In this case, a male patient was first reported, who had acute MCL secondary to previous primary malignant mediastinal GCT. As a group of neoplasms commonly occurring in gonads, GCT is a model of curable cancer that often has a satisfactory outcome with cisplatin-based chemotherapy [9]. However, primary mediastinal non-seminomatous GCTs (PM-NSGCTs) tend to have a poor prognosis due to the evolution of neoplasms in other somatic types and resistance to cisplatin, particularly hematologic malignancies with a median survival time of fewer than 6 months [10]. Moreover, the second hematologic malignancies preceded by or concurrent with PM-NSGCTs are mostly acute myeloid leukemia (AML) (the largest proportion is acute megakaryoblastic leukemia), sharing more genetic similarities characterized by i (12p) and/or TP53 mutations with PM-NSGCTs rather than primary AML. The most competing hypothesis is that a common cancer progenitor cell with the capacity of differentiation into germ cells and hematopoietic lineages may evolve into both tumors [11].
The diagnosis of acute MCL of this case was based on the implementation of SM with the major criteria (dense infiltration of over 15 aggregated MCs in marrow) and one minor criterion (>25% atypical MCs in the marrow aspirate and biopsy), with the addition of up to 22% atypical MCs in PBMCs and C-findings including splenomegaly and cytopenia. It was found that the typical morphology and positive results of chemical staining with Toluidine Blue played decisive roles in the differential diagnosis of MCs.
The leukemic cells of the patient were heterogeneous in morphology, varying widely in size and shape with a very rare phenotype and genetic manifestations. The neoplasm cells phenotypically had CD9, neither CD2 nor CD25. Previous reports of SM revealed that the expression levels of two markers gradually decreased along with malignant progression [12–14]. It has been reported that 38% of MCL cases have a double-negative CD2/CD25 immuno-phenotype [7]. Besides, the positive co-expression of CD2 and CD25 has a significantly higher proportion in MCL patients with KIT D816V than those with no missense variant (66% vs 25%) [7].
MCL has typical genetic characteristics. The proportion of gain-of-function somatic mutations in KIT, including KIT D816V and other KIT mutants at exons 8, 9, 10, 11, 13, and 17, is approximately 90% [3]. KIT D816V in MCL can activate the PI3K pathway and oncogenic STAT5 molecule by JAK2 and MEK/ERK1/2 pathways, which can consistently contribute to persistent IL-6 induction. The level of IL6 is related to the severity and prognosis of SM [15]. In addition, KIT D816V may be accompanied by other gene variants that jointly contribute to malignant expansion. The multivariable risk analysis of MCL patients indicated that the mutations of SRSF2, ASXL1, or RUNX1 (S/A/Rpos) were the only dependent risk factor, which is adversely associated with phenotype, therapeutic response, progression, and OS [16]. However, the secondary MCL as the patient developed in the setting of primary mediastinal GCTs, appeared a classical genetic aberrance of TP53 frameshift mutation, additionally with an FLT3 nonsense mutation, a SETBP1 missense mutation, and a JAK3 missense mutation (approximately 50% high VAF for three genes), which may be genetic alterations related to chemotherapy or radiotherapy.
TP53 has been reported as an anti-oncogene to induce mutations in about 50% of human tumors [17], mainly located in the region of the DNA binding domain (DBD, codons 94-297). Both DBD and non-DBD mutations of TP53 in AML are considered independent high-risk factors with a poor prognosis [18]. In this case, the mutation locus of TP53 P301Qfs*44 is adjacent to the DBD. Consistent with our findings, TP53 mutation is regarded as one of the typical genetic characteristics in secondary hematologic neoplasm preceded by mediastinal dysgerminoma [11]. Besides, it was previously reported that an ASM patient had a history of ovarian dysgerminoma, and TP53 developed a somatic nonsense mutation in the dysgerminoma and bone marrow of an ASM period. Therefore, it is speculated that both above-mentioned tumors may derive from the common cancer stem cell, or that ASM is evolved from the residual cancer cells of dysgerminoma [19]. FLT3 is an oncogene of the receptor tyrosine kinase family that is involved in the proliferation and differentiation of hematopoiesis. The activating mutations represented by FLT3-ITD are closely related to tumorigenesis, especially AML. As previously reported for the exclusive relationship of FLT3-ITD and TP53 mutations in AML [18], a rare nonsense mutation of FLT3 R973X occurred in our case, corresponding to the inactivating mutation of TP53.
