Mixed responses to targeted therapy driven by chromosomal instability through p53 dysfunction and genome doubling

Abstract

The phenomenon of mixed/heterogenous treatment responses to cancer therapies within an individual patient presents a challenging clinical scenario. Furthermore, the molecular basis of mixed intra-patient tumor responses remains unclear. Here, we show that patients with metastatic lung adenocarcinoma harbouring co-mutations of EGFR and TP53, are more likely to have mixed intra-patient tumor responses to EGFR tyrosine kinase inhibition (TKI), compared to those with an EGFR mutation alone. The combined presence of whole genome doubling (WGD) and TP53 co-mutations leads to increased genome instability and genomic copy number aberrations in genes implicated in EGFR TKI resistance. Using mouse models and an in vitro isogenic p53-mutant model system, we provide evidence that WGD provides diverse routes to drug resistance by increasing the probability of acquiring copy-number gains or losses relative to non-WGD cells. These data provide a molecular basis for mixed tumor responses to targeted therapy, within an individual patient, with implications for therapeutic strategies.

Fig. 1. TP53 pathway disruption is associated with shorter progression-free survival and mixed clinical responses to TKI therapy.

a Bar chart showing the percentage of responding patients with homogenous (gray) or mixed (red) responses to treatment with erlotinib or chemotherapy. b Mixed responses in the RECIST database were analysed using response criteria defined by ref. . Patients with at least two lesions where one shrank by at least 30% were included in the analysis. The number of patients with homogenous responses are shown in black for patients receiving erlotinib. The different patterns of progression seen in patients with a mixed response are shown in red. c Kaplan–Meier survival analysis of patients with E (n = 35, yellow line) and EP tumors (n = 82, green line), in the AURA2, AURA3, and AURA phase II expansion cohort, demonstrating the difference in PFS after osimertinib treatment (Log-rank test (two-sided) p = 4e⁻⁰⁴, HR 0.36, CI: 0.20–0.65). d Bar chart of the proportion of Homogenous (yellow) and Mixed (green) responses to osimertinib in patients with E or EP tumors (p = 0.0106 two-sided Fisher’s exact test). e Bar chart of the proportion of patients with E or EP mutant tumours with new lesions during osimertinib treatment (p = 0.0846 two-sided Fisher’s exact test). f Individual first tumor response within six months on osimertinib treatment, presented as % change in CT-measured tumor length. Each x-axis tick represents one patient (n = 21, E group of patients with 127 lesions and n = 39, EP group of patients with 246 lesions in total). The dotted lines show the Reiter et al. criteria for response (−30%) and progression (10%), respectively. Gray dots and whiskers represent the median change in tumor size and variability around the median value using the median absolute deviation (MAD) for each patient. Boxes underneath the graph indicate the occurrences of new lesions (red box), and mixed responses of existing lesions, as defined by ref. (gray box), and patients with mixed responses with or without the occurrences of new lesions (blue box). Source data are provided as a Source Data file.

Fig. 2. Trp53 loss results in mixed responses and therapy resistance in murine models of NSCLC.

a Mouse-to-human across genome synteny histograms. Upper panel E mice vs. patients with E tumors. Lower panel EP mice vs patients with EP tumors. Significantly changed regions in both species are colored pink (gain) and blue (loss). b Kaplan–Meier survival analysis of E (n = 10, yellow line) and EP (n = 17, green line) mice, demonstrating the difference in OS after erlotinib treatment (p < 0.0001, HR 3.72, 95% CI: 1.65–8.38, log-rank Mantel–Cox test). c Differences in tumor responses after one month of erlotinib treatment in E (n = 12 yellow) and EP (n = 16 green) mice, presented as % change in CT-measured tumor diameter. Each column represents one mouse, and each dot represents one tumor within the mouse (p = 0.006464, two-sided Mann–Whitney U-test). The dotted lines show the Reiter et al criteria for response (−30%) and progression (10%), respectively. d Bar chart showing the proportion of sensitive and resistant tumors in E (yellow) and EP (green) mice (p = 0.0082, two-sided chi-squared test). The total number of mice in each group are indicated in the bars. e Dot plot showing time to resistance in E (n = 13 yellow) and EP (n = 18 green) mice (P = <0.0001 two-sided Mann–Whitney U-test). f Bar chart showing identified single-nucleotide variant-related resistance mechanisms in E and EP mice. Source data are provided as a Source Data file.

