—  SPECIALTY CONFERENCE  —

Liver Pathology

Case 2 - "Acute Hepatitis" Developing 10 Months after
Allogeneic Hematopoietic Stem Cell Transplantation

Howard Shulman
Fred Hutchinson Cancer Center
Seattle, WA


Click on each slide thumbnail image for an enlarged view
Clinical History
A 40 year-old with ALL received an allogeneic matched sibling hematopoietic stem cell transplant (HCT) in 2002.The only posttransplant (PT) problem, mild skin graft-versus-host disease (GVHD), resolved with prednisone. At six months, tapering of prophylactic prednisone and cyclosporin were completed. At 8.5 months PT she finished a 2 week course of famciclovir for zoster. Two weeks later, the LFTs revealed a normal bilirubin with elevations of alkaline phosphatase (AP)406, SGOT 112, and SGPT 133. Within 2 weeks, SGOT had risen to 2,086, SGPT 1,641, AP 347. The first liver biopsy was performed 26 days after the onset of LFT elevations, 10 months PT. Foscarnet was given for three weeks for treatment of preemptive zoster hepatitis. In the ensuing month the SGOT/SGPT and AP levels markedly declined. However, the SGOT rose again to 475 and the bilirubin continued to rise, peaking at 30 mg/dl 33 days later when a second liver biopsy was done, 11 months PT. The patient had a history of hepatitis A reactivity pretransplant, but was currently IgM negative. The patient was CMV antibody positive but multiple PCR studies for circulating CMV were negative. Tests for hepatitis viruses B and C were negative. All cultures were negative. There was no indication of any potentially hepatotoxic drugs.


Case 2 - Figure 1 - Low power view showing marked lobular hepatitis with portal inflammation
Case 2 - Figure 2 - Marked lobular hepatitis with prominent sinusoidal lymphocytosis
Case 2 - Figure 3 - Shrunken interlobular bile duct with compression of lumen


Case 2 - Figure 4 - Lobular disarray with hepatocellular swelling
Case 2 - Figure 5 - Portal inflammatory infiltrate with plasma cells; interlobular bile duct with epithelial injury
Case 2 - Figure 6 - Portal tract without identifiable interlobular bile duct

Histologic Findings
The first liver biopsy performed 10 months PT had a striking lobular hepatitis with lymphocytes and Kupffer cells filling many of the sinusoids, associated with many scattered acidophilic hepatocytes. Portal spaces were expanded by an infiltrate of lymphocytes and plasma cells spilling over into the surrounding periportal hepatocytes. The larger bile ducts appeared unremarkable. Some of the interlobular bile ducts had swollen cytoplasms, with nuclear dyspolarity, anisonucleosis, segmental loss of nuclei and scattered intraepithelial lymphocytes.

The second liver biopsy displayed pronounced cholestatic changes with canalicular bile plugs and marked swelling of pigment-laden hepatocytes in zone 3, associated with focal hepatocytolysis and sinusoidal fibrosis. Inflammation within the acinus and portal spaces was less pronounced. The larger portal spaces contained proliferated ductules. The interlobular ducts were shrunken, having only two or three remaining nuclei within eosinophilic syncytium without a central lumen. The bile ducts in the smallest portal spaces were nearly destroyed.

Followup: Treatment with high dose steroids and cyclosporin led to rapid improvement in liver tests, achieving a normal bilirubin within two months.

Chronic Graft-versus-Host Disease of the Liver: Presentation as an Acute Hepatitis
The sudden development of marked elevations of serum aminotransferase enzymes following allogeneic HCT is a medical emergency since these tests may presage acute liver failure. Potential causes of the clinical presentation can be deduced from the PT time interval. Within the first three weeks, the most likely cause is severe veno-occlusive disease, also known as sinusoidal obstruction syndrome, resulting from hepatocyte ischemia caused by the severe congestion and obstruction of blood flow in zone 3. Beyond day 30 the most likely etiology is from cytopathic viruses such as adenovirus, varicella and herpes simplex occurring roughly in that frequency. Following immune reconstitution beyond 3 months PT, infections caused by hepatitis B, varicella zoster, adenovirus or hepatitis C may result in severe hepatitis and occasionally fulminant hepatic failure. Other than hepatitis caused by these viruses, extreme elevations are unusual after day 100. [1, 2]

