Here, we retrospectively analyzed the immunophenotypes, prognostic factors, and patient outcomes of transformed lymphoma. To further evaluate the genetic basis of HT, and to find genomic alterations and key pathways that could be exploited therapeutically, we performed a whole-exome sequencing study of 7 tMZL cases.
In our study, the 5-year OS rate of patients with DLBCL HT of FL was 79% compared to 46–87% reported in previous studies [6, 25]. The 5-year OS rate of patients with tMZL has been reported to be 33–94% [7, 8, 26, 27]. In the present study, the 5-year OS rate of tMZL was 57%. In the current study, the median OS of patients with tCLL/SLL was 22.5 months, which was longer than the OS reported by Elnair et al [28, 29]. This might be partly due to the young age and low IPI scores of most of the patients included in the present study.
Immunohistochemical results for Bcl-6 and CD10 were positive in 91% and 64% of patients, respectively, among those with tFL. Bcl-6 and CD10 demonstrated lower expression in tMZL (Bcl-6, 29%; CD10, 6%) and tCLL/SLL (Bcl-6, 33%; CD10, 0%). This indicates that although these three indolent lymphomas were transformed into DLBCL, their immunophenotypes are still different.
Prognostic factors for different types of transformed lymphomas have not been previously reported. Our data show that the OS of tFL was better than that of tMZL, while tCLL/SLL had the poorest OS. Synchronous versus asynchronous HT (P < 0.001), and type of transformed lymphoma (P = 0.018) were significant prognostic factors for OS. Asynchronous HT (P = 0.002) was a significant adverse factor for PFS. Previous studies have reported that patients with synchronous transformed lymphoma had a better OS and PFS than those with asynchronous transformed lymphoma [7, 30, 31], which is consistent with the results of the present study. The current study also showed that advanced-stage disease at HT was not an adverse prognostic factor for PFS or OS, which may be related to the fact that most cases of transformed lymphoma are at an advanced-stage. A previous study reported that the survival of patients with transformed lymphoma improved in the rituximab era [32]. However, HT treatment was not a significant adverse prognostic factor for OS and PFS in the present study. Because the majority of the transformed lymphomas were treated with R-CHOP, only five patients with transformed lymphoma received CHOP.
The present study revealed extensive genetic heterogeneity, there is a large number of genetic variations affecting many genes and pathways, reflecting the complexity of the HT process. Our results suggest that the genomic mutational landscape and CNA profiles of tMZL is much more complex than that of MZL [21, 33, 34]. We identified chromatin regulator genes were mutated in MZL (EP300) and tMZL (CREBBP and EP300), previous studies demonstrated that chromatin regulator genes mutations were early events in MZL [33, 35]. Somatic mutations within regions of CNAs detected in our study, such as those in TNFAIP3, BCL10, and CD79B are associated with NF-κB activation and frequently mutated in de novo ABC-DLBCL [23, 36]. Therefore, it is likely that they also have similar functional roles in tMZL. Previous studies strongly suggest that activation of the NF-κB pathway is closely associated with the transformation of Waldenström macroglobulinemia and FL [11, 37], it may have also played an important role in the transformation of MZL. Many therapeutic agents target the NF-κB signaling pathway, such treatment might be beneficial for MZL patients harboring mutations in this pathway. Consistent with the divergent evolution model, our analysis corroborates the dominant tMZL clone arises from common progenitor clone (CPC) through the acquisition of independent genetic events. The existence of CPC can be postulated based on the presence of a common set of lesions between MZL and tMZL, which is consistent with previous studies based on FL progression [38].
In conclusion, the 5-year OS and PFS rates after DLBCL HT were 50% and 26%, respectively. The OS and PFS of patients with tFL were higher than those of tCLL/SLL and tMZL. Furthermore, Bcl-6 and CD10 were positive in 91% and 64% of tFL, respectively, while Bcl-6 and CD10 were less expressed in tMZL and tCLL/SLL. Kaplan-Meier survival analysis revealed that asynchronous HT and tCLL/SLL were significant adverse prognostic factors for OS after DLBCL HT. In terms of the mutation landscape of tMZL, although this is only the first step, it seems likely that certain alterations contribute to the occurrence of aggressive disease, particularly the activation of the NF-κB pathway. More research is needed to better understand the genetic mechanisms underlying HT, this knowledge will help identify promising therapeutic targets in the future.