Parenchymal Alterations in Cirrhotic Livers in Patients With Hepatopulmonary Syndrome or Portopulmonary Hypertension

Transcription

1 LIVER TRANSPLANTATION 19: , 2013 ORIGINAL ARTICLE Parenchymal Alterations in Cirrhotic Livers in Patients With Hepatopulmonary Syndrome or Portopulmonary Hypertension Changqing Ma, 1 Jeffrey S. Crippin, 2 William C. Chapman, 3 Kevin Korenblat, 2 Neeta Vachharajani, 3 Kristen L. Gunter, 3 and Elizabeth M. Brunt 1 Departments of 1 Pathology and Immunology, 2 Internal Medicine (Division of Gastroenterology), and 3 Surgery, Washington University School of Medicine, St. Louis, MO Hepatopulmonary syndrome (HPS) and portopulmonary hypertension (PPH) are distinct pulmonary vascular complications of cirrhosis. Little is known about possible associated hepatic histopathological features. Explanted livers from patients clinically diagnosed with HPS (n 5 8) or PPH (n 5 7) and cirrhotic explants from controls (n 5 30) without HPS or PPH were evaluated with trichrome histochemistry, anti glutamine synthetase (anti-gs), and anti-cd34 immunohistochemistry (IHC). Trichrome stains were characterized by cirrhotic nodules (CNs) of various sizes, including incomplete septal cirrhosis (ISC). ISC was overrepresented in the HPS (4/8 or 50%) and PPH livers (3/7 or 43%); in addition, neither group had micronodular cirrhosis. The control explants showed the entire spectrum of nodules: micronodular, macronodular, mixed CNs, and ISC (P ). The variability of cirrhosis severity was shown with the Laennec grading system (0-6). The cirrhosis of the majority of the HPS (6/8) and PPH livers (6/7) was scored as mild, whereas the control explants were more evenly distributed across the mild (14/30) and moderate/severe grades (16/30). GS positivity was retained in a perivenular location as the dominant pattern in each explant group. CD34 staining detected capillarized sinusoids of CNs as well as vascular channels within septa, but no significant differences were found between the groups. None of the observed light microscopy or histochemistry and IHC patterns showed a correlation with the underlying liver disease. Although our results demonstrate variable architectural and vascular remodeling within and between explant livers regardless of the presence or types of pulmonary complications, there were differences in explants with HPS or PPH versus controls that correlated with less severe cirrhosis. Liver Transpl 19: , VC 2013 AASLD. Received November 5, 2012; accepted February 18, Hepatopulmonary syndrome (HPS) and portopulmonary hypertension (PPH) are 2 unique pulmonary vascular complications of advanced chronic liver disease. 1,2 HPS results from diffuse or localized intrapulmonary microvascular dilatation, which causes blood gas exchange abnormalities and ultimately lifethreatening arterial hypoxemia. It encompasses a clinical triad characterized by intrapulmonary vascular dilatations, arterial deoxygenation, and chronic liver disease/hepatic dysfunction. 3-6 HPS occurs most commonly in patients who have established cirrhosis and portal hypertension, but it may also occur in patients with noncirrhotic portal hypertension. 7 The reported prevalence of HPS in adult patients with cirrhosis ranges from 4% to 32%. 2,8-11 The risk of developing HPS is reportedly not associated with the Additional Supporting Information may be found in the online version of this article. Abbreviations: CN, cirrhotic nodule; GS, glutamine synthetase; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; HPS, hepatopulmonary syndrome; IHC, immunohistochemistry; ISC, incomplete septal cirrhosis; MELD, Model for End-Stage Liver Disease; OLT, orthotopic liver transplantation; PPH, portopulmonary hypertension. This study was funded by the Department of Pathology and Immunology at Washington University in St. Louis (St. Louis, MO). Address reprint requests to Elizabeth M. Brunt, M.D., Department of Pathology and Immunology, Washington University School of Medicine, 660 South Euclid Avenue, Box 8118, St. Louis, MO Telephone: ; Fax: ; ebrunt@wustl.edu DOI /lt View this article online at wileyonlinelibrary.com. LIVER TRANSPLANTATION.DOI /lt. Published on behalf of the American Association for the Study of Liver Diseases VC 2013 American Association for the Study of Liver Diseases.

