Human immunodeficiency virus-negative plasmablastic lymphoma: A comprehensive analysis of 114 cases

Human immunodeficiency virus-negative plasmablastic lymphoma (PBL) is an extremely rare entity. Its clinicopathological features, optimal treatment strategy and prognostic factors remain obsure. An extensive search was performed in the English language literature within the Pubmed database using the key words: ‘plasmablastic lymphoma and human immunodeficiency virus-negative or immunocompetent’. Data from 114 patients from 52 articles were analyzed. The mean patient age at diagnosis was 58.90 years (range, 2–86). HIV-negative PBL showed a predilection for elderly individuals (patients older than 60 years, 56.14%) and affected more males than females (M:F, 2.29:1). Ann Arbor stage IV patients accounted for 39.22% while bone marrow involvement was less frequent (12.79%). The Ki-67 index was high with a mean expression of 83%. Epstein-Barr virus (EBV) infection was common being positive in 58.70% of the patients while herpesvirus-8 (HHV-8) infection was rare being positive in only 7.55% of the patients. Immunosuppression was noted in 28.16% of patients. The median overall survival (OS) was 19 months. The 1- and 2-year survival rates were 52.3 and 45.3%, respectively. Age, gender and primary site showed no strong relationship with OS while Immunosuppression, Ann Arbor stage IV and EBV negativity were able to predict a poorer OS. Either complete remission (CR) or partial remission (PR) was superior to the refractory group in OS (P<0.0001 and P=0.0066, respectively). For stage I patients, the application of radiotherapy did not improve the OS. In conclusion, HIV-negative PBL is a distinct entity likely occurring in elderly and immunosuppressed individuals. Immunosuppression status, Ann Arbor stage IV, EBV negativity and refractory to treatment are poor prognostic factors of OS in HIV-negative PBL.


Introduction
Plasmablastic lymphoma (PBL) was firstly identified as a unique clinicopathological entity by Delecluse et al (1). For a long time, PBL was viewed as a disease exclusively involving the oral sites of human immunodeficiency virus-positive individuals. Recently, more and more cases of PBL in immunocompetent patients have been reported. However, there is little consensus concerning many aspects such as etiology, clinical findings, optimal treatment strategy and prognostic factors of HIV-negative PBL. Furthermore, Castillo et al demonstrated that HIV-negative PBL cases are rather different compared with their counterparts (2). Through an extensive literature search, 114 cases of HIV-negative PBL were described. To our knowledge, this is the most comprehensive analysis concerning PBL in HIV-negative patients. Our study provides a full-scale view and helps to expand our understanding of this unique lymphoma.

Patients and methods
Literature review. An extensive search was carried out in Pubmed using the key words: 'plasmablastic lymphoma and human immunodeficiency virus-negative or immunocompetent' in the English language literature. Only cases with definite pathologic diagnosis of PBL and description of no HIV infection were enrolled. A total of 114 cases of HIV-negative PBL were described in case reports or in small sample size case analyses between February 1997 and 2014 (1,. Data retrieval. Data were retrieved according to characteristics such as age, gender, stage, site, bone marrow involvement, Ki-67 expression, pathological findings, Epstein-Barr virus (EBV) infection, herpesvirus-8 (HHV-8) infection, immunosuppression, treatment strategy (chemotherapy, radiotherapy and surgery), treatment response, survival and prognosis. Chemotherapy included treatment with bortezomib, but excluded treatment merely with steroids. Surgery excluded incisional biopsy. Ki-67 expression with exact values was recruited, and data showing values of 1+, 2+, 3+ and 4+ were excluded. Immunohistochemistry (IHC) with ± was viewed as positive expression. Complete remission (CR) included near CR. The period from diagnosis to death or latest follow-up was considered as overall survival (OS).
Statistical analysis. Cumulative survival was expressed by Kaplan-Meier estimates and compared with the log-rank (Mantel-Cox) test. SPSS 15.0 statistical software was used for data analysis. P-value <0.05 was indicative of a statistically significant result.

