ASCO 2017 | DLBCL: Interim results of the phase III OPTIMAL>60 trial and identification of clinically relevant copy number alterations in patient plasma samples

At the 2017 American Society of Clinical Oncology (ASCO) Annual Meeting, an oral abstract session took place that was jointly chaired by John M. Pagel, MD, PhD, of the Swedish Cancer Institute, Seattle, WA, USA, and Ranjana H. Advani, MD, from the Stanford Cancer Institute, Stanford, CA, USA.

Two key oral abstracts were presented on Diffuse Large B-Cell Lymphoma (DLBCL).

Abstract 7506

This abstract presentation included the first interim analysis results of the phase III OPTIMAL>60 study (NCT01478542) and was given by Michael Pfreundschuh, MD, from University Saarland Medical School, Saarbrücken, Germany.

The presentation began by discussing a paper by Held et al. who compared cohorts of the phase III RICOVER-60 study: one cohort received R-CHOP x6 plus rituximab x2 plus involved-field Radiotherapy (RT), and the other did not receive RT (RICOVER-noRTh; prospective amendment to trial). Overall, 164/166 RICOVER-noRTh patients were evaluable after a median observation time of 39 months. In multivariate analysis of the ITT population, EFS of patients was worse without additive RT (HR, 2.1; 95% CI, 1.3–3.5; P = 0.005). Also, patients who did not receive RT demonstrated trends towards inferior PFS (HR, 1.8; 95% CI, 1.0–3.3; P = 0.058) and OS (HR, 1.6; 95% CI, 0.9–3.1; P = 0.127). In a per-protocol analysis of 11 patients who were excluded from RICOVER-noRTh due to undergoing unplanned RT, multivariate analysis demonstrated HRs of 2.7 (95% CI, 1.3–5.9; P = 0.011), 4.4 (95% CI, 1.8–10.6; P = 0.001), and 4.3 (95% CI, 1.7–11.1; P = 0.002) for EFS, PFS, and OS in patients not receiving RT to bulky disease, respectively.

The phase III OPTIMAL>60 study aimed to improve outcomes and reduce toxicity in elderly (61–80 years) patients with CD20+ aggressive DLBCL by administering an optimized schedule of rituximab and liposomal vincristine; followed by FDG-PET based reduction of therapy by substitution with vitamin D. Prospective analysis of whether RT can be spared in PET-negative patients after R-CHOP was also conducted.

The trial followed a randomized 2x2 factorial design, with patients assigned to:

• R-CHOP-14 (conventional vincristine 2mg)
• Opti-R-CHOP-14 (conventional vincristine 2mg)
• R-CHLIP (liposomal vincristine 2mg/m²)
• Opti-R-CHLIP (liposomal vincristine 2mg/m²)

Patients with bulk (≥7.5 cm) PET-positive disease after 6 cycles chemotherapy were assigned to RT (39.6Gy), while PET-negative bulk patients were observed. The planned interim analysis was performed after 40% of expected events to ensure that PET-based reduction of RT did not put patients at risk. Historical comparisons were made with patients with bulky disease (>7.5cm) included in the RICOVER-60 trial.

PET assessment was carried out in 166 of 187 patients with bulky disease in the OPTIMAL>60 study. Reasons for no PET scan included early death (n=5), excessive toxicity (n=3), protocol violation (n=1), non-compliance (n=4), change of diagnosis (n=6, and others (n=2). Of the 166 patients, 80 (48%) remained PET-positive after 6 cycles of chemotherapy and 62/187 patients (33%) were irradiated compared to 67/117 (57%) in RICOVER-60; resulting in a relative reduction in RT of 42%. Reasons for not undergoing RT in the OPTIMAL>60 and RICOVER-60 trials included progression (44% vs. 36%), medical reasons (toxicity, 50% vs. 33%), protocol violation (0% vs. 20%), negative biopsy (6% vs. 0%), and others (0% vs. 11%).

Michael Pfreundschuh went on to state that, even with the unfavorable characteristics, the outcome of the 187 OPTIMAL>60 patients with bulky disease was non-inferior to those in RICOVER-60. In patients who were administered 6xCHOP-14 plus 8xR in the OPTIMAL>60 (n=47) and RICOVER-60 (n=117) trials, 2-year PFS was 79% and 75%, respectively. OS in these patients was 88% and 78%, respectively.

Multivariate analysis adjusting for IPI, the HR of the OPTIMAL>60 vs. RICOVER-60 bulky disease patients was 0.7 (95% CI, 0.3–1.5; P = 0.345) for PFS and 0.5 (95% CI, 0.2–1.3; P = 0.154) for OS.

Lastly, when comparing PET-positive patients with no RT (n=18), PET-positive with RT (n=62), and PET-negative patients with no RT within the OPTIMAL>60 study, PFS at approximately 50 months was around 35% vs. 70% vs. 80%, respectively.

