EXPERIMENTAL AND CLINICAL TRANSPLANTATION

A 49-year-old white man with no past medical history was diagnosed with stage IVB, grade 3A follicular lymphoma with a Follicular Lymphoma International Prognostic Index score of 4. He received 6 cycles of R-CHOP (rituximab plus cyclophosphamide, hydroxydaunorubicin, vincristine, prednisone) and achieved complete remission. However, 6 months later, he noticed several rapidly growing subcutaneous nodules. Biopsy of a left anterior abdominal wall nodule showed transformation to diffuse large B-cell lymphoma (DLBCL), germinal center B-cell subtype, and was negative for MYC gene rearrangement. He was treated with 3 cycles of R-DHAP (rituximab plus dexamethasone, cytosine arabinoside, and cisplatin), with a partial remission after cycle 2. He subsequently underwent autologous stem cell transplant using carmustine, etoposide, cytarabine, and melphalan conditioning. A planned tandem matched unrelated donor allogeneic hematopoietic cell transplant (allo-HCT) was performed 3 months after the autologous transplant, using a reduced-intensity conditioning regimen of total lymphoid irradiation (800 cGy in 10 fractions) and antithymocyte globulin (1.25 mg/kg per day for 5 days)¹ and tacrolimus with mycophenolate mofetil for graft-versus-host disease (GVHD) prophylaxis. A positron emission tomography scan immediately before allo-HCT showed disease progression in the lung, lymph nodes, and subcutaneous nodules. Nineteen days after allo-HCT, the left anterior abdominal wall nodule was still palpable and 1 cm is size. Two weeks later, on day 33 after transplant, a positron emission tomography scan (Figure 1A) showed interval progression of the disease with increase in size and fluorodeoxyglucose avidity in multiple abdominal wall nodules (maximum size 1.4 cm), a new left axillary lymph node, and a new lung nodule (8 mm). A core biopsy of the left anterior abdominal wall nodule confirmed DLBCL. Bone marrow biopsy and short tandem repeat chimerism test showed trilineage maturation and full donor chimerism. The patient’s creatinine and lactate dehydrogenase levels at this time were 2.00 g/dL and 366 U/L (reference range, 100-250 U/L) with a therapeutic tacrolimus trough level. His only other immunosuppressive medication was mycophenolate mofetil, with no evidence of GVHD.

Considering the relatively small burden of the relapsed disease and elevated creatinine levels, both mycophenolate mofetil and tacrolimus were discontinued, with plans to initiate sirolimus once creatinine levels returned to reference ranges. The rationale for this plan was the previously described antilymphoma effect resulting from inhibition of the mammalian target of rapamycin pathway,^2-4 the well-established use of sirolimus as GVHD prophylaxis, and the acute kidney injury that necessitated at least a temporary discontinuation of tacrolimus. Five days after discontinuation of mycophenolate mofetil and tacrolimus, the patient presented with new oral ulcers, generalized nondesquamating erythroderma, hypotension, and fever. Liver function tests were normal. A clinical diagnosis of acute GVHD grade III was established. Interestingly, the abdominal wall nodule was no longer palpable. Given the apparent potency of the ongoing graft-versus-lymphoma effect, only a single intravenous dose of hydrocortisone (50 mg) was administered, which resulted in prompt resolution of hypotension, suggesting that the septic picture was due to cytokine release due to severe acute GVHD and, perhaps to a lesser extent, tumor cell kill due to the graft-versus-lymphoma effect. Although the patient was empirically treated with antibiotics, comprehensive infectious work-up was negative and he defervesced after the steroid pulse. However, mouth ulcers progressed on the following day, and he was treated with high-dose glucocorticoids (1 mg/kg/day intravenous methylprednisolone) and sirolimus. His acute GVHD showed no response to this treatment after 3 days, and photopheresis was initiated, resulting in a partial response. A positron emission tomography scan on day 101 showed no evidence of lymphoma (Figure 1B). The patient is now 1 year post-allo-HCT and in complete remission. He has moderate chronic GVHD of the skin and eyes and is maintained on tacrolimus (0.5 mg/day), a slowly tapering course of prednisone (currently 10 mg/day), and photopheresis every 4 weeks.

The graft-versus-lymphoma effect is generally considered to be weaker in DLBCL versus several other hematologic malignancies such as follicular lymphoma and chronic myeloid leukemia.^5-7 This differential responsiveness to immunomodulation may be related to relatively weak antigen presentation in DLBCL or more aggressive growth kinetics of the disease, which can outpace the immune effect. Bishop and colleagues reviewed responses to donor lymphocyte infusion (DLI) or immunosuppression withdrawal (and no chemotherapy) in 11 DLBCL patients who were not in complete remission on their day 100 evaluation (10 patients) or who had subsequently relapsed (1 patient).⁸ One patient who received a DLI achieved a sustained complete response and had no acute or chronic GVHD. Among the 10 patients who were treated only with withdrawal of immunosuppression, 5 patients achieved a complete response (1 with acute grade II-IV GVHD and extensive chronic GVHD and 2 with extensive chronic GVHD) and 1 patient showed a nonsustained partial response (with acute grade II-IV GVHD). One of the patients who achieved a complete response after immunosuppression withdrawal relapsed 23 months later, with others not relapsing during study follow-up. The intervention (DLI or immunosuppression withdrawal) in this study was performed on or later than day 100. Unfortunately, similar to most other hematologic malignancies, response to DLI or immunosuppression withdrawal tends to correlate with development of GVHD.

The case presented here represents a unique scenario from 2 aspects. First, the intervention following relapse post-allo-HCT was performed on day 33, significantly earlier than in previously reported cases. The intervention was discontinuation of both tacrolimus and mycophenolate mofetil, which is rarely done this early after transplant. Second, the graft-versus-lymphoma response was extremely rapid and profound, manifesting clinically in a few days and resulting in a radiographic complete response. We hypothesize that rapid expansion of alloreactive donor T cells resulted in hemodynamic instability and shock, which, given the prompt response to steroids, was probably a manifestation of a “cytokine storm.” Clinically significant response to DLI (or withdrawal of immunosuppression) usually takes several weeks to occur.⁹ The reason for the remarkably rapid response in our patient is a matter of speculation, but we can think of at least 1 plausible mechanism. At the time of early withdrawal of immunosuppression, a large number of preexisting alloreactive T cells in the graft may have still been present. Stimulation of these T cells caused a profound graft-versus-lymphoma effect and, unfortunately, a similarly powerful GVHD reaction. In contrast to the unusually rapid onset of the immune effect in our patient, GVHD after DLI or immunosuppression withdrawal typically occurs after a median of approximately 1 month.10 Although we demonstrated a potent graft-versus-lymphoma response in our patient by simply withdrawing immunosuppression, more sophisticated ways of immunomodulation to treat early relapse of aggressive lymphomas after allo-HCT need to be developed in future studies.

Volume: 14 Issue: 4 August 2016

FULL TEXT