Concurrent myeloid and lymphoid malignancies are quite uncommon. The difficulty of diagnosis can be illustrated in the present case. At presentation, generalized subcutaneous tumors and lymphadenopathies led to the diagnosis of a high grade lymphoma. The pathological diagnosis of T lymphoblastic lymphoma was made unequivocally by typical morphology and detailed immunophenotypical studies. The clinical course was complicated with spontaneous tumor lysis syndrome, which typically occurred in high grade lymphoma and therefore supported the diagnosis of T lymphoblastic lymphoma. Although lymphoblastic lymphoma may involve bone marrow or peripheral blood, the results of bone marrow studies revealed the leukemic cells did not belong to the T-lineage. The strongly positive myeloperoxidase staining suggested the diagnosis of AML. This was further confirmed definitively by its typical immunophenotypical profile. Some leukemic cells of AML may occasionally express T cell markers. In our case, however, the leukemia cells was found negative for CD2, CD5 and CD7, making aberrant expression of AML relatively unlikely. The diagnosis of concomitant AML and T lymphoblastic lymphoma, was confirmed.
Although it appears unusual, such concurrence is not coincidental. In English literature, similar cases have been reported. In 2007, Metzgerot et al. reported 7 cases of AML or T lymphoblastic lymphoma bearing FIP1L1-PDGFA fusion genes. In 2 cases, T lymphoblastic lymphoma was found synchronously with myeloid neoplasms (acute eosinophilic leukemia in one and MPN in the other) . In 2008, Capovilla et al. reported another case of synchronous chronic eosinophilic leukemia and T lymphoblastic lymphoma. Like the previous report, the leukemic cells were found to bear FIP1L1-PDGFRA fusion genes . It appears from these reports that such genetic aberrations may lead to mixed lineage differentiation. In contrast to the previous reports with PDGFRA gene rearrangements, the present case was found to bear t(5;12)(q33;p13). This translocation most likely involves PDGFRB genes mapped on chromosome 5p. This assumption was further confirmed by studies of FISH. Based on chromosomal mapping and prior literature reports, the most likely partner gene is ETV6 .
In 2008 WHO classification, a new class of hematopoietic diseases, myeloid and lymphoid neoplasms with eosinophilia and abnormalities in PDGFRA, PDGFRB and FGFR1 genes, were introduced . In such a classification system, all cases of concurrent myeloid neoplasms and T lymphomblastic lymphoma, including the present case, belong to the same category. While most cases of this entity presented with MPNs, the genetic aberrations have a wide spectrum of differentiations, including eosinophilia, myelodysplastic syndrome, chronic myelomonocytic leukemia, AML, B or T cell acute lymphoblastic leukemia, T lymphoblastic lymphoma, and disorders with mast cell proliferation . Most cases with PDGFRA gene rearrangements manifest as MPNs, usually chronic eosinophilic leukemia. PDGFRB gene rearrangement cases typically present as MPNs which are almost always accompanied with eosinophilia. Aberrations of FGFR1 genes are also commonly associated with MPNs with eosinophilia. In particular, this genetic mutation is often associated with differentiation into both myeloid and lymphoid lineages. The most common lymphoid neoplasm is T lymphoblastic lymphoma. The concurrence of myeloid and lymphoid malignancies was reported both in human cases and animal models .
While these reports revealed it is not uncommon for PDGFRA and FGFR1 genetically rearranged cells to differentiate into lymphoid neoplasms, hematologic diseases bearing PDGFRB mutations are rare [2, 5]. The two largest series reported in literature included 22 and 9 cases respectively. PDGFRB rearranged cells are known to develop into B and T lineage lymphoma . This differentiation pattern was well described in transgenic mice models [11, 12]. However, according to an extensive review by Holroyd et al, patients with PDGFRB mutations have not been known to develop lymphoid malignancies. This is somewhat surprising and may be explained by the rarity of such mutations among humans . As a result, the case in the present article is the first case report of lymphoma and concurrent AML bearing PDGFRB gene mutations.
Treatment of concurrent AML and T lymphoblastic lymphoma and AML is a clinical challenge. There is mounting evidence that PDGFRA, PDGFRB and FGFR1 entities respond well to tyrosine kinase inhibitors [2, 5, 9, 10, 13]. However, tyrosine kinase inhibitors are expensive and still off-label for lymphoma. Therefore, we did not consider tyrosine kinase inhibitors as the frontline treatment in this case. In view of the simultaneous expression of myeloid and lymphoid lineage features, consideration of chemotherapeutic regimens is similar to that of bi-phenotypic leukemia. Bi-phenotypic leukemia is an uncommon disorder. No consensus protocol is available for treatment. Some studies suggest that bi-phenotypic leukemia has good responses to treatment directed against AML [14–16]. As a result, we chose treatment that was based on cytarabine and daunomycin. Steroids were included in the induction phase and complete remission was achieved for both AML and lymphoma. The patient subsequently underwent sibling-matched, allogeneic stem cell transplantation, which was quite successful until the development of chronic GVHD. Although the patient died of GVHD and related infections, he remained free from both malignancies for at least 11 months after transplantation. We believe allogeneic stem cell transplantation is an effective, albeit risky treatment option. Clinicians should weigh the risk against potential benefits before deciding treatment strategies, especially when tyrosine kinase inhibitors have become relatively safe and effective options.