Graft-Versus-Host Disease

Liver graft-versus-host disease after donor lymphocyte infusion for relapses of hematologic malignancies post allogeneic hematopoietic stem cell transplantation

Article metrics


Graft-versus-host disease (GVHD) is the commonest complication after donor lymphocyte infusion (DLI). In 19 patients undergoing DLI for relapses of hematologic malignancies post hematopoietic stem cell transplantation (HSCT), 11 developed GVHD, of whom nine had isolated liver involvement, and two had liver and skin involvement. The clinical diagnosis of liver GVHD was hepatitic in six patients (55%) and classical in five patients (45%). Patients with GVHD post-DLI showed a different clinical pattern when compared to a cohort of 106 cases of GVHD post-HSCT, in having significantly more isolated liver involvement (9/11 vs 17/106, P<0.001), and less skin (2/11 vs 80/106, P<0.001) and gut (0/11 vs 28/106, P<0.001) involvement. However, liver GVHD post-DLI and post-HSCT had comparable patient characteristics, underlying diseases, clinical subtypes (classical and hepatitic) and response to treatment.

Donor lymphocyte infusion (DLI) is an important treatment for hematologic malignancies relapsing after allogeneic hematopoietic stem cell transplantation (HSCT).1 Currently, DLI is used in post-HSCT relapses in many hematologic malignancies, with chronic myeloid leukemia (CML) showing the best response,2, 3 followed by multiple myeloma4, 5 and acute myeloid leukemia.2

The commonest complication after DLI is graft-versus-host disease (GVHD), occurring in 20–60% of patients.1, 5, 6, 7 The best predictor for GVHD post-DLI is a high T-cell dose.6, 7, 8 Few studies have compared the pattern of GVHD after HSCT with post-DLI. Recently, a hepatitic variant of liver GVHD9 has been found to be more prevalent after DLI, which is apparently associated with a poor response to treatment and prognosis.10

In this study, we evaluated 19 patients treated with DLI for leukemic relapses post-HSCT, and analyzed the patterns and treatment outcome of liver GVHD in these patients.

Materials and methods


The records of patients undergoing DLI between January 1996 and June 2002 at the Department of Medicine, Queen Mary Hospital, were reviewed. Patients with CML in molecular/cytogenetic relapse, whose blood group was compatible with the donor, were given 1 U of blood venesected from the donor as DLI. For all other patients, as well as CML patients not responding to the first infusion, mononuclear cells (MNC) were collected from donors by processing 7 L of blood through cytapheresis equipment (CS-3000 Plus, Baxter, Deerfield, IL, USA), and were infused directly into the patients. There was no specific target MNC count for each infusion. Depending on the response, patients might receive one or more additional MNC infusions on a weekly basis. The total cellular dose was defined as the cumulative MNC dose/kg transfused.

Clinical and laboratory evaluation

A thorough physical examination was performed and investigations were carried out.11 In patients with impaired liver function tests (LFT), potential viral causes of hepatitis were investigated with the following laboratory tests: hepatitis B surface antigen (HBsAg) (microparticle enzyme immunoassays, Abbott Laboratories, Chicago, IL, USA), anti-hepatitis C virus (HCV) antibody (enzyme immunoassays EIA-3, Abbott Laboratories), cytomegalovirus (CMV)-DNA in white blood cells by polymerase chain reaction (PCR) and standard serological tests for herpes simplex virus and varicellar zoster virus infection. The highest value of bilirubin, alkaline phosphatase (ALP), aspartate aminotransferase (AST) and alanine aminotransferase (ALT) were recorded and analyzed. GVHD affecting other organs were recorded and graded according to standard criteria.12

Study design and definition

Biochemical parameters were used to classify liver GVHD into the classical type, defined as elevation of bilirubin and ALP, with AST and ALT less than 10 times normal, and the hepatitic type, defined as elevated bilirubin and ALP, with AST and/or ALT more than ten times normal.9, 10 To investigate if liver GVHD post-DLI and post-HSCT might be different, patients undergoing HSCT and developing GVHD within the same study period were used for comparison.13

Statistical analysis

Difference between groups was analyzed by t-test for continuous data and the χ2 test or Mann–Whitney test for categorical data (SPSS statistical program, version 10.0). A P-value of less than 0.05 was considered to represent a statistically significant difference.



