Pulmonary extramedullary hematopoiesis in patients with myelofibrosis undergoing allogeneic stem cell transplantation

We examined the lung function of 11 patients with intermediate/high risk myelofibrosis undergoing allogeneic hematopoietic stem cell transplantation (HSCT). In 3 patients, chest computerized tomography (CT) scans revealed multiple pulmonary nodules with extramedullary hematopoiesis that disappeared after transplantation. Pulmonary extramedullary hematopoiesis in patients with myelofibrosis rapidly regresses after allogeneic HSCT.

Primary myelofibrosis (PMF) or myelofibrosis secondary to polycythemia vera (PV-MF) or essential thrombocythemia (ET-MF) are clonal myeloproliferative disorders often characterized by pancytopenia, bone marrow fibrosis, leukoerythrocytosis, teardrop poikilocytosis and splenomegaly.1 Splenomegaly, in particular, is the result of extramedullary hematopoiesis. Due to the extramedullary hematopoiesis, patients with myelofibrosis can develop pulmonary hypertension secondary to hematopoietic infiltration, portal hypertension, thrombocytosis, hypercoagulability, and left ventricular failure.2,3 Radiographic findings such as ground glass appearance, effusions, septal thickening on chest computerized tomography have been also described in PMF patients.4 In patients with intermediate/high risk myelofibrosis,5 allogeneic hematopoietic stem cell transplant (HSCT) is the only known curative therapy and the development of reduced intensity conditioning has allowed a decrease in transplant related mortality while inducing long-term remission, especially in older patients In particular, HSCT can restore a normal hematopoiesis, and allows the resolution of marrow fibrosis10 as well as the progressive reduction of splenomegaly.11