2 742 MA ET AL. LIVER TRANSPLANTATION, July 2013 etiology or severity of cirrhosis or with the severity of portal hypertension. 2-4,8,12,13 The only effective treatment for HPS is orthotopic liver transplantation (OLT), after which complete resolution has been well documented; 10,14,15 without OLT, the prognosis of HPS is dismal. Swanson et al. 16 reported 5-year survival rates of 76% and 23% for HPS patients who underwent OLT and HPS patients who were not candidates for OLT, respectively. PPH, defined as pulmonary arterial hypertension in association with portal hypertension, is due to vasoconstriction and remodeling of the resistance vessels in the lung parenchyma, which lead to pulmonary hypertension. 6,17 PPH also occurs mostly in patients with cirrhosis, but it is less common than HPS. The reported prevalence of PPH in patients with cirrhosis is 2% to 16%. 1,6,18,19 The severity of PPH correlates poorly with the degree of portal hypertension and the severity of liver disease as measured by the Model for End-Stage Liver Disease (MELD). 1,6,18,20 A multicenter, prospective study by Kawut et al. 1 showed (1) an increased risk of developing PPH for women and for patients with autoimmune hepatitis and (2) a decreased risk for patients with hepatitis C virus (HCV) infection. 1 A retrospective study by Swanson et al. 21 showed that without treatment, the 1-year survival rate for PPH patients was 46%, and the 5- year survival rate was 14%. With medical treatment alone, the 5-year survival rate was 45%; it improved to 67% with a combination of medical treatment and OLT. As anticipated, the histopathological features of the pulmonary vasculature in HPS and PPH differ. 6,12,22 Focal or diffuse dilatation of precapillary and capillary vessels in an otherwise intact alveolar and airway architecture characterizes HPS, 6,22 whereas the features of pulmonary vessels in PPH patients are indistinguishable from other forms of pulmonary arterial hypertension. 6,12,23 Morphological findings in livers associated with either HPS or PPH, however, have not been reported. Therefore, we undertook this study of our institution s series of explants from patients affected by HPS or PPH to search for possible histological features that were different between them or between them and control cirrhotic explant livers without pulmonary complications. PATIENTS AND METHODS This study was approved by the Human Research Protection Office of Washington University. The liver transplant database supported by the Department of Surgery at Barnes-Jewish Hospital provided the names, demographics, and clinical characteristics of patients who were diagnosed with HPS or PPH and a set of age- or sex-matched and underlying liver disease matched cirrhotic controls. All study groups had undergone liver transplantation. The diagnostic criteria for HPS included evidence of intrapulmonary shunting on contrast-enhanced echocardiography (ie, saline bubble positivity after 4-5 cardiac cycles) or bubbles seen in the pulmonary vein on transesophageal echocardiography and evidence of a shunt by an alveolar arterial gradient on arterial blood gases. 6,14,24 The criteria for diagnosing PPH were based on evidence of pulmonary hypertension by pulmonary artery catheter measurements (mean pulmonary arterial pressure 25 mm Hg) with a normal left ventricular end-diastolic pressure and on evidence of portal hypertension on ultrasound or other imaging modalities. 6,24 Weights and volumes of liver explants were collected and calculated with measurements documented in pathology reports. Volumes of cirrhotic livers calculated before OLT were retrieved via reviews of radiology reports. All histological slides were coded before evaluation; both pathologists (C.M. and E.M.B.) were blinded to the patient category. Hematoxylin and eosin stained tissue sections of liver explants from patients and controls were reviewed. Two representative sections/ tissue blocks per explant (1 from the left and 1 from the right) were selected; if an explant had hepatocellular carcinoma (HCC), none of the selected sections included tumor or treated tumor. Morphological evaluations of the explanted livers were focused on the architectural and vascular remodeling in cirrhosis. Architectural remodeling was primarily characterized via trichrome histochemistry staining for collagen. The following definitions were applied. A cirrhotic nodule (CN) was surrounded by nearly or completely circumferential fibrosis. A micronodule (ie, a micro-cn) had a diameter < 3 mm; a macronodule (ie, a macro-cn) had a diameter 3 mm. Incomplete septal cirrhosis (ISC) was defined as indistinct parenchymal nodularity with slender fibrous septa that extended into the parenchyma but ended blindly without connections. 