Nodular Regenerative Hyperplasia of the Liver

Nodular Regenerative Hyperplasia of the Liver

69:746-751, 1975 Copyright© 1975 by The Willia ms & Wilkins Co. GASTROENTEROLOGY Vol. 69, No. :1 Printed in U.S.A. NODULAR REGENERATIVE HYPERPLASI...

11MB Sizes 0 Downloads 91 Views

69:746-751, 1975 Copyright© 1975 by The Willia ms & Wilkins Co.

GASTROENTEROLOGY

Vol. 69, No. :1

Printed in U.S.A.

NODULAR REGENERATIVE HYPERPLASIA OF THE LIVER DANIEL

M.

KNOWLES

II, M.D.,

GoRDON

I.

KAYE,

PH.D.,

GABRIEL

C.

GooMAN,

M.D.

Department of Pathology, College of Physicians and Surgeons of Columbia Universitv New York, New York ·'

Nodular regenerativ~ hype~p_lasia of the liver, an uncommonly reported and poorly defmed chmcopathological entity, obscured clinical diagnosis and was misdiagnosed on hepatic biopsy in a recent case. Approximately 19 cases are recorded in the English literature. Six patients had Felty's syndrome, about 12 patients had congestive heart failure, and the patient under discussion had subacute bacterial endocarditis. Light- and electron-microscopic examination was utilized to define nodular regenerative hyperplasia pathologically. Features common to all reported cases are discussed but elucidation of the pathogenesis of nodular regenerative hyperplasia must await further investigation. Nodular regenerative hyperplasia is a diffuse hepatic regenerative process of rare occurrence, distinguished from cirrhosis by the absence of true fibrous septal formation .' Because of the infrequency of this lesion and confusion over its true nature, clinically and anatomically, report of an instructive instance appears warranted.

peared chronically ill , febrile, and had left upper quadrant tenderness, abdominal distention, and hepatosplenomegaly but was without jaundice, adenopathy, cardiac murmurs, or signs of heart failure or chronic liver disease. The pertinent laboratory data included: hemoglobin, 10.7 g per 100 ml ; hematocrit, 33.6%; white blood cell count, 33,000; platelet count, 120,000; erythrocyte sedimentation rate, 45; blood urea nitrogen, 10; calcium, 8.5; phosphorus, ~.9; cholesterol , 124; albumin , 3.8; Case Report SGOT, 62; bilirubin, 1.1 mg per 100 ml; alkaline A 42-year-old asymptomatic business execu- phosphatase, 527; prothrombin time, 14.2 sec tive, without a history of hepatitis or hepato- (control 13.7 sec) ; and negative heterophil, toxin exposure, was found to have abnormal lupus erythematosus preps, febrile agglutinins. liver function tests on routine examination in hepatitis-associated antigen, and serum a -fetoMarch 1973. Five months later he began to protein. A chest film revealed bilateral pleural effusions. Blood cultures were sterile. The cliniexperience episodic shaking chills and fever. Hepatosplenomegaly was discovered and he was cal impression was an occult carcinoma and hospitalized in December 1973. Evaluation, hypersplenism. Paracentesis yielded 200 ml of chylous ascites including laparotomy and liver biopsy, was negative and no diagnosis was made. Following with negative cytology. A liver scan demondischa rge he continued to have febrile episodes strated reduced clearance of radiocolloid. The associated with shaking chills and drenching patient underwent exploratory laparotomy, night sweats, weight loss, increasing girth, and splenectomy, and wedge biopsy of the liver; the latter was misinterpreted as diffuse, well differleft upper quadrant tenderness. On admission to the Columbia-Presbyterian entiated hepatoma and chronic passive splenic Medical Center on January 19, 1974 he ap- congestion. Postoperatively , he was intermittently feb rile but blood cultures were negative. He experienced several postoperative episodes Received November 19, 1974. Accepted April 4, of sudden transient confusion , lethargy, and 1975. diaphoresis, occasionally associated wi t h Address requests for reprints to: Dr. Daniel Knowles, Department of Pathology, College of Physi- slurred speech. The white blood cell count rose cians and Surgeons of Columbia University, 630 West to 25,000. The alkaline phosphatase rose to 1025 168 Street, New York, New York 10032. but t he bilirubin only increased to 2.0 mg per 746

September 1975

CASE REPORTS

100 mi. It was believed that the fever was secondary to carcinoma and the chylous effusions secondary to either undetected tumorous lymphatic obstruction or an iatrogenic complication of the original laparotomy. On February 21, 1974, 5-fluorouracil was initiated. Five days later he suffered a 1-hr long headache, after which he was found unresponsive with a fixed and dilated right pupil and was pronounced dead. Postmortem findings. At autopsy the liver weighed 2100 g; the capsular surface was cobbled by numerous closely spaced 0.1- to 0.4-cm nodules. On cut section the liver was replaced

FIG. 1. The enlarged liver is diffusely permeated by small yellow-brown nodules, some of which are abut-

ting against the bile ducts and vessels.

