Vigorous inflammatory responses in noninfectious pulmonary complication induced by donor lymphocyte infusion

Abstract

BACKGROUND

Donor lymphocyte infusion (DLI) is used for treatment of hematologic malignancy relapse or mixed chimerism after allogeneic hematopoietic stem cell transplantation. Although graft‐versus‐host disease is well recognized as one of the adverse effects of DLI, there are limited reports on noninfectious pulmonary complications (NIPCs) after DLI.

CASE REPORT

A 55‐year‐old woman with acute myeloid leukemia received DLI for conversion from recipient predominant to complete donor chimerism on Day +193 after allogeneic HSCT. Eight weeks later, she complained of dyspnea with fever; chest computed tomography revealed diffuse, bilateral, ground glass opacity and reticular appearance. She was diagnosed as having NIPC based on serum and bronchoalveolar lavage fluid (BALF) findings. She was successfully treated with prednisolone (PSL) and completely recovered.

DISCUSSION

We analyzed the cell profile from the BALF and 27 cytokines and chemokines in the serum using the Bio‐Plex platform. The cells consisted of recipient predominant macrophages and T cells. The serum cytokine and chemokine profile showed significant elevation of interleukin (IL)−1β, IL‐6, IL‐8, tumor necrosis factor‐α, macrophage inflammatory protein (MIP)−1α, and MIP‐1β, which declined with the improvement of symptoms after initiation of PSL treatment.

CONCLUSION

Inflammatory effectors by recipient cells, rather than allogeneic responses by donor cells, played an important role in the pathogenesis of NIPCs after DLI in the present case.

ABBREVIATIONS

BALF

bronchoalveolar lavage fluid

DLI

donor lymphocyte infusion

IPS

idiopathic pneumonia syndrome

MIP

macrophage inflammatory protein

NIPC

noninfectious pulmonary complication

PSL

prednisolone

VEGF

vascular endothelial growth factor

Allogeneic hematopoietic stem cell transplantation (HSCT) is used to treat hematologic malignant diseases, often providing prolonged   survival; however, relapse is the major cause of treatment failure. Withdrawal of immunosuppressive drugs is the usual initial step, and donor lymphocyte infusion (DLI) is performed for intervention of disease reinduction.1 DLI is also used to convert from mixed to complete donor chimerism for prevention of graft failure or disease relapse.2 The most significant adverse effect of DLI is graft‐versus‐host disease (GVHD), which typically develops 32 to 42 days after infusion in approximately 50% to 60% of patients.3, 4, 5

Among numerous DLI clinical trials,3, 4, 5, 6, 7, 8 only two reports on pulmonary complications induced by DLI have been published. Lee and colleagues9 reported that 9 of 38 patients receiving prophylactic DLI died of idiopathic pneumonia syndrome (IPS). They speculated that the high rate of fatal IPS may be associated with prior radiation to the chest. In another report, Badros and colleagues10 found that interstitial pneumonitis developed in two patients after DLI. Neither of these reports included any detailed clinical information, and the underlying pathophysiologic mechanisms of this pulmonary complication remain unknown. In this report, we present detailed clinical information of a noninfectious pulmonary complication (NIPC) after DLI and further investigated the pathophysiologic mechanisms using serum and bronchoalveolar lavage fluid (BALF) findings.

CASE REPORT

A 55‐year‐old Japanese woman with acute myeloid leukemia received allogeneic bone marrow transplantation (BMT) from an HLA‐DRB1 allele 1 locus‐mismatched unrelated donor in the first remission phase. The conditioning regimen consisted of 3.2 mg/kg busulfan for 4 days and 30 mg/m² fludarabine for 5 days. GVHD prophylaxis consisted of 0.03mg/kg tacrolimus and short‐term methotrexate. Rapid engraftment was observed, and chimerism analysis demonstrated that 89% of whole blood cells were donor derived by Day +29 after BMT. Marrow aspiration on Day +53 revealed no blasts, confirming complete remission. No acute GVHD signs or symptoms were observed.

