The impact of genomic markers on survival outcomes in patients with CLL treated with obinutuzumab plus either venetoclax or chlorambucil

The phase III CLL14 trial (NCT02242942), previously covered on the Lymphoma Hub, compared survival outcomes of adult patients with previously untreated chronic lymphocytic leukemia (CLL) who were given a combination of obinutuzumab plus venetoclax (VenG) or chlorambucil (GClb). Genomic analysis of this study was the first to identify the effect of markers on prognostic outcomes of these non-chemotherapeutic agents. Markers included del(17p), del(11q), unmutated IGHV, and mutated TP53, BIRC3, and SF3B1, which were all found to predict poorer survival for GClb-treated patients, while only del (17p) was independently associated with poorer outcomes in venetoclax-treated patients. Additionally, patients with unmutated IGHV particularly benefited from venetoclax, which was identified as a strong predictor of overall survival.¹

At the 16th International Conference on Malignant Lymphoma (16-ICML) and the European Hematology association (EHA)2021 Virtual Congress, Eugen Tausch provided an updated analysis, which included 2 more years of follow up, an assessment of the prognostic impact of minor mutations (<10% variant allele fraction), and analysis of resistance mutations in 113 relapse samples. We summarize key results below.^2,3

Study design

A total of 432 patients with CLL were randomized to two treatment arms:

• GClb: 12 cycles of chlorambucil, including obinutuzumab for the first 6 cycles.
• VenG: 12 cycles of venetoclax, including obinutuzumab for the first 6 cycles.

Analyses included IGHV mutational status by DNA sequencing, cytogenetics via fluorescence in situ hybridization, and gene mutations screened via next-generation sequencing. A custom next-generation sequencing panel was used, targeting the 11 most frequently mutated genes in CLL. In total, 421/432 patients were analyzed, with the eligibility criteria being patients with a Cumulative Illness Rating Scale score >6 and/or a creatinine clearance <70 mL/min.

Results

Incidence of markers 

Previous analysis revealed IGHV was unmutated in 60% of patients, while cytogenetic analysis revealed that 35% of patients had del(13q), 18% had del(11q), and 7% had del(17p). The highest incidence of major mutations—that is, mutations with a variant allele fraction (VAF) >10%—and minor mutations (VAF, 2–10%) are shown in Table 1.

Table 1. Prevalence of genomic markers*

VAF, variant allele fraction. *Data from Tausch, et al.2,3
Marker, %	Major mutations (VAF >10%)	Minor mutations (VAF, 2–10%)
NOTCH1	23	5
SF3B1	15	5
ATM	13	4
TP53	10	3
XPO1	6	—
RPS15	5	—
POT1	5	—
BRAF	<5	—
BIRC3	<5	5
NFKBIE	<5	—
EGR1	<5	—
MYD88	<5	—
FBXW7	<5	—

Prognostic impact of genomic markers

The impact of genomic markers on progression-free survival (PFS) and overall survival (OS) is summarized in Table 2. Unmutated IGHV and del(17p) were revealed to be prognostic factors, conferring significantly shortened PFS in both treatment arms. Major mutations in del(11q), TP53 and BIRC3 were additional prognostic factors associated with PFS in patients receiving GClb, while TP53 was the only other prognostic factor associated with PFS in patients receiving VenG.

Table 2. Prognostic impact of genomic markers on PFS and OS*

CI, confidence interval; GClb, obinutuzumab plus chlorambucil; HR, hazard ratio; OS, overall survival; PFS, progression-free survival; VenG, obinutuzumab plus venetoclax. *Data from Tausch, et al.2,3
Genomic marker	GClb			VenG
	HR	95% CI	p value	HR	95% CI	p value
PFS
IGHV unmutated	3.07	2.07–4.53	<0.001	2.14	1.15–3.98	0.02
Del(17p)	3.15	1.77–5.62	<0.001	3.19	1.66–6.14	0.001
Del(11q)	1.84	1.20–2.79	<0.01	1.34	0.73–2.44	—
TP53	1.66	0.95–2.90	—	2.42	1.28–4.57	<0.01
BIRC3	2.98	1.38–6.43	<0.01	1.98	0.72–5.47	—
OS
IGHV unmutated	5.67	2.59–12.41	0.01	3.52	1.53–8.17	0.23
Del(17p)	2.83	1.28–6.18	<0.01	1.61	0.74–3.52	<0.01
SF3B1	2.04	1.01–4.10	0.05	0.76	0.23–2.49	0.65
NOTCH1	1.21	0.59–2.50	0.60	1.51	0.70–3.26	0.30

In terms of OS, unmutated IGHV, del(17p), and mutated SF3B1 were prognostic factors for GClb treatment, while only del(17p) was a prognostic factor for VenG treatment.

