times, and 3 to 4 per cent 9 times. Correspondingfigures for atypical cirrhosis were 10 below 3 percent, 6 above 4 per cent, and 9 between 3 and 4per cent. In addition, globulin exceeded 4 percent 5 times in biliary cirrhosis, 3 times in activeacute hepatitis, 3 times after recovery, and once inbiliary obstruction. Hyperglobulinemia is not,therefore, a reliable diagnostic sign of portal cir-rhosis. Serum albumin is reduced (less than 4per cent) in most patients with liver disease (seeFigure 2B). Although the lowest figures coincidewith low cholesterol, there is no direct correlationbetween the 2 variables, nor is albumin consis-tently depressed in any particular disease. It isprobably determined rather by the nutritive stateof the patient.

Although attention in the past has been confinedchiefly to cholesterol, phospholipid (12) is quiteas much disturbed in diseases of the liver. Ingeneral, the 2 follow one another, rising and fall-ing together. Figure 1 suggests that diseases are,perhaps, a little more sharply differentiated intocategories by means of phospholipids than by cho-lesterol. The parallelism between phospholipidsand cholesterol is well illustrated in Figures 4and 5.

About the concentration of neutral fat in theserum, both in diseases and in experimental dis-orders of the liver, little can be found in the litera-ture (12). Statements about this lipid compo-nent are based chiefly upon determinations of totallipids or total fatty acids. The latter are peculiarlyunreliable measures of neutral fat in conditionsin which, as in liver disease, the concentrations ofphospholipids and cholesterol esters are greatlyaltered. From Figure 1, it appears that concen-trations of neutral fat are not high in the serumof most patients with liver disease. In only 2instances did neutral fat rise to the maximumencountered in normal subjects (4). Neutral fatis, however, high in biliary obstruction and in in-fectious hepatitis, as compared with cirrhosis andtoxic hepatitis in which it is usually comparativelylow. From Figures 4 and 5, it will be seen thatin both biliary obstruction and in some cases ofinfectious hepatitis neutral fat rises, to fall againas the obstruction is relieved or the icterus sub-sides. Neutral fat is, of course, estimated on the

629

630 MAX., LARTIN, DURLACHER., AND PETERS

9001x.

5 12610 83260 8 36866 849208' 800-

Q .

0 0

Q.

40400-

301300.

' 20 -200

0~~~~~~~~~~~~~00t ___

C510o14 27 I 6ZZ 2 25 12713tDay of dise8se

FIG. 4. THE COURSEOF THE SEuum LipiDs IN PATIENTS WITH BILIARY OBSTRUCTION

For interpretation, see text.

*350.o ~~10225 B43698 IB 500888 0 I,o

07251 250 H

+zozo

L.15 J150-

o"0L 00I ,c- I 0 V

9 17 24

12430 117 643 5 12 21 24 40u ay *f uu@a3eo

FIG. 5. THE COURSEOF THE SERUmLIPIDS IN PATIENTS WITH INFECTIOUS HEPATITIS



assumption that the proportions of phospholipidsremain constant. If, as lipid phosphorus rose, theincrements consisted entirely of phospholipids with2 fatty acids, while those with 1 fatty acid dimin-ished, fatty acids could increase out of proportionto phosphorus without addition of neutral fat.Even if it were assumed that every phosphoruswere combined with 2 equivalents of fatty acid, theneutral fat figures in this study would not be sig-nificantly altered. If, on the other hand, thephospholipids with only 1 fatty acid increaseddisproportionately, neutral fat would be muchgreater than Figure 1 indicates in patients withhigh lipid phosphorus.

Some idea of the effect this would have can begained from Figure 3 in which neutral fat is com-pared with lipid phosphorus and cholesterol.There is a general tendency for all 3 fractions tovary together, but the correlation is not consistent.The most striking exception is the patient (44700)with chronic biliary tract disease, diabetes, andnecrobiosis who is responsible for all the diagonalcrosses above 300 mgm. per cent of cholesteroland 15 mgm. per cent of lipid phosphorus. Thereis reason to believe that neutral fat in this casewas underestimated, since on 1 occasion calcula-tion yielded a value of - 8.5 m.eq. per liter offatty acid. The general tenor of the data wouldnot be altered if the neutral fat figures of this pa-tient and those with obstructive jaundice wereraised. It would still remain true that elevationof neutral fat is confined almost entirely to pa-tients with obstructive jaundice and infectioushepatitis. Among the latter it is those with earlyintense jaundice that are chiefly affected. Thehighest figures of all, however, were observed inthe patient with postabortional E. coli septicemia.From Figures 1, 3, 4 and 5 it can be seen thatneutral fat rose proportionally more in hepatitisthan it did in obstructive jaundice. The only caseof cirrhosis with neutral fat greater than 10 m.eq.per liter was the patient who had atypical portalcirrhosis with dilated pancreatic ducts.

The interrelationships of the serum lipids

It has been reported repeatedly that the ratioof free to total cholesterol is frequently elevated inhepatic disease (8, 9). Published data have, how-ever, been obtained almost entirely by colorimetric

procedures that yield higher and more variableratios than does the digitonin method employedfor the present study. For example, one worker(8) gave as the normal range of the cholesterolratio 0.30 to 0.50. On the other hand, in a largeseries of normal subjects several workers ( 13 to 15,4) found by digitonin methods that the ratio, freecholesterol: total cholesterol, never departed fromthe narrow limits of 0.24 to 0.32. Furthermore,this relation was not disturbed in diseases of thethyroid gland (6), nephritis (16), and a variety-of other disorders, in spite of the fact that theseconditions greatly affected the concentration ofcholesterol.

In Figure 6, total cholesterol and cholesterolesters of the patients of our series are compared.In only a few of the patients with cirrhosis is theratio of esters to total cholesterol normal; in theothers, it is invariably low. The grouping ofpoints on the chart indicates that the distortionarises largely from a deficiency of ester forms. Inonly a few instances, when total cholesterol isgreatly increased, does the concentration of estersexceed the upper normal limits; sometimes estersare almost extinguished. In patients with biliaryobstruction and extreme hypercholesterolemia bothfractions are sometimes elevated. In this condi-tion and in acute hepatitis and biliary cirrhosis,the ratios are particularly reduced. From Figures4 and 5, it can be seen that cholesterol esters fallat the onset of obstruction or acute hepatitis, evenbefore total cholesterol rises. Conversely, duringrecovery, esters rise and the ratio is restored be-fore the total cholesterol falls. This is particu-larly evident in B49208, Figure 4. Before opera-tion, ester cholesterol was only 80 and 85 mgm.per cent, with total cholesterol of 782 and 572mgm. per cent. Normally, with cholesterol estersas low as this, total cholesterol could not exceed118 and 125 mgm. per cent. Fifteen days afteroperation, with relief of the obstruction, the estershad risen to 236 mgm. per cent when the total cho-lesterol had fallen to only 485 mgm. per cent. Itis this tendency for the ratio to approach normalbefore the concentration does that accounts forthe obstruction and hyperlipemic acute hepatitiscases that have normal ratios in Figure 6.