On Day +74, chimerism analysis of whole blood cells showed decreased donor‐derived cells (66.3%). Concurrent lineage‐specific chimerism analysis demonstrated that only 10.5% of CD3+ cells were donor‐derived cells, although 73.5% of CD3– mononuclear cells (MNCs) and more than 95% of granulocytes were donor derived. Based on these findings, tacrolimus was tapered off and then discontinued; however, neither recovery of donor‐derived CD3+ cells nor GVHD symptoms were observed. Approximately 1 month after discontinuation of the immunosuppressant, DLI (CD3+ dose of 4.5 × 10⁶/kg recipient body weight) was performed from the same donor on Day +193 after allo‐BMT. On Day 14 after DLI, the patient complained of spontaneous itchy and dry skin, with eruption on the face suspected to be GVHD. The skin eruption continued until Day 42 without progression and improved without treatment by Day 50. On Day 56, the patient complained of very mild dyspnea with a low‐grade fever, which gradually worsened until Day 60. Chest radiography on Day 62 showed diffuse shadows in both lower lung fields. Chest computed tomography (CT) on Day 63 revealed diffuse, bilateral, ground glass opacity and reticular appearance predominantly in the lower lung fields (Fig. 1A). Pulmonary function tests showed a decline in forced vital capacity, from 3.21 L to 1.98 L, as well as in percentage of vital capacity from 128% to 80% and almost no change (78% to 80%) in forced expiratory volume in 1 second (FEV1), indicating restrictive lung dysfunction. On Day 78, fiber optic bronchoscopy and a bronchoalveolar lavage were performed. Due to the small sample, histologic findings of the transbronchial lung biopsy only showed trivial alveolar edema without evidence of leukemia infiltration; the specimen did not include bronchiole for differential diagnosis of COP/BOOP or for identification of the BO as chronic GVHD. Analysis of the BALF showed a total cell count of 3.4 × 10⁵ cells/mL, with differential cell counts of 36% macrophages, 63% lymphocytes, and 1% granulocytes. Of the lymphocytes, 98% were CD3+ T cells, and the CD4/8 ratio was 1.72 (CD4, 62.4%; CD8, 36.3%). Bacterial, mycobacterial, and fungal cultures of the BALF were negative. A cytomegalovirus antigenemia test was also negative. The BALF was also shown to be negative for influenza A (including subtypes H1 and H3), influenza B, RSV A and B, HMPV, PIV 1‐4, coronavirus OC43, 229E, NL63, HKU1, rhinovirus/enterovirus, adenovirus, and bocavirus by a respiratory viral panel fast assay (Luminex xTAG, Luminex Diagnostics, Toronto, Canada). Based on these findings, we diagnosed NIPC induced by DLI. On Day 88, prednisolone (PSL) was administered at a dose of 1 mg/kg/day, which led to rapid clinical improvement. The follow‐up CT 12 days after initiation of PSL treatment showed remarkable improvement of the abnormal shadows (Fig. 1B). The PSL dose was tapered off and discontinued 9 months after onset, without recurrence of NIPC.

Figure 1.

Chest CT scan. (A) Scan on Day 63 after DLI. (B) Scan on Day 100 after DLI (Day 12 after start of PSL treatment).

MATERIALS AND METHODS

Collection of sera and BALF

All blood samples for cytokine determination were collected from the patient before PSL treatment, as well as at four weekly time points after the initiation of PSL treatment, with written informed consent. Cells from the BALF were also collected for phenotypic and chimerism analysis; BALF serum and cells were separated immediately and stored at −20°C until the assays were performed.

Chimerism analysis of peripheral blood cells and invaded cells in BALF

To clarify whether the NIPC after DLI was induced only by the infused donor T cells, chimerism of peripheral blood mononuclear cells (PBMNCs) and BALF was assessed by polymerase chain reaction analysis of short tandem repeat loci using a cell line identification system (Cell ID, Promega Corp., Madison, WI), according to the manufacturer's instructions.

Multiplex immunoassay analysis of cytokines

Cytokines in sera were quantitated using a suspension array system with a pro human cytokine group 27‐plex panel (Bio‐Plex, Bio‐Rad Laboratories Inc., Hercules, CA). This panel includes the cytokines and chemokines interleukin (IL)‐β, IL‐1ra, IL‐2, IL‐4, IL‐5, IL‐6, IL‐7, IL‐8, IL‐9, IL‐10, IL‐12p70, IL‐13, IL‐15, IL‐17A, eotaxin, fibroblast growth factor (FGF)‐basic, granulocyte–colony‐stimulating factor (G‐CSF), granulocyte‐macrophage–colony‐stimulating factor (GM‐CSF), interferon (IFN)‐γ, IFN‐γ inducible protein‐10, MCP‐1, macrophage inflammatory protein (MIP)−1α, MIP‐1β, platelet‐derived growth factor (PDGF)‐bb, RANTES, tumor necrosis factor (TNF)‐α, and vascular endothelial growth factor (VEGF), according to the manufacturer's specifications.11

RESULTS

T‐cell chimerism in BALF was predominantly patient T cells

As shown in Table 1, approximately 10.7% of CD3+ MNCs and 80.5% of CD3– MNCs were of donor origin in the BALF, suggesting that T‐cell chimerism was predominantly patient T cells. The BALF chimerism was approximately comparable to that of the peripheral blood of the patient.