When comparing VenG and GClb survival outcomes dependent on genetic subgroups, VenG was superior for patients with mutated (hazard ratio [HR], 0.25; 95% confidence interval [CI], 0.17–0.37; p < 0.001) and unmutated IGHV (HR, 0.36; 95% CI, 0.19–0.68; p < 0.001). VenG was also superior to GClb in patients with del(17p) (HR, 0.30; 95% CI, 0.12–0.74; p < 0.001).

Prognostic impact of minor vs major mutations

Patients with minor TP53 mutations (n = 11) had similar outcomes to patients with wild type TP53, which was perhaps a result of 10/11 patients not carrying del(17p). However, major mutations (n = 43) resulted in significantly reduced PFS in patients treated with GClb (HR, 2.01; p = 0.02).

Patients with both major (HR, 2.20; 95% CI, 1.19–4.06) and minor (HR, 2.28; 95% CI, 1.34–3.88) mutations in BIRC3 had shorter PFS. Major (HR, 1.67; 95% CI,1.21–2.30) and minor (HR, 2.25; 95% CI, 1.27–3.99) mutations in NOTCH1 also had an adverse impact on PFS. Finally, while major mutations in SF3B1 had an adverse impact on PFS (HR, 1.60; 95% CI, 1.12–2.28), minor mutations had no significant effect (HR, 0.74; 95% CI, 0.35–1.58).

Impact of G-Clb vs VenG on acquisition of high-risk variants

Analysis of VAF at baseline and relapse of 88 paired samples after GClb treatment showed an acquisition of high-risk variants, including TP53 (n = 4), BIRC3 (n = 5), SF3B1 (n = 3), and RPS15 (n = 1). Analysis of 25 paired samples before and after VenG treatment revealed a single variant each was gained for high-risk TP53 and BIRC3 at relapse. Notably, there was no detection of venetoclax resistance mutations in BCL2 or other genes of interest.

Multivariate analysis of PFS and OS

Regarding PFS, multivariate analysis revealed VenG as a favorable prognostic factor, while IGHV unmutated, del(17p), complex karyotype, and tumor lysis syndrome (risk category high) were all adverse prognostic factors (Table 3).

For OS, del(17p), serum β2 >3.5 mg, and age ≥75 years were all reported as significantly adverse prognostic factors (Table 3).

Table 3. Significant prognostic factors associated with PFS and OS following multivariate analysis in both treatment arms*

CI, confidence interval; CKT, complex karyotype; HR, hazard ratio; OS, overall survival; PFS, progression-free survival; TLS, tumor lysis syndrome; VenG, obinutuzumab plus venetoclax. *Data from Tausch, et al.2
Factor	HR	95% CI	p value
PFS
VenG	0.24	0.17–0.34	<0.001
IGHV unmutated	2.51	1.77–3.58	<0.001
Del(17p)	1.98	1.20–3.29	0.008
CKT	1.92	1.28–2.87	0.002
TLS risk high	1.53	1.08–2.16	0.016
OS
Del(17p)	4.58	2.59–8.15	<0.001
serum β2 >3.5 mg	2.46	1.36–4.45	0.003
Age ≥75 years	1.79	1.11–2.90	0.02

Conclusion

This updated analysis at 52 months provided further evidence for the prognostic value of IGHV mutation status and del(17p) in patients treated with obinutuzumab and venetoclax or obinutuzumab and chlorambucil. VenG continued to produce superior survival outcomes dependent on such markers and was shown to be a favorable independent factor for PFS in a multivariate analysis. Additionally, treatment with VenG had less impact on VAF burden, with only two high-risk mutations identified at relapse, while patients treated with GClb were found to have acquired several high-risk variants at relapse. Finally, no BCL2 resistance mutations were reported for VenG-treated patients at relapse, supporting its use for re-treatment.