25 Ectatic veins protruding into the parenchyma could be present and adjacent to the septum. When 1 type of CN or ISC was seen in >50% of all CNs in a tissue section, it was labeled the predominant pattern. A mixed-size CN pattern had equal amounts of micro- CNs and macro-cns. For each explant, the predominant pattern was determined via the averaging of the observations from both tissue sections. The fibrosis of each cirrhotic liver was evaluated with trichrome-stained tissue sections and was scored semiquantitatively with the Laennec system, which is a system used for scoring cirrhosis severity and known to correlate with the severity of portal hypertension. 26 Briefly, the Laennec system divides fibrosis into 7 grades (0-6) based on the width and number of fibrous septa in a liver tissue section. Laennec fibrosis scores of 4 to 6 correspond to cirrhosis and reflect the variability and severity of cirrhosis. Score 4/stage 4A indicates mild/definitive cirrhosis with marked septation and visible nodules, although most septa are thin; score 5/stage 4B indicates moderate cirrhosis with at least 2 broad septa and less than half of the tissue section composed of micronodules (<3 mm); and score 6/stage 4C indicates severe cirrhosis with at least 1 very broad septum or more than half of the tissue section composed of micronodules. 26

3 LIVER TRANSPLANTATION, Vol. 19, No. 7, 2013 MA ET AL. 743 Parenchymal remodeling was additionally studied through the evaluation of the loss of lobular zonality with glutamine synthetase (GS) immunohistochemistry (IHC). GS, one of the key enzymes in nitrogen metabolism, is expressed exclusively in the pericentral hepatocytes most proximal to the terminal hepatic venules in normal livers 27 and has been reported to be undetectable in CNs. 28,29 In the current study, GS reactivity was absent in nodules or was observed in CNs in residual perivenular hepatocytes, periseptal hepatocytes, or both. When it was positive, GS staining was typically strong/intense and cytoplasmic. Positive periseptal reactivity was defined as full circumferential periseptal reactivity to GS (ie, at least 1 layer of periseptal hepatocytes showed strong GS staining). Positive GS staining of an explant was defined as GS reactivity in >50% of all CNs in the 2 tissue sections of an explant. Although 50% is an arbitrary cutoff, it was used to exclude high background GS staining. GS can also be diffusely positive in HCC. 30 By design, however, no tissue sections used for analysis in this study included HCC. Vascular shunts in septa and capillarization of sinusoids were studied by IHC with anti-cd34. Endothelial cells of hepatic arterioles, terminal hepatic venules, and portal venules are positive for CD34; lymphatic endothelial cells are negative or only rarely and weakly positive for CD In normal/ noncirrhotic livers, only the immediate periportal sinusoids are positive for CD The degree of CD34 reactivity in septa was used as a surrogate for the number of vascular channels in septa in this study, and it was graded through the evaluation of the density of CD34 reactivity in the total number of septa available in a particular tissue section in a quartile fashion: minimal, mild, moderate, or marked. Diffuse hepatic sinusoidal reactivity to CD34 is pathological and is seen in HCC 29 and capillarized sinusoids 32 in patients with cirrhosis. In this study, when >50% of all CNs in the 2 tissue sections of an explant showed CD34 reactivity in sinusoids beyond the immediate periseptal region (3-4 layers of hepatocytes), it was considered abnormal. CD34 was further classified as confluent when all sinusoids of the nodule were positive or partially confluent when only some were. G om ori trichrome staining was performed on 4-lm sections cut from formalin-fixed, paraffin-embedded tissue blocks in a 1-step procedure with Bouin s solution (Sigma-Aldrich, St. Louis, MO), Weigert s iron hematoxylin, and a G om ori trichrome staining solution (Newcomer Supply, Inc., Middleton, WI). A section of a normal liver was used as a positive control. CD34 and GS IHC was performed on a Ventana Benchmark LT automated immunostainer (Ventana Medical Systems, Inc., Tucson AZ) according to standard protocols. Briefly, IHC was performed on representative 4-lm sections cut from formalin-fixed, paraffinembedded tissue blocks with a monoclonal anti- CD34 antibody (clone name: QBEnd/10; Ventana; 1:1 dilution) and a monoclonal anti-gs antibody (BD Transduction Laboratories, San Jose, CA; 1:200 dilution). The detection system was Ventana s ultra- View universal 3,3 0 -diaminobenzidine detection kit. No biotin was involved for anti-cd34, whereas Ventana s ultrawash was used for anti-gs as a blocking reagent. Antigen retrieval was performed with the standard procedure on the machine with the Ventana cell conditioning 1 solution. Sections of tonsils and HCC were used as positive controls for anti- CD34 and anti-gs, respectively; negative reagent controls were also used for both antibodies. Statistical tests were used for comparisons of clinical characteristics among the 3 groups (controls, HPS cases, and PPH cases) and for comparisons of categorical data for staining patterns. Raw data (ie, the numbers of HPS, PPH, and