In 2000, Strasser, et al, reported a series of 14 long lived allogeneic HCT recipients, from our institution and several others sent in consultation, who had a clinical presentation of chronic GVHD that resembled acute hepatitis. [3] Characteristics of these 14 patients are summarized from Tables I and II of that report. All patients were transplanted for malignancy; the majority received HLA sibling identical allografts. Six had no acute GVHD, and 4 had only skin and gut involvement. Beyond day 80, half of these patients had chronic GVHD but in only 2 of 14 was there liver involvement. The hepatitic presentation occurred from day 74 to day 749, median day 294. Elevations in alkaline phosphatase were roughly 3x (peak 11x) in comparison to the marked median and peak elevations of aminotransferases, SGOT 1,100 and SGPT of 1,600 u (700-2500), and total bilirubin of 12mg/dl(1-56), respectively.

Clinical events preceding the hepatitic onset: The typical clinical presentation was preceded by the recent cessation or tapering of the anti-GVHD immunosuppressive drugs to minimal doses. Withdrawal from immunosuppression routine tapppering in 9, however, in 4r patients the drugs were abruptly discontinued, 1 patient stopped the medicine of her own accord, and 3 patients who developed leukemic relapse had immunosuppression withdrawn rapidly, aiming to induce a graft-versus-leukemia effect (GVL). Two of these patients then rapidly developed jaundice and transaminitis soon after receiving of donor lymphocytes infusions (DLI).

In 2002, Akpek et al from Johns Hopkins reported that 30% of their patients who had received DLI for a GVL developed hepatitic onset of GVHD an average of 35 days following DLI. They further noted a marked difference in the maximum aminotransferases between this hepatitic variant and "classical" GVHD occurring in the usual PT setting. [4]

Histologic Features of Hepatitic Onset of Chronic GVHD
The characteristic features of the 11 patients with liver biopsies in Strasser's study was a striking lobular hepatitis with a marked necroinflammatory infiltrate, numerous acidophilic bodies, moderate to marked hepatocyte arrest and disarray of the cords, prominence of sinusoidal Kupffer cells and lymphocytes. Biopsies from 4 of the patients contained striking confluent perivenular hepatocyte dropout with collections of pigmented macrophages and a mixed lymphocytic/plasmacytic infiltrate. Portal spaces in most were expanded by a mixed inflammatory infiltrate of lymphocytes, many plasma cells, pigmented macrophages and a few eosinophils. Mild changes of portal endothelialitis were present in 2 patients. The portal inflammation frequently spilled out into the surrounding periportal hepatocytes. The damaged bile ducts present in all of the biopsies were characterized by shrunken, withered outlines with compression or loss of the lumen. The epithelial nuclei display dyspolarity, anisonucleosis, hyperchromatism and focal loss of contiguous nuclei, resulting in segments of swollen eosinophilic vacuolated anucleate cytoplasms. Intraepithelial lymphocytes were frequent but apoptotic epithelial cells were rarely identified. These bile duct changes are characteristic of those produced by GVHD. [1, 5]

Treatment and Clinical Course
In those patients in whom the diagnosis was promptly made at presentation, the institution of cyclosporin, or sirolimus plus steroids and ursodeoxycholate, led to progressive and often rapid improvement with eventual normalization of liver tests. When the diagnosis was not made immediately and the administration of appropriate immunosuppressive therapy delayed, at least one of the patients developed progressive cholestatic abnormalities, including jaundice with bilirubin of 23. This patient had four liver biopsies between days 184-580 showing a sequence of ductopenia followed by marked ductule proliferation. He was eventually salvaged with high-dose immunosuppressive therapy, ultimately achieving normal hepatic function by 12 months.. A similar sequence of events has also been reported by Freese, et al, following successful treatment of rejection after orthotopic liver transplantation. [6] However, if the diagnosis of chronic GVHD is delayed some patients may fail to respond to late introduction of high-dose immunosuppressive therapy and later die with progressive cholestasis. [2] These patients had portal fibrosis with marked cholestasis and loss of small bile ducts without cirrhosis The few reports of cirrhosis developing after intractable chronic hepatic GVHD were completed prior to the identification of hepatitis C. Even if the reports of cirrhosis caused by GVHD are correct, they are of largely of historical note because with earlier diagnosis and more effective therapies for GVHD cirrhosis has not since been reported. [2, 7, 8, 9]

