Anti-tumor necrosis factor-α for the treatment of steroid-refractory acute graft-versus-host disease

Allogeneic stem cell transplantation has been increasingly performed for a variety of hematologic diseases. Clinically significant acute graft-versus-host disease (GVHD) occurs in 9 to 50% of patients who receive allogeneic grafts, resulting in high morbidity and mortality. There is no standard therapy for patients with acute GVHD who do not respond to steroids. Studies have shown a possible benefit of antiTNF-α (infliximab) for the treatment of acute GVHD. We report here on the outcomes of 10 recipients of related or unrelated stem cell transplants who received 10 mg/kg infliximab, iv, once weekly for a median of 3.5 doses (range: 1-6) for the treatment of severe acute GVHD and who were not responsive to standard therapy. All patients had acute GVHD grades II to IV (II = 2, III = 3, IV = 5). Overall, 9 patients responded and 1 patient had progressive disease. Among the responders, 3 had complete responses and 6 partial responses. All patients with cutaneous or gastrointestinal involvement responded, while only 2 of 6 patients with liver disease showed any response. None of the 10 patients had any kind of immediate toxicity. Four patients died, all of them with sepsis. Six patients are still alive after a median follow-up time of 544 days (92-600) after transplantation. Considering the severity of the cases and the bad prognosis associated with advanced acute GVHD, we find our results encouraging. AntiTNF-α seems to be a useful agent for the treatment of acute GVHD. Correspondence


Introduction
Transplantation with allogeneic bone marrow or peripheral blood has been increasingly performed for a variety of hematologic diseases.Clinically significant acute graft-versus-host disease (GVHD) occurs in 9 to 50% of patients who receive allogeneic grafts, and remains an important cause of morbidity and mortality.Patients with severe, grades III to IV, acute GVHD are usually treated with intensification of their im- munosuppressive regimen, typically with high-dose steroids.However, there is no standard therapy for patients with acute GVHD who fail to respond to an initial course of corticosteroids (1).
We report here on the outcomes of 10 patients treated with infliximab against severe acute GVHD at our institution.

Patients
Between July 2003 and January 2006, 10 patients with a median age of 13 years (range: 4-36 years) with steroid-refractory acute GVHD received infliximab at our institution.These patients were retrospectively analyzed.The characteristics of the patients and transplants are shown in Table 1.

Transplants
GVHD prophylaxis was done with cyclosporine and prednisone in all patients who received cord blood, and with cyclosporine and methotrexate in those who received either bone marrow or peripheral blood stem cells.The characteristics of the transplants are shown in Table 1.

Graft-versus-host disease
Acute GVHD was diagnosed within 100 days of allogeneic transplantation or donor lymphocyte infusion.Acute GVHD was staged and graded according to modified Glucksberg criteria (11).GVHD that developed 100 days after allogeneic transplantation was considered to be chronic GVHD.At the start of infliximab treatment, 2 patients had developed grade II acute GVHD, while 8 had developed either grade III or IV acute GVHD.The liver was compromised in 6 (grades 1-4; median = 2), the gastrointestinal tract (GIT) in 8 (grades 1-4; median = 3.5), and the skin in all (grades 1-4; median = 3).Sicca syndrome was present in one patient, and isolated xerophthalmia in another.The data about GVHD are shown in Table 2.

Treatment of graft-versus-host disease
Before starting infliximab, all 10 patients were receiving a combination of methylprednisolone, 2 mg kg -1 day -1 , iv, with a calcineurin inhibitor (9 on cyclosporine, 1 on tacrolimus).Patients 1, 3, and 10 were also on mycophenolate mofetil, which is the drug of choice in our service, once steroid refractory acute GVHD occurs.Patient number 3 also received basiliximab and was started on extracorporeal photopheresis for severe cutaneous involvement.Patients 5, 6, 7, 8, and 9 were switched from cyclosporine to tacrolimus in view of the reports of better results in unrelated bone marrow transplants with the latter drug.Since patient number 6 had developed hemolyticuremic syndrome attributed to tacrolimus and was not able to start mycophenolate mofetil because of pancytopenia, he was started on thalidomide.All 10 patients received infliximab, 10 mg/kg, iv, once weekly for a median of 3.5 doses (range: 1-6).
After the start of infliximab, all patients continued to receive methylprednisolone, 2 mg kg -1 day -1 , iv, in divided doses.Cortico-steroids were progressively tapered and discontinued as GVHD improved.The median time between the introduction of corticosteroids and the start of infliximab was 20.5 days (10-33) (Table 2).Details are shown in Table 3.

Response
The overall response represented the responses of the skin, GIT and liver.Complete response was defined as resolution of GVHD in all evaluable organs, partial response as any improvement in at least one evaluable organ without deterioration of others, and no response as the absence of any change or any situation other than complete or partial response, or progressive disease (7).

Evaluation of the response
Overall, 9 patients responded and one patient had progressive disease.Among the responders, 3 had a complete response and 6 had a partial response.Only 2 of 6 patients with liver disease showed any grade of liver improvement, while all patients with either cutaneous or GIT involvement had responses in these organs.Seven patients had progressive chronic GVHD, 3 were not evaluable (2 died before day 100, and 1 was on day 92 by the time of the analysis).The data are summarized in Table 2.

