Delays to Diagnosis of Peripheral T-Cell Lymphoma and Clinical Utility of Molecular Profiling: A UK Multicentre Real-World Study

Background: Peripheral T-cell lymphomas (PTCL) comprise a group of heterogeneous and aggressive lymphomas, often associated with poor outcomes. Histological heterogeneity and variability in clinical presentation and tissue biopsies are postulated to influence timely diagnoses and outcome, although this has not been formally evaluated. Knowledge of the mutational profiles of PTCLs is well established, however, these have not translated into routine diagnostics tumour profiling, with poor uptake despite guideline recommendations in the National Comprehensive Cancer Network and UK National Genomic Test Directory (NGTD). Here, we evaluated the diagnostic pathway for the two most common PTCLs - angioimmunoblastic T-cell lymphoma (AITL) and peripheral T-cell lymphoma, not otherwise specified (PTCL-NOS) and the utility of mutational profiling.

Methods: Clinicopathologic data, including number, type and timing of biopsies and outcome, from consecutively diagnosed AITL and PTCL-NOS patients (pts) between January 2013 and June 2022 from 13 UK centres were collected. Progression-free survival (PFS) and overall survival (OS) were estimated using Kaplan-Meier method.

A custom 113-gene panel comprising recurrently mutated PTCL genes, including the 10 genes in the NGTD (RHOA, DNMT3A, IDH2, TET2, PLCG1, CARD11, IRF4, POT1, VAV1, FYN) was used for mutational profiling (mean sequencing depth >500x) on a subset of the cohort (N=39 samples from 31 pts, 19 AITL, 12 PTCL-NOS).

Results: In total, 281 pts (AITL, N= 162, PTCL, N= 119) were included. The median age was 69 years (range 20 to 94 years). The majority were male (62%). Among the evaluable cohort, 72% had an International Prognostic Index (IPI) score >2 and 91% had stage ≥3 disease. The most common 1st biopsy was a core biopsy (54% of pts), followed by a surgical excisional biopsy (16%) and the remainder included bone marrows, needle aspirates, skin and organ-specific biopsies. The median number of biopsies required for diagnosis was 2 (range: 1-6 days) with 52% of pts requiring >1 biopsy. The median time from first biopsy to diagnosis was 20 days (5 – 869 days).

If the first biopsy was an excision, pts required fewer biopsies for diagnosis (median: 1) and had a shorter time from first biopsy to diagnosis; 17 days (6- 869 days) compared to 20 days (5 – 672 days) with other 1^st biopsy types. 71% of patients who underwent an excision first did not require additional biopsies, compared to only 56% of patients who underwent a core biopsy.

Comparing the diagnostic timelines to Hodgkin lymphoma (HL), a disease often associated with diagnostic delays, between 2018-2022 (HL, N=79, AITL/PTCL-NOS, N=111), the median time to diagnosis was 12 days (2 – 237 days) for HL and 21 days (3 – 672 days) for AITL/PTCL-NOS. The median number of biopsies required for diagnosis was 1 (1-4) in HL compared to 2 (1-6) in AITL/PTCL-NOS.

Most pts received treatment (86%), of whom 73% had first-line CHOP-like chemotherapy. With an estimated median follow up time of 56.8 months (0 – 114.6 months), the median PFS and OS were 9.2 and 18 months respectively, with the 5-year OS estimate of 25% (95% CI: 20-32%). There was no significant difference in survival between pts who had prolonged diagnostic pathways (>1 biopsy or >28 days from presentation to diagnosis) versus those who did not, perhaps reflecting the overall adverse outcomes in PTCL.

Amongst the 39 samples (31 pts) that underwent gene panel sequencing, 68% of patients (21 of 31 pts) had detectable mutations in at least 1 NGTD gene and 30% had >1 gene affected. Thirteen of 21 of these pts had multiple biopsies. For pts with multiple biopsies sequenced, 5 out of 7 patients had the same mutations in a NGTD gene detectable in both the earliest and subsequent biopsies. A higher proportion of AITL biopsies had detectable mutations compared to PTCL-NOS patients (16/19 vs 5/12). For the remaining 10 pts without detectable mutations in NGTD genes, a subset had mutations in other genes from the broader PTCL panel.

Conclusions: This represents the first real-world PTCL dataset providing clear evidence of significant diagnostic delays and the need for multiple biopsies. An excision biopsy clearly provides superior diagnostic accuracy but may not always be practical. Our pilot genomic profiling highlights its complementary clinical diagnostic utility in PTCL that might circumvent the need for multiple biopsies; this approach should be prospectively evaluated.