control cases) rather than their percentages within groups were used in all statistical analyses. A Kruskal-Wallis 1-way analysis of variance was performed with SPSS (IBM, Armonk, NY), and Fisher s exact test was performed in R. 33 RESULTS Clinical Results Eight HPS cases and 7 PPH cases from 597 consecutive adult patients undergoing OLT for chronic liver disease between 2002 and 2010 together with 30 ageor sex-matched and disease-matched controls without HPS or PPH were retrieved from the database for evaluation. Matching for disease was prioritized. Clinical characteristics are provided in Table 1. HPS patients were predominantly women (6/8), whereas PPH cases were equally distributed between men and women. Chronic HCV infection was the major etiology for both HPS (4/8) and PPH cases (5/ 7). Other etiologies for HPS included cryptogenic cirrhosis (2/8), primary biliary cirrhosis (1/8), and nonalcoholic steatohepatitis (1/8). The etiologies for the remaining 2 PPH cases were cryptogenic cirrhosis and autoimmune hepatitis. More than 1 etiology was present for 2 of the 8 HPS cases and for 5 of the 7 PPH cases. Alcoholic liver disease was listed as the second diagnosis for 1 patient with HPS and for 1 patient with PPH, and both had a primary diagnosis of chronic HCV infection. Five patients also had HCC: 1 HPS case with a primary diagnosis of chronic HCV infection and 4 PPH cases with primary diagnoses including chronic HCV infection, autoimmune hepatitis, and cryptogenic hepatitis. Three of the 4 PPH cases with HCC had been treated with chemoembolization before OLT; the HPS case with HCC and 1 PPH case had HCC identified as incidental lesions < 2cm in size during the evaluation of their explanted livers. The MELD scores of the controls were higher than those of the HPS or PPH cases because individuals with HPS or PPH were given priority exception points for transplantation. However, this did not reach statistical significance (P ). The waiting time on the transplant list, a somewhat arbitrary index for the

4 744 MA ET AL. LIVER TRANSPLANTATION, July 2013 TABLE 1. Clinical Characteristics of the Patients Control Group (n 5 30) HPS Group (n 5 8) PPH Group (n 5 7) P Value Age (years)* 56 (48-68) 55 (39-66) 57 (48-65) Sex (n) Female Male Underlying liver disease [n (%)] Chronic HCV infection 18 (60.0) 4 (50.0) 5 (71.4) Cryptogenic cirrhosis 6 (20.0) 2 (25.0) 1 (14.3) Autoimmune hepatitis 2 (6.7) 0 1 (14.3) Primary biliary cirrhosis 2 (6.7) 1 (12.5) 0 Nonalcoholic steatohepatitis 2 (6.7) 1 (12.5) 0 HCC as a secondary diagnosis (n) Incidental Treated with transarterial chemoembolization Alcohol as a secondary diagnosis (n) MELD score* 21 (10-37) 16 (11-18) 17 (11-21) 0.08 Waiting time on transplant list (days)* 251 (1-1654) 130 (4-350) 233 (69-530) 0.20 *The data are presented as medians and ranges. TABLE 2. Weights, Volumes, and Laennec Fibrosis Stages of Liver Explants Control Group (n 5 30) HPS Group (n 5 8) PPH Group (n 5 7) P Value Explant weight (g)* 1100 ( ) 826 ( ) 1530 ( ) 0.03 Explant volume (cm 3 )* 2268 ( ) 1924 ( ) 2729 ( ) 0.46 Liver volume calculated by radiology before OLT (cm 3 )* 1397 ( ) 1058 ( ) 1739 ( ) 0.09 Laennec scoring system for fibrosis [n (%)] Stage 4A (score 4): mild 14 (46.7) 6 (75.0) 6 (85.7) Stage 4B (score 5): moderate 7 (23.3) 0 1 (14.3) 0.19 Stage 4C (score 6): severe 9 (30.0) 2 (25.0) 0 *The data are presented as medians and ranges. chronicity of liver disease, showed a wide range of variation between controls and cases. The HPS cases had the shortest average waiting time on the transplant list (Table 1); however, these differences also were not statistically significant (P ). The weights and volumes of liver explants extracted from pathology reports and the volumes of livers from the last radiological assessments before OLT were highest in PPH explants, and these were followed by control explants and then HPS explants (Table 2). However, only the weight difference was significant (P ) according to an analysis of variance of the 3 groups. When pairwise comparisons were performed, the weights and the liver volumes calculated by radiology before OLT were significantly different between HPS and PPH cases [P and P , respectively (t test)]. Two or more calculated volumes from radiological assessments were available for 11 controls, 5 HPS patients, and 6 PPH patients. Radiological assessments were not performed at standardized intervals (Supporting Table 1). Controls had volume data available from 12 to 1695 days before OLT, HPS patients had data available from 48 to 1735 days before OLT, and PPH patients had data available from 2 to 681 days before OLT. Among these, 8 controls, 4 HPS cases, and 3 PPH cases showed overall decreases in their calculated liver volumes over time, whereas the rest had overall volume increases (Supporting Fig. 1A- C). Volume changes in HPS livers showed a weak correlation with time (correlation coefficient r ), whereas volume changes in control and PPH livers did not correlate with time (r and r , respectively). PPH livers had the highest volume change over time and the fastest rate of volume change over time; control livers had the lowest and slowest changes (Supporting Table 1). Histopathological Results All controls and study cases had Laennec fibrosis scores of 4 or greater (Table 2). Fourteen controls and the