This abstract presentation was concluded by stating that this planned interim analysis confirmed that PET-based RT to bulky disease is safe. Additionally, RT to bulky disease can be spared in PET-negative patients after immune-chemotherapy without negatively affecting outcome.

While at the 2017 ASCO Annual Meeting, we interviewed Professor Soubeyran on new approaches for older patients with DLBCL, watch it here.

Abstract 7507  

Michael C. Jin from the Stanford University School of Medicine, Stanford, CA, USA, presented the second oral abstract on DLBCL, titled “Non-invasive detection of clinically relevant copy number alterations in Diffuse Large B-Cell Lymphoma.”

Circulating Tumor DNA (ctDNA) is an emerging biomarker through a range of methods including fraction of total Cell Free DNA (cfDNA), immunoglobulin high-throughput sequencing (single gene), and targeted Next Generation Sequencing (NGS). ctDNA can be used to profile tumors according to cell of origin, identify resistance mechanisms, and stratify patient outcomes.

Jin stated that Somatic Copy Number Alterations (SCNAs) are described as the gain or loss of regions, or entire, chromosomes. They have been found to occur frequently in DLBCL (Lenz et al. 2008). In addition, SCNAs have been found to be prognostic for patient outcomes and response to treatment. For example, extra copies of MYC and BCL2 in tissue is prognostic for poor OS in patients who have received R-CHOP (Ennishi et al. 2017), and Classical Hodgkin Lymphoma patients with amplifications of PD-L1 had the highest responses to PD-L1 checkpoint inhibitors (Ansell et al. 2015). However, non-invasive detection of SCNAs remains a challenge.

Using CAPP-Seq, a targeted high-throughput sequencing platform, the group developed a method to profile both focal and broad SCNAs from plasma with the aim of determining if clinically relevant SCNAs can be identified genome-wide.

Overall, plasma from a cohort of 75 patients with DLBCL from five different sites (Stanford, MD Anderson, NCI, Italy, and Germany) were assessed pre-treatment for ctDNA by CAPP-seq, as well as 48 healthy controls. Focal SCNAs were evaluated at ultra-high depths (around 10,000x), allowing for detection of lesions at approximately 1% ctDNA fraction. The median age of patients was 61 years (range, 25–84) and 62% were male. Twenty-three percent of patients had high-risk IPI, and most patients (78%) were treatment naïve. Non-GCB and GCB subtype made up 47% and 53% of patients, respectively. Nearly three-quarters (73%) of patients had received DA-EPOCH-R or R-CHOP, and a ctDNA fraction >1% was found in 70% of patients.

In terms of performance against tissue genotyping, 23 tumors were analyzed for genome-wide SCNAs. Concordance as a function of ctDNA plasma burden was assessed in 101 matched plasma samples. When ctDNA >1% (70% of samples): sensitivity was 77% and specificity was 97%. When ctDNA >5% (38% of samples): sensitivity was 95% and specificity was 97%. All future analyses only used samples if ctDNA >1% (n=117). It was observed that pre-treatment ctDNA allele fraction increased with stage (stage I–II vs. stage III–IV; P < 0.05).

The clinically relevant genes recurrently found to have SCNAs are listed in the table below:

Moreover, amplifications of BCL2 (39% vs. 17%), BCL6 (31% vs. 15%), and PD-L1 (39% vs. 25%) were more frequently observed in the non-GCB subtype of DLBCL compared to the GCB subtype of DLBCL. This group’s findings regarding copy number gains in PD-L1 were consistent with the literature:

Jin drew attention to the fact that 26.7% of the cohort demonstrated amplification of both PD-L1 and PD-L2 and amplifications of only PD-L2 occurred in 13.3% of patients. Amplifications of PD-L1 were more common in relapsed than those who were newly diagnosed (43.5% vs 19.2%; P = 0.02). Most PD-L1 amplifications were focal (65%) while the rest typically involved >80% of chromosome 9p.

Gains in MYC and BCL2, as well as deletion of TP53, were found to predict poor outcomes in patients with DLBCL. SCNAs in MYC, BCL2, and TP53 are significant independent of translocations; these findings are consistent with what has been reported with tissue biopsies.

This abstract presentation was concluded by stating that for most patients with DLBCL, SCNAs can be detected in a non-invasive way at diagnosis. Sensitivity and specificity of alterations in tissue compared to plasma was high, especially in advanced disease with high tumor burden (ctDNA >5%). Clinically significant SCNAs were identified and monitored in plasma using a genome-wide approach which identified: PD-L1 gains in DLBCL, cHL, and PMBCL; a high frequency of PD-L1 aberrations in DLBCL (potentially relevant to patient selection for checkpoint immunotherapy trials); and ctDNA alterations in MYC, BCL2, and TP53 predict poor patient survival.

We interviewed Prof. Gilles Salles, the chair of our Executive Steering Committee, asking what the highlights in DLBCL were at the 2017 ASCO Annual Meeting: watch the video interview here.