Fig. 3. Trp53, together with WGD, results in therapy resistance associated with increased CIN and cell-to-cell variability.

a Dot plot of mean cancer cell fraction per tumor of early losses in naïve (E, n = 23; EP n = 20) and resistant E (n = 9 yellow) and EP (n = 10 green) tumors. b Frequency of copy-number gains (positive y-axis) and losses (negative y-axis) in treatment naïve (yellow) vs resistant (green) tumors with either E (upper panel) or EP (lower panel) genotypes. c Ploidy-relative copy-number gains (red colors) are reported across all mouse tumors, separating treatment naïve vs resistant and E vs EP (ploidy represented in gray colors) for 18 genes whose amplification is known to have an impact on TKI resistance. d For every group (triangle shape) of single cells obtained from the same FACS ploidy peak (red colors) from either naïve E (top row), naïve EP (second row), resistant E (third row), or resistant EP (bottom row) mouse tumors, the fraction of the genome affected by different SCNAs (yellow-to-blue colors) was computed between every pair of cells (square within a triangle) as a proxy to measure cell-to-cell diversity. e Dot plot of average Shannon evenness index measured per tumor from naïve (n = 21) and resistant (n = 7) E (yellow) vs naïve (n = 19) and resistant (n = 710) EP (green) mouse tumors (naïve p = 0.0066, resistant p = 0.0004, two-sided Mann–Whitney U-tests). Bar charts showing WGD frequencies in lesions from patients with E and EP tumors from f Tx421 (using WES data, ns, two-sided chi-squared test) and g OncoSG cohort (using WES data, p = 0.0080, two-sided chi-squared test). Dot plot showing weighted Genome Instability Index (wGII) of tumors with or without WGD in patients with E or EP lesions from the h Tx421 (non-WGD: E n = 3, EP n = 1. WGD: E n = 9, EP n = 11 p = 0.0310, two-sided Mann–Whitney U-test) and i OncoSG cohorts (non-WGD: E n = 26, EP n = 10, p = 0.8758. WGD: n = 9 E, n = 11 EP, p = 0.0392, two-sided Mann–Whitney U-test). Source data are provided as a Source Data file.

Fig. 4. Genome doubling permits elevated ploidy and promotes multiple avenues to therapy resistance in the presence of p53 pathway dysfunction.

a Plot showing the number of resistant subclones generated from each of the 24 triploid and 24 hexaploid progenitor clones after 5 weeks of culture in 1.5 μM erlotinib. b Number of triploid (blue) and hexaploid (red) progenitor PC9 clones that generated at least one erlotinib-resistant subclone (p = 0.0346, chi-squared test). c Left panel: Presence of somatic mutations in genes related to the EGFR pathway (black squares) is reported across all triploid (upper blue row) and hexaploid (upper red row) resistant daughter clones derived from either triploid (lower blue row) or hexaploid (lower red row) parental clones. Right panel: Ploidy-relative copy-number gains (red colors) are reported for resistant daughter clones that have changed their ploidy state for 13 genes whose gain is known to have a role in TKI resistance. d Frequency of copy-number gains (positive y-axis) and losses (negative y-axis) are reported across either triploid (blue) or hexaploid (red) resistant daughter clones, highlighting events affecting oncogenes and tumor suppressors. e Clone-to-clone diversity measured by computing the copy-number difference (fraction of genome with different copy numbers reported in blue colors) between either left all pairs of triploid and hexaploid resistant daughter clones derived from the same parental clone (triangles), or right all pairs of triploid and hexaploid resistant daughter clones. f Impact of siRNA mediated repression of gained genes on re-sensitization of erlotinib-resistant hexaploid PC9 subclones. By factoring in effects on viability, the effect of gene silencing on erlotinib resistance was scored. Tiles corresponding to genes exhibiting a significant treatment-varying response upon knockdown (p < 0.05) in a hexaploid subclone are colored. Hue corresponds to the direction of change, and brightness to the erlotinib treatment status. Gene names depicted in bold did not impact parental PC9 viability. g Model depicting factors contributing to tumor resistance in E and EP tumors. Source data are provided as a Source Data file.