5 LIVER TRANSPLANTATION, Vol. 19, No. 7, 2013 MA ET AL. 745 Figure 1. Trichrome histochemistry stains and GS reactivity. (A) The hepatic parenchyma comprises mixed-size CNs isolated by fibrous septa and areas of extinction, which are highlighted by trichrome. Each CN is surrounded by full circumferential fibrosis. (B) ISC may have histological features composed of inconspicuous parenchymal nodularity, and slender fibrous septa that extend into the parenchyma but end blindly without making connections. (C) GS reactivity is restricted to 1 layer or a few layers of hepatocytes immediately adjacent to the terminal hepatic venule. (D) GS reactivity can be seen in periseptal hepatocytes, which form a ring at the periphery of each CN. majority of the HPS (6/8) and PPH cases (6/7) had a Laennec fibrosis score of 4 (fibrosis stage 4A). The remaining controls were nearly evenly distributed between scores of 5 (stage 4B; 7/30) and 6 (stage 4C; 9/ 30). No PPH case had a score of 6; no HPS case had a score of 5. Thus, the cirrhosis of the majority of the HPS and PPH cases was scored as mild according to the Laennec system. These results were not significantly different from one another statistically (P ). Cirrhotic patterns that were observed with trichrome staining included micro-cn predominance, macro-cn predominance, mixed-size CNs, and ISC with or without mixed-sized CNs (Fig. 1A,B). All 4 patterns were observed in the controls (Fig. 2A); the majority had either mixedsize CNs (13/30 or 43%) or ISC with or without mixedsized CNs (9/30 or 30%). In contrast, 4 of 8 HPS cases showed ISC with or without mixed-sized CNs, 3 of 8 showed mixed-sized CNs, 1 of 8 showed macro-cns, and none had micro-cns. The PPH cases had either macro- CN (4/7) or ISC with or without mixed-size CNs (3/7); none had micro-cns. The differences between the trichrome staining patterns of the controls, HPS cases, and PPH cases were statistically significant (P ). GS staining was observed in both perivenular hepatocytes and periseptal hepatocytes to various extents in all explant sections (Fig. 1C,D). Similar proportions of controls (8/30 or 27%), HPS cases (2/8 or 25%), and PPH cases (2/7 or 29%) were considered to have no GS reactivity because most of the tissue section was negative (Fig. 2B). When they were present, GS staining patterns were grouped into 3 categories: perivenular, periseptal rimming, and mixed (perivenular and periseptal rimming). All 3 patterns were observed with similar distributions in the controls and HPS cases. The PPH cases were either perivenular or periseptal; none had the mixed GS staining pattern. Perivenular GS was observed in 10 of 30 controls, 3 of 8 HPS cases, and 3 of 7 PPH cases. The perivenular GS pattern in the HPS and PPH cases versus the controls was not statistically significant (P ). CD34 staining patterns in CNs were periseptal or confluent/partially confluent (Fig. 3A,B). Confluent or partially confluent CD34 reactivity was the dominant pattern, and it was observed in 16 of 30 controls (53%), 5 of 8 HPS cases, and 4 of 7 PPH cases (Fig. 2C). Rarely, sinusoids showed minimal CD34

6 746 MA ET AL. LIVER TRANSPLANTATION, July 2013 Figure 2. (A) Trichrome, (B) GS, and (C,D) CD34 staining patterns among controls, HPS cases, and PPH cases. Study cases and controls are grouped according to the staining patterns observed. The percentage at each bar corresponds to the proportion of either cases or controls, and it is followed by the number of cases or controls associated with a staining pattern. reactivity, and this was observed only in controls (4/ 30). CD34 reactivity in septa was scored as minimal to marked (Fig. 3C-F). Reactivity was relatively evenly distributed among the controls and the PPH cases (Fig. 2D); in contrast, marked reactivity was observed in 4 of the 8 HPS cases, and none had minimal CD34 in septa. The CD34 patterns of CNs and septa seen in the controls, HPS cases, and PPH cases were similar for the controls and study cases (P and P , respectively). Correlations between underlying liver diseases and staining patterns of trichrome, GS, and CD34 were negative (Supporting Fig. 2). DISCUSSION Little is known about the hepatic histopathological features associated with the pulmonary vascular complications of HPS and PPH. Both processes occur predominantly in the setting of chronic liver disease induced cirrhosis; 1,2 HPS has also been reported in patients with noncirrhotic portal hypertension. 7 In the present study, we compared the explanted livers of 8 HPS patients and 7 PPH patients from our institutional experience with 30 age-, sex-, and diseasematched explanted cirrhotic controls. Statistically significant differences between controls, HPS cases, and PPH cases were noted only for the types of CNs. Specifically, although the controls showed all forms of CNs, neither HPS nor PPH explants had micronodular CNs. In addition, although this was not statistically significant, ISC was overrepresented in HPS (50%) and PPH cases (43%) versus controls (30%). ISC is an injury pattern included in the spectrum of noncirrhotic portal hypertension and is presumably due to portal venopathy in the absence of underlying liver disease. 7 In the current study, all patients had diagnosed chronic liver disease and thus did not fit