A total of 19 adult patients received DLI for relapses of hematologic malignancies post-HSCT (Table 1). None of these HSCT was T-cell depleted. Chemotherapy was not given before DLI. DLI was given at a median of 797 (294–2053) days post-HSCT. Six patients received one infusion, nine patients two infusions and four had three or more infusions. The median total dosage of MNC infused was 6.16 × 108 (3.66 × 107 – 6.68 × 108/kg). HBsAg was positive in only one patient. No patient was positive for anti-HCV. A total of 11 patients developed GVHD at a median of 66 (22–138) days post-DLI. Nine patients had isolated liver GVHD, and two had liver and skin GVHD.

Table 1 Clinicopathologic features of 19 patients with or without liver GVHD after donor lymphocyte infusion (DLI)

Clinical correlations of liver GVHD post-DLI

The clinicopathologic features of patients with and without liver GVHD post-DLI are shown in Table 1. None of the parameters examined, including sex, age, underlying disease, types of donor, timing and indications of DLI, pre-existing GVHD, total cell dose infused and the disease response, were related to the development of liver GVHD post-DLI.

Comparison of liver GVHD post-DLI and post-HSCT

During the same period, 320 adults underwent allogeneic HSCT, of whom 106 developed liver GVHD.13 To assess whether liver GVHD post-DLI differed from that post-HSCT, their clinicopathologic features were compared (Table 2). The two groups differed significantly in the patterns of GVHD. Liver GVHD post-DLI compared to post-HSCT showed more frequent isolated hepatic involvement (9/11 vs 17/106, P<0.001), less concomitant skin involvement (2/11 vs 80/106, P<0.001), and no gut involvement (0/11 vs 28/106, P<0.001). At the onset of GVHD, liver GVHD post-DLI and post-HSCT did not differ in bilirubin, AST, ALT and ALP. However, at peak LFT derangement, liver GVHD post-DLI showed a lower bilirubin level than that post-HSCT (42 vs 130 μmol/l, P=0.015). Otherwise, the two groups were comparable with respect to demographic features, underlying diseases, patterns of LFT derangement, clinical types (hepatitic or classical) of liver GVHD and the duration of LFT impairment.

Table 2 Clinicopathologic features and treatment outcome of liver GVHD in 11 post-DLI patients and 106 post-HSCT patients

Treatment outcome of liver GVHD after DLI and HSCT

Patients were treated with cyclosporine and steroid-based immunosuppressive regimens, with azathioprine and mycophenolate mofetil added at the discretion of the attending physicians. Cyclosporine was used in fewer patients and at a lesser cumulative dose in patients post-DLI, but this might reflect a more conservative approach of the attending physicians in treating GVHD, in an attempt to maximize the putative graft-versus-disease effect. The same reason might account for the use of prednisolone at a lower cumulative dose. However, resolution of liver GVHD was comparable in both groups (10/11 vs 88/106, P=not significant). The time for onset, peaking and resolution of LFT derangement was also comparable in the two groups (107 vs 69 days, 33 vs 21 days and 92 vs 75 days, respectively, P=not significant). LFTs returned to normal in all patients post-DLI, but remained abnormal although improved in five post-HSCT patients. A total of one (9%) patient in the post-DLI group and 13 (12%) patients in the post-HSCT group died of GVHD.


GVHD is one of the major causes of morbidity and mortality after HSCT and DLI. GVHD of the skin and gut is clinically evident and easily confirmed. However, the diagnosis of GVHD of the liver may be problematic, as other confounding factors including viral hepatitis, and liver injury related to drugs, treatment regimen and sepsis are often present.14, 15, 16 The problem is accentuated post-DLI, as the pattern and prognosis of liver GVHD in this setting are not clearly delineated. Although liver biopsy is helpful in diagnosis, it is not entirely safe, particularly with low platelet counts. Even with the use of transjugular liver biopsy, the rate of complications has ranged from 1.3 to 20.2%, with a consequent mortality of 0.1–0.5%.17, 18 Therefore, liver GVHD is often diagnosed on clinical grounds, relying traditionally on the presence of cholestatic LFT derangement with marked elevation of bilirubin and ALP as sensitive markers.19, 20 However, with the recent description of a hepatitic type of GVHD after both HSCT9 and DLI,10 the diagnostic algorithm of LFT derangement becomes complicated, as patients with predominantly hepatitic changes may also have to be evaluated for GVHD.

