Despite combination antiretroviral therapy, high-grade malignant non-Hodgkin's lymphoma (NHL) is still one of the most frequently acquired immunodeficiency syndrome (AIDS)-defining disorders in the end stage of infection with human immunodeficiency virus (HIV). NHL can also be observed in rhesus macaques infected with the simian immunodeficiency virus (SIV). Thus, they represent a useful model to study morphological characteristics and oncogenetic mechanisms of NHL in humans.
When reviewing the occurrence of lymphoma at the German Primate Center over the past 25 years within the context of pathogenic SIV infection we noticed a strikingly high incidence (four out of seven animals) of these tumors in rhesus macaques infected with ex vivo derived SIVmac251/32H/spleen in AIDS-defining end-stage disease. Polymerase chain reaction analysis of this virus stock revealed the co-presence of rhesus lymphocryptovirus (rhLCV), which represents the monkey homologue to human Epstein–Barr virus (EBV), suggesting an association between co-application of SIV and rhLCV and increased tumorigenesis.
In addition, we present two cases of NHL in rhesus macaques infected with a
SIVmac239
Lymphomas are linked to human immunodeficiency virus (HIV) infection in that
they are known to occur in association with acquired immunodeficiency
syndrome (AIDS) and are therefore classified as an AIDS-defining disease.
Lymphoid neoplasms can be divided into Hodgkin's and non-Hodgkin's lymphomas
(NHL). Most of the AIDS-associated lymphomas are high-grade malignant NHL of
Simian immunodeficiency virus (SIV) infection of rhesus macaques
(
So far, reports evaluating the incidence of lymphoma depending on the used SIV challenge stocks are not available. Some authors describe the overall incidence of lymphoma development in cohorts of rhesus macaques ranging between 4 and 19 % (Habis et al., 1999; Kahnt et al., 2002) but associations with different SIV stocks are missing. Our report describes the relation between differently prepared SIVmac stocks and the incidence of lymphoma at the German Primate Center over the last 25 years. In addition, we present two cases of NHL in SIV-infected rhesus macaques, one because of its unusual location and the other because of its rare immunophenotype.
The monkeys were housed at the German Primate Center under standard
conditions complying with
This study comprised 165 rhesus macaques infected with SIVmac; 163 animals were included in the evaluation of SIV stock-dependent incidence of lymphoma. The selection criterion for this cohort was euthanasia because of early AIDS-defining illness as described (Siddiqui et al., 2009).
The two remaining animals are presented as case reports of SIV-associated
lymphomas and were infected with a SIVmac239
Animal 2 (5.5 years, male) died unexpectedly at week 45 pi. Besides mild gastrointestinal symptoms, the animal did not show overt clinical signs prior to death. Tissue samples were processed as described for animal 1.
Flow cytometric data from the two case report animals were compared with
those of seven other rhesus macaques infected in parallel with SIVmac239
(
Sampling of blood, lymph nodes and mononuclear cells (MNCs) from lymph nodes by fine-needle aspiration before and after the experimental SIV infection was performed as previously described (Klippert et al., 2015). As reported, colonic biopsies were collected and MNCs isolated (Schultheiss et al., 2011).
Blood samples, a peripheral lymph node and tissue from the colon were taken
from animal 1 during necropsy. The lymphoid and colon tissues were placed in
RPMI1640 medium (PAN-Biotech, Aidenbach, Germany) on ice enriched with 10 % fetal bovine serum (FBS) Gold (PAA, Pasching, Austria) and 1 %
penicillin, 1 % streptomycin (both PAN-Biotech), 1 % gentamycin (Life
Technologies GmbH, Darmstadt, Germany), 50
Immunological and virological analyses for animal 2 were performed in blood and lymph node 1 week before death and in colon biopsies at 36 pi.
Whole blood (50
After fixation for at least 48 h, tissue samples of various organs and
of grossly visible neoplastic lesions were paraffin-embedded, sectioned at 3
Immunophenotyping of lymphocytic tumors was performed by immunohistochemical
staining of paraffin-embedded sections using an anti-CD3 antibody labeling
IHC was performed with an automated immunostaining system (Discovery XT, Roche Diagnostics GmbH, Mannheim, Germany) using the SABC (streptavidin–biotin complex) method and DAB (diaminobenzidine tetrahydrochloride) for signal detection (DAB Map Kit, Roche Diagnostics GmbH, Mannheim, Germany). Rhesus lymph node or lymphoma tissue from a rhesus macaque with confirmed rhesus lymphocryptovirus infection served as positive controls. Pure antibody diluent was applied to the negative control sections.