7 LIVER TRANSPLANTATION, Vol. 19, No. 7, 2013 MA ET AL. 747 Figure 3. CD34 reactivity in capillarized sinusoids in CNs and septa. (A) Minimal reactivity to CD34 can be seen only in periseptal sinusoids (ie, periseptal restriction). (B) Confluent reactivity to CD34 can be seen in all sinusoids of the CN in the center of the photomicrograph. (C-F) Endothelial cells of vascular channels in septa are also positive for CD34. The CD34 reactivity in the septa can be scored as (C) minimal, (D) mild, (E) moderate, and (F) marked. the criteria of noncirrhotic portal hypertension. However, cases within HPS, PPH, and control livers were found to have features described for ISC and particularly blindly ending septa and parenchymal nodules without fibrous bands. This feature has also been described as one of the findings in the spectrum of regression of cirrhosis, 25,32,34 which may actually have been the situation in our study because all the patients had known chronic liver disease, and the explanted livers were nodular and had evidence of septal formation and remodeling. The fact that none of the HPS or PPH cases had micro-cns is interesting and raises 2 possible interpretations. The first is that micronodular cirrhosis is a form of remodeling that potentially protects the pulmonary vasculature from the complications of HPS or PPH. More likely, however, is the broader concept that because subjects with the pulmonary vascular complications of chronic liver disease and cirrhosis undergo liver transplantation on account of lung complications rather than fully developed end-stage liver disease, their livers show injury and remodeling but not end-stage

8 748 MA ET AL. LIVER TRANSPLANTATION, July 2013 micronodular cirrhosis. Mechanisms that initiate and perpetuate the development and progression of pulmonary complications in the setting of chronic liver injury, fibrosis, and remodeling cannot be determined from this type of study, but these observations for our small set of subjects deserve further investigation. As noted previously, 2-4 no particular underlying etiology of cirrhosis predisposed patients to the development of HPS in our series, nor did we confirm the previously reported associations of female sex and autoimmune hepatitis with an increased risk of PPH. 1 Finally, HCV infection has been reported to be protective against PPH, 1 but all of our study groups reflected our patient population in general, which is rich with HCV subjects. The current literature also suggests that there is no association between the severity of liver disease or portal hypertension and either pulmonary condition. 13,20,35 However, according to both MELD and Laennec fibrosis scores, the HPS and PPH study cases had evidence suggestive of milder liver cirrhosis than the control cases. The inference from our combined findings supports 2 possible speculations: (1) the hepatic vascular and parenchymal remodeling of the subjects affected by HPS and PPH never progressed to micronodular or mixed CN cirrhosis in comparison with the control cohort with end-stage liver disease, and (2) the explanted livers of the HPS and PPH cases underwent scarring and subsequent regression of fibrosis as the pulmonary complications evolved. These may not be mutually exclusive speculations. Another interesting finding of our study was the maintenance of GS in cirrhosis. GS has previously been reported as undetectable in CNs, 28,29 but it was present in 73% of the explants (33/45) in this study in 3 possible patterns: perivenular, periseptal, or a combination. Periseptal rimming, not previously reported, was observed in 29% of the explants (13/ 45); none of the nodules was dysplastic. We have hypothesized that GS reactivity among hepatocytes immediately adjacent to septa indicates marked vascular remodeling with reorganization of the hepatocyte zonal designation. This could result either from local alterations of structural, vascular, pressure, or flow environments within the liver or from modulation of the regulation of GS expression by upstream signals such as beta-catenin. 36 What the stimulus/stimuli would be in cirrhosis remains to be shown. The maintenance of a central location of outflow veins within CNs in 40% of the study cases, as demonstrated by GS IHC, is also interesting. This highlights the concept that cirrhotic remodeling is a complex, dynamic process. 