NEW YORK (GenomeWeb) – Researchers reported in Science today that they were able to culture circulating tumor cells for genomic analysis and drug susceptibility testing.
Harvard Medical School's Daniel Haber and his colleagues isolated CTCs from 36 patients and were able to establish cell lines for six of those samples. CTC cultures present an opportunity, the researchers said, to noninvasively monitor new mutations that arise in the tumors as well as the drug susceptibility of those tumors.
"We now can culture cells from the blood that represent those present in metastatic deposits, which allows testing for drug susceptibility as the tumor evolves and acquires new mutations," co-senior author Shyamala Maheswaran from the Massachusetts General Hospital Cancer Center said in a statement. "We need to improve culture techniques before this is ready for clinical use, and we are working on doing that right now."
To capture circulating tumor cells, which are thought to be cells that break off from either primary or metastatic tumors and enter the bloodstream at low levels, the researchers used a microfluidic chip dubbed the CTC-iChip that they presented in Science Translational Medicine last year.
The microfluidic platform combines magnetic labeling and inertial focusing to sniff out tumor cells circulating in the bloodstream.
In the current study, Haber and his colleagues used the device to collect CTCs from blood samples from 36 patients with ER-positive breast cancer.
After testing a number of culture conditions and finding that CTCs grew best in serum-free media supplemented with EGFR and FGF under hypoxic conditions, the investigators were able to establish six cell lines.
These cell lines, they noted, shared cytological features with matched primary CTCs, and five of the six lines retained ER positivity in culture.
Additionally, RNA sequencing and characterization of the six CTC cell lines along with 29 uncultured CTCs and 13 common established breast cancer cell lines showed that CTC cultures clustered with one another, though both the CTC and established cell lines had increased proliferation signatures.
Testing in an animal model further showed that three of the five CTC cell lines retained their tumorigenicity.
Using a hybrid capture-based next-generation sequencing platform, the researchers screened the CTC lines for mutations, drawing on a panel of 1,000 annotated cancer genes. This screen, they noted, is more comprehensive than the current SNaPShot panel currently in clinical use at Massachusetts General Hospital. That screen covers some 140 mutations in 25 genes, and was also performed on primary tumor samples or pre-treatment metastatic lesion biopsies from these six patients.
From the broader screen, the investigators uncovered more than a dozen mutations within the CTC lines, including a handful of mutations shared by different lines.
For instance, mutations in the estrogen receptor gene ESR1 were in three of the six CTC lines. Each of these patients had been treated with aromatase inhibitors, and the researchers noted that recent work found that ESR1 mutations occurred in between 18 percent and 54 percent of patients who had been treated with AIs.
Other common mutations included ones in KRAS and TP53.
Haber and his colleagues also examined to which drug treatments these cultured CTCs were sensitive.
In particular, they focused on CTC lines that remained ER-positive, but that also acquired an ESR1 mutation — a combination for which the best treatment is unknown.
For these cells, they found that the selective estrogen receptor modulators tamoxifen and raloxifene, and the selective ER degrader fulvestrant were not effective, either alone or in combination with everolimus, an inhibitor of the PI3K- mTOR pathway.
However, the HSP90 inhibitor STA9090 was cytotoxic in these cells, both alone and in combination with raloxifene and fulvestrant. ER, they noted, is a client protein for HSP90.
Further studies of HSP90 inhibitors and ER inhibitors, the researchers added, are needed to find the best treatment for patients with these mutations.
Additionally, Haber and his colleagues also found that a CTC cell line with activating mutations in PIK3CA and FGFR2 was highly sensitive to the PIK3CA inhibitor BYL719 and the FGFR2 inhibitor AZD4547, though it was moderately responsive to PD173074, a FGFR1 inhibitor.
The combined inhibition of PIK3CA and FGFR2 suggested to the researchers that both mutations might be oncogenic drivers in this tumor.
In a panel of established breast cancer cell lines, the investigators found that six of the seven PIK3CA-mutant lines responded to BYL719.
Two of these lines also had mutations of unknown significance in FGFR4 and FGFR2. The FGFR4 mutation showed cooperative cytotoxicity by BYL719 and AZD4547, while the FGFR2 mutation did not respond to FGFR inhibition.
Of the five PIK3CA-mutant lines in the panel without an FGFR mutation, one was modestly sensitive to AZD4547, while the other four were resistant.
This, the researchers said, underscores that both genotyping and function testing of drug susceptibility are needed to define therapeutically relevant driver mutations in cell lines and CTC cultures.
"This approach of culturing circulating cancer cells in the blood, analyzing them for new mutations that have developed during therapy, and testing the utility of drugs targeting those mutations could become the essence of individually adjusted cancer therapy in the future," Haber, the director of the MGH Cancer Center, said.