In those patients who develop hypercholestero-lemia, free cholesterol rises strikingly at the onset,as esters diminish. At the height of the hyper-

631

6MAN, KARTIN, DURLACHER,AND PETERS

x

x

xx)x

Total cholesterol, mrg. per centFIG. 6. THE RELATION OF CHOLESTEROLESTERS TO ToTAL CHOLESTEROLIT

PATIENTS WITH LIVER DISEASEThe diagonal lines defining ratios of 0.76 to 0.68 bound the zone of normal

variation. For interpretation of the symbols, see Figure 2.Solid circles = typical, open circles = atypical portal cirrhosis. Diagonal

crosses = typical, vertical crosses = atypical biliary cirrhosis. Solid tri-angles = active, open triangles = recovered infectious hepatitis. Solidsquares = active, open squares = recovered toxic hepatitis. Solid diamonds= active, open diamonds = relieved biliary obstruction.

lipemia, free cholesterol, which is normally a smallfraction of the total, may exceed the normal valuefor total cholesterol. In 1 instance, of a total of730 mgm. per cent, 530 mgm. per cent was in thefree form. In patients with extreme hypocholes-terolemia, on the other hand, free cholesterol,though high in proportion to the esters, may liebelow the normal limits.

Just as the ratio of free cholesterol to totalcholesterol is more constant than the concentra-tion of total cholesterol or either of its fractionsin the serum of normal persons, so the proportionof phospholipid to cholesterol is less variable thanis the concentration of either phospholipid or cho-lesterol (6). The ratio, cholesterol: lipid phos-phorus, however, is not constant, as is the ratioof free to total cholesterol, but increases with thecholesterol in a manner that has been described inan earlier paper (6).

In liver disease, lipid phosphorus, in general,parallels total cholesterol. The normal relations

between the 2 are not, however, preserved. Inmost cases, the ratio of cholesterol to lipid phos-phorus is lower than normal. Especially in thehyperlipemic cases, this distortion implies thatphospholipids rise proportionally more than totalcholesterol. In a review of the normal relationbetween these 2 lipid fractions, it was discoveredthat when lipid phosphorus is plotted directlyagainst cholesterol, the distribution of points above100 mgm. per cent of cholesterol is indistinguish-able from a straight line. The distribution ofpoints in thyroid diseases coincides with the dis-tribution in normal persons or a rectilinear exten-sion of this distribution. The best straight linecalculated from 856 observations in which choles-terol varied from 108 to 911 mgm. per cent, wasdefined by the equation:

lipid phosphorus = 0.0294 cholesterol + 3.62,S.D. = 1.043 mgm. per cent

This line meets the y axis at a point indicating that

4-I

0)IL)

L.

e1~

04-C.)Z

L IC

II)LI I

.-632


cholesterol should disappear when the lipid phos-phorus falls to 3.62 mgm. per cent. Actually cho-lesterol does not disappear, but falls at a slowerrate than lipid phosphorus when the latter ap-proaches this concentration. The curvilinear rela-tion of the ratio, cholesterol:lipid phosphorus, tocholesterol, defined in an earlier paper (6) arisesfrom the fact that the line defining the relations ofthese 2 fractions does not pass through the origin.By means of the linear relation defined by the newequation, it is possible to construct a graph some-what like that used for the evaluation of the cho-lesterol ratio. This has been done in Figure 7.The radial lines define ratios of lipid phosphorus- 3.62 to total cholesterol. The lines defining thenormal limits, with slopes of 0.0205 and 0.0383 re-spectively have not as precise significance as thosedefining the limits of the cholesterol ratio, becausethe latter is more constant than the ratio of lipidphosphorus to cholesterol. These lines are drawnto include twice the mean deviation of the ratioat 200 mgm. per cent of cholesterol, which is about+ 0.009. The deviations are smaller at low than

at high concentrations of cholesterol, which justi-fies, to some extent, the use of radial lines.

It is evident from Figure 7 that the ratio, lipidphosphorus: cholesterol, is unmistakably larger

than normal in a large proportion of patients withliver disease. The most striking elevations arefound in subjects with biliary obstruction andacute hepatitis. Certain exceptions are to befound. Among these are patients on the road torecovery from obstruction or hepatitis. This ratio,like the cholesterol ratio, seems to return to nor-mal more rapidly than the concentrations of lipidsdo. Other exceptions are subjects with choles-terol below 100 mgm. per cent. To these, for rea-sons given above, the chart does not properly ap-ply. The cirrhosis cases, for the most part, havenormal ratios.

In Figure 8, the ratio of free to total choles-terol is compared with that of lipid phosphorus tocholesterol. The horizontal and vertical lines de-fine the normal limits of the 2 ratios. Area 5,therefore, presents the normal zone. In this area,lie patients who have recovered from acute hepa-titis or have been relieved of biliary obstructionand a few cases of typical or atypical cirrhosis.There are no points in Areas 1, 4, and 7 whichrepresent cholesterol ratios less than 0.24. Area2 is also clear. This means that the ratio of lipidphosphorus to cholesterol is not altered if thecholesterol ratio is normal. Biliary obstructionand acute hepatitis tend to congregate in Area 3,

X x

** 'xX x

X XX

x*0.020

wto 200 300 400 500 600 700 800 900

Chboeste rol,wmg. pepr centFIG. 7. THE RELATION OF LiPm PHOSPHORUSTO CHOLESTEROL

The diagonal lines bound the zone of normal variation of the ratio of lipidphosphorus-3.62 to cholesterol. For interpretation of the symbols, see Fig-ure 6.

701

4 60c0

0

E 40'9

3-0° 300.00.20l.J lo. A

M

633


(free clolesterol: Total c7olesterol) x 100FIG. 8. THE RELATION OF THE LIPID PHOSPHORUS:CHOLESTEROLRATIO TO THE

CHOLESTEROLRATIO IN PATIENTS WITH LIVER DISEASEThe vertical and horizontal parallel lines define the normal limits of variation to

these ratios. For interpretation of the symbols, see Figure 6.

where both ratios are high. In this same areaappear some subjects with atypical cirrhosis. Asacute hepatitis and obstruction improve they fre-quently move through Area 6; that is, the lipidphosphorus: cholesterol ratio is restored earlierthan the cholesterol ratio. In this area are alsofound most of the observations on typical andatypical cirrhosis. The few points in Areas 8 and9 are almost all derived from patients with ex-treme hypolipemia.