Nomenclature, Grading and Prognostic Features
The "natural" history of GVHD, in reality, applies only to the experimental models where the allogeneic strain differences between donor and host can be tightly controlled and the process is untreated. [10] In the usual clinical setting, patients receive multiple potent immunosuppressive drugs to prevent or combat established GVHD. The typical syndrome of "classical" clinical hepatic GVHD is progressive cholestasis with elevations of bilirubin, alkaline phosphatase, and gamma glutamyl peptidase with only modest elevations of aminotransferases. Hepatic GVHD often occurs with other organs involved in the acute or chronic phase. However, as in this case, it may present as the solitary organ involved. This is particularly vexing in the setting of an allogeneic HCT using umbilical cord blood or T-cell depleted donor cells and the expectation that since there was no acute GVHD there will be no chronic GVHD. [2]

Clinically, GVHD is administratively defined as acute or chronic depending on whether the symptoms and signs are present before or after a certain date, typically between day 80 and 100. Though acute and chronic GVHD share certain histologic features they differ in that the latter has sclerodermatous skin manifestations, keratoconjunctivitis sicca with involvement of the lacrimal glands, salivary glands, and upper aerodigestive tract and small airways. However, unlike cutaneous chronic GVHD, there is not a clear dichotomy between the histology of acute and chronic hepatic GVHD. Histologic analysis of sequential liver biopsies has shown that the degree of bile duct damage/destruction, portal fibrosis and cholestasis is proportionate to the duration of GVHD [5], and the absence of immunosuppression. [3] The degree of allogenicity likely influences histology; however, this is difficult to assess in the clinical complex setting. [2]

Several different schema have been used to grade and prognosticate the manifestations of acute and chronic GVHD. However, as treatment options have improved, the utility of such schema has been questionable, with refractoriness to corticosteroids refractory gut disease, or the presence of extensive chronic GVHD remaining as the most important prognostic variables. Initial attempts at grading liver GVHD by Lerner [11] were based on the number of portal spaces with damaged bile ducts. However, serial liver biopsies indicate that, the number of portal spaces affected as well as the extent of bile duct damage reflect the duration of GVHD [5]. Analogous to the grading and staging of chronic viral hepatitis C, some of the features that might portend severity also indicate chronicity. For example, the degree of hepatocellular-cholestasis, ballooning is a reflection of the duration of GVHD while the periportal-cholangiolar bile thrombi, due to the coexistence of gut GVHD showering the portal blood with lipopolysacccharide stimulating release of TNF- a and inducing bile ductule proliferation. [12]. At the December 2003 American Society of Hematology meeting Lie et al presented data on 38 additional cases of hepatitic onset GVHD [13] and Couriel et al from 41 biopsies with liver GVHD including 12 cases of Chronic GVHD. [14] These two studies confirmed that response to therapy was unrelated to the degree of lobular inflammation, acidophilic body formation, and fibrosis. But, in the study by Couriel, as well as Strasser's study [3], patients with more severe bile duct injury or loss had a slower response to therapy, but it did not reach significance [14]. To summarize, no multivariate analysis has ever demonstrated that any histologic grading schema of hepatic GVHD has significant prognostic value independent of response to treatment, though delay will result in a longer recovery period.

The nomenclature developed for use in liver allografting has been applied to the diagnosis and implied in the grading of GVHD. Although both processes involve alloimmune-related destruction of the bile ducts, there are several striking dissimilarities. First, the time course is quite different. The term, vanishing bile duct, or paucity of interlobular bile ducts, is typically applied to chronic rejection. In contrast, this occurs both early and late after HCT. Second, endothelialitis is an uncommon feature for identifying GVHD as it was seen, in 7-15% of biopsies. Last, the vasculopathy that occurs in some long-lived survivors with liver allografts is never seen following hematopoietic stem cell transplantation.