Quantification of the plasma viral load for SIV was carried out as previously described (Klippert et al., 2015).
Baseline characteristics of the SIV-infected rhesus macaques
(
Abbreviations: DPZ is German Primate Center, Germany; CPRC is Caribbean Primate Research Center, Puerto Rico; MIC is Morgan Island Colony, USA; DSTL is Defence Science and Technology Laboratory, GB. The rhesus macaques of the DPZ originated from the CPRC, MIC, and the Centre de Primatologie, France.
The SIVmac251/32H/spleen stock was analyzed for the co-presence of LCV. For
this purpose, total DNA was prepared from 0.2 mL of the virus stock using
the First-DNA all-tissue kit (GEN-IAL GmbH, Germany) according to the
manufacturer's instructions. A pan-herpesvirus consensus PCR
targeting the DNA polymerase gene was carried out in a nested format
(Chmielewicz et al., 2003) using 0.1
A total of 163 SIV-infected rhesus macaques were included to calculate the incidence of lymphoma development in SIV-infected rhesus macaques at the German Primate Center over the past two and a half decades. It should be noted that this report includes only macaques that were euthanized with early AIDS-defining illness. The animals were infected with different cell-free strains of SIVmac, i.e., 251, 239, 32H and 251/32H/spleen. SIVmac251 and SIVmac239 represent the two prototypes of SIV from macaques (Daniel, 1985). SIVmac215/32H was derived from an in vivo titration of SIVmac251 (Rud et al., 1994). SIVmac251/32H/spleen had been prepared as a homogenate from a spleen of a SIVmac251/32H-infected monkey without any in vitro passage (Dittmer et al., 1995). Details about the cohort are given in Table 1.
As shown in Fig. 1a, 7 of 67 animals infected with SIVmac251 developed
lymphoma, accounting for an incidence of 10 %. Infection with SIVmac239
yielded a comparable incidence with 8 of 70 animals with lymphoid
tumors. For the rhesus macaques infected with SIVmac251/32H, the incidence
was about the same (16 %) with a total of 3 of 19 animals affected by
lymphomas. The incidence was significantly higher in the animals infected
with SIVmac251/32H/spleen (57 %) with four out of seven animals
developing lymphoid tumors (
Animals presenting with lymphoma at the time of euthanasia survived the
infection on average longer than those without (median survival
without lymphoma is 37 wpi and with lymphoma is 82.5 wpi;
Since we suspected the spleen-derived SIV stock to contain
lymphocryptovirus, a pan-Herpes virus PCR was performed with the cell-free
SIV-containing fluid and the DNA sequence of the amplicon determined. A
comparison of the 175 base pair DNA sequence with published sequences of
lymphocryptovirus from macaques revealed 99 % identity to the published
genome of the
Percentages of
Here, we report about an in-depth analysis of two special cases of lymphomas
in animals originating from the DPZ breeding colony and infected with a
CD20
Percentage of expression of different surface and intracellular markers in SIV-associated NHL based on immunohistochemistry.
Abbreviations: SIV is simian immunodeficiency virus; NHL is non-Hodgkin's lymphoma; EBNA2 is Epstein–Barr nuclear antigen 2.
Gross findings in animals 1 and 2.
Animal 1 revealed an ulcerated tumorous thickening of the gastric body wall measuring approximately 2 cm in diameter (Fig. 3a). Adjacent pancreaticoduodenal lymph nodes were severely enlarged.
Major gross findings in animal 2 were observed in the urogenital tract.