Lipotropic agentsCholine chloride, in doses varying from 10 to

25 grams a day, was given for varying periods to3 patients presumably suffering from portal cir-rhosis. In none of them did it have any demon-strable effect upon the clinical course nor upon theserum lipids. Two of these subjects developeddiarrhea when they took choline. This reaction

occurred in 4 of the 8 patients who received thiscompound, but in 3 the diarrhea was transitory.The compound was also given to 3 patients withacute hepatitis. All three recovered, but it is im-possible to say that their clinical courses were in-fluenced by the choline. Another patient (TablesI and II, 28639) was given 10 grams subcutane-ously on the second and third days preceding herdeath.

The course of the lipids in the eighth patientwho received choline is illustrated in Figure 9This is the patient with diabetes, necrobiosis, andchronic icterus with partial biliary obstructionand cholangitis (Protocol 44700). The serumlipids were greatly increased throughout hercourse, with strikingly high ratios of free to totalcholesterol and of lipid phosphorus to cholesterol.The lipid picture resembled that of obstructivejaundice, with the exception that neutral fat was

634

LIPIDS OF SERUMAND LIVER IN HEPATIC DISEASES63

not elevated. Nevertheless, biliary obstructionwas never complete. From the 96th to the 113thday in the hospital, she was given 1 gram of cho-line chloride daily. Under this treatment choles-terol and lipid phosphorus fell and continued todecline in the next 2 weeks without choline. Atthe same time, the lipid ratios fell towards normal.Choline was again resumed on the 128th day andwas taken irregularly until the 167th day. On the186th day., the -serum lipids were almost normal.Subsequently they rose again, falling only shortlybefore death. No apparent improvement in theclinical condition accompanied the reduction ofthe serum lipids. Whether the latter can be at-tributed to the choline, it is impossible to state.

One of the patients with portal cirrhosis whohad reacted to choline with diarrhea was given 3grams of methionine daily for a month withoutany demonstrable change of clinical signs andsymptoms or of serum lipids. Another patientwho developed nephritis and hepatic enlargementwith icterus during the puerperitun, folliowing

erysipelas, was given 5 grams of methionine dailyfor 4 days shortly before death.

Lipocaic was given to a patient who developedfatty diarrhea and multiple vitamnin deficienciesafter a subtotal pancreatectomy for pancreatic car-cinoma that had completely obstructed the com-mon bile duct. Although she took as large dosesof the material as she was able (up to 12 capsulesa day), together with pancreatin and vitamins,she went rapidly and uninterruptedly downhilluntil her death. During her postoperative course,her lipids, which had been depressed, rose to thenormnal range despite exctreme wasting. The lipidratios, however., remained abnormal throughout.

Liver analysesResults of the liver analyses are shown in Ta-

ble I. For comparison in the same table are analy-ses of livers from 3 patients presumably free fromhepatic disease and other disorders known to af-fect serum lipids., and a summary of similar analy-ses published by Ralli (17). The milliequivalents

TABLE I

Lipids of the liver

CholesterolCase number teursdath Lieigr Solids Fatty* Lipid _______ _______Neutraltterdeathweightadids phosphorus.-____ ___ fatTotal Free Ester Free:Ester

gasgrams Per m.eq. Per grams Per grams per grams per grams per m.eq. pergas kgm. kgm. kgm. kgm. kgm. kgm. kgm.

Liver disease

B43797 5.5 319 532.0 0.963 2.40 1.27 1.13 0.53 473.2B36812 16.0 4675 535 1327.0 0.729 2.05 1.43 0.62 0.70 1283.1

28639 12.5 2325 197 63.9 1.207 2.63 2.51 0.12 0.95 -6.5A57138 22.5 2300 532 1036.5 0.642 6.76 2.96 3.80 0.44 988.4

19694 21.0 650 228 53.0 0.832 2.77 2.80 0.00 1.00 4.7A47621 16.0 1525 225 190.6 0.690 3.09 2.57 0.52 0.83 149.1B43658 6.5 .1050 197 55.0 0.803 2.64 .2.05 0.59 0.78 6.9A45460 7.5 850 228 95.8 1.001 3.49 2.79 0.70 0.80 35.9B43742 16.5 184 61.4 0.713 3.03 2.70 0.33 0.89 19.1

Normal

6269 19.0 2045 297 219.9 1.139 3.40 2.09 1.31 0.62 150.36270 8.0 1650 274 188.5 1.180 3.74 2.59 1.15 0.69 116.96366 14.0 1450 284 183.7 1.284 2.97 2.26 0.71 0.76 107.1

Ralli's (17) normals

Minimum 59.1 0.560 2.40 1.97 0.43 0.82 25.5Maximum 264.0 1.167 3.88 2.17 1.71 0.56 191.8Average 128.0 10.800 2.83 2.04 0.79 0.72 79.5

* Concentrations of all lipid components are referred to weight of fresh tissue. t In termns of fatty acid.

635


TABLE II

Serum lipids of patients with liver analyses

Serum protein Cholesterol LipidCase __________Fatty Lipid _____-_________phosphorus Neutral

number Date acid phosphorus -3.6:chol- fat*Total Albumin Total Free Free:Total esterol

per cen.t per cent m.eq. Per mgm. mgm. mgm. m.eq. Perliter per cent per cent per cent liter

1943B43797 June 21 5.15 3.08 26.2 27.3 350 320 0.91 0.049 9.6

July 20 4.94 2.81 9.5 7.2 138 46 0.33 0.068 2.9

B36812 Jan. 21 8.19 2.93 18.4 12.3 166 106 0.64 0.052 9.7Feb. 1 6.22 2.72 18.2 11.1 155 108 0.70 0.049 10.5Feb. 3 21.8 10.2 234 133 0.57 0.028 13.3

28639 May 29 ,5.00 3.30 37.5 14.5 132 108 0.82 0.082 28.5May 30 4.64 3.28 36.2 133 120 0.90May 31 5.03 3.01 34.2 16.7 157 127 0.81 0.083 23.7June 1 33.6 15.8 127 120 0.95 0.096 24.2

