NEW YORK – In a study published in Nature Cancer on Monday, researchers confirmed that small-cell lung cancers tend to become heterogeneous and rapidly develop resistance over the course of treatment. The findings demonstrate the importance of maximizing initial responses in treatment-naïve patients using rational therapy combinations, they concluded.
In the study, a team led by MD Anderson Cancer Center's Lauren Byers confirmed through single-cell RNA sequencing of circulating tumor cells in xenograft models and from a patient's blood sample that untreated SCLC tumors are molecularly heterogenous to start with. They further showed that treatment relapse is associated with increasing variation in the expression of therapeutic target genes and epithelial-to-mesenchymal transition (EMT) genes in clusters of cancer cells within a tumor.
Single-cell RNA profiling of CTCs from a patient's blood samples taken before, during, and after relapse on platinum-based treatment provided additional data that intra-tumor heterogeneity increases in tandem with treatment resistance.
"These findings suggest that, in response to treatment, SCLC develops increasing transcriptional [intratumor heterogeneity] marked by concurrent, diverse resistant cell clusters," Byers and colleagues wrote in Nature Cancer. "Clinically, these findings underscore the importance of maximizing and maintaining the initial response in platinum-sensitive SCLC tumors, and highlight the intrinsic transcriptional fluidity underlying the profound treatment resistance of SCLC following initial therapy."
More than 60 percent of SCLC patients respond to frontline therapy. However, these patients usually rapidly relapse, after which, less than 20 percent of patients will respond to approved therapies. The median survival for these patients is only about one year.
Recent analyses have defined molecular subtypes of SCLC characterized by the expression of ASCL1, NEUROD1, POU2F3, NKX2-1, Notch pathway genes, MYC family genes, neuroendocrine genes, and others. However, researchers have previously observed that genes are transcribed differently across populations of cancer cells within a tumor, and they suspected that this heterogeneity may be why SCLC evolves quickly to develop resistance.
However, genomic analysis of tumor tissue samples usually occurs at diagnosis, which allows for sampling of only a very small portion of the tumor and doesn't reflect the molecular heterogeneity driving cancer cells. SCLC patients also rarely under surgical resection, which limits access to tumor samples for longitudinal analysis of the type Byers' group did in this study using CTCs.
To investigate the evolution of tumor heterogeneity in SCLC, Byers and colleagues generated CTC-derived xenograft (CDX) models that would mirror SCLC patients' tumor genetics and response to chemotherapy and performed baseline single-cell RNA sequencing analysis of both drug-responsive and drug-resistant models. Researchers also performed longitudinal RNA-sequencing on CDX models and directly on a patient's CTCs at different treatment time points.
To quantify heterogeneity, Byers' group calculated an intratumoral heterogeneity (ITH) score for each CDX, which factored in the average distance between the normalized expression profiles of each cell and all other cells in the sample. They reported that platinum-resistant CDX models had a higher ITH score than platinum-sensitive models. Further, increased ITH scores were associated with emerging treatment resistance as seen in CDx models and in patient blood samples collected over the course of treatment with cisplatin or DNA damage response inhibitors.
The team used single-cell RNA-seq data to establish molecular subtypes of the CDX models based on expression of neuroendocrine and non-neuroendocrine markers, ASCL1/NEUROD1/ POU2F3/YAP1 transcription factors, MYC family genes, and EMT score. They found that CDXs expressed predominantly neuroendocrine markers. High levels of ASCL1 were detected in both chemo-sensitive and chemo-resistant CDXs models. One chemo-resistant model had high NEUROD1 expression. No CDX models had large populations of cells which expressed POU2F3 or YAP1.
MYC was expressed in a moderate number of cells in two CDX models. MYCL was expressed in three models and MYCN was abundantly expressed in one model. Additionally, the team found that MYC activation in one model was associated with spontaneous leptomeningeal metastasis, which they noted was consistent with the role of MYC in driving central nervous system metastases. EMT has previously been associated with treatment resistance and metastasis. The team found high expression of epithelial genes like CDH1 and EPCAM in the majority of CDXs at the single-cell level. Like for the MYC genes, the EMT score was not uniform across cells within a single tumor.
Pathways like MYC, mammalian target of rapamycin, EMT, and WNT were known to be connected to chemotherapy resistance in SCLC. However, the fact that these pathways can coexist within a single tumor is a key new finding, the researchers noted.
Further, they confirmed that ITH is driven by transcriptional changes rather than genomic or gene copy-number changes. Additionally, unique resistance mechanisms occurred in response to targeted therapies like poly ADP-ribose polymerase (PARP) inhibitors or checkpoint kinase 1 (CHK1) inhibitors.
"These results suggest that not only are the mechanisms of cisplatin resistance distinct between tumors but, even within a single tumor, targeting multiple pathways (for example, MYC and EMT) may be necessary to overcome platinum resistance in SCLC," the authors wrote.
They conclude that the data indicates the importance of maximizing the efficacy of response in the first line, and diversifying combination treatment strategies before the emergence of tumor heterogeneity.