747

by similar yellowish-brown nodules which compressed the intervening parenchyma (fig. 1). These nodules occasionally abutted against bile ducts or vessels but were never invasive. Microscopically, these nodules were identified as circumscribed collections of normal appearing hepatocytes, without bile ducts or central veins. The nodules were often contiguous, compressing the normal hepatic parenchyma between them (fig. 2). Reticulin stains showed the condensation of reticulin fibers at the margins of the expanding nodules (fig. 3). Minimal collagen fiber formation was only occasionally present in these areas. The hepatocytes composing these nodules were sometimes larger than normal, occasionally binucleate, and often contained a prominent central nucleolus (fig . 4). Mitoses were not seen. There was no evidence of invasion, necrosis, or inflammation. There was no evidence of hepatocellular carcinoma after thorough sectioning of the liver. Microscopic examination of numerous lymph nodes revealed no evidence of malignant lymphoma. The common bile duct was patent and the gallbladder was grossly and histologically unremarkable; there was no evidence of extrahepatic obstruction to explain the elevated alkaline phosphatase. The un'derlying disease responsible for the clinical course was subacute bacterial endocarditis (SBE) of the aortic valve. A friable,

748

CASE REPORTS

Vol. 69, No.3

FIG. 3. A reticulin stain demonstrates the variably sized regenerative nodules and the internodular reticulim fiber condensation (Laidlaw reticulin stain, x 40) ..

hemorrhagic, vegetation was firmly attached to each aortic valve cusp. There was no evidence of valvular damage from the endocarditis or of preexisting valvular disease . There were large bacterial clumps within the vegetations but neither antemortem nor postmortem cultures could identify the causative organism . Splenomegaly was attributable to the bacterial endocarditis ; there was no evidence of portal hypertension. There was no evidence of thoracic duct obstruction at the second laparotomy or at autopsy. The development of chylous ascites 2 weeks after the first laparotomy, combined with the absence of thoracic duct obstruction , prompted the clinicians and pathologists to consider this an iatrogenic complication of the first laparotomy . The transient cerebral episodes may have' been secondary to emboli from the aortic valve. Massive fresh hemorrhage in the central white matter of the right cerebral hemisphere with marked herniation of the right cingulate and parahippocampal gyri accounted for the terminal clinical findings and death. Electron microscopically, the cytoarchitecture of the nodules was consistently normal (fig. 5) and showed none of the morphological aberrations of organelles hitherto described in human hepatocellular carcinoma. z-• The hepatocytes were closely packed, but normal sinusoidal spaces and Kupffer cells were always present. Normal bile canaliculi with pericanalicular desmosomes were present. There was some increase in the amount of collagen present in

the sinusoidal spaces. Glycogen and ribonucleoprotein was abundant and distributed as in normal active hepatocytes. In summary, there was no microscopic evidence of abnormality of either the cytology or relationships of the hepatocytes in the nodules , which were indistinguishable from those of normal human liver.'

Discussion Hepatic cirrhosis is characterized by hepatocellular necrosis, lobular collapse, an increase in connective tissue, and irregular nodular regeneration of the liver. This results in the formation of pseudolobules delineated by fibrous septae and distortion of the vascular and architectural relationships of the liver. 6 In nodular regenerative hyperplasia there is neither apparent necrosis nor fibrous connective tissue proliferation, and it can readily be distinguished from the nodular regeneration of cirrhosis. 1 The absence of inflammation or necrosis suggests that the stimulus to micronodular regeneration occurred in the past, without causing symptoms. The resulting nodules may occasionally be larger than normal hepatic lobules. They are recognizable as consisting of new regenerative hepatocytes by their abnormal relationship to portal fields, irregularity of the liver cords, enlargement of the hepatocytes, increased