In this study, 11/19 patients (58%) had clinical evidence of GVHD post-DLI, a frequency that was comparable with previous studies.2, 7 Interestingly, all of these cases showed LFT derangement, with only two cases showing concomitant skin involvement. A limitation of this study is the absence of liver biopsy for confirmation of GVHD. However, we had excluded other possible etiologies, including drugs and known viral infections. Furthermore, we have shown previously that, in the absence of a suspicion of hepatitis B and C infection, the clinical diagnosis of GVHD correlated well with histopathologic findings.11 Finally, successful treatment with immunosuppression strongly suggested that the LFT derangement was due to GVHD. Predominant liver involvement in GVHD post-DLI has not been reported before. De Lima et al21 reported that of eight patients developing GVHD post-DLI, all had skin and gut involvement, but only two had liver involvement. Marks et al22 reported that of 27 patients with GVHD post-DLI, skin involvement was the commonest (78%), followed by gut (41%) and liver (26%). In these studies, it was uncertain if hepatitic LFT derangements were also considered as a form of liver GVHD, which might have affected the relative frequency of liver GVHD as compared to skin and gut GVHD. Finally, when our 11 patients with liver GVHD post-DLI were compared with the 106 patients with liver GVHD post-HSCT, they were similar in practically all clinical and biochemical aspects. However, an interesting difference was the preponderance of liver GVHD and the apparent absence of concomitant gut involvement post-DLI. These results suggested that the pathogenetic mechanisms or effector cells of liver GVHD post-DLI might be different from those post-HSCT. However, it must be noted that in HSCT, chemoirradiation is used in conditioning, and GVHD prophylaxis is also administered; neither of which are given with DLI. Furthermore, CML patients constituted more than 70% in the DLI group as compared with about 20% in the HSCT group. Finally, GVHD post-HSCT usually occurs in the setting of continued administration of immunosuppression, and the possible existence of residual toxicity from the preceding conditioning regimen; both of these are absent in patients receiving DLI. These factors might also have contributed to the different patterns of GVHD.

Recently, Akpek et al10 reported 22 cases of liver GVHD post-DLI, with 11 cases (50%) being classified as hepatitic GVHD. They reported a poor overall prognosis for hepatitic GVHD, with about 60% showing progression despite high-dose combination immunosuppression, and death attributable to GVHD in more than 30% of cases. In our smaller series, the frequency of hepatitic GVHD at 55% was similar. However, hepatitic changes in our patients had apparently little impact on treatment and prognosis, as more than 80% of our cases responded to standard immunosuppression, an outcome that was comparable to liver GVHD post-HSCT. These results were also similar to those described by Strasser et al,9 who reported in their original study on hepatitic GVHD that patients showed good responses to immunosuppression. Our results were therefore at variance with those of Akpek et al,10 but further studies in larger numbers of patients are required to validate our observations. Whether a difference in patient populations might be relevant has to be determined. Our study comprised mainly CML patients, as we treated relapses of other hematologic malignancies with a second full or nonmyeloablative transplant. Furthermore, scheduling and dosage of DLI might also affect the outcome of hepatitic GVHD.6 The exact dosage and schedule of DLI in the studies of Akpek et al10 have not been stated. Previous studies have shown that an escalating dose regimen might be associated with less GVHD than the bulk dose regimen. How this might have altered the outcome of liver GVHD remains unclear. In our study, the dose of DLI was gauged by response, an approach similar to an escalating dose regimen. This might have led to a more favorable outcome.

Finally, because only a few cases of hepatitic GVHD post-HSCT and post-DLI have been reported, the prognostic impact of predominantly hepatitic LFT derangement requires further clarifiication. This is of clinical importance, because the grading of liver GVHD is currently only dependent on bilirubin level. Whether aminotransferase levels will also need to be taken into consideration in clinical grading of liver GVHD warrants further investigations.


  1. 1

    Kolb HJ, Mittermuller J, Clemm C et al. Donor leukocyte transfusions for treatment of recurrent chronic myelogenous leukemia in marrow transplant patients. Blood 1990; 76: 2462–2465.

  2. 2

    Collins Jr RH, Shpilberg O, Drobyski WR et al. Donor leukocyte infusions in 140 patients with relapsed malignancy after allogeneic bone marrow transplantation. J Clin Oncol 1997; 15: 433–444.

  3. 3

    Dazzi F, Szydlo RM, Cross NC et al. Durability of responses following donor lymphocyte infusions for patients who relapse after allogeneic stem cell transplantation for chronic myeloid leukemia. Blood 2000; 96: 2712–2716.

  4. 4

    Lokhorst HM, Schattenberg A, Cornelissen JJ et al. Donor leukocyte infusions are effective in relapsed multiple myeloma after allogeneic bone marrow transplantation. Blood 1997; 90: 4206–4211.

