Introduction

Primary mediastinal large B-cell lymphoma (PMLBL) is a rare subtype of non-Hodgkin lymphoma (NHL) in childhood arising from mature thymic B cells. In the World Health Organization classification, PMLBL was originally described as a subtype of diffuse large B-cell lymphoma (DLBCL), but is currently classified independently according to the results of recent analyses, including molecular genetics [1]. Clinically, PMLBL occurs in the mediastinum and invades aggressively, often causing superior vena cava syndrome. Reports from Western countries showed that it is predominant in adolescents and females [2, 3]. In addition, recent clinical trials of pediatric B-cell NHL (B-NHL) showed inferior outcome of pediatric PMLBL patients under standard therapy for pediatric B-NHL [4–7], therefore, stratified to the other treatment option. On the other hand, little is known about PMLBL of Asian children. Herein, we report a retrospective study of PMLBL in Japanese children and adolescents.

Methods

The clinical data of children and adolescents with PMLBL in this study consisted of two separate cohorts. Cohort 1 includes the patients (diagnosed between 1991 and 2004) from the registries of four local clinical study groups in Japan, the Tokyo Children’s Cancer Study Group (TCCSG), the Japan Association of Childhood Leukemia Study (JACLS), the Children’s Cancer and Leukemia Study Group (CCLSG), and the Kyushu Yamaguchi Children’s Cancer Study Group (KYCCSG). The pathological diagnosis of patients in Cohort 1 was based on each institutional pathological review. Cohort 2 includes the patients (diagnosed between 2004 and 2014) from the nationwide central pathological review registration of the Japanese Pediatric Leukemia/Lymphoma Study Group (JPLSG), a national study group established in 2003 after merger of the 4 existing groups as mentioned before. Patients’ data in Cohort 2 were collected as part of the retrospective study focusing on the rare subtype of pediatric NHL conducted by the JPLSG lymphoma committee. We combined the information, and analyzed their clinical characteristics, treatments received, and outcome. The current study was approved by the JPLSG and the ethics committee of the Sapporo Hokuyu Hospital in Hokkaido, Japan. Overall survival (OS) and event-free survival (EFS) rates were analyzed using the Kaplan–Meier method. SPSS 22.0 software (IBM SPSS, Inc. in Chicago, IL, USA) was used for analysis of data.

Results

We collected 17 PMLBL patients: 10 from Cohort 1 and seven from Cohort 2. According to the current nationwide registration data, PMLBL occupied 1.0% of NHL (7 out of 696 NHL patients) and 2.2% of B-NHL patients (7 of 314 B-NHL patients) (Table 1). Detailed characteristics of the 17 cases are shown in Table 2. All patients were teenagers, including seven adolescents, with median age of 14 years (range 12–18 years). Among them, 10 patients were male and seven were female. Regarding initial Murphy’s staging, 15 were in stage III and 2 were in stage IV. Treatments for the patients were heterogeneous, especially in Cohort 1 [8–10]. In Cohort 2, three out of seven patients received NHL-BFM95 [5] and three received JPLSG B-NHL03 [11], which was the current nationwide clinical trial for B-NHL in Japan. Two patients who received B-NHL03 treatment switched to rituximab combination therapy after diagnosis because of the methotrexate-related toxicity in one patient and the residual tumor after chemotherapy in the other patient. One patient received R-CHOP (rituximab, cyclophosphamide, vincristine, and prednisolone) from the beginning. The best response achieved after the initial treatments was complete remission (CR) in 15 and induction failure in two. Among them, two patients received autologous peripheral blood stem-cell transplantation (SCT) in CR and one patient with Wiskott–Aldrich syndrome received cord blood transplantation to correct the underlying disease. Two patients who could not achieve CR received SCT but no detailed information was available. All the patients who could achieve first CR were alive, although one patient suffered from secondary leukemia. Both patients who could not achieve CR died of lymphoma in spite of receiving SCT without CR. The overall 5-year OS rate and EFS rates were 84.4 and 81.9%, respectively, with median follow-up period of 4.5 years (range 2.7–9.3 years) (Fig. 1).

