S-ALCL is a lymphoma with an aggressive and heterogeneous clinical course, primarily occurring in children and young adults. Most investigators reported that the response of ALCL to chemotherapy was good, especially in children, ranging from 60%–90% [2, 9, 29]. However, for patients at high-risk for treatment failure, other therapies such as high-dose chemotherapy followed by stem cell therapy may be needed to improve the long-term survival . Therefore, identification of prognostic factors would aid clinicians in selecting an optimal therapeutic regimen.
In this study, we analyzed clinical characteristics and prognostic factors in Han Chinese patients with S-ALCL from a single referral center. We found some unique clinical features in our S-ALCL patients. The median age of our patients was 26 years regardless of ALK expression status, and even when pediatric patients were excluded, the median age of ALK- S-ALCL was 34, which is much younger than other reports [6–10, 30–32]. The percentage of patients with clinical stages III/IV in our study was also less than previous reports (50 vs. 56%–72%) [6–10, 30–32]. Since some studies [8, 33] excluded pediatric patients, we also analyzed the subset of patients aged 19 years or older and compared the results to those reported by Savage et al.. We found that some of the clinical features of ALK+ S-ALCL in our analysis were similar to those in Savage et al., but Chinese patients with ALK- S-ALCL are younger (median age 34 vs. 58), with a lower percentage of B symptoms (35 vs. 57%), and different sites and frequency of extranodal involvement, compared to those reported by Savage et al.. Our study found that skin, spleen, lung, and bone were common sites in ALK- S-ALCL, while in Savage’s report bone marrow, subcutaneous tissue, bone, liver, and spleen were often involved. The 5 year-OS of our S-ALCL patients was also lower than Savage’s report; however, their patient population had varied ethnic backgrounds (25% North American, 24% European, and 40% Asian), which might account for some of these differences.
Our study found that patients who had a CR following induction chemotherapy had a more favorable prognosis, suggesting that achieving a CR during induction chemotherapy is a principal factor of longer OS. In our cohort of patients, most deaths occurred in the first year after diagnosis, which indicates that S-ALCL is a rapidly fatal disease and effective induction chemotherapy is essential for long-term survival.
Although previous studies [6, 7] identified several clinical characteristics, such as advanced clinical stage, older age, extranodal involvement, and B symptoms with a high-risk disease, clinical stage was the only independent clinical prognostic marker in our study. None of the other clinical factors had statistical significance whether by univariate analysis or by multivariate analysis, consistent with the report by Savage et al.. In addition to the clinical characteristics, we also analyzed the expression of several tumor biomarkers. We found for the first time that Ki-67 had prognostic significance in S-ALCL. Furthermore, the expression of Ki-67 also defined different risk categories in patients with ALK- S-ALCL, although not in ALK+ S-ALCL.
ALK+ S-ALCL may have a more favorable clinical course than ALK- S-ALCL [6–8], but recent study found that activation of ALK could provides oncogenic addiction to tumors harboring activating mutation or translocation of ALK such as in non-small cell lung cancer, so ALK inhibitor may be a potent novel targeted therapeutic in some solid tumor . In our Han Chinese patients, the 5-year OS was 58% in ALK+ S-ALCL patients versus 36% in ALK- S-ALCL patients, which was lower than other reports showing 70%-80% survival in ALK+ versus 33%-49% in ALK- S-ALCL [6, 7, 31, 32]. Though poorer outcome for ALK- S-ALCL patients, they may be divided into different groups by stage or expression of Ki-67. Recently, Savage et al. found that the favorable outcome of ALK+ S-ALCL was restricted to younger patients and no outcome differences were seen between the ALK+ and ALK- groups in patients over 40 years old. In contrast, Sibon et al. reported that there was no impact of ALK status on survival in patients younger than 40 years old and Park et al. also reported no predictive value for ALK expression. Wang et al. thought that suppressor Tregs in ALK+ S-ALCL could suppress the anti-tumor immune response induced by effector T cells, which maybe weaken the role of ALK. Besides, Beltran et al. found the favorable prognostic role of ALK expression in DLBCL and the clinical course of ALK- DLBCL was aggressive. Compared to the previous reports, in our study, the favorable prognostic influence of ALK expression seems to be confined to pediatric patients. In univariate analysis, in patients who are 14 years old or younger, the ALK+ group had better survival rate than ALK- group, but this survival difference was not present in patients older than 14 years. Nevertheless, in multivariate analysis, ALK was no longer a significant prognostic factor whether in pediatric or adult patients. These results suggest that the prognostic value of ALK expression may be affected by other as yet unknown factors. Clinical stage or biomarkers may ultimately be more important predictors of prognosis.
It is not known why some patients with ALK+ S-ALCL have better outcomes. One study showed that the tumor cells of ALK+ S-ALCL could be associated with higher levels of apoptosis by chemotherapeutic drugs than those of ALK- S-ALCL . Stress-induced apoptosis is a pathway involved in chemotherapy-induced cell death, and BCL-2 may inhibit the cell death pathway by suppressing the function of pro-apoptotic molecules [23, 24]. Previous studies found that BCL-2 expression is almost completely restricted to ALK- S-ALCL and is correlated with a poor outcome [13, 14], however, in our study, BCL-2 expression did not have a significant influence on long-term survival, although the expression of BCL-2 was more frequently found in ALK- patients than in ALK+ S-ALCL patients. Nevertheless, to fully evaluate the role of BCL-2 in S-ALCL, a larger number of samples would be needed.
The role of WT1 as an unfavorable prognostic marker has been confirmed in acute leukemia and some of solid tumors [20, 21]. Unlike its role in acute leukemia, WT1 was not a significant prognostic factor in S-ALCL in our study. Over-expression of WT1 was more frequently found in ALK+ S-ALCL and there was a significant positive correlation between the expression of WT1 and ALK. This may indicate that ALK expression is affected by multiple factors and may explain why ALK was no longer a prognostic marker in our multivariate analysis. The functional interaction of ALK and WT1 in S-ALCL is a worthwhile target for future studies, although larger studies are required to validate the prognostic value of WT1 expression in S-ALCL.
In conclusion, the Han Chinese patients with S-ALCL in our study had some unique clinical features, including younger age distribution and a slightly higher percentage of early stage. Our data add to other reports stressing a potential contribution of ethnic and racial background on clinical and biological characteristics of hematological malignancies. Similar to other reports, in our study the expression of ALK and Ki-67 and clinical stage are significant prognostic factors for S-ALCL patients in univariate analysis. However, clinical stage is the only independent prognostic marker in multivariate analysis. Nevertheless, it is possible that these prognostic factors may play a role in different age groups and different populations. A larger scale and preferably multi-institutional study will be needed to confirm the prognostic role of geographical or ethnic differences. Further characterization of these and other prognostic factors in S-ALCL patients will provide better prognostic guidance for stratifying patients for future therapeutic trials.