In this study, we evaluated the chest computerized tomography (CT) and pulmonary function tests (PFTs) of 11 consecutive patients with primary myelofibrosis (n=5), PV-MF (n=3), or ET-MF (n=3) who received an allogeneic hematopoietic stem cell transplantation (HSCT) to assess the presence of pulmonary extramedullary hematopoiesis. An informed consent was signed prior to transplant and the transplants were performed according to the protocol approved by the University of Illinois at Chicago Institutional Review Board. All patients received an HLA matched graft from related (n=7) or unrelated (n=4) donors. Prior to transplantation, patients underwent cardiac and pulmonary testing in addition to other standard laboratory testing. Cardiac testing was performed by either a transthoracic echocardiogram (TTE) or by nuclear multiple gated acquisition scan (MUGA). Pulmonary evaluation was made by CT of the chest and functional tests. In particular, values of corrected diffusion limiting capacity of oxygen (DLCO), forced expiratory volume in one second (FEV1), forced vital capacity (FVC), and FEV1/FVC ratio were determined within 30 days prior to transplant. Patients were conditioned with a fludarabine based regimen including melphalan (n=10) or targeted i.v. busulfan (n=1) as previously described12 Graft-versus-host disease (GVHD) prophylaxis consisted of standard methotrexate and tacrolimus (started on day -2). Tacrolimus levels were maintained between 5 and 15 ng/mL until day +180 unless GVHD or recurrence of disease occurred. A total of 4 patients who received a transplant from unrelated donors received thymoglobulin (rabbit antithymocyte globulin, ATG, Genzyme Inc.) at 7 mg/kg total dose as additional GVHD prophylaxis on day -3 to day -1. Myeloid growth factors were not used in the peri-transplant setting. Post transplantation, patients’ pulmonary function was re-evaluated at three, six, and 12 months. Of 11 patients, 3 (27%) showed bilateral small pulmonary nodules, which varied in size from 5 mm to 15 mm by chest CT. The nodules were contrast enhancing and were located diffusely throughout the lungs. Due to the limited size of the nodules and the lack of symptoms suggestive of an infectious etiology, in 2 patients biopsy was not performed. In one patient, who showed a progressively increasing number of these nodules in the two months prior to transplant, an open biopsy of one nodule was performed to rule out an infectious or extra-hematologic neoplastic process (Figure 1A). The biopsy revealed areas of hematopoiesis with scattered megakaryocytes consistent with extramedullary hematopoiesis (Figure 1B). We then analyzed the pulmonary function testing in all the 11 patients in this study. The analysis of pulmonary function in the 3 patients with nodules in the lung showed a reduced DLCO (Figure 1C, left panel) (p=0.04) and a trend for reduced FEV1 and FEV1/FVC ratio (Figure 1C, central and right panel) compared to the other 8 patients. The age of the patients also correlated with DLCO values since 4 out of 4 patients (2 with nodules) older than 60 years of age had a median DLCO of 56% (range 52–59%) compared to 7 patients younger than 60 years (one with nodules) who had a median DLCO of 72% (range 59–83) (p=0.006). However, older patients did not have a significantly longer duration of disease compared to the younger ones (p=0.2). None of the patients had clinical symptoms of pulmonary hypertension and a transthoracic echocardiogram (TTE) was performed in 7 of them showing a median pulmonary systolic pressure of 24 mmHg (range:20–34). We analyzed whether the degree of splenomegaly in the 3 patients with nodules differed from that of the patients without nodules. The patients with pulmonary nodules had extensive splenomegaly with spleen size of 25, 30.5, and 34 cm. In the 8 patients without pulmonary nodules or pulmonary disease, the median spleen size was 21 cm (range 12.5–34 cm) (n=7 patients) (p=0.2) with one patient who had previously undergone splenectomy. Lung CT scan was repeated within three months after an allogeneic HSCT and the nodules resolved in all of the 3 patients. This also correlated with a reduction in spleen size, which occurred in all the 10 patients with initial splenomegaly. Pulmonary hematopoiesis has rarely been reported in patients with primary or secondary MF. Nevertheless pulmonary findings at the time of autopsy are quite common.4 In addition, the clinical course when pulmonary hematopoiesis is present is unknown and one study suggested that pulmonary hematopoiesis correlated with progression of the disease.4 In our series, 27% of patients with PMF, ET-MF, or PV-MF and with a Lille score ≥1 had multiple nodules in the lungs. In one case, these nodules were biopsied and showed extramedullary hematopoiesis. The other two cases had similar nodules and although biopsies were not performed, we believe they also represented alveolar foci of hematopoiesis due to the rapid clearance after allogeneic HSCT. Our findings are consistent with the observation of progressive reduction of bone marrow fibrosis and splenomegaly following reduced intensity conditioning allogeneic HSCT.1,10,11 In conclusion, pulmonary extramedullary hematopoiesis may be a more common finding in patients with myelofibrosis than has been previously reported. However, it is not a contraindication for allogeneic HSCT.

Table 1.Characteristics of patients with primary myelofibrosis or myelofibrosis secondary to polycythemia vera or essential thrombocythemia in the study.

Figure 1.Pulmonary extramedullary hematopoiesis in patients with intermediate/high risk myelofibrosis undergoing allogeneic HSCT. (A) Pulmonary nodule in the right lung of a patient with PMF detected by CT scan (see arrow). (B) Section of lung showing hematopoietic precursors (nucleated erythroid and myeloid precursors and megakaryocytes) in the alveolar space. Some pigment laden macrophages are also noted. (H&E stain; original magnification x 400). (C) Differences in pulmonary function testing (DLCO, left panel; FEV1, central panel; FEV1/FVC, right panel) in patients with (n=3) or without (n=8) pulmonary nodules. The difference in DLCO was statistically different (p=0.04) by two-tailed Mann Whitney test.

Footnotes

• Funding: this study was supported in part by the National Cancer Institute grant P01CA108671 (R.H.), Department of Defense grant MP048010 (R.H.), and a grant from the Myeloproliferative Diseases research Foundation (R.H.).