25 The cirrhotic liver volume decreases gradually with the progression of liver disease, and volume changes vary according to the etiology. A small liver volume is a proven poor prognostic factor in cirrhosis Attempts at correlations with liver volumes and weights were made, but they were largely unsuccessful because of the lack of prospective standardized data collection for the subjects (see the Results section). Most cirrhotic livers in this study lost volume closer to the time of OLT, and this was likely due to normal parenchymal structure replacement by condensed connective tissue components. However, the PPH livers consistently had the highest values for both weights and volumes in the 3 groups. The significance of this is unclear. The primary limitation of this study is the small number of HPS and PPH cases to study. Eight HPS explants and 7 PPH explants were collected from 597 consecutive adult explants for chronic liver diseases over a period of 9 years ( ) at our center. The number of HPS and PPH explants is thought to be lower than the actual number of patients diagnosed with these 2 complications during this time period at our institution because HPS and PPH patients may die while waiting for transplantation or may not meet the criteria for OLT for other reasons. However, our database does not list the actual number of patients diagnosed with HPS or PPH during this period who did not undergo transplantation. In our OLT database, the incidence of transplantation is 1.3% for HPS patients and 1.2% for PPH patients. Although these are small numbers, the fact that they represent our entire series is meaningful in terms of the overall prevalence of these complications. In summary, our study has demonstrated architectural and vascular remodeling of varying degrees in all cirrhotic explants, regardless of the presence or absence of pulmonary vascular complications. A significant difference between the controls and the HPS and PPH explant hepatectomies was related to the nodule size; none of the HPS or PPH explanted livers had micronodules. Although this was not statistically significant, ISC was overrepresented in the HPS and PPH groups; in addition, lower MELD scores and a decreased severity of cirrhosis according to Laennec scoring were noted in the HPS and PPH groups. Whether these changes preceded and contributed to the pulmonary complications or resulted from them could not be determined by this morphological analysis, but they do provide histological evidence that the livers of subjects with HPS and PPH are less diseased according to standard analyses than matched controls at the time of OLT. It is apparent from our study that pulmonary vascular complications requiring liver replacement may occur in patients with chronic liver disease without complete cirrhotic remodeling of the liver parenchyma and regardless of the underlying liver disease etiology. Finally, although our study was not designed to evaluate the underlying mechanisms of either the pulmonary or hepatic structural alterations in these patient groups, it has provided histological insight into the livers of subjects affected by HPS or PPH who require OLT. ACKNOWLEDGMENT The authors thank the Anatomic and Molecular Pathology Core Lab and specifically Lab Manager Neha Dahiya, M.D., M.B.A., Senior Research Technician

9 LIVER TRANSPLANTATION, Vol. 19, No. 7, 2013 MA ET AL. 749 Jianping Li, B.S., Research Lab Supervisor Autumn Watson, B.A., and Research Technician I Vernetta Layton for their excellent work in preparing the histological sections and immunohistochemical stains. REFERENCES 1. Kawut SM, Krowka MJ, Trotter JF, Roberts KE, Benza RL, Badesch DB, et al.; for Pulmonary Vascular Complications of Liver Disease Study Group. Clinical risk factors for portopulmonary hypertension. Hepatology 2008; 48: Lange PA, Stoller JK. The hepatopulmonary syndrome. Ann Intern Med 1995;122: Krowka MJ, Cortese DA. Hepatopulmonary syndrome. Current concepts in diagnostic and therapeutic considerations. Chest 1994;105: Rodrıguez-Roisin R, Agustı AG, Roca J. The hepatopulmonary syndrome: new name, old complexities. Thorax 1992;47: Rodriguez-Roisin R, Krowka MJ. Is severe arterial hypoxaemia due to hepatic disease an indication for liver transplantation? A new therapeutic approach. Eur Respir J 1994;7: Rodrıguez-Roisin R, Krowka MJ, Herve P, Fallon MB; for ERS Task Force Pulmonary-Hepatic Vascular Disorders (PHD) Scientific Committee. Pulmonary-hepatic vascular disorders (PHD). Eur Respir J 2004;24: Schouten JN, Garcia-Pagan JC, Valla DC, Janssen HL. Idiopathic noncirrhotic portal hypertension. Hepatology 2011;54: Martınez GP, Barbera JA, Visa J, Rimola A, Pare JC, Roca J, et al. Hepatopulmonary syndrome in candidates for liver transplantation. J Hepatol 2001;34: Gupta D, Vijaya DR, Gupta R, Dhiman RK, Bhargava M, Verma J, Chawla YK. Prevalence of hepatopulmonary syndrome in cirrhosis and extrahepatic portal venous obstruction. Am J Gastroenterol 2001;96: Machicao VI, Fallon