There was bilateral hydroureter (diameter 0.5 mm) and bilateral
hydronephrosis with moderate atrophy of the renal cortex and medulla. A
large tumorous white mass (approximately 5
Histopathological examination of the gastric body mass of animal 1 revealed a densely cellular, infiltrative, unencapsulated, poorly demarcated round cell neoplasm composed of sheets of medium-sized to large neoplastic lymphocytes separated by a collagenous stroma. Tumor cells had indistinct cell borders with a small amount of finely granular eosinophilic cytoplasm. Nuclei were irregularly round, often indented, vesicular and revealed one to three prominent nucleoli. There was mild anisokaryosis and anisocytosis of the centroblastic tumor cells. The mitotic rate ranged from 6 to 10 per high power field (HPF). There were multifocal aggregates of infiltrating and residual small mature lymphocytes throughout the neoplasm (Fig. 4a). Infiltrative tumor growth led to focally extensive effacement of the gastric mucosa. The tumor extended into the outer muscular layer with occasional evidence of tumor cells in the gastric serosa. Immunohistochemically, the tumor cells equally expressed CD3 and CD20 (Fig. 4b and c, Table 2). While a small amount of tumor cells were only positive for CD3 or CD20, the majority of neoplastic lymphocytic cells expressed both CD3 and CD20. The enlarged pancreaticoduodenal lymph nodes revealed a comparable lymphocytic proliferation, which obscured normal lymph node architecture and extended into the perinodal adipose tissue. However, there were less mature lymphocytes and fewer amounts of collagenous stroma within the neoplastic tissue, and compared to the gastric neoplasm tumor cells were slightly larger (Fig. 4d). Almost all lymphocytic cells of the nodal tumor expressed CD20 admixed with only few CD3 positive cells (Fig. 4e and f, Table 2). In both tumors of animal 1, the majority of cells showed nuclear expression of Ki67, confirming the high proliferative activity of the lymphocytic neoplasms.
Light microscopic images of the gastric (
The tumorous mass in the urinary bladder neck region of animal 2 was represented by an unencapsulated neoplasm composed of sheets of round cells accompanied by a sparse collagenous stroma. Neoplastic cells were medium-sized to large and had variably distinct borders, scant amounts of eosinophilic granular cytoplasm and round to oval vesicular nuclei with one to three distinct nucleoli compatible with centroblasts. There was mild anisocaryosis and anisocytosis. The mitotic rate averaged five to seven per HPF (Fig. 5a). The neoplastic lesion led to effacement of the luminal mucosa, showed transmural expansion and infiltrated the caudal segments of the seminal vesicles (Fig. 5b). Almost all lymphocytic cells of the bladder tumor expressed CD20 and Ki67 (Fig. 5c, Table 2). Only a few CD3 positive cells could be observed within the tumor of animal 2 (Table 2).
Based on the histological appearance together with the immunophenotype and
Ki67 expression, the tumor of animal 2 was diagnosed as a centroblastic
DLBCL according to the WHO lymphoma classification. In animal 1,
immunophenotyping and histological features of the nodal tumor were also
compatible with a centroblastic DLBCL. In contrast, final diagnosis for the
gastric neoplasm remained questionable due to the extensive expression of
CD3 in this tumor. CD3 expressing lymphocytic cells partly represented
infiltrating non-neoplastic
EBNA-2 expression was present in all examined tumors indicating co-infection with rhLCV in both animals (Fig. 5d, Table 2).
Light microscopic images of the urinary bladder tumor of animal 2.
First, we analyzed the incidence of SIV-associated lymphoma in relation to
the infecting virus stock. The highest incidence of lymphoma was observed
after infection with the ex vivo prepared SIVmac251/32H/spleen
stock leading to lymphomagenesis in almost 60 % of the macaques. When
using the viruses SIVmac251, SIVmac239 or SIVmac32H for infection of rhesus
macaques, we found lower incidences ranging between 10 and 16 %,
similar to those reported by others. These overall incidences of lymphoma in
cohorts of SIV-infected rhesus macaques varied between 4 % (Habis et
al., 1999) and 19 % (Matz-Rensing et al., 1999; Kahnt et al., 2002). A
detailed association of lymphoma development in relation to the infecting
virus has not been performed in those studies. The high incidence of
lymphoma linked to infection with our SIVmac251/32H/spleen stock was
striking and also significant when compared to the total of the other
groups. This virus stock was derived from a tissue homogenate prepared from
the spleen of an adult SIV-infected monkey without any in vitro culture (Dittmer et al., 1995). As it is well known that almost all
macaques older than 2 years of age are latently infected with
lymphocryptovirus (Fujimoto and Honjo, 1991; Wang, 2013), we therefore
suspected the co-presence of this virus in our SIV stock. Homogenization of
the spleen could have released not only SIV from
In addition, we focused on two cases of lymphoma in chronic SIV infection
because they were special with respect to either the localization or the
phenotype. Both animals presenting with tumorous masses had the
co-infection with oncogenic rhLCV in common. Similarly, CD4
Still, both lymphoma cases differed in certain aspects. First, lymphoma in animal 1 arose at an extranodal as well as a nodal site. It is not clear whether a primary extranodal or a primary nodal NHL development was present in this case. The appearance of a primary extranodal NHL in the stomach with metastasis to the adjacent lymph node is conceivable, but a secondary NHL with first appearance in the lymph node and metastasis to the stomach is also possible.