Immunopathogenesis of Hepatic GVHD
A principle unanswered question remains: why there is selective targeting of certain epithelia by GVHD, including the small bile ducts and adjacent ductules, In murine studies by Howell et al, where strain differences could be tightly controlled, the V b receptors of the T cell infiltrating the liver were polyclonal and directed at multiple non-MHC antigens selective for the liver. [10] The dynamics of the GVHD process begins when the donor T cells recognize host allogeneic-antigens presented on host dendritic presenting cells. [10] Experimentally, this initiating step is enhanced by the chemo/irradiation therapy given for conditioning, which results in release of lipopolysaccharide into the circulation, which in turn activates host antigen-presenting cells and stimulates monocytes and macrophages to secrete inflammatory cytokines. Ichiba and Ferrara et al at the University of Michigan, looked at the expression of genes in a murine model of GVHD after two time points. [15] Beyond day 7 there was marked up-regulation of IFN- d, MHC II molecules, and genes related to leukocyte trafficking in the spleen, but not the liver. In related studies, Braun et. al. with John Vierling at UCLA studied the gene expression of immortalized biliary epithelial cells(BEC) exposed to pro-inflammatory cytokines, TNF- a and IFN-g. [16] The up-regulated genes expressed in the immortalized BEC included adhesion molecules, receptors for IL-1, IL-4 and interleukin-1, pro-apoptotic and anti-apoptotic genes MHC class 1 and II, matrix metaloproteases and caspases. Braun, et al, hypothesize that the cytokine-chemokines secreted by the BEC lining only the small to medium-sized caliber ducts lead to the attraction, and susceptibility to death by cytotoxic T cells. Unlike the selective targeting of small bile ducts, the individual hepatocyte necrosis seen particularly in our cases of hepatitic GVHD is likely explained by the intensity of the lymphocytic inflammation with cytokine up-regulation of the death receptor pathway on hepatocytes. [3]

Differential Diagnosis

Cytopathic virus infection Even prior to liver biopsy, patients with the clinical features of this case should be given empiric treatment for acute herpes simplex, varicella zoster or adenovirus with high-dose anti-viral agents such as acyclovir and cidofovir until the biopsy has excluded them. The histologic features of these infections show random punched-out areas of necrosis, nicely demonstrated on the PAS stain. Viral inclusions can be seen at the edge of the necrotic areas with adenovirus or zoster, as well as in the bile ducts. In our experience, herpes simplex hepatitis may not be associated with obvious viral inclusions. Immunostains and rapid shell vial culture rapidly confirm the diagnosis. Recently, Bründler et al, described adenovirus ascending cholangiohepatitis in three children, two with liver transplants and one with acquired HIV infection. [17] These patients presented with marked elevations of AST and GGT to nearly 2,000 units. Histologically, the biopsies revealed cholangiohepatitis with progressive loss of interlobular bile ducts with adenovirus inclusions noted within the biliary tract. It was hypothesized that the adenovirus occurred as a result of ascending infection from the gastrointestinal tract to the biliary tree. To my knowledge, no such cases have been reported after HCT.

Hepatitis C and B The frequency of hepatitis C has been markedly reduced since the initiation of effective blood screening. Whereas 15 years ago one-third of all HCT patients became hepatitis C positive. Today, the incidence is under 0.3%. [14] However, there are still many earlier survivors with chronic hepatitis C, and given the high prevalence of chronic GVHD, it is likely that some patients may harbor both conditions. The distinction of these two entities can be difficult to discern, especially when there is marked portal fibrosis, though destruction of the bile ducts is much more a feature of chronic GVHD. [5] Chronic viral hepatitis C does not lead to expression of MHC class II on the bile duct epithelium whereas it does in primary biliary cirrhosis (PBC) [18] and GVHD (unpublished observations}

HCT recipients with chronic C are susceptible to a severe, potentially fatal rebound hepatitis following withdrawal of immunosuppressive therapy, during which time immunosuppression allows for marked viral proliferation unimpeded by the host immune system. [1] As hepatitic flares of GVHD may also be seen during tapering of immunosuppressive drugs used for GVHD prophylaxis, deciding whether the transaminitis is related to a flare of viral hepatitis or GVHD is crucial, since GVHD requires immunosuppressive treatment to prevent progressive liver injury The acute exacerbations of hepatitis C are usually self-limited and rarely lead to fulminant rebound hepatitis C. However, with followup in long- term HCT survivors, the cumulative incidence of cirrhosis increased from 0.6% at 10 years to 3.8% at 20 years. [9, 19] Unless there is good evidence of active GVHD in other organs, liver biopsy will be required to access the need for anti viral therapy. PCR viral copy number is useful in determining the impact of the hepatitis C infection.

