Next-Generation High Throughput Sequencing Analysis In The Search Of Infectious Pathogens In Suspect Lymphoproliferations

Infectious etiologies have been established for a variety of lymphomas whereby viral, bacterial or ricketssial organisms can induce or promote lymphomagenesis either indirectly through antigen stimulation or by inflicting immune dysregulation or by directly infecting lymphocytes mimicking or co-opting signaling pathways leading to cellular transformation.

EBV infection represents the most well studied pathogen that can directly infect lymphocytes, and induce neoplasia via cellular immortalization.

An infectious etiology for a variety of EBV negative (EBV-) lymphoproliferative neoplasms/disorders, occurring in either immunocompromised or immunocompetent individuals has long been considered, but never been established.

To investigate the possibility of novel lymphotropic pathogens as causative agents of lymphomagenesis, we used next-generation, high throughput sequencing (HTS) to analyze subsets of suspect T- and B-cell lymphomas for the presence of non-human genetic material.

Forty-nine lymphomas, representing 5 different entities were evaluated: 9 EBV- negative post-transplant lymphoproliferative disorders, monomorphic type (EBV- PTLD), 10 EBV-negative classical Hodgkin lymphomas (cHL), 10 peripheral T-cell lymphomas (PTCL), 10 marginal zone lymphomas (MZL) and 10 chronic lymphocytic leukemia/small lymphocytic lymphomas (CLL/SLL).

Frozen sections of non-Hodgkin lymphoma tumor blocks were evaluated and samples with tumor representation >70% were selected for analysis.

EBV status was determined by using in situ hybridization (ISH) for EBV encoded small RNAs (EBER).

For next-generation sequencing, RNA (0.5 µg) was extracted from frozen tumors, DNase I digested (DNA-free; Ambion, Austin, TX) and reverse transcribed using Superscript II kit (Invitrogen) with random octamer primers (MWG, Huntsville, AL).

The cDNA was RNase H treated prior to random amplification by PCR.

Products of 70 bp were purified (MinElute, Qiagen) and ligated to linkers for sequencing using a GS FLX sequencer (454 Life Sciences, Branford, CT).

Primer sequences were removed, followed by multiple filtering steps and sequences obtained were compared with those of known infectious agents using software available at the BLAST website (www.ncbi.nlm.nih.gov/BLAST).

Sequencing was successful in 46 cases. Microbial sequences were detected in 5 specimens (9%).

In the remaining 41 cases, including all EBV- PTLDs and cHLs no non-human genetic material was identified.

Human herpes virus 4 (EBV) was detected in one PTCL harboring an EBV+ B-cell lymphoma which embodied 20% of the total tumor mass in the specimen (as evaluated by ISH).

EBV sequences were not detected in 4 other PTCL exhibiting EBV+ B-cells (range 1-10% involvement by ISH). These 4 cases represented angioimmunoblastic T-cell lymphomas.

Human immunodeficiency virus -1 sequences (HIV) were detected in a lung MZL occurring in a known HIV+ patient.

Sequences corresponding to propionebacterium sp., tetracyclin resistant streptococcus sp. and acinetobacter sp., were identified in one case each: MZL, EBV-PTLD and CLL; and were considered contaminants, likely acquired during biopsy procurement.

No novel lymphotropic microbial pathogens were identified in non-EBV associated T- and B-cell lymphoproliferations.

Our findings argue against a clonal infectious etiology, which has previously been hypothesized for subsets of the lymphomas analyzed.

Inability in detecting EBV sequences in samples containing low levels of EBV infected cells, suggests that this methodology might not be suitable for investigating lymphoproliferations with low tumor burden (e.g. cHL) or those arising as a consequence of chronic antigen stimulation due to a low-frequency intratumoral microbial pathogens (e.g. MZL).

Further studies in a larger cohort of lymphoproliferative neoplasms will be helpful to further validate our results.

Schecter:Seattle Genetics: Honoraria, Research Funding.