1944A57138 Nov. 18 5.46 3.92 10.4 7.9 91 47 0.52 0.042 5.0

194319694 Mar. 5 6.62 3.37 7.1 4.2 78 52 0.67 0.007 4.0

A47621 Jan. 12 6.14 2.03 6.5 5.6 79 42 0.53 0.024 3.2Jan. 26 8.39 2.18 7.5 7.7 122 49 0.40 0.033 3.0

B43658 June 4 5.85 2.91 6.1 5.5 104 42 0.40 0.018 2.9

1944A45460 Nov. 14 6.25 1.77 7.4 6.1 139 63 0.45 0.018 1.9

1943B43742 Apr. 3 7.08 1.90 10.3 7.7 93 78 0.84 0.044 5.4

* In terms of fatty acid.

of fatty acid are calculated from this worker'sfigures, which are given in grams, by assumingthat the fatty acids were composed of equal partsof stearic, oleic, and palmitic acids. The valuesfor lipid phosphorus have been recalculated bymeans of the factor by which she converted lipidphosphorus to grams per cent of phospholipid(18). From these figures, neutral fat has beenestimated by the formula used for the estimationof neutral fat from our own analyses. Her figuresfor neutral fat were not employed because theywould not be comparable to ours. The results ofanalyses of sera from the patients whose liverswere analyzed are shown in Table II. Abstractsof protocols of the cases are given at the end ofthe paper. The three "normal specimens" of liverwere taken from the livers of: (1) a man of 50years who committed suicide by taking cyanide,(2) a-man of 31 years who died 6 hours after acoronary occlusion, (3) a man of 29 years whodied from a stab wound of the heart.

Interpretation of analyses of tissues from hu-man autopsy material presents several knottyproblems. In the first place, it is not possible tosecure the specimens in such a fresh state thatautolytic changes can be excluded. Fortunately,lipids appear to be less subject to autolysis thancarbohydrate and proteins are. Ralli (17) foundthat total cholesterol and fatty acids did notchange, though lipid phosphorus decreased slightlyin a dog's liver kept in the refrigerator for 23hours. The differences in the "normal" analysesin Table I are not related to the interval be-tween death and examination. It is difficult toobtain suitable normal control specimens. The 3analyses in Table I are not sufficient to establishthe range of normal variation. In composition,however, all three are relatively uniform. Ralli's(17) figures for all constituents are far morevariable, and her series of analyses is far largerthan ours. It may be that her material is morerepresentative or that ours is more fortunately

636


selected. In the general magnitude of all frac-tions, the two sets of analyses agree. Many othershave reported liver analyses that differ widelyfrom hers (17, 19) and ours, presumably becauseof the analytical techniques employed (20 to 23).It may be assumed that the structure of the liversof adults is relatively uniform, that the propor-tions of liver cells to connective tissue and otherelements are approximately the same. It is quiteotherwise in diseased livers. In cirrhosis, for ex-ample, the major part of a sample may consist,not of liver cells, but of undifferentiated connec-tive tissue. The proportion of liver cells to con-nective tissue cannot be estimated, but it must

have a great influence upon the chemical composi-tion of the specimens taken for analysis.

There is no feature common to all the patho-logical livers; but the most frequent abnormalityis a relatively low concentration of phospholipids.The first, second, and fourth cases are distin-guished from all the others by the high proportionsof solids they contain. This is associated with thepresence of excessive quantities of fatty acids, ofwhich an unusually large fraction belongs to neu-tral fats. In B43797 and B36812 pancreatic tissuewas almost entirely destroyed, though neither haddiabetes. These presumably are examples of truedietary fatty livers comparable to those that de-

FIG. 9. THE COURISE OF THE SERUM LiE5Ds OF A PATIENT WITH PARTIAL BILIARY OBSTRUCTION, CIR-RHOSIS, DIABETES AND NECROBIOSIS AND THE REACTION OF THE LIDS TO CHOLINE


637


velop in depancreatized animals treated with in-sulin. Phospholipids are reduced in both. Totalcholesterol is also low, the deficiency affectingchiefly the free cholesterol fraction. B43797, whenfirst observed, had high serum cholesterol, phos-pholipid, and moderately-elevated neutral fat inher serum, all of which fell before death, possiblybecause she had a biliary fistula. Cholesterol,lipid phosphorus, and neutral fat rose steadily inthe serum of B36812, and the cholesterol ratiowas high throughout. An erroneous diagnosis ofportal cirrhosis was made because of an alcoholichistory, a large liver, ascites, and high serumglobulin. The serum lipid pattern, however, dif-fers widely from that usually seen in portal cir-rhosis, which is characterized by low concentra-tions of lipids and little distortion of the choles-terol ratio.

In some respects A57138 resembles the 2 pre-ceding cases: total solids, fatty acids, and neutralfat are extremely high, lipid phosphorus greatlyreduced, but cholesterol is enormously increasedby accumulation of esters. This correspondsclosely to the pattern of lipids described in thelivers of animals with purely dietary fatty livers.Certainly from his history, the patient was en-titled to such a condition. It is, however, impos-sible to evaluate or to discount the poisonous ef-fects of the rubbing alcohol and the renal tubulardestruction.

All the other livers of the series contain lessthan the normal complement of solids and in allbut one, A47621, the deficiency can be attributedpartly to a reduction of the concentration of fattyacids in the organ. Fatty infiltration in a grosssense is not, therefore, a consistent feature of he-patic disease. Case 28639 is peculiarly intriguing.Whether as a result of some unknown drug orpuerperal sepsis, the patient had jaundice andother evidences of profound injury to the liver, isnot known, and post mortem examination revealedno gross nor microscopic disturbance of hepaticstructure. The chemical composition of the organis profoundly altered, as might have been sur-mised from the serum analyses. The most strik-ing change in both serum (last observation) andliver is the almost complete extinction of choles-terol esters and the relative preservation of thephospholipids. In fact, serum lipid phosphorus ishigh. This patient received choline for the last 3

days of her life. On the other hand, no neutral fatcould be detected in the liver, although its con-centration in the serum is extremely high, as if itsutilization had been blocked. The calculation ofneutral fat is, of course, subject to error becausethe composition of the phospholipids is unknown.If these contained an unusually large proportionof compounds containing only a single fatty acid,the neutral fat would be elevated somewhat.

The features common to this liver and that of19694, who died of a toxic hepatitis, are the ab-sence of cholesterol esters and the deficiency ofneutral fat. Phospholipid is also low in 19694.All the lipid components in his serum were alsoreduced. These distinctions may arise from dif-ferences in the duration of the disease. Case28639 died only a week after the onset of her ill-ness, while 19694 had been exposed for months tothe poison which killed him and had jaundice forat least 3 weeks before his death.