Distinguishing a flare of chronic GVHD from Hepatitis B in an HCT recipient who develops abnormal AST or ALT at the time of tapering of immunosuppressive drugs, particularly if there is evidence of GVHD and the presence of HbsAg or HBV DNA in the serum poses a similar diagnostic dilemma with a far greater risk of fatal rebound hepatitis. Liver biopsy with immunohistology to access the amount of staining of core and surface antigens is crucial to determine the dominant process. However, both conditions may coexist and require treatment. [2]

Cytomegalovirus With of the development of effective antiviral therapies and effective screening of CMW negative blood products stem cells, disseminated CMV disease has become uncommon except among recipients of HCT from unrelated donors, T-cell depleted grafts or those with severe GVHD. [1] Liver involvement with CMV virtually always accompanies disseminated disease detectable by antigenemic tests, almost never appears as an isolated liver disease, and only rarely results in significant liver dysfunction. [20] CMV hepatitis in immunocompetent patient resembles anicteric mononucleosis.The usual histology of CMV liver infection in immunoincompetent (HCT)patients include sinusoidal neutrophilic microabscesses, intranuclear and intracytoplasmic inclusions, including those found in bile ducts with sparse inflammation [21]. PCR of liver tissue, the most sensitive test for detection, cannot differentiate among viremia, latent infection or true hepatic infection. Informative diagnostic modalities include in situ hybridization, shell vial culture and immunocytochemistry using monoclonal antibodies against the viral proteins. [2]

Drug liver injury: This must always be considered in this setting and should also include a careful clinical history to exclude herbal medicines. Any drugs with known or suspected hepatotoxicity should be withdrawn.

Autoimmune Cholangiopathies: Epstein et al. with Dame Sheila Sherlock published an article in Lancet 24 years ago in which they posited that chronic GVHD, a dry gland disease that shares many features with PBC) , may share common a immunopathogenesis. [22] The initial reports of chronic hepatic GVHD noted a variety of low titer auto-antibodies to ANA and AMA and ASMA. However, a 1998 study by Quaranta et al, using sera and liver biopsies from patients with chronic hepatic GVHD indicated that the AMA likely represented false positive passive transfer of antibody. Furthermore, when liver biopsies with chronic GVHD were immunostained with a specific monoclonal antibody for PBC epithelium, there was no reactivity. [23]

Chronic GVHD is associated with several autoimmune disorders including myasthenia gravis, bullous pemphigoid, and polymyositis/polyserositis. At this time, I am unaware of any clearly documented cases of primary biliary cirrhosis or primary sclerosing cholangitis occurring after HCT. In contrast, a few anecdotal cases of autoimmune hepatitis(AIH) have been reported following HCT. [9] Certainly the lymphoplasmacytic infiltrate in our patient requires consideration. Such patients with AIH, however, would be expected to have the typical antibody profile and would be unlikely to have such high levels of aminotransferases. [24] None of the chronic GVHD sera evaluated by Quaranta et al, were reactive to liver/kidney or thyroidal microsomes. [23]

Concluding Remarks
The incidence of chronic GVHD has increased with the widespread use of peripheral blood stem cells as the source of the allogeneic graft, the increasing use of mismatched related and matched unrelated donors, and the use of non-myeloablative(NMAB) conditioning regimens in older populations. [25] This can result in a 70% cumulative incidence of chronic GVHD requiring treatment. In addition, NMAB HCT is associated with the use of fewer systemic immunosuppressants in the first three months and delayed institution of steroid treatment for GVHD, leading to the syndrome of severe late onset skin and gut GVHD following tapering of immunosuppression. As a result we are now seeing many more cases with hepatitic onset of chronic GVHD. What was once rare is now commonplace and needs be considered within the setting of "hepatitis" developing late after HCT

