In this study, we analyzed a total of 61 patients with newly diagnosed FL grade 3 treated with frontline RCHOP regimen, including 42 patients with FL grade 3A and 19 patients of aggressive FL. The 3-year OS was better in patients with FL grade 3A (97.1%) than aggressive FL (81.9%). The 3-year PFS was not significantly different between two groups, with 69.1% and 71.1%, respectively. Of note, patients of aggressive FL reached a plateau in the PFS curve after 2 years.
FL grade 3 is a rare disease. The outcomes of FL grade 3 patients have not been well evaluated. The linear progression of FL grade 3A to 3B and then to grade 3 with areas of DLBCL represents an evolution of FL from an indolent to more aggressive form. This study included a relatively large number of patients treated uniformly by frontline RCHOP regimen. Our results reflect this feature of FL, with a better OS in FL grade 3A but a plateau of PFS curve in aggressive FL.
Generally, FL grade 3B is considered as aggressive lymphoma, whereas FL grade 3A is controversial. Biologically, FL grade 3B, compared with grade 3A, is often IRF4/MUM1 positive but CD10 negative and less likely to harbor Bcl2 rearrangement[10]. Moreover, patients with FL grade 3B or 3B with DLBCL were associated with a higher level of Ki-67 than those of FL grade 3A[11]. This support the concept the underlying genetic events were different between these two subtypes. FL grade 3A is relatively indolent lymphoma.
Clinically, our study showed that a longer OS in patients with FL grade 3A than those with aggressive FL when treated with frontline RCHOP regimen. In the rituximab era, most investigators found a higher OS rate in FL grade 3A versus aggressive FL. However, no significant difference was observed[8, 12–14]. Consistent with our results, studies from Wahlin et al and Durate et al suggested that, compared with FL grade 3A, grade 3B was related to significantly worse OS[15–17]. The discordant results may be due to a relatively small number of patients were included in these studies. Besides, our results were in line with previous studies that the PFS was similar between FL grade 3A and aggressive FL[3, 8, 14, 18–20]. Of note, three studies analyzed FL grade 3 patients with RCHOP treatment all found that grade 3A and 3B had a similar outcome in term of PFS[3, 8, 9]. Whereas we do observe a plateau in the PFS curve in aggressive FL. This could be explained by the fact that aggressive FL is a curable disease which was confirmed in a recent study showed a better PFS in aggressive FL compared with FL grade 3A[11].
The patients included in our study represent a group of patients between low grade FL and DLBCL. Our results suggested histological subtype is a prognostic factor for OS but not for PFS. FL and DLBCL had different prognostic models. Up to now, it is unclear which prognostic model, such as FLIPI-1, FLIPI-2, IPI, and PRIMA-prognostic index could better differentiate high-risk patients with FL grade 3. Our data indicated PRIMA-prognostic index may be a better PFS prognostic index than others. Besides, we found patients with more than 2 extra nodal sites involvement trend to have inferior PFS. Whereas we did not find rituximab maintenance could prolong PFS, maybe due to small number of patients received rituximab maintenance in this study. Although, some investigators have showed that certain factors, such as older age, MYC breaks, higher ECOG scores, bone marrow involvement, larger nodal dimension, B symptoms, more than 4 lymph node sites involvement, elevated serum β2 microglobulin or LDH levels were related with poor outcome in FL grade 3, these factors were not consistent in different studies [7, 8, 20].
Several limitations of this study should be addressed. Firstly, the sample size is small in this study. On the one hand, it is due to the rarity of FL grade 3 patients which only accounts for less than one fifth of the whole FL entity. One the other hand, various treatment strategies were used in these group of patients and we only included uniformly RCHOP regimen treated patients. However, compared with previous studies, our study analyzed relatively large number of FL grade 3 patients. Secondly, the median follow-up time was 40.2 months in our study which is not long enough. Further long-term follow up is needed to observe more events.