Colorectal cancer (CRC) relies on a population of cancer stem cells that express the stem cell marker gene LGR5 on the cell surface. Through functional screenings in patient-derived organoids, we generated MCLA-158, a biospecific antibody that inhibit the growth of LGR5+ cancer stem cells. We showed that MCLA-158 targets cancer stem cells but spare healthy stem cells. This bi-specific antibody showed anti-tumor activity in multiple in vitro and in vivo preclinical models. It is currently being tested in phase 1 clinical trials for different types of solid tumors with very significant responses.
In CRC patients with overt metastases, chemotherapy initially halts tumor growth, but almost inevitably, the disease progresses after some cycles of treatment. Previous studies have shown that patient-derived organoids predict responses to chemotherapy. Whereas CRC organoids expand from highly proliferative cancer stem cells, we discovered that lack of optimal growth conditions specifies a latent cancer stem cell population. These cells expressed the gene Mex3a.
Using lineage-tracing analysis combined with single-cell profiling, we showed that drug-tolerant persister Mex3a+ cells downregulate the cancer stem cell program immediately after chemotherapy and adopt a transient regenerative state reminiscent of that of fetal intestinal progenitors. In contrast, Mex3a-deficient tumor cells differentiate towards a goblet cell-like phenotype and are unable to resist chemotherapy. Overall, our findings reveal how pre-existing cell heterogeneity imposed by adaption to different stem cell niches shape chemotherapy responses in CRCs and may help develop strategies to improve the outcome of current treatments by targeting the Mex3a+ drug-persister cell population.
In addition, we studied the identity and features of the residual tumor cells responsible for CRC relapse. By analyzing the transcriptomes of individual tumor cells in multiple primary CRC patient samples, we discovered that genes associated with an elevated risk of metastatic relapse are expressed by a defined subset of tumor cells that we named High Relapse Cells (HRCs).
To investigate HRCs, we established a human-like CRC mouse model that undergoes metastatic relapse after the surgical resection of the primary tumor. We also developed a methodology to isolate residual disseminated tumor cells before metastases are detectable. Using these advances, we demonstrated that residual tumor cells occult in mouse livers after primary CRC surgery resembled the HRC population present in patients. Over time, HRCs gave rise to multiple cell types and generated macrometastases that can kill the host. Genetic ablation of HRCs prior to extirpation of the primary CRC prevented metastatic recurrence and mice remained disease-free after surgery.