Clinicopathologic features of plasmablastic lymphoma: Single-center series of 8 cases from Saudi Arabia

PBL is a rare aggressive B-cell NHL that is often related to EBV infection; it accounts
for 2.6 % of all AIDS-related lymphomas with overall survival rates between 6 and
12 months 18], 19]. The available pathological, immunohistochemical, molecular, and genetic data suggests
that plasmablastic lymphoma arises from post-germinal centre, terminally differentiated,
activated B-cells that are in transition from immunoblasts to plasma cells 20]. Array comparative genomic hybridization (involving 16 cases of PBLs) demonstrates
that despite the high degree of immunophenotypical similarity between PBL and plasma
cell myeloma (PCM) 21], the genomic aberration pattern of PBL is more similar to DLBCL than to PCM 22].

Similar to previously published studies 3], our patients are characterized by male predominance (6/8). Though the reason for
this observation is not clear, it may be due, in part, to the over-representation
of HIV infection in men compared to women in developed countries 23]. Another contributing factor could be the fact that a number of cancer-related genes,
known as cancer/testis antigens (CTAs) that are located on the X-chromosome (MAGEA1,
SSX1, SSX4, and CTAG1/2) 24]–26], are highly upregulated in PBL, which mainly affect male individuals 27]. However, CTAs can be aberrantly expressed in different types of cancer 28], 29].

In PBL, extranodal involvement occurs much more frequently than nodal involvement
with a predilection for the oral/nasal cavity 12]. In this study, nodal involvement (either primary or secondary) is common among patients.
(6/8) However, extra-oral sites were more common than oral sites. These differences
may be attributed to the small sample size.

EBV has been found to be associated with AIDS related lymphomas and is thought to
play a key role in its pathogenesis. In one study, there was a 70 % association between
PBL and EBV 12], 30]–32]. As reported by Delecluse et al., 1], 9 of 15 cases were found to be associated with EBV using ISH; of these 9 cases, 5 demonstrated expression of latent membrane proteins (LMP-1).
In another study of oral cavity PBL, 10 of 12 cases tested positive using EBV-EBER
in situ hybridization, but none of the 12 were reactive for EBV-L MP-1 13], 33]. In our study, EBV-LMP-1 also failed to react in all 8 cases examined. Although this
technique is less sensitive than ISH, our study failed to detect the presence of EBV in all 8 cases. Six cases tested
by EBER-1, were non- reactive for EBV, thus conflicting with previous reports that
demonstrated a strong association and a role of EBV infection in tumorigenesis in
a portion of PBL cases 1], 12], 33]. However, this does not hold true for all cases of PBL, as demonstrated in our study
and other reported cases of EBV-negative PBLs 1], 33]–35]. The significant value of association between EBV expression and PBL cases is unclear.
Some studies have shown a better outcome in immunocompetent patients with PBL, while
others have shown that EBV expression is not associated with the outcome of HIV-associated
PBL 36]–38]. In brief, our results (though the sample size is small) shows that EBV infection
is not the sole cause of the disease.

Although PBL was originally associated with AIDS, an increasing number of reports
has described cases arising in immunocompetent individuals 2], 36], 39]. It is currently unclear if HIV status alone confers a better prognosis in patients
with PBL. In a recent review of literature, patients with PBL and HIV infection were
found to have an OS of 14 months compared to 9 months in HIV-negative patients 36]. A review of 112 HIV positive patients with PBL showed a median overall survival
(OS) of 15 months and a 3-year OS rate of 25 % 12], compared to a review of 76 HIV-negative PBL patients that showed a median OS of
9 months with a 2-year OS rate of 10 % 40].

The reasons for the better outcome are unclear, but a potential explanation for this
finding is that the use of highly active antiretroviral therapy (HAART) may restore
immune surveillance to combat the tumor more efficiently. However, other studies do
not support this result 27], 36], 41].

In our PBL study, 2 male patients of the 8 cases (25 %) were positive for HIV and
had an average OS of 6.5 months which is similar to HIV negative patients (6 months).
This may be attributed to the fact that our patients were not on HAART at presentation
and the small sample size.

PBL is associated with rearrangements in the c- MYC gene in more than 40 % of cases 42], 43]. Valera and colleagues reported c- MYC rearrangement in 50 % of the cases reviewed and noted that EBV positive patients
are more likely to have the c- MYC abnormality and c-myc rearrangement occurs less frequently in EBV negative patients 44]. The presence of a MYC/IgH rearrangement is associated with a worse overall survival
38], 42], 44]. In our study, the c- MYC gene is not common among our PBL series as FISH analysis
failed to detect the presence of MYC translocation in 7 patients and only one case (case 7) was positive for MYC translocation. The discrepancy between our results and those previously reported
may be due to a small sample size, ethnic characteristics, or molecular characteristics
of this lymphoma in our population. In addition, all cases were EBV negative and may
serve as a contributing factor since c-myc rearrangement occurs less frequently in
EBV negative patients as reported by previous studies 44].