The 4 remaining patients died with cirrhosis ofvarious degrees and kinds. The low solids andlipids cannot, therefore, be attributed altogetherto changes in composition of the liver cells; re-placement of these cells by connective tissue mustbe a contributory factor. The lipid constituentsare not, however, uniformly affected. The con-centrations of total cholesterol are quite well pre-served. If the reductions of lipid phosphorus andfatty acids derive from replacement of liver cellsby connective tissue, the liver cells must containexcessive amounts of cholesterol. Both morpho-logically and chemically the livers differed fromone another. The most classical picture of portalcirrhosis was seen in B43658. With the exceptionof neutral fat, which is disproportionately low, theinterrelationships of the lipid components in hisliver are relatively well retained. The pattern oflipids in the liver of A45460 is very much thesame, except for larger proportions of cholesteroland neutral fat. Both these patients had the char-acteristic hypolipemia of portal cirrhosis.

When allowance is made for the fibrous tissuein the liver of A47621, neutral fat and cholesterolappear to be elevated, while lipid phosphorus isdepressed. Much central necrosis was found inthis liver. The cholesterol ratio in both serumand liver was greatly elevated in B43742, whoseliver showed a more diffuse fibrosis than the oth-ers with some evidence of bile-stasis.

638


DISCUSSION

Although the amount of work involved in theseanalyses is large, the definite conclusions that canbe drawn from them are comparatively meager.The material studied was so large and varied thatno single condition was exhaustively examined.The investigation can be regarded only as a pre-liminary survey. The results are highly sugges-tive. They raise the hope that possibly such par-titions of lipids may prove a more accurate meansthan the rather empirical tests thus far employedfor the differentiation of hepatic disease and dis-orders. They may give more information thanmeasurements of cholesterol and its fractions, towhich attention has been largely confined. This isto be expected from experimental studies of die-tary fatty livers and toxic hepatitis, in which ithas been found that the formation of a balancedmixture of phospholipids is essential to the properfunction of the liver and the orderly metabolismof fat (3).

Obstruction of the biliary passages appears tolead regularly to the accumulation in the serum ofcholesterol, of phospholipids and, to a lesser ex-tent, of neutral fat. The character of the choles-terol is also altered. Instead of consisting chieflyof esters, free forms predominate; esters are notonly relatively, but absolutely, reduced. Avail-able information indicates that cholesterol estersoccur in the cells of only those organs or tissuesin which there is active sterol metabolism: theliver, testes, adrenals, etc. This suggests that themetabolism of cholesterol is linked with the for-mation of esters. It has now been demonstratedthat cholic acid is produced from cholesterol (24).In addition, large amounts of cholesterol arepoured into the bile, chiefly in the free state. It isconceivable that obstruction to the escape of bilebacks up these materials and retards the processesby which they are formed in the liver. Thiswould explain the accumulation of free cholesterolin the serum and the reduction of esters. It hasbeen demonstrated that administration of exces-sive amounts of cholesterol to rats is one of themost effective methods of producing fatty livers(25 to 29). The proportions of cholesterol estersin the livers of such rats are unusually large (25to 29), while in the blood serum they are reduced(30, 31). The formation of phospholipids is dis-

turbed, and the turnover of these compounds inthe liver is retarded in this condition (32, 33).Cholesterol may compete with phospholipids forsuitable fatty acids. Both cholesterol esters andphospholipids in proper proportions may be re-quired for the orderly conduct of hepatic lipidmetabolism. Whether all the phospholipid com-ponents of the serum are equally disturbed orwhether the proportions of these components arealtered by biliary obstruction remains to be deter-mined. If the condition is similar in origin to thecholesterol fatty liver, it should respond to the ad-ministration of choline and other lipotropic factors(34). Some workers (35, 36) have, indeed,shown that in experimental biliary obstructionfatty infiltration of the liver can be minimized bythe administration of high protein diets.

When the parenchyma of the liver is profoundlyinsulted by poisons or other destructive agents,the serum lipids appear to decrease. At the sametime, the pattern of lipids is distorted in much thesame manner as it is in biliary obstruction. Thissuggests that the maintenance of proper concen-trations of lipids in the serum depends upon thepresence of an adequate quantity of liver tissue,while the preservation of the normal pattern oflipids depends on the orderly function of the livercells. This hypothesis would explain the fact thatin cirrhosis concentrations of lipids are usuallylow, while patterns are little disturbed. Althoughthe quantity of functioning liver tissue in cirrhosesis greatly reduced, the remaining cells may becomparatively normal. Distortions of patternsobserved in particular cases would suggest that inthese there is a more active process at work. Itmay be of some significance that such distortionswere encountered more frequently in patients withatypical than they were in patients with typicalcirrhosis.

In hepatitis 2 distinct types of disorders wereseen. Certain cases had initial hyperlipemia, re-sembling in character, but seldom equalling inmagnitude, that of obstructive icterus. Increasesof cholesterol and lipid phosphorus were moremoderate, while increases of neutral fat weregreater in hepatitis. In other patients with in-fectious hepatitis and in the patients with toxichepatitis, cholesterol and lipid phosphorus werelow with an extreme deficiency of cholesterolesters. Clinical evidence suggests that the hyper-

639


lipemic cases had a greater degree of biliary ob-struction than the others. If the hypotheses ad-vanced above are sound, this should yield a hyper-lipemia like that of biliary obstruction, but ofsmaller magnitude because of the injury to theliver cells. It is possible that the difference be-tween the hyperlipemic and hypolipemic casesmay depend upon etiological factors; it may proveto have prognostic significance.

The experiments with lipotropic agents offerlittle encouragement to the use of these agents,although they may have been of some benefit to 5of the patients. This does not mean that theseagents may not be useful therapeutic aids, butonly that the cases or the dosage may have beenimproperly selected. There is experimental evi-dence that choline or methionine may prevent oralleviate the hepatic lesions produced by certainliver poisons and diseases (1, 2, 37). These andother lipotropic agents will preserve the lives ofpancreatectomized animals (38, 39). They willprevent or eliminate fatty infiltrations that resultfrom dietary deficiencies. It has been suggestedthat similar fatty infiltrations characterize dis-eases of the liver arising from other causes. Theseries of liver analyses is all too small to warrantany generalizations or attempts to correlate lipidpatterns of serum and liver closely. Like theserum analyses, it suggests that such correlationsmay be found by further studies of a similar na-ture. The diversity of the disorders in both seraand livers may explain the inefficacy of cholineand other lipotropic agents in the cases to whichthey were given in this study. The benefits thatmay accrue from adding chemical analyses tomorphological techniques is especially evident insuch cases as 28639. This is more feasible withrespect to lipids than to other compounds, sincethe lipids appear to be comparatively stable. Itshould be particularly informative when appliedto the liver because this organ plays such an im-portant r6le in the metabolism of lipids.

