Abstract
Introduction
The diagnosis of B and T cell malignancies relies on the demonstration of B-cell (BCR) or T-cell (TCR) antigen receptor clonality. This can be studied through the analysis of V(D)J rearrangements of BCR and TCR genes by PCR (van Dongen Leukemia 2003) or, more recently, by high-throughput sequencing (HTS). Amplification of a clonal population with a "primers approach" could fail in case of hybridization problems due to too fragmented DNA, somatic mutations or polymorphic variations. Here we evaluated the performance of a HTS capture system for the analysis of B and T-cell clonality in clinical samples from mature T or B malignancies. We further combined this technology to concomitant sequencing of oncogenes of interest.
Patients and Methods
DNA was extracted from 58 tumoral samples from fresh/frozen (FF) cells or tissues or formalin-fixed paraffin-embedded tissue (FFPE) (n=19). These samples comprised various T-cell [i.e. 1T-cell prolymphocytic leukemia, 1 T large granular lymphocytic leukemia, 2 Sézary syndrome, 4 peripheral T-cell lymphoma not otherwise specified, 14 angioimmunoblastic T-cell lymphoma] or B-cell [i.e. 14 chronic lymphocytic leukemia, 1 mantle cell lymphoma, 5 diffuse large B-cell lymphoma, 1 grey-zone lymphoma, 13 Hodgkin lymphoma, 1 Poppema, 2 Waldenström and 1 multiple myeloma] malignancies. The Biomed-2 PCR technique was used as standard for assessing the performance of TRG, IGH and IGK clonality analysis. An extensive panel of capture probes was designed (SureSelect XT HS2 DNA system, Agilent Technologies) that covered the variable (V), + diversity (D) and junction (J) segments of the IGH, IGK, TRG, TRB loci and diagnostic/theranostic genes of interest i.e. B2M, BTK, CARD11, CD28, DNMT3A, IDH2, JAK3, PLCG1, PLCG2, ROHA, SOCS1, STAT3, STAT5B, STAT6, TET2, TNFAIP3, TP53. Paired-End sequencing was performed on a MiSeq system (Illumina) in 300, 500 and 600 cycles. Analysis of clonality profiles was performed using Vidjil software and SeqOne.
Results
HTS runs resulted in a median total read count of 1,6M (0.7-2.9) per sample. V(D)J rearrangements were identified with a median of 1503 reads (189-6824) per sample. Five samples were excluded because less than 300 rearranged reads were obtained. The number of rearranged reads and of clonotypes identified are influenced by the number of sequencing cycles (300<500 or 600) but not by the quality of DNA (FFPE vs FF).
Analyses of tumoral samples with HTS versus PCR were compared. For the IGH locus (n=47), comparable PCR/HTS clonal (n=22) and polyclonal (PCL, n=20) profiles were identified. One discordant case showed a clonal PCR profile and a PCL HTS profile but the IGK was clonal. For the IGK locus (n=23), 10 clonal and 12 PCL cases were similar with both techniques. One case appeared discordant with a PCL PCR profile but a clonal HTS profile. For the TRG locus (n=31), PCR and HTS profiles were similar in 14 clonal, 5 oligoclonal and 9 PCL cases respectively. Three cases were discordant with oligoclonal PCR profiles but a clonal HTS profile.
Overall in the 38 cases of B-cell malignancies, 27 and 11 cases had a concordant B-cell clonal or PCL profile with PCR and HTS. Among PCL cases, only one was discordant with a clonal HTS profile. This case and 3 other PCL cases were Hodgkin lymphomas which all disclosed another mutation (i.e. TP53, TNFAIP3, SOCS1). Of the 20 cases of T-cell malignancies, 14 displayed a clonal TRG profile with PCR and HTS. Among them, 13 showed oncogene mutations that confirmed the oncogenic nature of the clonal proliferation. Among 6 patients with a non-clonal PCR TRG profile, two cases of AITL and T-LGL had a discordant clonal TRG HTS profile and both also had specific mutations (SOCS1, RHOA and STAT3 respectively). Two other AITL samples showed a T-PCL profile with PCR and HTS but also had a mutation/CNV (RHOA, SOCS1).
Conclusion
A very good performance of B and T cell clonality assessment was obtained here with capture-HTS compared to Biomed-2 PCR. The combined identification of mutation/CNV allowed to confirm the malignant character in cases of clonal or PCL lymphoproliferations, while concomitantly specifying the type of lymphoproliferative disorder. The combined capture-HTS of B and T repertoires and oncogenes of diagnostic or theranostic interest thus appears as an efficient, accurate and useful approach for the diagnosis of mature B and T lymphoid malignancies in clinical practice.
Disclosures
No relevant conflicts of interest to declare.
Analysis of a single-institution cohort of patients with Felty's syndrome and T-cell large granular lymphocytic leukemia in the setting of rheumatoid arthritis
