Rapid tumor DNA analysis of cerebrospinal fluid accelerates treatment of central nervous system lymphoma

Abstract

Delays and risks associated with neurosurgical biopsies preclude timely diagnosis and treatment of central nervous system (CNS) lymphoma and other CNS neoplasms. We prospectively integrated targeted rapid genotyping of cerebrospinal fluid (CSF) into the evaluation of 70 patients with CNS lesions of unknown cause. Participants underwent genotyping of CSF-derived DNA using a quantitative polymerase chain reaction-based approach for parallel detection of single-nucleotide variants in the MYD88, TERT promoter, IDH1, IDH2, BRAF, and H3F3A genes within 80 minutes of sample acquisition. Canonical mutations were detected in 42% of patients with neoplasms, including cases of primary and secondary CNS lymphoma, glioblastoma, IDH-mutant brainstem glioma, and H3K27M-mutant diffuse midline glioma. Genotyping results eliminated the need for surgical biopsies in 7 of 33 cases (21.2%) of newly diagnosed neoplasms, resulting in significantly accelerated initiation of disease-directed treatment (median, 3 vs 12 days; P = .027). This assay was then implemented in a Clinical Laboratory Improvement Amendments environment, with 2-day median turnaround for diagnosis of CNS lymphoma from 66 patients across 4 clinical sites. Our study prospectively demonstrates that targeted rapid CSF genotyping influences oncologic management for suspected CNS tumors.

Graphical abstract

Figure 1.

Clinical integration of rapid CSF genotyping for diagnosis of suspected CNS neoplasms. (A) Retrospective patient cohorts from before rapid genotyping introduction were included in this study to serve as benchmarks for diagnostic trajectories of CNS neoplasms and potential for detection of hallmark variants in CSF. Diverse neoplastic and nonneoplastic diagnoses were encountered in 62 patients with new CNS lesions with radiographic suspicion for neoplasms (historical cohort, left panel). Among 52 patients with PCNSL (PCNSL cohort, middle panel) and 21 patients with nonhematologic CNS malignancies (CID cohort with known variants present in diagnostic biopsy site tissue, right panel), analyzing variant status and safety of lumbar puncture revealed the maximum proportion of patients with potentially detectable variants in CSF. (B) Prospective clinical workflow for returning TetRS rapid genotyping assay results to providers during real-time diagnosis of suspected CNS neoplasms. Excess CSF was used to perform the TetRS assay in a research laboratory setting. Results were reported to the clinical team, who then included them in tumor board discussions and collaborative treatment decisions with patients. Conventional CSF assays included chemistries, cell counts, cytology, flow cytometry, cultures, microbiologic, serologic (antibody), and antigen testing. (C) The categories and proportions of final diagnoses secured in the 70 patients in the rapid genotyping cohort. The inner circles adjacent to each final diagnosis display the detection of variants in CSF by TetRS.

Figure 2.

Clinical impact of rapid CSF genotyping. (A) This shows the method of securing final diagnoses among 33 patients diagnosed with a neoplasm in the rapid genotyping cohort. Each illustrated person represents 1 patient; the color of the head represents whether a variant was detected in CSF by rapid genotyping, and the color of the body represents the final diagnosis reached. (B) Among 14 patients in the rapid genotyping cohort with a variant detected in CSF by TetRS, the variant detected is displayed below the final diagnosis established in each case. (C) Plots show the number and percentages of patients with clinical effects of positive variant detection by TetRS. (D) The clinical course of a representative patient with a new CNS neoplasm diagnosed by rapid CSF genotyping without neurosurgical biopsy is shown. A 69-year-old woman (patient RG003) presented with deep periventricular-enhancing lesions concerning for CNS lymphoma, glioma, or metastasis. TetRS detected the MYD88 L265P variant and was negative for TERT promoter variants. Positron-emission tomography computed tomography scan, magnetic resonance imaging of the spine, and ophthalmologic examination were negative for other sites of disease. Primary CNS lymphoma was diagnosed on the basis of TetRS variant detection, enabling methotrexate and rituximab initiation within 4 days. After completion of induction chemotherapy and during consolidation, lumbar punctures were negative for MYD88 L265P, correlating with radiographic response to treatment. MYD88 L265P was again detected in CSF at the time of recurrence; retreatment with radiation and ibrutinib was initiated, with ongoing response 1 month later. Additional TetRS and orthogonal sequencing results are displayed in the supplemental Data. IgH, immunoglobulin heavy chain rearrangement; TERT denotes the TERT C228T and C250T variants.

Figure 3.

Clinical implementation of rapid genotyping of cell-free DNA in CSF accelerates diagnosis of CNS lymphoma (CNSL). (A) Analysis of a historical cohort of 1007 patients with new CNS lesions of unknown cause for which CNSL was in the differential diagnosis showed that only 14.7% are ultimately diagnosed with CNSL. Securing a CNSL diagnosis required 95.6% of patients to undergo neurosurgical biopsy, resulting in a median time to diagnosis of 10 days (top pathway, adapted from Gupta et al²). Because 53% of CNS lymphoma cases in our institutional experience were positive for the MYD88 L265P variant (Figure 1), we estimated that 7.7% of the initial cohort would harbor this variant. Following clinical implementation of the Clinical Laboratory Improvement Amendments–certified cell-free DNA assay, a similar percentage of CNSL bearing the MYD88 L265P mutation was identified, with a median turnaround time of 2 days (bottom pathway). (B) The rate of accrual of CSF specimens was 1 sample/5.3 days in the prospective rapid genotyping cohort (middle panel) vs 1 sample/3.2 days following clinical implementation (right panel) (P = .002). The frequency of MYD88 L265P detection in these study populations was 8 of 70 patients (11.4%) and 5 of 66 patients (7.6%), respectively, similar to the 78 of 1007 patients (7.7%) found in the historical cohort (left panel).