Comparative analysis of survival and mutation spectrum according to Ki-67 (low [< 50%] vs high [≥50%]) in AH-MCL and genomic profile in patients with AH-MCL (AH-DN and AH-t) compared with nonaggressive MCL (classic variant). (A) Median survival was significantly longer in patients with low Ki-67 (118 months) vs high Ki-67 in AH-t (20 months) (P < .001). The cutoff point of 50% was based on classification and regression tree analysis. (B) Median FFS was significantly longer in patients with low Ki-67 (27 months) vs high Ki-67 (10 months; P < .001). The cutoff point of 50% was based on classification and regression tree analysis. (C) Pattern of somatic mutation distribution in AH-MCL (n = 42) vs nonaggressive MCL (n = 39). All alterations were identified by using WES. Mutation frequencies are shown as nonaggressive (light blue) and aggressive (dark blue). Genes with nonsynonymous mutations or copy number alterations in ≥2 patients are listed. CCND1 gene was mutated more frequently in the aggressive group. Mutations of UBR5, NOTCH2, and NOTCH3 were exclusively observed in the aggressive group. (D) Violin plots depicting remarkable variation in the degree of aneuploidy in AH-t vs AH-DN vs nonaggressive MCL groups. Significantly higher degree of aneuploidy was observed in AH-t and AH-DN MCL, compared with the nonaggressive category (P < .0021). Kruskal-Wallis test. Boxes in the box plot indicate interquartile range and the center line the median. (E) Pattern of somatic mutations in high (n = 30) and low (n = 10) Ki-67 categories. Differences in the 2 groups were statistically significant with a distinct somatic mutation profile in patients with high Ki-67. The bar graphs on either side show the accumulated counts of somatic alterations for each specific gene in their group. Almost all of the somatic mutations were exclusive in patients in the high Ki-67 group. CCND1 mutations were significantly higher in the high Ki-67 group. (F) The schematic diagram shows the CCND1 and NSD2 protein domains and the positions of specific mutations. The length of the line that connects the mutation annotation to the protein is directly proportional to the number of samples with the mutation. The most recurrent mutations are shown in the diagram.