The gastrointestinal tract (GIT) is the most common extranodal site for NHL, and the stomach represents the most affected organ within the GIT (Chen et al., 2010; Yang et al., 2011; Ding et al., 2016). A primary GIT NHL is rare. Usually, it appears as a secondary tumor (Ghimire et al., 2011). In scientific literature, the definition of primary extranodal NHL is debatable. Krol et al. (2013) subdivided cases of NHL into primary nodal, primary extranodal and NHL with extensive involvement. A correlation between primary extranodal NHL and longer disease-free periods in humans was found (Krol et al., 2003). In our case, it is impossible to trace back the exact chronological development of tumorigenesis, especially since both a nodal as well as an extranodal site in close anatomic proximity were involved.
The immunophenotype of the gastric lymphoma of animal 1 is also debatable.
We diagnosed the nodal lymphoma as a classic DLBCL because of the dominant
CD20 expression with only minor CD3 staining. As stated in the
introduction, DLBCL is the most common type of NHL tumor
(Armitage and Weisenburger, 1998). In contrast, levels of
CD20 and CD3 expression were similar in the gastric tumor. A minority of CD3
positive cells probably represented pre-existing or recruited non-neoplastic
Animal 2 presented with lymphoma of the DLBCL type located at the urinary bladder neck leading to bilateral urethral obstruction and death due to acute uremia. Lymphoma of the urogenital tract is rare, and a primary extranodal lymphoma in the urinary bladder is even less common (Venyo, 2014). The most common type is the extranodal marginal zone lymphoma of mucosa-associated lymphoid tissue type (MALT lymphoma) and the DLBCL (Bates et al., 2000). The latter develops through transformation of the MALT lymphoma. In an earlier study at our center, just 1 out of 16 SIV-infected rhesus macaques with lymphoma developed a tumor in the urinary bladder (Kahnt et al., 2002).
To summarize, our cases of lymphoma in two SIV-infected rhesus macaques represented special entities regarding either the immunophenotype or localization of the tumor.
In conclusion, we correlated the incidence of SIV-associated lymphoma that occurred during the last 25 years at the German Primate Center with the different virus stocks used for infection. Particularly an ex vivo derived SIV, designated SIVmac251/32H/spleen, was associated with a high incidence of lymphoma development which might be due to the presence of an additional pathogen, i.e., the tumorigenic rhLCV, in this virus stock.
Two rhesus macaques with lymphomas developing in the chronic phase of
SIV-infection were investigated in more detail in this study. Lymphoma
development is a common complication of SIV infection. Both cases revealed
unusual characteristics. One animal presented a DLBCL at an extranodal as
well as nodal site with co-expression of the
All relevant data are presented in the paper. Please contact the corresponding author for further details.
Antonina Klippert and Martina Bleyer wrote the manuscript with contribution of Maria Daskalaki, Berit Neumann, Ulrike Sauermann and
Christiane Stahl-Hennig. Antonina Klippert
and Maria Daskalaki collected the samples. Martina Bleyer performed the macroscopical, histological
and immunohistochemical examinations. Ulrike Sauermann calculated the incidence and
prepared the associated figures and table. Artur Kaul performed the
lymphocryptovirus PCR. Antonina Klippert and Berit Neumann analyzed the flow cytometry data. Frank Kirchhoff
provided the SIVmac239
The authors declare that they have no conflict of interest.
The authors would like to thank Wolfgang Henkel for the assistance during necropsy and Larissa Hummel, Nadine Schminke, Corinna Boike, Judith Hampe, Sandra Heine and Kerstin Eckelmann for excellent laboratory assistance. Edited by: E. Fuchs Reviewed by: G. Koopman and one anonymous referee