Results
Clinical features. HIV-negative PBL occurred in a wide spectrum of patients, aged from 2 to 86 years, with a mean age at diagnosis of 58.90 years. notably, HIV-negative PBL mostly occurred in the elderly population. As Table I shows, patients older than 60 years accounted for 56.14% of all the cases. PBL was rarely present in young immunocompetent individuals especially children and teenagers (only 3 patients). HIV-negative PBL was more common in males with a maleto-female ratio of 2.29. With respect to clinical stage, stage Ⅳ was most common and Ⅰ was secondary. Although stage Ⅳ was noted in 39.22% of the patients, bone marrow involvement was present in only 12.79% patients. As far as the primary site was concerned, the majority was extranodal. The oral cavity, nasal cavity and sinus were 2 most common sites of involvement. A total of 15.79% of the cases involved the gastrointestinal tract. Table II, EBV infection was common in HIV-negative PBL, involving 58.70% of the patients, while HHV-8 infection was rather rare, being positive in only 4 cases. Immunosuppression status including post-transplantation, immune-related disease and current or previous malignancy was noted in quite a number of patients (28.16%). Table III, plasma cell markers CD38, VS38c, CD138 and MuM1 were expressed universally in PBL without HIV infection. EMA and CD45 were variably expressed, being positive in 59.26 and 40% of the cases, respectively. B-cell marker CD20 was rarely noted (only 1 case +, 4 cases ±). A total of 37.93% of the cases had CD79a expression. notably, a small number of cases expressed T-cell markers CD3 and CD5; 14.89 and 13.04%, respectively. Furthermore, a high number of patients (21.43%) expressed nK-cell marker CD56. Ki-67 expression was universally high with a mean value of 83%, indicating an aggressive behavior. Table IV    Prognostic factors of OS. As shown in Fig. 1, the median OS was 19 months. The 1-and 2-year survival rates were 52.3 and 45.3%, respectively. Immunosuppression was a poor prognostic factor of OS. As Fig. 2A demonstrates, patients without immunosuppression had an apparently better OS than their immunosuppression counterpart with a median survival of 36.5 vs. 6.5 months. As far as Ann Arbor stage was concerned, stage Ⅰ showed an OS advantage over stage Ⅳ (P=0.0079, Fig. 2B). EBV negativity was another poor prognostic factor of OS. As Fig. 2C shows, HIV negative PBL patients without EBV infection had a much poorer OS than EBV positive patients (P=0.0046).

Treatment and prognosis. As
Treatment response had a strong association with OS. As shown in Fig. 3, either CR or PR was superior to the refractory group in OS (P<0.0001 and P=0.0066, respectively). Furthermore, the CR group showed an OS benefit when compared with PR (P= 0.001).
univariate analysis (Table V) revealed that age, gender and primary site had no strong relationship with OS. Ann Arbor stage IV, immunosuppression, EBV negativity and refractory to treatment were poor prognostic factors of OS. For stage Ⅰ patients, the application of radiotherapy did not improve the OS.
Regarding the aspect of etiology, some studies have indicated that PBL is closely related with EBV infection (55). Our study also confirmed that 58.70% cases of HIV-negative PBL exhibited EBV infection. Furthermore, univariate analysis showed that EBV infection was a good prognostic factor of OS (P= 0.0046). The concrete mechanism need to be clarified in the future. Furthermore, only 7.55% of patients were HHV-8positive, indicating that HHV-8 infection has a negligible role in HIV-negative PBL. On the contrary, we found that immunosuppression contributed to 28.16% of the cases including post-transplantation, immune-related disease and current or previous malignancy. Additionally, immunosuppression was a poor prognostic factor for OS. Patients with immunosuppression had a poorer OS with a median survival of 6.5 months    (54). Elderly individuals often have immune senescence. Decline in immune function may be critical in the disease evolution of HIV-negative PBL.
Since PBL without HIV infection usually has a poorer response to chemotherapy compared with HIV-positive counterparts (2), there is urgent need to develop novel therapies to improve treatment efficacy. Immunotherapy may be promising.
For stage Ⅰ patients, we first explored the role of radiotherapy. However, the application of radiotherapy did not improve the OS.
In our research, statistical analysis revealed that treatment response had a defined correlation with OS. Either CR or PR was superior to the refractory group in OS (P<0.0001 and P= 0.0066, respectively). Furthermore, the CR group showed an OS benefit when compared to PR (P= 0.001). Castillo et al (54) demonstrated that CR to chemotherapy had increased OS when compared with patients without CR. Differently, our data elucidated that not only CR, but PR also could bring an OS benefit. The different treatment response was derived from tumor heterogeneity. Extensive research on the heterogeneity of HIV-negative PBL may find the causes of treatment resistance and contribute to the improvement of OS.