Since there is no standard chemotherapy protocol for treatment of this type of lymphoma,
3 of our PBL cases were treated by anthracycline based protocols (Cyclophosphamide,
Adriamycin, Vincristine, Prednisolone and Rituximab or CHOP-R), and 2 cases received
a regimen combination of chemotherapeutic drugs including etoposide, methylprednisolone
(solumedrol), high-dose cytarabine (ara-C) and cisplatin (ESHAP). Both of these regimens
were combined with radiotherapy when indicated. Two cases received palliative chemotherapy
(due to the patients’ clinical condition), and one case was treated by bortezomib
which is a proteasome inhibitor and a cornerstone in myelomas therapy and relapsed
or refractory mantle cell lymphoma 45]. Although CD20 expression was nonreactive in all cases, we used anti-CD20 monoclonal
antibody rituximab with CHOP protocol, though its use in such cases are of uncertain
utility 46].

Among the 6 patients who received chemotherapeutic regimens, none of them (0 %) achieved
complete remission (CR), 4 (66 %) had a partial response (PR) and 2 (33 %) did not
respond (NR) or progress. Since the sample size is small, we cannot give conclusive
results, but it appears that the more intensive regimen ESHAP does not show a survival
advantage when compared to the less intensive CHOP-R regimen. The treatment responses
were partial and temporary in both regimens which is similar to the results that have
been published in several past studies 38], 47]. Patients with PBL that were not treated with chemotherapy invariably died with a
median survival of 3 months 36]. Similarly, 2 of our patients received palliative therapy and died with a median
survival of 3 months. The patient (case 4) that was treated with bortezomib, a proteasome
inhibitor with dexamethasone, was without improvement and died after one month of
final diagnosis due to disease progression.

Generally, PBL has a poor prognosis with very few long term survivors and most patients
dying within 2 years from initial presentation 48]. The survival outcomes in this study are similar to those published in other studies
12], 49]. The median OS was 5.5 months (average survival time 6.4 months) from diagnosis but
varied considerably (range, 1 to 13 months). The average survival time reported by
Delecluse et al.1], was a few months, and half of the 16 patients with available follow up, died within
12 months of diagnosis. In one series of 13 patients, all died within 34 months with
a median survival of 7 months. 11] Of the 90 cases in literature that mention survival data, almost half (47 %) died
within 1 year of diagnosis; others report a very short survival 11], 35].

More recent reports have reported improved survival in HIV infected PBLs when treated
with both HAART and appropriate chemotherapy. This is similar to outcomes of HIV infected
patients with other NHLs 7], 12]. However, the outcome presented here is inferior to that recently reported 7], 12], 16]. The difference may be attributed to several factors, including ethnic and molecular
differences. Most of the cases (7/8, 95 %) presented with Ann Arbor stage IV and 6/8
are HIV negative which are associated with poor prognosis 21], 26].

Because the clinical and histopathological features are usually ambiguous, PBL remains
a diagnostic challenge. Rendering the correct diagnosis can be quite difficult in
the absence of an exhaustive integration of clinical, morphological, phenotypic and
molecular features. The diagnosis of such neoplasms can be even more challenging in
the setting of extra-oral localizations and in immunocompetent patients. The differential
diagnosis with the activated B-cell-like (ABC-like) subgroup of DLBCL and PCM with
plasmablastic morphology is still a common problem due to the lack of a distinctive
phenotype. The differential diagnosis includes immunoblastic DLBCL and other lymphoid
neoplasms with plasmacytic features such as ALK-positive DLBCL, primary effusion lymphoma
(PEL), BL with plasmacytoid differentiation and plasmablastic plasmacytoma/myeloma
(Table 3). Immunoblastic DLBCL and BL can be excluded on the basis of the characteristics
CD20 and LCA positivity in combination with negative markers of plasma cells, such
as CD138 1], 8]. PBL is distinguished from ALK-positive DLBCL by its lack of expression of the ALK
protein, and the absence of HHV8 co-infection distinguishes PBL from PEL which usually
manifests as pleural or pericardial effusion and rarely associates with lymphadenopathy
or mass. PBL is also variably positive for CD30, epithelial and endothelial markers
such EMA and CD31, posing some problems in differential diagnosis with poorly differentiated
solid tumors 50].

Table 3. Differentiating PBL from other Neoplasms by Immunophenotyping

Similar to previously reported studies, the main differential diagnosis in our study
was plasmablastic PCM particularly if extramedullary. Even though it is difficult,
it is clinically important and critical to differentiate between these two entities,
as treatment for the twodiseases is significantly different 21].

In practice, morphologic distinction is not always possible. PBL, particularly when
associated with paraproteinemia and bone lesions, (as in case 1, 3 and 4) is very
difficult to distinguish from PCM with transformation. The designation of these neoplasms
is based mainly on clinical presentation like de novo presentation as aggressive tumors
with plasmablastic morphology, predominantly extramedullary, often with mucosal involvement,
and associated with HIV and EBV. These characteristics favor PBL. PCM rarely present
de novo as high-grade lesions and the association with EBV and HIV, although previously
reported, remains quite rare 51], 52]. Recently, an immunohistochemistry stain for BLIMP1 and XBP1; markers of terminal
B-cell differentiation have been proposed to identify PBL; however, this finding remains
investigational 53] and these markers are often not routinely available. There is some controversy whether
PCM and PBL represent two separate entities or are part of a common pathway. Moreover,
Qing and colleagues reported a case where PBL occurred as a result of transformation
from a plasmacytoma 54]. However, the distinction between the two has important clinical and therapeutic
implications.