Clinical Utility of Ctdna Detection By NGS for Diagnosis of CNS Lymphoma

INTRODUCTION

Central nervous system (CNS) involvement in lymphoma patients (CNS-L) is a finding with a poor prognosis. Although infrequent in general, the risk is considerably higher in cases with certain clinical and biological features at the time of diagnosis, so the identification of this subgroup is key to guide prophylaxis strategies. However, despite efforts to stratify lymphoma patients according to the risk of CNS infiltration, all the factors involved have not yet been clearly defined. Currently, to rule out CNS involvement in patients at high risk of CNS relapse, magnetic resonance imaging (MRI) along with brain excisional biopsy and cerebrospinal fluid (CSF) analysis by cytology and flow cytometry (FMC) should be performed at the time of lymphoma diagnosis. However, these methods often provide inconclusive results, either due to their low sensitivity or the limited number of cells present in CSF. Therefore, diagnosis of Primary CNS lymphomas is challenging and often delays the diagnosis or force to use empirical treatments. Recently, the detection of circulating tumor DNA (ctDNA) by NGS in CSF has emerged as a valuable and promising tool to diagnose and evaluate the response to treatment of lymphoma affecting the CNS.

METHODS

This study assesses the clinical utility of ctDNA detection and monitoring by NGS in CSF in a cohort of lymphoma patients at risk of or diagnosed with CNS infiltration.

CSF supernatants from 11 patients were studied: 3 primary CNS-L; 1 patient without anatomopathological diagnosis but suspicion of Primary CNS-L; 6 systemic diffuse large b cell lymphoma (DLBCL) and 1 follicular lymphoma (FL) with isolated CNS relapse. A custom NGS panel containing the 56 most relevant mutated genes in lymphoma B malignancies was applied. NGS result was compared with available gold standard methods (MRI, FC, cytology and biopsy) at diagnosis and follow-up.

RESULTS

NGS panel detected lymphoma-related mutations in all samples at diagnosis. An average of 12 variants per patient with VAFs ranging from 1.1 to 95.7% were detected. The most frequently mutated genes were EP300, KMT2D, CD79B, PIM1, MYD88, NOTCH1, CIITA and ARID1A.

Of the three primary CNS-L, cytology and FMC detected malignant cells in only 1 of 3 patients. However, in these 2 patients, ctDNA amount was abundantly found in the CSF. Therefore, the diagnosis of these 3 patients was based on a positive MRI result.

Regarding the patient with high radiological and clinical suspicion of primary CNS-L, a biopsy could not be performed to confirm the diagnosis. FMC was not able to detect lymphoma cells, but we detected lymphoma mutations by NGS. The patient started MATRIX treatment and MRI confirmed a partial response. Nevertheless, he died because of the toxicity.

Of the six patients with DLBCL, infiltration by FMC was only demonstrated in 2 of them. Regarding imaging analysis, 4 of them had MRI compatible with CNS infiltration.

There was one patient with a FL who reach a clinical response after R-CHOP treatment. However, after 2 months he developed neurological symptoms. Then, an MRI analysis confirmed a lesion in the CNS. The presence of lymphoma cells was also confirmed by FMC and NGS.

Finally, we also evaluated NGS-ctDNA to monitor response to treatment. One of the patients with PCNS lymphoma, who achieved CR after MATRIX treatment, was monitored for 3 months, after which he presented an MRI result compatible with clinical relapse. However, NGS performed simultaneously was negative for lymphoma. All MRIs performed thereafter showed clinical improvement.

The patient with DLBCL who showed positive FMC, RMI and NGS-ctDNA results at diagnosis was also monitored and tested positive by NGS-ctDNA and CMF at the end of treatment. He eventually died of the lymphoma. The last patient monitored with LF was negative by NGS-ctDNA after 4 months of treatment and is still in CR.

CONCLUSION

These results highlight the importance of CSF analysis by NGS, postulating as a transformative tool for improving diagnosis, monitoring and outcome of CNS-L patients. Since our data demonstrate that NGS surpassed in some cases the capabilities of traditional methods, CSF should be preferably analyzed by NGS in the clinical practice.

Jiménez Ubieto:Regeneron Pharmaceuticals, Inc.: Consultancy; Roche: Consultancy, Speakers Bureau; Sandoz: Speakers Bureau; Lilly: Consultancy; Incyte: Speakers Bureau; Genmab: Consultancy; Kite-Gilead: Consultancy, Speakers Bureau; AbbVie: Consultancy, Speakers Bureau. Ruiz-Heredia:Altum Sequencing: Current equity holder in private company. Baumann:Abbvie; Incyte; Janssen; Lilly; MSD; Novartis: Consultancy; Abbvie; Astra Zeneca; Gilead Kite; Janssen; MSD; Roche: Honoraria. Rodriguez Izquierdo:Takeda Pharmaceutical, BMS, AstraZeneca: Honoraria. Cedena Romero:JANSSEN: Honoraria. Martínez-López:Pfizer: Honoraria; Altum Sequencing: Current equity holder in private company; Janssen: Honoraria. Ayala:BMS: Speakers Bureau; Altum Sequencing: Current equity holder in private company; Astellas: Speakers Bureau; Incyte: Consultancy; Novartis: Consultancy, Speakers Bureau.