Toxicity and infections
None of the 10 patients had any kind of infusion-related, immediate toxicity.Nevertheless, all of them had a diagnosis of infection.Four had febrile neutropenia, 4 had cutaneous infections, 3 had a clinical diagnosis of sinusitis confirmed by computed tomographic scan, 3 had pneumonia, 1 had hemorrhagic cystitis, 1 had encephalitis, and 1 had oral candidiasis.Seven bacterial infections were microbiologically documented in four patients: Pseudomonas sp (N = 1), Streptococcus sp (N = 2), Staphylococcus sp (N = 2), and Enterococcus sp (N = 2).One patient had pneumonia, with Pneumocystis carinii identified in the sputum.Eight patients developed cytomegalovirus infection detected by antigenemia, but none progressed to cytomegalovirus disease.Fungal infections occurred in 3 patients: 1 patient had a subcutaneous abscess due to Aspergillus sp, which resolved after surgical drainage and specific antifungal therapy.The other 2 patients died of sepsis due to Fusarium sp; 1 had only cutaneous lesions and the other had cutaneous as well as cerebral involvement.Data about infectious complications are summarized in Tables 4 and 5.

Outcomes
Four patients died.Two of them, who had controlled GVHD, died of fungal infections.One patient whose GVHD had also been controlled died with bacterial sepsis.The last one had progressive GVHD and also died due to bacterial sepsis.Six patients are still alive, with chronic GVHD, at a median of 544 days (92-600) after the transplant.All 6 had either a partial or complete response to infliximab.Other level (ng/mL) (mg kg -1 day -1 ) (g/day) (mg kg -1 day -1 ) level (ng/mL) (mg kg -1 day -1 ) (g/day) (mg kg -1 day -1 ) Anti-TNF-α and steroid-resistant GVHD www.bjournal.com.br

Discussion
Despite the use of prophylactic immunosuppressive regimens, grades II-IV acute GVHD occur in 30 to 80% of allogeneic transplant recipients, with greater frequency after transplantation from HLA non-identical or unrelated donors (12)(13)(14).
The pathophysiology of acute GVHD is triphasic (15,16).The initial phase results from the damage induced by the conditioning regimen in host tissues.Damaged tissues secrete inflammatory cytokines, including interleukin-1, TNF-α, and interferon-gamma.In the second phase, recipient antigenpresenting cells trigger the activation of donor-derived T cells, which expand and differentiate into effector cells (17,18).In the third phase, the effector phase, activated donor T cells mediate cytotoxicity against target host cells through Fas-Fas ligand interactions, perforin-granzyme B and cytokine production, including production of TNF-α (19,20).This sequence of events leads to the tissue damage that is characteristic of acute GVHD.
The standard of primary therapy for acute GVHD is considered to be methylprednisolone for 14 days, followed by a steroid taper (21).However, reports from the 1990's indicated that 80% of responses to corticosteroids, incomplete or complete, were not sustained.Failure was associated with a 75% non-relapse mortality rate (22,23).When steroid therapy fails to control the manifestations of acute GVHD, there is no standard approach to secondary GVHD therapy.
TNF-α is a key cytotoxic and proinflammatory cytokine involved in the pathogenesis of GVHD.TNF-α induces apoptosis, activates macrophages, granulocytes and lymphocytes, and produces a cascade of other inflammatory cytokines.TNF-α also increases the expression of HLA molecules and facilitates cytolysis mediated by Tlymphocytes (24).High levels of TNF-α have been correlated with the occurrence of acute and chronic GVHD (3,24).
In our series, 9 of 10 patients responded to infliximab, 3 with complete and 6 with partial responses.When considering organ involvement, 9 of 10 patients with cutaneous involvement responded, 7 of 8 with intestinal GVHD, and only 2 of 6 with liver disease.Couriel et al. (7) reported a 67% overall response rate.These were primarily complete responses, and the highest response rates were observed in the GIT and skin.In the study of Patriarca et al. (9) there was a 59% response rate to infliximab, with 19% complete and 40% partial responses.Ac- cordingly, the best responses were also observed in the GIT.In the case series of Kobbe et al. (10), 3 of 4 patients responded to infliximab.None of our patients had any kind of toxicity attributable to infliximab.Some studies have described adverse events, which include acute infusional toxicity, serum sickness, development of autoantibodies (21), and neurological (30) or hepatic complications (31), but most of the series do not report adverse events (7,9,10).
Fungal infections, which were severe, occurred in 3 of our patients.TNF-α is essential for immune defense, playing a major role in the recruitment of inflammatory cells to the site of infection and in the formation and maintenance of granulomas (32).Marty et al. (1) reported on 5 cases of fungal infection in 11 patients who had received infliximab for the treatment of acute GVHD.In other studies, the incidence of fungal infections ranged from approximately 6% (2 of 32 patients) (9) to 28% (6 of 21 patients) (7).The possibility that infliximab contributed to a higher risk of lethal infections cannot be ruled out; however, acute GVHD and its immunosuppressive treatment already put these patients at higher risk for these complications.Tuberculosis has also been described after the use of infliximab (32), but none of our patients developed this complication.
Our study has important limitations.First, the number of patients is small and the distribution of strong risk factors for GVHD such as the use of unrelated donors is not homogeneous.There was no consensus on the choice of different immunosuppressive agents once resistance to corticosteroids was recognized.Second, the rationale for the introduction of infliximab was not the same in every patient.Third, it is only a retrospective, uncontrolled study.Although the response to infliximab appeared to be significant, one cannot exclude the possibility of a late response to corticosteroids or any of the other immunosuppressive treatments simultaneously employed.Nevertheless, in view of the high morbidity and mortality of steroidresistant GVHD, we believe that these results may help justify the design of controlled, randomized clinical trials to investigate the subject further.
In summary, infliximab was well tolerated and appears to have a role in the treatment of acute GVHD, especially when the gastrointestinal tract is compromised.Randomized studies on the use of infliximab for the treatment of acute GVHD are underway (7) and the results are eagerly awaited.

Table 1 .
Characteristics of patients and transplants.

Table 5 .
Sites of infection.