Polymorphic alleles of ABCB1 or VEGF were predictive of lenalidomide efficacy in patients with mantle cell lymphoma who received lenalidomide maintenance after autologous stem cell transplant.
Polymorphic alleles of ABCB1 or VEGF were predictive of lenalidomide (Revlimid) efficacy in patients with mantle cell lymphoma (MCL) who received lenalidomide maintenance after autologous stem cell transplant (ASCT), according to findings from a pharmacogenetic analysis of germline biomarkers in the phase 3 MCL0208 trial (NCT02354313).1
Polymorphisms of ABCB1 or VEGF appeared in 69% and 79% of 278 patients included in the pharmacogenetic study, respectively, and they predicted favorable progression-free survival (PFS) outcomes vs those with homozygous wild-type ABCB1 or VEGF in the lenalidomide arm. In those who received lenalidomide, the 3-year PFS rate for patients with ABCB1 variants was 85% vs 70% in those with ABCB1wild-type (P = .05), and the 3-year PFS rate for those with VEGF variants was 85% vs 60% in those with VEGF wild-type (P = .0021).
The prospective, multicenter MCL0208 trial investigated the efficacy and safety of lenalidomide maintenance vs observation after frontline treatment with rituximab (Rituxan) followed by high dose chemotherapy and ASCT in patients younger than 66 years with previously untreated MCL. The 3-year PFS rate in the enrolled population (n = 300) was 67% and the 3-year overall survival (OS) rate was 84%.2
Of 248 patients who received ASCT, 205 were randomly assigned to receive lenalidomide maintenance (n = 104) or undergo observations (n=101). The 3-year PFS rate was 80% (95% CI, 70%-87%) in the lenalidomide arm vs. 64% (95% CI, 53%-73%) in the observation arm (stratified HR, 0.51; 95% CI, 0.30-0.87; P = .013).
“We investigated specific germline polymorphisms of transmembrane transporters, metabolic enzymes, and cell surface receptors [ABCB1, ABCG2, VEGF-A, FCGR2A, NCF4, GSTP1, CRBN] that might predict the efficacy and safety of immunochemotherapy and lenalidomide maintenance,” lead study author Simone Ferrero, MD, of the University of Torino in Turin, Italy, and colleagues, wrote in a paper of the data that was published in Blood Advances.1
This analysis used bone marrow and peripheral blood samples collected at diagnosis and during follow-up from patients enrolled in MCL0208. MCL0208 enrolled 300 patients with no clinically significant comorbidities who received 3 cycles of R-CHOP-21 (rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone), followed by rituximab and high-dose cyclophosphamide, 2 cycles of rituximab and high-dose cytarabine, and ASCT. Patients who were randomly assigned after ASCT received either lenalidomide at 15 mg on days 1 through 21 every 28 days or observation for 24 months.
Among all enrolled patients, 93% (n = 278) had adequate data available for inclusion in the pharmacogenetic study. In total, 96% (n = 197/205) of the randomized population were genotyped, including 97% of the patients in the lenalidomide arm (n = 101/104) and 95% of those in the observation arm (n = 96/101).
The frequency of the 3 ABCB1 single nucleotide polymorphisms—wild type homozygous, heterozygous polymorphic, and polymorphic homozygous—was 18.8%, 51.5%, and 29.7%, respectively, in the lenalidomide arm, and 16.7%, 50.0%, and 33.3%, respectively, in the observation arm. The investigators reported no significant differences in haplotype distribution between the 2 arms or between the enrolled and randomized populations (χ2 = 0.264; P = .876).
Among the investigated polymorphisms, ABCB1 rs2032582 and VEGF rs699947 were associated with PFS and OS benefits with lenalidomide. In the randomized population, 31% (n = 60), 54% (n = 107), and 15% (n = 30) of patients had wild-type homozygous, heterozygous, or polymorphic homozygous ABCB1 variants, respectively, and 21% (n = 42), 49% (n = 96), and 30% (n = 59) of patients were had wild-type homozygous, heterozygous, or polymorphic homozygous VEGF variants, respectively.
In total, 70% (n = 137) of the randomized population had at least 1 ABCB1 rs2032582 polymorphic allele. Patients with heterozygous or polymorphic homozygous variants had better outcomes vs those with wild-type homozygous variants in the lenalidomide arm, but not in the observation arm. The 3-year OS rate was 98% in patients with heterozygous or polymorphic homozygous variants vs 90% in those with wild-type homozygous variants (P = .026).
In the randomized population of patients with VEGF-A variants, the 3-year OS rate was 90% in those with heterozygous or polymorphic homozygous variants vs 86.5% in those with wild-type homozygous variants (P = .094).
Patients with ABCB1 or VEGF-A variants trended toward deeper minimal residual disease clearance in the bone marrow after 6 months of lenalidomide treatment vs those with wild-type homozygous ABCB1 or VEGF-A.
When the randomized population was stratified by pharmacogenomic background, a greater PFS benefit was observed in patients with at least 1 variant of ABCB1 (HR vs ABCB1 wild-type, 0.41; 95% CI, 0.22-0.75) or VEGF-A (HR vs VEGF-A wild-type, 0.51; 95% CI, 0.30-0.87; P = .012).
Patients with both wild-type ABCB1 and VEGF-A variants (n = 17) did not benefit from lenalidomide over observation (P = .632).
Lenalidomide dose reduction was significantly associated with the CRBN rs1705814 genotype. In the overall population, 41% (n = 113), 33% (n = 91), and 26% (n = 73) of patients had wild-type homozygous, heterozygous, or polymorphic homozygous CRBN variants, respectively. Additionally, 28% (n = 28) of the patients in the lenalidomide arm had polymorphic homozygous variants. These patients were at a higher risk of requiring a lenalidomide maintenance dose reduction of more than 66% or lenalidomide discontinuation compared with those with wild-type homozygous or heterozygous CRBN variants (odds ratio, 3.24; CI, 1.69-6.21; P = .013). However, in the lenalidomide arm, the investigators observed no statistically significant effects of the CRBN rs1705814 single nucleotide polymorphism on infections or hematological toxicities.
In the overall population, the 3 polymorphisms associated with hematological toxicity were ABCB1 rs2032582, NCF4 rs1883112, and GSTP1 rs1695. Seventeen percent (n = 46) of patients with ABCB1 TT/AT/AA genotypes were found to be at lower risk of hematological toxicity after their first R-CHOP cycle compared with patients with heterozygous or wild-type homozygous genotypes (odds ratio for grade 3 or higher toxicity, 0.39; CI, 0.15-0.88; P = .033). Additionally, 58% (n = 162) of patients had NCF4 AG/GG genotypes and were at a lower risk of developing hematological toxicities during induction vs those with wild-type homozygous genotypes (odds ratio, 0.56; CI, 0.34-0.92; P = .024). Moreover, 10% (n = 28) of patients had the GSTP1 GG polymorphic homozygous genotype and had a lower risk of toxicities than those with heterozygous or wild-type homozygous genotypes, with an odds ratio of 0.35 (CI, 0.15-0.79; P = .014).
The investigators also identified a subgroup of 48 patients with wild-type homozygous NCF4 and GSTP1genotypes who were at increased risk of hematological toxicities during induction therapy (odds ratio, 2.26; CI, 1.09-4.83; P = .031). However, these single nucleotide polymorphisms did not influence hematologic toxicities during later treatment phases or post-ASCT hematological recovery.
Additionally, in the overall population, ABCB1 rs1045642 and CRBN rs1705814 were associated with infections during chemoimmunotherapy. For ABCB1 c.3435.C>T, 25% (n = 70), 51% (n = 143), and 23% (n = 65) of patients had wild-type homozygous (CC), heterozygous (CT), and polymorphic homozygous (TT) genotypes, respectively. For CRBN rs1705814 T>C, 41% (n = 113), 33% (n = 91), and 26% (n = 73) of patients had wild-type homozygous (TT), heterozygous (TC), and polymorphic homozygous (CC) genotypes, respectively. A total of 65 patients with the ABCB1 wild-type homozygous genotype had a lower infection risk than those with CT or CC genotypes (odds ratio, 0.53; CI, 0.30-0.95; P = .030). Furthermore, 73 patients with the polymorphic homozygous CRBN rs1705814 genotype had a lower infection risk than those with wild-type homozygous or heterozygous genotypes (odds ratio for grade 3 or higher AE, 0.39; CI, 0.22-0.68; P = .001). No single nucleotide polymorphisms predicted infection onset during lenalidomide maintenance.
“Our main findings suggest a key role of ABCB1 and VEGF-A polymorphisms in enhancing the clinical activity of lenalidomide maintenance after ASCT in MCL,” study authors concluded.