  5. 5

    Sehn LH, Alyea EP, Weller E et al. Comparative outcomes of T-cell-depleted and non-T-cell-depleted allogeneic bone marrow transplantation for chronic myelogenous leukemia: impact of donor lymphocyte infusion. J Clin Oncol 1999; 17: 561–568.

  6. 6

    Dazzi F, Szydlo RM, Craddock C et al. Comparison of single-dose and escalating-dose regimens of donor lymphocyte infusion for relapse after allografting for chronic myeloid leukemia. Blood 2000; 95: 67–71.

  7. 7

    Guglielmi C, Arcese W, Dazzi F et al. Donor lymphocyte infusion for relapsed chronic myelogenous leukemia: prognostic relevance of the initial cell dose. Blood 2002; 100: 397–405.

  8. 8

    Mackinnon S, Papadopoulos EB, Carabasi MH et al. Adoptive immunotherapy evaluating escalating doses of donor leukocytes for relapse of chronic myeloid leukemia after bone marrow transplantation: separation of graft-versus-leukemia responses from graft-versus-host disease. Blood 1995; 86: 1261–1268.

  9. 9

    Strasser SI, Shulman HM, Flowers ME et al. Chronic graft-versus-host disease of the liver: presentation as an acute hepatitis. Hepatology 2000; 32: 1265–1271.

  10. 10

    Akpek G, Boitnott JK, Lee LA et al. Hepatitic variant of graft-versus-host disease after donor lymphocyte infusion. Blood 2002; 100: 3903–3907.

  11. 11

    Ma SY, Au WY, Ng IO et al. Role of liver biopsy in the management of liver dysfunction after hematopoietic stem cell transplantation in a hepatitis B virus prevalent patient population. Transplantation 2003; 76: 169–176.

  12. 12

    Glucksberg H, Storb R, Fefer A et al. Clinical manifestations of graft-versus-host disease in human recipients of marrow from HLA-matched sibling donors. Transplantation 1974; 18: 295–304.

  13. 13

    Ma SY, Au WY, Ng IO et al. Hepatitic graft-versus-host disease after hematopoietic stem cell transplantation: clinicopathologic features and treatment outcome. Transplantation 2003; (in press).

  14. 14

    Farthing MJ, Clark ML, Sloane JP et al. Liver disease after bone marrow transplantation. Gut 1982; 23: 465–474.

  15. 15

    McDonald GB, Shulman HM, Sullivan KM, Spencer GD . Intestinal and hepatic complications of human bone marrow transplantation. Part 1. Gastroenterology 1986; 90: 460–477.

  16. 16

    McDonald GB, Shulman HM, Sullivan KM, Spencer GD . Intestinal and hepatic complications of human bone marrow transplantation. Part 2. Gastroenterology 1986; 90: 770–784.

  17. 17

    McAfee JH, Keeffe EB, Lee RG, Rosch J . Transjugular liver biopsy. Hepatology 1992; 15: 726–732.

  18. 18

    Lebree D, Goldfarb G, Degou C et al. Transvenous liver biopsy: an experience based on 1000 hepatic tissue samplings with this procedure. Gastroenterology 1982; 83: 338–340.

  19. 19

    Strasser SI, McDonald GB . Gastrointestinal and hepatic complications. In: Thomas ED, Blume KG, Forman SJ (eds.). Haematopoietic Cell Transplantation. Blackwell Science: Boston, MA, 1999; pp 627–658.

  20. 20

    Yasmineh WG, Filipovich AH, Killeen AA . Serum 5′ nucleotidase and alkaline phosphastase-highly predictive liver function tests for the diagnosis of graft-versus-host disease in bone marrow transplant recipients. Transplantation 1989; 48: 809–814.

  21. 21

    de Lima M, Bonamino M, Vasconcelos Z et al. Prophylactic donor lymphocyte infusions after moderately ablative chemotherapy and stem cell transplantation for hematological malignancies: high remission rate among poor prognosis patients at the expense of graft-versus-host disease. Bone Marrow Transplant 2001; 27: 73–78.

  22. 22

    Marks DI, Lush R, Cavenagh J et al. The toxicity and efficacy of donor lymphocyte infusions given after reduced-intensity conditioning allogeneic stem cell transplantation. Blood 2002; 100: 3108–3114.

Download references


This study is supported in part by the Kadoorie Charitable Foundation.

Author information

Correspondence to Y L Kwong.

Rights and permissions

Reprints and Permissions

About this article


  • liver GVHD
  • DLI
  • HSCT

Further reading