Table 1 Reports of children and adolescents with PMLBL
Full size table
Table 2 Characteristics, treatment, and outcome of Japanese children and adolescents with PMLBL
Full size table
Fig. 1
figure 1

Kaplan–Meier curve for probability of overall survival (OS) and event-free survival (EFS) in 17 patients enrolled in this study

Full size image

Discussion

To the best of our knowledge, this is the first case series describing PMLBL in Asian children and adolescents. The reported pediatric case series of PMLBL in Western countries was listed in Table 1 [2, 3, 12]. In our cohort, all cases were teenagers similar to previous series. On the other hand, female predominance was not observed in Japan. In addition, it seemed that the incidence of PMLBL in Japan was low. The same trend could be seen in the registration of the recent large clinical trials for pediatric B-NHL; JPLSG B-NHL03 included only 0.6% of PMLBL cases (two patients in 321 all enrolled cases), whereas NHL-BFM95 trial and FAB/LMB96 trial included 2.9% (15 patients in 515 all enrolled cases) and 3.8% (42 patients in 1111 enrolled cases) of PMLBL cases, respectively [3–5, 11].

It is well recognized that the incidence of certain subtypes of pediatric lymphomas is different between Japan and Western countries [11, 13, 14]. For example, higher frequency of DLBCL and fewer Hodgkin lymphoma (HL) is observed in Japan. From the analysis of gene expression profile, the features of PMLBL were completely different from those of DLBCL, but similar to those of HL [15, 16]. The rarity of PMLBL in Japan might be associated with that of HL.

In terms of the prognosis, PMLBL has been reported to have worse prognosis than other subtypes of pediatric B-NHL despite the improvement in the overall treatment results [5, 6]. Considering these results, PMLBL is separated and treated with another strategy, and dose adjusted (DA)-EPOCH-R (etoposide, prednisone, vincristine, cyclophosphamide, doxorubicin, and rituximab) in the Inter-B-NHL ritux 2010, which is the ongoing international clinical trial for pediatric B-NHL to investigate the effectiveness of rituximab combination therapy. Dunleavy et al. [17] reported that DA-EPOCH-R was effective for PMLBL in adults. Although confirmation in large clinical trials is necessary, DA-EPOCH-R regimen is expected to be a promising treatment even for pediatric cases [18]. In our cohort, about 90% of the patients who achieved CR after primary treatment survived with no relapse and two patients in Cohort 1 died with the refractory disease. On the other hand, all the seven recent cases in Cohort 2, whose diagnoses were confirmed at the central pathological review, were alive. As treatments in Cohort 1 were too heterogeneous, it is difficult for discussions on the reason of inferior outcome. However, with regard to Cohort 2, all the patients received modern intensive chemotherapy, of which three patients received rituximab combination therapy. Considering that the EFS and the OS rates were 87.4 and 92.7% of JPLSG B-NHL03 study [11], the prognosis of PMLBL patients of children and adolescents in Japan did not seem inferior compared with that of other B-NHL patients, especially in the recently diagnosed cohort. In addition, rituximab combination therapy might contribute to the good prognosis.

On the other hand, eight patients with primary mediastinal mass were registered on B-NHL03 protocol. Their diagnosis was BL in three patients, DLBCL in one, and the aggressive B-NHL with insufficient material in two, other than two patients with PMLBL. As one of two patients with ambiguous diagnosis suffered relapse and died of disease, the true diagnosis might impact on the prognosis of Japanese patients with PMLBL.

In conclusion, this is the first case series describing PMLBL patients in children and adolescent from Asia. The incidence, characteristics, and outcome varied considerably from those of Western populations. Further studies, including molecular analysis, will be needed.