SUMMARYAND CONCLUSIONS

The serum lipids have been fractionated on 174occasions in 70 patients with diseases of the liverand bile ducts.

Total cholesterol and lipid phosphorus are ele-vated in the serum of patients with biliary ob-

struction, returning to normal when the conditionis relieved by operation. They are also high in acertain proportion of patients with biliary cirrhosisand with infectious hepatitis. The latter group iscomposed chiefly of subjects who have had in-tense jaundice at the onset of their illness, withacholic stools. This suggests that hyperlipemiais a product of biliary obstruction.

Cholesterol and lipid phosphorus are normal orsubnormal in most patients with portal cirrhosis,tending to fall as the disease advances. In pa-tients with toxic hepatitis and some with infectioushepatitis, these lipid components are also reduced.This suggests that hypolipemia arises from ex-tensive degeneration or destruction of liver pa-renchyma.

Neutral fat in liver disease seldom rises abovethe upper normal limits. It is, however, usuallyabove the average normal, and occasionally greatlyelevated, in obstructive icterus and infectious he-patitis, falling when obstruction has been relievedor when the hepatitis subsides.

The most frequent disorder of serum lipids inliver diseases is an increase of the ratio of free tototal cholesterol. Free cholesterol is usually notonly relatively but absolutely increased, whilethere is usually a deficiency of esters.

The relation of phospholipids to cholesterol innormal subjects has been redefined. In liver dis-ease, especially in those disorders attended byhyperlipemia, the ratio of lipid phosphorus tocholesterol is greater than normal.

The two ratios, free cholesterol: total cholesteroland lipid phosphorus: total cholesterol, are differ-ently and somewhat characteristically altered invarious diseases of the liver.

Choline chloride and other lipotropic agentswere given to a small series of patients with dis-eases of the liver, without definitely beneficialeffects.

The lipids of the sera of 9 patients have beencompared with the lipids of samples of liver ob-tained at autopsy.

PROTOCOLS

44700, in addition to her hepatic condition, had diabetesand necrobiotic lesions of the extremities. These began aspurplish nodules, the centers of which became necrotic andlater ulcerated. Cultures of the necrotic material beforeulceration were sterile; biopsies of the ulcers revealed

640


only chronic inflammation. On the legs, the ulcers be-came confluent and quite extensive. Two years beforeshe came to the New Haven Hospital, she had a periodof intermittent jaundice without abdominal pain. Herliver was considerably enlarged. She died after entero-stomy for symptoms suggesting intestinal obstruction.Post mortem examination revealed chronic cholecystitisand both extra- and intrahepatic cholangitis, stones par-tially obstructing the common bile duct, adhesions andabscesses around the gall-bladder, and abscesses in theliver. During her course of 2% years in the hospital, shehad jaundice of variable intensity, usually slight. Thediabetes was satisfactorily controlled.

B43797, female, born in 1875. For 20 years before ad-mission on June 14, 1943, she had suffered from attacksof right upper quadrant pain, distention and eructations,lasting 1 or 2 days. Three weeks before admission shehad sharp pain in the same area, followed shortly bychills, fever, jaundice, light-colored stools, and dark urine.On admission the temperature was 103.40 F., the liverconsiderably enlarged and tender with a mass below itsuggesting the gall-bladder. At operation on June 21, ahard tumor of the pancreas was reported and a cholecysto-jejunostomy was performed. Because fever and symp-toms did not subside, an external choledocho-duodenostomywas performed on July 20. She died 4 days later with abiliary fistula. Autopsy revealed a carcinoma of the pan-creas which had obstructed the common bile duct and in-vaded the liver. No normal pancreatic tissue except afew islands of Langerhans could be found.

B36812, female, born in 1914. In 1941, she was ad-mitted to another hospital because of excessive menstrualbleeding, fever, and icterus. January 19, 1943, she wasadmitted to the New Haven Hospital because of increas-ing weakness and insomnia, continued menorrhagia, withtingling and numbness of the extremities, obviously alco-holic. She confessed to drinking a pint of whiskey a daybut contended that she had developed this habit only afterthe attack of jaundice mentioned above. Because of thealcoholism, a moderately-enlarged liver, ascites and telan-giectases, a diagnosis of portal cirrhosis was made. Thecephalin flocculation test was 4 plus, the prothrormbintime normal, there was 28 per cent retention of brom-sulfalein at the end of 30 minutes, and she was slightlyicteric. She was readmitted March 25, in shock, almostexsanguinated from vaginal bleeding, and died beforeeither diagnostic or therapeutic measures could be insti-tuted. Post mortem examination revealed only the slight-est fibrosis of the liver which was tremendous. Thehepatic cells were profusely infiltrated with fat. Thepancreas, except for some islands of Langerhans, was en-tirely replaced by fibrous tissue, the remnants of the ductswere dilated, and the main duct was occluded near its en-trance into the duodenum, for no discoverable cause.

28639, female, born 1913. On May 25, 1943, the patientaborted after insertion of a catheter into the uterus. Theabortion was followed by profuse hemorrhage. On May27, she was admitted to the hospital in coma and shock,with a blood pressure of 50/30. She was given two 500

ml. transfusions of whole blood and some sulfadiazine.Fifteen hours later, deep jaundice was noted. In spite offurther transfusions and all other therapeutic measures,she did not respond, dying on June 2. She was given,,among other things, 10 grams of choline parenterally onMay 30 and May 31. The blood culture yielded a puregrowth of E. coli. On May 29, the cephalin flocculationtest was 3 plus and the icterus index 100. Notwithstand-ing these evidences of liver damage and the abnormalserum lipids, the liver post mortem appeared altogethernormal both grossly and microscopically. Acute endo-metritis and pelvic cellulitis were discovered. It is sus-pected that the patient had used drugs as well as me-chanical means to promote abortion.