SETBP1 mutations, firstly identified in Chinzel-Giedion syndrome characterized by severe mental retardation, are considered as a biomarker of the myelodysplasia /myeloproliferative neoplasm overlap syndrome [20]. Additionally, SETBP1 hotpot mutations within a conserved 11-nucleotide region (amino acids 868-871) are often detected in secondary AML and chronic myelomonocytic leukemia. It has been also reported that only hotpot mutations can induce the resistance and poor prognosis of myeloid neoplasm [21, 22]. The role of SETBP1 non-hotpot mutations is currently elusive, such as SETBP1 N272D in our case. Along with SETBP1 mutations, a non-receptor tyrosine kinase JAK3 mutation is identified as the secondary mutation in juvenile myelomonocytic leukemia (JMML), while the latter is involved not in the initiation but the progression of JMML, indicating a poor prognosis [23, 24]. JAK3 mutation is regarded as a driver mutation that is frequently reported in T lineage acute lymphoblastic leukemia [25, 26]. Recurrent JAK3 V722I is reported in malignant struma ovarii, a specific ovarian teratoma [27]. Consistent with JAK3 V722I, the mutation site of JAK3 I688F is located in the same protein domain, protein kinase 1, potentially harboring a similar function.
In this case, it was first considered as acute promyelocytic leukemia (APL) when blood and marrow smears showed a mass of coarse granules in the cytoplasm of abnormal cells, although the Auer body was not found. However, the subsequent laboratory results of marrow aspirate or biopsy were all negative for MPO molecules, t (15; 17) (q22; q21), and RARα rearrangement, which were the classical diagnosis criteria for APL [28, 29], so that the disease was excluded. Myelomastocytic leukemia (MML) is another major differential diagnosis to MCL and is also considered an advance myeloid neoplasm with excess blasts accompanied by abnormal MCs morphologically, which fails to meet the criteria of SM to distinguish from MCL [14, 30].
Patients with MCL always have typical clinical symptoms. MC activation symptoms (MCAS, also known as mediator-related symptoms) include pruritus, flush, fever, malaise, diarrhea, etc., occurring more frequently among patients without KIT D816V mutations than positive mutations [7]. Besides, end-organ failure is another common clinical symptom of MCL patients. Similarly, our patient developed MCAS at the onset of MCL and died of respiratory failure in the end.
The treatment of MCL is limited up to now, and there is no accepted standard first-line therapy. The application of corticosteroids can reduce the MC burden and improve clinical symptoms, but the effect is often transient [7]. Cladribine is considered an effective agent to induce a meaningful response in half of the patients with advanced SM, followed by a high percentage of resistance [31]. The effect of allogeneic hematopoietic stem cell transplantation has a fine result in non-transformed SM, but it is uncertain in overt MCL [32, 33]. Midostaurin is a targeted drug that blocks the kinase activity of wild-type KIT and its activation variants including KIT D816V, thus prolonging the median OS to 28.7 months [34]. Since 2017, the combination therapy of midostaurin with chemotherapy has been listed as the standard first-line therapy by FDA. Dasatinib as another selective KIT inhibitor may be the second choice because of its available application in China. The patient in this case was originally effective to corticosteroids and dasatinib-based treatment, with the identical knowledge that the KIT D816V mutant is more resistant to tyrosine kinase inhibitors [35]. The patient even became more optimistic about the progression of neoplasm, however, he died of complications from chemotherapy. Personalized therapy of MCL remains challenging.
There are some defects in this case. Due to clinical laboratory limitations, the concentration of serum tryptase was not examined to investigate the correlation between tryptase expression and leukemic development. In addition, the patient’s relatives refused to perform gene sequencing, and specimens in the period of primary GCT were also not examined for genetic alterations. Thus, it is difficult to determine whether these gene mutations are germline or somatic.