Table 1.

Chimerism analysis of PBMNCs and BALF cells before PSL treatment

Cells	Peripheral blood	BALF
CD3+ MNCs	9.4%	10.7%
CD3– MNCs	86.8%	80.5%
Neutrophils	100%	NA

Cytokine profile

Serum analysis revealed a significant elevation of nine cytokines and chemokines (IL‐1β, IL‐2, IL‐6, IL‐7, IL‐8, TNF‐α, MIP‐1α, MIP‐1β, and VEGF) at the onset of NIPC, which declined soon after initiation of PSL treatment (Fig. 2A). The kinetics after PSL treatment are shown in Fig. 2B; gray shadow indicates the range of normal concentration calculated by median ± 2SD, using five healthy donors. The IL‐2, IL‐7, and VEGF levels were not significantly different from the normal range of the healthy controls. IL‐5, IL‐15, and GM‐CSF were below the detection limit. The concentrations of the other 15 cytokines and chemokines were also measured, but did not change during the clinical course after PSL treatment (figure not shown).

Figure 2.

Multiplex immunoassay analysis of 27 cytokines and chemokines. (A) Black line indicates the levels before PSL treatment. Gray line indicates the levels 4 weeks after PSL treatment. (B) Weekly kinetics of the elevated serum cytokines and chemokines after PSL treatment. 0w indicates the levels before PSL treatment. Gray shadow indicates the range of normal concentration calculated by median ± 2SD using five healthy donors.

DISCUSSION

Since the first report of DLI for treatment of CML was published by Kolb and colleagues in 1990,3 numerous clinical trials of DLI against various hematological malignancies have been reported.4, 5, 6, 7, 8 Although the efficacy of DLI depends on the disease type and status, DLI has been shown to have a direct graft‐versus‐leukemia effect by infusion of alloreactive donor T cells. On the other hand, DLI may induce severe adverse effects, including severe GVHD. When our patient complained of dyspnea with fever, we suspected that the complication was pulmonary GVHD by alloreactive donor T cells recognizing alloantigens. Previous studies have suggested that infused lymphocytes lead to lung injury in murine GVHD models, although the lungs have not been generally considered a classic target organ of acute GVHD.12, 13, 14 To clarify whether NIPC of the patient was pulmonary GVHD induced by infused donor lymphocytes, we performed chimerism analysis of T cells in the BALF and demonstrated that the cells recruited to the patient's lung were not exclusively derived from donor T cells. These results suggested that the pathogenesis of DLI‐induced NIPC was not due to attack by donor‐derived cytotoxic T cells directly recognizing the patient's pulmonary tissue, but by migrated patient T cells that were activated after DLI.

Analysis of the cytokine profile revealed the kinetics of inflammatory mediators, such as proinflammatory cytokines and chemokines in the serum. IL‐1β, IL‐6, IL‐8, TNF‐α, MIP‐1α, and MIP‐1β were elevated at the onset of the patient's pulmonary symptoms and decreased after treatment with PSL. This phenomenon has previously been reported in IPS occurring in the early period after transplantation in humans.15, 16 The importance of chemokines in IPS has also been reported in murine models.17, 18

IPS has frequently been a fatal complication after allo‐HSCT, with a 3% to 15% incidence and median onset time between the second and third week after HSCT.19, 20 The clinical outcome for patients with IPS is poor, with mortality ranging from 50% to 80% and median time from diagnosis to death as short as 13 days. Risk factors for classical IPS include full‐intensity conditioning with total body irradiation and acute GVHD.21 Based on previous findings, it is thought that potential etiologies for IPS include the direct toxic effects of conditioning regimens and the release of inflammatory cytokines. The cytokine profile of the present case suggested that NIPC was a result of systemic immune response stimulated by DLI.

In conclusion, the elevated cytokines and chemokines measured in the present case may have played an important role in the pathogenesis of NIPCs after DLI. Further study of cytokine and chemokine profiles after DLI in a number of patients may clarify the pathophysiology of NIPC.

CONFLICT OF INTEREST

The authors have disclosed no conflicts of interest.