September 1975

CASE REPORTS

749

FIG. 4. Several features of regenerat ive hepatocy tes a re evident : numerous binucleate cells, large nuclei, and prominent nucleoli . Such features may mistakenly lead to a diagnosis of hepatoma (hematoxylin and eosin, x 256) .

frequency of binucleated cells , the presence of prominent nuceoli, and the compression exerted by the nodules on the contiguous parenchyma. Compression atrophy at the margins allows partial collapse of the preexisting fibrillar stroma and an apparent stromal increase around the nodules. However, increased collagen formation and true fibrous septae do not occur. Steiner 1 did not believe that nodular regenerative hyperplasia was a stage in the development of cirrhosis. Considering the frequency of hepatic cirrhosis and the scarcity of nodular regenerative hyperplasia, this conclusion is reasonable. Ranstrom 7 described a histologically identical lesion in a 36-year-old female with dermatitis, polyarthritis, subcutaneous rheumatic nodules , and Felty's syndrome who received gold therapy . Ranstrom 7 labeled the entity miliary hepatocellular adenomatosis, but we believe that it represents nodular regenerative hyperplasia. Blend is et at. 8 recently reported 5 patients with Felty's syndrome and multisystem involvement with nodular regenerative hyperplasia of the liver. One patient had

Sjogren 's syndrome and generalized lymphadenopathy and 2 had diffuse pulmonary fibrosis and rheumatoid nodules. Aside from palmar erythema, no cutaneous stigmata of chronic liver disease were present. Hepatomegaly was present in all fi patients and 3 had portal hypertension. Serum bilirubin was at the upper limit of normal in 3 patients and alkaline phosphatase was elevated in 3 cases, as in our patient. The elevated alkaline phosphatase was not readily explainable in our patient nor in Blendis' patients, as there was no evidence of biliary tract disease or extrahepatic obstruction. The etiology of the diffuse nodular regen eration is unknown. Although Steiner 1 felt that it may be associated with congestive cardiac failure the latter was not present in Ranstrom's 7 or Blendis' 8 patients who had Felty's syndrome (rheumatoid arthritis , leukopenia, and splenomegaly 9 ) or our patient with SBE. Ranstrom 7 interpreted the lesion as an idiosyncratic reaction to gold therapy. Blendis 8 suggested that changes in volume or distribution of blood f1ow through the liver, related to Felty's syndrome, might be responsible for the devel-

750

CASE REPORTS

Vol.69, No.3

FIG. 5. The hepatocytes are closely packed and sinusoidal spaces are seen (S) . Bile canaliculi (B) with pericanalicular desmosomes are evident. Each hepatocyte contains a full complement of normal appearing organelles including rounded mitochondria, smooth and rough endoplasmic reticulum, glycogen, and lysosome· like bodies. The rough endoplasmic reticulum is often organized into stacked cisternae (arrows). (x 6500).

opment of nodular regenerative hyperplasia. Uniformly present in all cases was splenomegaly, and in no instance is it necessary to invoke nodular regenerative hyperplasia as the causative factor. Splenomegaly was explained in Steiner's 1 cases as a manifestation of congestive heart failure, in Ranstrom 7 and Blend is ' 8 patients as a component of Felty's syndrome, and in our patient as a manifestation of SBE. Although it might initially appear that such splenomegaly is secondary to hepatic ar-

chitectural changes splenomegaly, with increased splenic blood flow, might result in augmented portal and intrahepatic blood flow such as, in some way, to promote reparative nodular proliferation. Liver function tests were abnormal in Ranstrom 's 7 patient, and alkaline phosphatase was elevated in 3 of Blendis' 8 cases. Hepatic abnormalities were not detected clinically in Steiner's 1 cases. No patient was jaundiced or had stigmata of chronic liver disease_ In one case abnormal liver function tests were discovered inci-