A57138, male, born in 1889, had been heavily alcoholicsince youth and had been seen in the hospital and emer-gency ward on various occasions in a state of acute alco-holism. His eating habits were extremely irregular, hisdiet poor; on one occasion he was suspected of vitamindeficiency. On November 8, 1944, he initiated an alco-holic spree by drinking a pint of rubbing alcohol on a bet.On November 14, he came to the Emergency ward of theNew Haven Hospital, weak and tremulous, complainingof a headache and epistaxis. The latter had recurred atintervals for 2 days. On November 16 he was admittedto the hospital after vomiting blood. He had eaten littlesince the beginning of his drinking bout. He appearedseriously ill and dehydrated, his liver was slightly en-larged, and he had some tenderness in the calves. Thenext day he became deeply icteric. Vomiting and epistaxiscontinued at intervals and ecchymoses appeared at the sitesof parenteral fluid injections. He lapsed into coma in thecourse of the day. From the time of admission until hisdeath he passed no urine and only 120 ml. could be ob-tained by catheter. This contained albumin, casts, andleukocytes. He died November 18 after a convulsion,despite all therapeutic measures, including frequent par-enteral administration of large amounts of fluid contain-ing carbohydrate in adequate quantities. The cephalinflocculation test was strongly positive, the icterus index100, and bromsulfalein retention more than 50 per centat the end of 30 minutes. The liver was grossly fatty,the pancreas normal, spleen not enlarged. There was ex-tensive degeneration and necrosis of the epithelium of theconvoluted tubules of the kidneys.

19694, male, born in 1913, was admitted to the NewHaven Hospital on March 5, 1943. For about 8 monthshe had been exposed in his occupation to an organichalide preparation. He continued to work for a weekafter the appearance of jaundice, 3 weeks before admis-sion, but was then forced to desist because of weaknessand vomiting which increased steadily. On March 2,blood was noted in the vomitus, and on March 4 he be-came stuporous. On admission he was comatose, deeplyjaundiced, and retching. There was a papular rash on hisback. The icterus index was greater than 125. He diedwithin 24 hours. At autopsy, the liver was the site ofprofuse hemorrhages and extensive necrosis, with littleevidence of regeneration.

641


A47621, female, born in 1908, had been treated forsyphilis from 1935 to 1937. Both Kahn and Wassermanntests were negative. Enlargement of the liver and spleenwere discovered when she was admitted to the hospital in1939 because of alcoholism and lysol poisoning. She wasreadmitted in 1942 because of a fissure in ano, externalhemorrhoids and a secondary anemia. The last probablyarose from rectal bleeding together with malnutrition.The liver was again noted to be slightly enlarged. OnJanuary 4, 1943, she again entered the hospital witherysipelas of the left leg. This time she had slight ascitesand subcutaneous edema. The lipid studies were madeafter the erysipelas had subsided. On April 29, she re-turned to the hospital in shock, almost exsanguinatedfrom a ruptured esophageal varix, and died an hour laterdespite transfusion and infusions. Autopsy revealed adistinct portal cirrhosis with rather extensive acute necro-sis in the central portions of the islands of liver tissue.

B43658, male, born in 1884, was first seen in the Dis-pensary on March 31, 1943, with a history of edema ofthe feet and legs for 2 to 3 weeks, with exertional dyspnea,unproductive cough, and scleral icterus, with light-coloredstools for a week. He had an enlarged heart, ascites, alarge liver, and marked edema of the legs. There was noalcoholic history. The icterus index was 20. He wasadmitted to the New Haven Hospital May 29, somewhatstuporous, with edema up to his nipples. On May 31, hisgums began to bleed, and on June 2 gross rectal bleedingwas noted. The next day his temperature rose sharply to104.60 F. He died June 4. The icterus index was 35, thecephalin flocculation test 2 plus. Autopsy revealed a clas-sical picture of portal cirrhosis with enormous amounts offibrous tissue and small islands of liver tissue in whichthe cells appeared relatively normal.

A45460, male, born in 1883, was in the habit of con-suming 4 to 5 quarts of beer daily in addition to an oc-casional glass of whisky. In 1939, he was admitted to theNew Haven Hospital because of nausea, vomiting, diar-rhea, and jaundice. The liver was greatly enlarged, theicterus index 55. He ran a febrile course for 2 weeks,after which he gradually improved, the liver diminishingin size and jaundice receding. He was again admitted tothe New Haven Hospital November 12, 1944, with ascitesof 8 weeks' duration, productive cough, exertional dyspnea,and bleeding from the rectum. His abdomen was greatlydistended with fluid. After removal of 10 liters, the livercould just be felt beneath the costal margin; the spleenwas never palpable. Skin and sclerae were moderatelyicteric, the icterus index was 50, and bromsulfalein reten-tion 24 per cent at the end of 30 minutes. On November15, he suddenly vomited bright-red blood, his blood pres-sure fell progressively, and he lapsed into coma. He diedlater in the day. His blood sugar shortly before deathwas 46 mgm. per cent. Post mortem examination re-vealed an advanced portal cirrhosis of the liver in whichthe connective tissue and, to a lesser extent, the islands ofparenchyma were densely infiltrated with mononuclearcells and with numerous polymorphonuclear leukocytes.

B43742, male, born in 1889. In 1926, during a cho-lecystectomy, some generalized fibrosis of the liver was

noted. In 1939 or 1940, enlargement of the liver was re-ported. From November, 1942, he noticed increasingmalaise, loss of weight, anorexia and flatulence. Earlyin March, 1943, after a coryza, he was seized with stab-bing pain in the right axilla, accompanied by a cough,and a temperature of 1010 F. Five days later, jaundiceappeared and increased steadily. Shortly after this, thestools became black from changed blood. He was ad-mitted to another hospital, deeply jaundiced, with a largeliver and fever. Because of anorexia, vomiting, andhiccough he was unable to eat and deteriorated rapidly.On March 3, he was transferred to the New Haven Hos-pital in a stuporous state, extremely emaciated, with tenseascites, and a huge liver. He died on March 6. Theicterus index was 60, cephalin flocculation 3 plus, pro-thrombin time 120 seconds (normal control 12.8). Au-topsy revealed diffuse fibrosis of the liver without largescars, some necrosis of the liver cells, some proliferationof the bile ducts, and moderate evidence of biliary stasis.Little sound liver tissue remained. In addition, there wascarcinoma of the liver. The sample for analysis wastaken from an area free from carcinoma.

BIBLIOGRAPHY

1. Barrett, H. M., Best, C. H., MacLean, D. L., andRidout, J. H., The effect of choline on the fattyliver of carbon tetrachloride poisoning. J. Physiol.,1939, 97, 103.

2. Goodell, J. P. B., Hanson, P. C., and Hawkins, W. B.,Methionine protects against mepharsen liver injuryin protein-depleted dogs. J. Exper. Med., 1944, 79,625.

3. McHenry, E. W., and Patterson, J. M., Lipotropic fac-tors. Physiol. Rev., 1944, 24, 128.

4. Peters, J. P., and Man, E. B., The interrelations ofserum lipids in normal persons. J. Clin. Invest.,1943, 22, 707.

5. Mann, F. C., Diet in relation to hepatic physiologyand pathology. A review of pertinent data. Col-lected papers, Mayo Clinic and Mayo Foundation,1943, 35, 34.