MB. Hepatopulmonary syndrome. Semin Respir Crit Care Med 2012;33: Schenk P, Fuhrmann V, Madl C, Funk G, Lehr S, Kandel O, M uller C. Hepatopulmonary syndrome: prevalence and predictive value of various cut offs for arterial oxygenation and their clinical consequences. Gut 2002; 51: Krowka MJ, Edwards WD. A spectrum of pulmonary vascular pathology in portopulmonary hypertension. Liver Transpl 2000;6: Schenk P, Sch oniger-hekele M, Fuhrmann V, Madl C, Silberhumer G, M uller C. Prognostic significance of the hepatopulmonary syndrome in patients with cirrhosis. Gastroenterology 2003;125: Rodrıguez-Roisin R, Krowka MJ. Hepatopulmonary syndrome a liver-induced lung vascular disorder. N Engl J Med 2008;358: Taille C, Cadranel J, Bellocq A, Thabut G, Soubrane O, Durand F, et al. Liver transplantation for hepatopulmonary syndrome: a ten-year experience in Paris, France. Transplantation 2003;75: Swanson KL, Wiesner RH, Krowka MJ. Natural history of hepatopulmonary syndrome: impact of liver transplantation. Hepatology 2005;41: Krowka MJ. Portopulmonary hypertension: diagnostic advances and caveats. Liver Transpl 2003;9: Benjaminov FS, Prentice M, Sniderman KW, Siu S, Liu P, Wong F. Portopulmonary hypertension in decompensated cirrhosis with refractory ascites. Gut 2003;52: Colle IO, Moreau R, Godinho E, Belghiti J, Ettori F, Cohen-Solal A, et al. Diagnosis of portopulmonary hypertension in candidates for liver transplantation: a prospective study. Hepatology 2003;37: Hadengue A, Benhayoun MK, Lebrec D, Benhamou JP. Pulmonary hypertension complicating portal hypertension: prevalence and relation to splanchnic hemodynamics. Gastroenterology 1991;100: Swanson KL, Wiesner RH, Nyberg SL, Rosen CB, Krowka MJ. Survival in portopulmonary hypertension: Mayo Clinic experience categorized by treatment subgroups. Am J Transplant 2008;8: Martınez-Palli G, Rodrıguez-Roisin R. Hepatopulmonary syndrome: a liver-induced oxygenation defect. Eur Respir Monogr 2011;54: Mooi WJ, Gr unberg K. Histopathology of pulmonary hypertensive diseases. Curr Diagn Pathol 2006;12: Krowka MJ, Mandell MS, Ramsay MA, Kawut SM, Fallon MB, Manzarbeitia C, et al. Hepatopulmonary syndrome and portopulmonary hypertension: a report of the multicenter liver transplant database. Liver Transpl 2004;10: Wanless IR, Nakashima E, Sherman M. Regression of human cirrhosis. Morphologic features and the genesis of incomplete septal cirrhosis. Arch Pathol Lab Med 2000;124: Kim MY, Cho MY, Baik SK, Park HJ, Jeon HK, Im CK, et al. Histological subclassification of cirrhosis using the Laennec fibrosis scoring system correlates with clinical stage and grade of portal hypertension. J Hepatol 2011;55: Gebhardt R, Baldysiak-Figiel A, Kr ugel V, Ueberham E, Gaunitz F. Hepatocellular expression of glutamine synthetase: an indicator of morphogen actions as master regulators of zonation in adult liver. Prog Histochem Cytochem 2007;41: Racine-Samson L, Scoazec JY, D Errico A, Fiorentino M, Christa L, Moreau A, et al. The metabolic organization of the adult human liver: a comparative study of normal, fibrotic, and cirrhotic liver tissue. Hepatology 1996;24: Coston WM, Loera S, Lau SK, Ishizawa S, Jiang Z, Wu CL, et al. Distinction of hepatocellular carcinoma from benign hepatic mimickers using glypican-3 and CD34 immunohistochemistry. Am J Surg Pathol 2008;32: Di Tommaso L, Franchi G, Park YN, Fiamengo B, Destro A, Morenghi E, et al. Diagnostic value of HSP70, glypican 3, and glutamine synthetase in hepatocellular nodules in cirrhosis. Hepatology 2007;45: Pusztaszeri MP, Seelentag W, Bosman FT. Immunohistochemical expression of endothelial markers CD31, CD34, von Willebrand factor, and Fli-1 in normal human tissues. J Histochem Cytochem 2006;54: Wanless IR, Crawford JM. Cirrhosis. In: Odze RD, Goldblum JR, eds. Surgical Pathology of the GI Tract, Liver, Biliary Tract and Pancreas. 2nd ed. Philadelphia, PA: Saunders; 2009: R Development Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing; Desmet VJ, Roskams T. Cirrhosis reversal: a duel between dogma and myth. J Hepatol 2004;40: Krowka MJ, Wiseman GA, Burnett OL, Spivey JR, Therneau T, Porayko MK, Wiesner RH. Hepatopulmonary syndrome: a prospective study of relationships between severity of liver disease, PaO(2) response to 100% oxygen, and brain uptake after (99m)Tc MAA lung scanning. Chest 2000;118:

10 750 MA ET AL. LIVER TRANSPLANTATION, July Cadoret A, Ovejero C, Terris B, Souil E, Levy L, Lamers WH, et al. New targets of beta-catenin signaling in the liver are involved in the glutamine metabolism. Oncogene 2002;21: Zoli M, Cordiani MR, Marchesini G, Abbati S, Bianchi G, Pisi E. Ultrasonographic follow-up of liver cirrhosis. J Clin Ultrasound 1990;18: Zoli M, Cordiani MR, Marchesini G, Iervese T, Labate AM, Bonazzi C, et al. Prognostic indicators in compensated cirrhosis. Am J Gastroenterol 1991;86: Lin XZ, Sun YN, Liu YH, Sheu BS, Cheng BN, Chen CY, et al. Liver volume in patients with or without chronic liver diseases. Hepatogastroenterology 1998;45: