Clinical relevance of molecular characteristics in Burkitt lymphoma differs according to age

Abstract

While survival has improved for Burkitt lymphoma patients, potential differences in outcome between pediatric and adult patients remain unclear. In both age groups, survival remains poor at relapse. Therefore, we conducted a comparative study in a large pediatric cohort, including 191 cases and 97 samples from adults. While TP53 and CCND3 mutation frequencies are not age related, samples from pediatric patients showed a higher frequency of mutations in ID3, DDX3X, ARID1A and SMARCA4, while several genes such as BCL2 and YY1AP1 are almost exclusively mutated in adult patients. An unbiased analysis reveals a transition of the mutational profile between 25 and 40 years of age. Survival analysis in the pediatric cohort confirms that TP53 mutations are significantly associated with higher incidence of relapse (25 ± 4% versus 6 ± 2%, p-value 0.0002). This identifies a promising molecular marker for relapse incidence in pediatric BL which will be used in future clinical trials.

Fig. 1. Mutational landscape of pediatric and adult BL.

In total, 191 pediatric and 97 adult patient samples were analyzed. Mutations with ≥10% cohort frequency are displayed. A In total, 14 recurrently mutated genes were found in the pediatric cohort, with ID3 being the most frequently altered gene. The color code indicates mutation types or clinical data classes (see legend). B In the adult cohort, 17 recurrently mutated genes were detected. In particular, YY1AP1 and BCL2 were unmutated or less frequently mutated in the pediatric cohort.

Fig. 2. Somatic copy number aberrations (SCNAs).

Significant recurring SCNAs are displayed as blocks, based on 72 pediatric cases (55 BL and 17 B-AL cases). Curves indicate residual q-values by GISTIC 2.0 after explaining broad aberrations (Supplementary Data 7). A Focal deletions affecting E2F2 and ID3 were detected. B The most consistently observed SCNA contains GPC5 (q = 1.8e-9). Notably, the non-protein-coding MIR17HG cluster directly precedes this gene and is also affected by these gains and amplifications (Supplementary Data 6). *: might represent benign (germline) copy number variants.

Fig. 3. Age dependency of the mutation profile.

A Comparison of mutation counts between age groups, using the usual age cutoff of 18 years for adult patients. Mutations in ID3, DDX3X, SMARCA4, GNA13, ARID1A, and PTEN are significantly more frequent in pediatric patient samples, while BCL2, YY1AP1, PIM1, CREBBP, CARD11, SOCS1, and DTX1 are significantly more frequently altered in the adult cohort (P values by one-sided exact Fisher tests, see Supplementary Data 8 for exact P values and FDRs, *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001). B A cutoff-free enrichment analysis of mutation densities over age largely confirms cutoff-based results but reveals that the main biological transition of the mutational profile occurs later between 25 and 40 years (P values by one-sided enrichment analysis and permutation test with 1e5 permutations per tail).

Fig. 4. Mutational frequencies associated with clinical characteristics.

For the pediatric cohort, gene mutation rates were associated with the following clinical characteristics: age dependency (A); sex (B); BM (bone marrow, C); B-AL vs. BL (D); and, central nervous system (CNS, E). F shows the sex-specific differences in the adult cohort. FOXO1, BCL2, and CREBBP are predominantly found in the female cohort, while MYC, SMARCA4, and DDX3X are overrepresented in the male cohort (P values by one-sided exact Fisher tests, see Supplementary Data 8 for all P values and FDRs, *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001.

Fig. 5. Mutations associated with survival and cumulative incidence of relapse in the pediatric cohort.

A The overall survival for pediatric patients with SNV/indels in TP53 was significantly inferior to patients with TP53 wild-type (Kaplan–Meier survival estimation, log-rank test). B TP53^mut cases also showed an increased risk of relapse (cumulative incidence, Gray’s test, no adjustments for multiple comparisons in this descriptive context). C Mutations in PCBP1 show a tendency toward prognostic relevance. D Cases with lymphomas harboring MYC and PCBP1 mutations are characterized by an increased risk of relapse. Gains or amplifications on 13q31 specific to GPC5/MIR17HG were significantly associated with a higher risk of relapse (E). The lowest incidence of relapse (0/38 cases) was observed for the TP53^wt sub-group and mutations in FBXO11/FOXO1 (F).

Fig. 6. Mutational overview of TP53, PCBP1, FBXO11, and FOXO1 at the amino acid level in the adult (top) and pediatric (bottom) cohort.

A TP53 shows an accumulation of known hotspots at R175, R248, R273, R282, and G245 as found in other malignancies. B PCBP1 shows a hotspot truncation mutant (Y183*) identified in both age groups. C FBXO11 alterations are unique except for W112*. Most mutations were found in the β-helix domain. D In total, 74% of all pediatric FOXO1 mutations span the Akt phosphorylation motif on R19, R21, and T24. Exon boundaries are indicated using dashed lines.