September 1975

CASE REPORTS

dentally during routine examination. Since it was not until 5 months later that our patient became symptomatic with SBE, the two entities appear to be unrelated. However, an association between splenomegaly and nodular regenerat ive hyperplasia appears in escapab le . Moreover, there is no proof that splenomegaly and SBE were not present at the time of discovery of the abnormal liver function tests or that the patient did not suffer, prior to discovery of the abnormal liver function tests, from SBE. The biopsy diagnosis of well differentiated hepatoma was based on the clinical impression of a diffuse in filtrative hepatic process, and reflects lack of ex perience with nodular regenerative hyperplasia and the sampling error inherent in a small fragmented liver biopsy. In only 1 of Blendis' 8 cases was the biopsy diagnosis correct. Akin to nodular regenerative hyperplasia, and equally obscure pathogenetically, are partial nodular transformation, 10 - 14 focal nodular hyperplasia, 15 and hepatocellular adenomatosis. 16 Although all of these entities appear related, Sherlock et al. believe that they represent distinctive clinicopathological entities; they deserve brief comparison with nodular re·generative hyperplasia . Partial nodular transformation is characterized by portal hypertension owing to intrahepatic circulatory obstruction by perihilar, noncirrhotic hepatocyte nodules with concomitant atrophy of the uninvolved hepatic parenchyma. 10 - 14 Focal nodular hyperplasia is most commonly an asymptomatic, subcapsular, well circumscribed, multinodular, cirrhotic-like mass, with a distinctive histology, in an otherwise normalliver. 1 5 Hepatocellular adenomatosis 16 is characterized by multiple, widely spaced, discrete hepatocyte nodules with histologically normal intervening parenchyma. The nodules do not totally replace the hepatic parenchyma as in nodular regenerative hyperplasia and are histologically distinguishable from those of focal nodular hyperplasia. Recogn ition of nodular regenerative hyperplasia as a distinctive clinicopathologi-

751

cal entity, distinguishable from cirrhosis, and more importantly, from hepatocellular carcinoma, is necessary . Since the clinical diagnosis of this entity is unlikely, the onus is upon the pathologist to make the correct diagnosis . We have attempted to define this entity pathologically to assist in making the distinction. Elucidation of the pathogenesis and natural history of this entity must await further investigation . REFERENCES 1. Steiner PE: Nodu lar rege ne rative hype rplasi a of the liver. Am .J Pathol :l!i:94:l- 9f>:l, 19!i9 2. Toker C, Trevino N: Ultrastructure of human primary hepat ic carcinoma. Cancer If): lf>~H - !Gtm. 1966 3. Theron ,JJ , Mekel RC: Electron microscopical studies of human malignant hepatoma ce lls. T Gastroenterol 7: 1!i2- 1!i4. 19G4 4. Yasutake S, Nakau K. Matsunaga Y: Fine structure of human hepatoma. Kururne Med ,J 9:19:3-204 , 19()2 5. Ma MH. Hiempica L: The normal human liver cell: cytochemica l and ultrastru ctural studies. Am J Pathol G2::l!i:l- :l7G. 1971 6. Anderson WAD: Pathology. Sixth edition . St Louis, CV Mosby, 1971 7 . Ranstrom S: Miliar.v hepatocellular adenomatosis. Acta Pnthol Microbiol Scam! :l:l:22!i - 229. 195:1 8. Blendis LM. Parkinson MC. Shilkin KB. et al.: Nodu la r regen erative hyperplasia of the li ve r in Felty's syndrome . q ,J Med 4:l:2f> -:l2, IH?•I 9. Felty AR: Chronic arthritis in the adult , associated with splenomegaly and leukopenia. Hull Hopkins Hosp :lf>: IG- 20, 1924 10. Sherlock S, Feldman CA. Moran H, et al.: Partial nodular transformation of the liver with portal hypertension . Am ,J Med 40: 19f> - 20:l, 19titi 11. Maillard JN. Potet F, Henhamon ,JJ>: Transformation nodulnire part ielle du i(>i c avec hype rt e nsion portale. Presse Med 7!i:27~lfl - 2802, 1%7 12. Classen M, Elster K. Pesch H.J, et al.: Portal hypertension caused by partial nodular transformation of the liver. Gut 11 :24!i- 249, 1970 13 . Dick AI', Gresham GA: Partial nodular transformation of the liver presenting with ascites . Gut 13:289-292, 1972 14 . Lurie 8 , Novis 8, HankS. et al.: CRSTsynd rome a nd nodular transfi>rmation of the liver. Gastroenterolo!(y 64:4!i7 -4 61. 197:3 15. Knowles II OM, Wolff M .: Focal nodular hyper plasia of the liver : A clinicopathologic study. Hum Pnthol (in press) 16. Stumpf HH, Liher AF: Hepatoc ellular adenomatosis. Am ,J Med 17 :RH7 -H90, l!l!i•l