6. Peters, J. P., and Man, E. B., The interrelations ofserum lipids in patients with thyroid disease. J.Clin. Invest., 1943, 22, 715.

7. Epstein, E. Z., Cholesterol of the blood plasma inhepatic and biliary diseases. Arch. Int. Med., 1932,50, 203.

8. Epstein, E. Z., and Greenspan, E. B., Clinical sig-nificance of the cholesterol partition of the bloodplasma in hepatic and in biliary diseases. Arch.Int. Med., 1936, 58, 860.

9. Gardner, J. A., and Gainsborough, H., Blood choles-terol studies in biliary and hepatic disease. Quart.J. Med., 1930, 23, 465.

10. Hawkins, W. B., and Wright, A., Blood plasma cho-lesterol: fluctuations due to liver injury and bileduct obstruction J. Exper. Med., 1934, 59, 427.

642

LIPDS OF SERUMAND LIVER IN HEPATIC DISEASES

11. Heinlein, H., Die Rolle der Leber im Cholesterin- undPhosphatidstoffwechsel. Ztschr. f. d. ges. exper.Med., 1933, 91, 638.

12. Boyd, E. M., and Connell, W. F., Lipopenia associatedwith cholesterol Estersturz in parenchymatous he-patic disease. Arch. Int. Med., 1938, 61, 755.

13. Brun, G., Changes in the lipide contents of serum inpatients with manic-depressive psychosis. H. K.Lewis, London, 1940.

14. Sperry, W. M., The concentration of total cholesterolin the blood serum. J. Biol. Chem., 1937, 117, 391.

15. Jones, C. A., A clinical and laboratory study of plasmalipids in obstructive jaundice and several types ofhepatic disease. Am. J. Digest. Dis., 1942, 9, 1.

16. Peters, J. P., and Man, E. B., The interrelations ofserum lipids in patients with diseases of the kidneys.J. Clin. Invest., 1943, 22, 721.

17. Ralli, E. P., Rubin, S. H., and Rinzler, S., The liverlipids in normal human livers and in cases of cir-rhosis and fatty infiltration of the liver. J. Clin.Invest., 1941, 20, 93.

18. Rubin, S. H., Present, C. H., and Ralli, E. P., Theliver lipids in normal dogs on different types of fatwith and without added lecithin. J. Biol. Chem.,1937, 121, 19.

19. Ralli, E. P., Paley, K., and Rubin, S. H., The liverlipids and their distribution in disease. An analysisof 60 human livers. J. Clin. Invest., 1941, 20, 413.

20. Cowie, D. M., and Magee, M.C., Lipoids and lipoiddiseases. III. Lipoid content of tissues in Schiuller-Christian's disease (xanthomatosis) and review ofliterature on lipoid content of human tissues. Arch.Int. Med., 1934, 53, 391.

21. Thannhauser, S. J., and Reinstein, H., Fatty changesin the liver from different causes. Arch. Path.,1942, 33, 646.

22. Rourke, G. M., and Stewart, J. D., Composition of theliver. Its uniformity with respect to the concentra-tion of certain biochemical constituents in differentparts of the same liver. Arch. Path., 1942, 33, 603.

23. Fagin, I. D., Sahyun, M., and Pagel, R. W., Cirrhosisof the liver: the lipotropic action of parenterallyadministered amino acids. J. Lab. and Clin. Med.,1943, 28, 987.

24. Bloch, K, Berg, B. N., and Rittenberg, D., The bio-logical conversion of cholesterol to cholic acid. J.Biol. Chem., 1943, 149, 511.

25. Best, C. H., Channon, H. J., and Ridout, J. H.,Choline and the dietary production of fatty livers.J. Physiol., 1934, 81, 409.

26. Best, C. H., and Ridout, J. H., The effects of choles-terol and choline on deposition of liver fat. J.Physiol., 1933, 78, 415.

27. Cook, R. P., Cholesterol feeding and fat metabolism.Biochem. J., 1936, 30, 1630.

28. Ralli, E. P., Rubin, S. H., and Present, C. H., Theliver lipids and fecal excretion of fat and nitrogenin dogs with ligated pancreatic ducts. Am. J.Physiol., 1938, 122, 43.

29. Stetten, DeW., Jr., and Grail, G. F., Effect of dietarycholine, ethanolamine, serine, cystine, homocysteine,and guanidoacetic acid on the liver lipids of rats.J. Biol. Chem., 1942, 144, 175.

30. Chaikoff, I. L., and Kaplan, A., The blood lipids incompletely depancreatized dogs maintained withinsulin. J. Biol. Chem., 1934, 106, 267.

31. Kaplan, A., and Chaikoff, I. L., The effect of raw andautoclaved pancreas on the liver lipids of the com-pletely depancreatized dog maintained with insulin.J. Biol. Chem., 1937, 119, 435.

32. Perlman, I., and Chaikoff, I. L., Radioactive phos-phorus as an indicator of phospholipid metabolism.V. On the mechanism of the action of choline uponthe liver of the fat-fed rat. J. Biol. Chem., 1939,127, 211.

33. VII. The influence of cholesterol upon phospholipidturnover in the liver. J. Biol. Chem., 1939, 128,735.

34. Gavin, G., Patterson, J. M., and McHenry, E. W.,Comparison of the lipotropic effects of choline,inositol, and lipocaic in rats. J. Biol. Chem., 1943,148, 275.

35. Johnson, J., Ravdin, I. S., Vars, H. M., and Zintel,H .A., Effect of diet on composition of the liver inthe presence of obstruction of the common bile duct.Arch. Surg., 1940, 40, 1104.

36. Ravdin, I. S., Thorogood, E., Riegel, C., Peters, R.,and Rhoads, J. E., The prevention of liver damageand the facilitation of repair in the liver by diet.J. A. M. A., 1943, 121, 322.

37. Sellards, A. W., and McCann, W. S., Choline hydro-chloride in experimental yellow fever in rhesusmonkeys. U. S. Nav. M. Bull., 1944, 43, 420.

38. Entenman, C., and Chaikoff, I. L., Is choline the fac-tor in the pancreas that prevents fatty livers in de-pancreatized dogs maintained with insulin? J. Biol.Chem., 1941, 138, 477.

39. Ralli, E. P., and Rubin, S. H., The effect of meat andmeat fractions on the fatty liver of the depan-creatized and pancreatic-duct-ligated dog. Am. J.Physiol., 1942, 138, 42.

643