References

  1. Sale GE, Shulman HM, Hackman RC. Pathology of hematopoietic cell transplantation. In:Blume KG, Forman SJ, Appelbaum FR, eds.Thomas' Hematopoietic Cell Transplantation (Third Edition). Blackwell Scientific Publishers, MA, in press.
  2. Strasser SI, McDonald GB. Gastrointestinal and hepatic complications. In:Blume KG, Forman SJ, Appelbaum FR, eds.Thomas' Hematopoietic Cell Transplantation (Third Edition). Blackwell Scientific Publishers, MA, in press.
  3. Strasser SI, Shulman HM, Flowers ME, Reddy, R, Margolis DA, Prumbaum M, Seropian SE, McDonald GB. Chronic graft-versus-host disease of the liver: presentation as an acute hepatitis. Hepatology 2000;32(6):1265-1271.
  4. Akpek G, Boitnott JK, Lee LA, Vogelsang GB, et al. Hepatitic variant of graft-versus-host disease after donor lymphocyte infusion. Blood 2002;100(1):3903-3907.
  5. Shulman HM, Sharma P, Amos D, et al. A coded histologic study of hepatic graft-versus-host disease after human marrow transplantation. Hepatology 1988;8:463-470.
  6. Freese DK, Snover DC, Sharp HL, Gross CR, Savick SK, Payne WD. Chronic rejection after liver transplantation: a study of clinical, histopathological and immunological features. Hepatology 1991;13:882-891.
  7. Stechschulte DJJr, Fishback JL, Emami A, Bhatia P. Secondary biliary cirrhosis as a consequence of graft-versus-host disease. Gastroenterology 1990;98:223-225.
  8. Shulman HM. Graft-versus-host disease-induced cirrhosis: possible mechanisms (editorial). Hepatology 1987:7(6):1385-1386.
  9. Strasser SI, Sullivan KM, Myerson D, Spurgeon CL, Storer B, Schoch HG, Murakami CS, McDonald GB. Cirrhosis of the liver in long-term marrow transplant survivors. Blood 1999;93(10):3259-3266.
  10. Howell CD, Li J, Roper E, Kotzin BL. Biased liver T cell receptor Vβ repertoire in a murine graft-versus-host disease model. J Immunol 1995;155:2350-2358.
  11. Lerner KG, Kao GF, Storb R, Histopathology of graft-versus-host reaction in human recipients of marrow from HLA matched sibling donors. Transplant Proc 1974; 6 398-393
  12. Fox ES, Broitman SA. Thomas P. Biology of disease. Bacterial endotoxins and the liver. Lab Invest 1990:63:733-741
  13. Lie KW, Ma SY, Au WY, et al. Hepatitic graft-versus host disease after hemopoietic stem cell transplantation: clinicopathological features and prognostic implications. Blood ;2003: 102:A2653
  14. Couriel D, Carrasco A, Saliba R, et al. Graft-versus-host disease of the liver: a clinicopathological analysis. Blood;2003: 102:A2634
  15. Ichiba T, Teshima T, Kuick R, Misek DE, Ferrara JLM, et al. Early changes in gene expression profiles of hepatic GVHD uncovered by oligonucleotide microarrays. Blood 2003;102(2):763-771.
  16. Braun M, Wang H, Chai NN, Wang C, Vierling J. Proinflammatory cytokines IFNГ and TNFα induce gene expression in mouse immortalized biliary epithelial cells (IBECS) that increases susceptibility to immunologic destruction by T cells. Hepatology October 2003; 38(4)(supp 1):600A.
  17. Bründler, MA, Rodriguez-Baez N, Jaffe R, Weinberg AG, Barton Rogers B. Adenovirus ascending cholangiohepatitis. Pediatric and Developmental Pathology 2003;6:156-159.
  18. Kaji K, Nakanuma Y, Sasaki M, Unoura M, Kobayashi K, Nonomura A, Tsuneyama K, Van de Water J, Gershwin ME. Hepatitic bile duct injuries in chronic hepatitis C: histopathologic and immunohistochemical studies. Modern Pathology 1994;7(9):937-945.
  19. Strasser SI, Myerson D, Spurgeon CL, et al. Hepatitis C virus infection after bone marrow transplantation: A cohort study with 10 year followup Hepatology 1999;29:1893-1899.
  20. Rees GM, Sarmiento JI, Myerson D, Coen D, Meyers JD, McDonald GB. Cytomegalovirus hepatitis in marrow transplant patients: clinical, histologic and histochemical analysis. Gastroenterology 1990:98:A470 Abstract.
  21. Snover DC, Hutton S, Balfour HHJr, et al. Cytomegalovirus infection of the liver in transplant recipients. J Clin Gastroenterol 1987;9:659-665.
  22. Epstein O, Thomas HC, Sherlock S. Primary biliary cirrhosis is a dry gland syndrome with features of chronic graft-versus-host disease. Lancet 1980;1:1166-1168.
  23. Quaranta S, Shulman H, Ahmed A, Shoenfeld Y, et al. Autoantibodies in human chronic graft-versus-host disease after hematopoietic cell transplantation. Clinical Immunology 1999;9(1):106-116.
  24. Vergani D, Mieli-Vergani G. Autoimmunity after liver transplantation. Hepatology 2002; 36(2):271-276.
  25. Mielcarek M, Martin PJ, Leisenring W, Flowers MED, Storb R, et al. Graft-versus-host disease after nonmyeloablative versus conventional hematopoietic stem cell transplantation. Blood 2003;102(2):756-762.