NGS: Driving a Revolution in Cancer Diagnostics and Treatment Management

What numbers and types of cases does your laboratory handle?

Massimo Barberis: The Molecular Pathology lab in the IEO runs >2,500 mutational assays/year and 1,800 FISH assays. Most tests are for non-small cell lung cancer (NSCLC), colorectal cancer (CRC) and malignant melanoma. Others include gastrointestinal stromal tumors, soft tissue sarcomas and ovarian cancers.

Ian Cree: We handle about 6,000 bloods/day and 70,000 histopathology cases/year. These include ~500 molecular tests for cancer annually, mainly lung cancer, CRC and melanoma. We are sometimes asked to test other tumor types – we did a Langerhans cell histiocytosis recently – and I think the range of tumors we assess will expand rapidly as targeted drugs become more widely used.

Is it advantageous to assay multiple biomarkers per patient sample?

MB: Yes. As more driver mutations are identified, it is increasingly necessary to assay multiple molecular markers on the same sample. For NSCLC patients, we must assess EGFR, ALK and ROS1 status at a minimum, but this is expanding to include tests for BRAF, HER2 and MET. For CRC patients, we routinely analyze KRAS, NRAS and BRAF, and in melanoma we evaluate BRAF, NRAS and C-KIT.

IC: I don’t believe the one diagnostic to one biomarker approach is sustainable. To guide lung cancer treatment, we need to know the EGFR / ALK status. We also test KRAS, because it avoids unnecessary ALK testing, which can needlessly delay treatment. In CRC, we test KRAS and NRAS because they guide use of anti-EGFR drugs; we also check BRAF status. For melanoma we assay BRAF, and quite often C-KIT also, again because particular treatments are relevant to those mutations.

How does NGS address challenges associated with the multiple biomarker approach?

MB: The main challenge is that the small tissue samples typical of today's biopsies are not compatible with running multiple single tests. NGS, however, allows unbiased interrogation of multiple cancer-related genes from a very small tissue sample.

IC: Samples are reducing in size. For example, endobronchial ultrasound biopsies are commonly used in lung cancer, but are tiny – if you're doing multiple single-gene tests, you quickly use up the sample. But the Oncomine™ Solid Tumor kit only requires 10 ng of DNA – a huge advantage. Panel-based PCR methods are a halfway house between NGS and PCR, but when you get beyond two to three genes assayed per patient, the only solution will be to use NGS. Another big plus of NGS is its coverage. The Oncomine™ Solid Tumor kits include everything you need to assay in colorectal and lung cancer. This allows us to consolidate tests, and get a lot of information at once. 

How easy is it to incorporate NGS into the laboratory workflow?

MB: We recently adopted the Oncomine™ system. There were training costs, but NGS was readily accepted by all staff and provided economies of scale compared with single-gene testing. And although the turnaround time increases compared with other techniques, the time and costs of single gene evaluation 
are reduced.

One issue is that reimbursement is based on gene assays requested, not on the panel used; we still use PCR-based techniques if physicians only need a single-gene test for a specific targeted therapy. However, we find NGS is cost-effective when three or more gene assays are requested. Given the amount of NGS data output, interpretation and communication of the results to oncologists and patients represent additional challenges of the NGS technique. We have developed a three-tiered system for gene variant report based on the clinical relevance of a single gene variant identified.

IC: Integrating new systems is never easy. You have to make a business case for the introduction of a new technology, which can be difficult. NGS can be perceived as slow and expensive – it takes three days to do, which doesn't leave much time to interpret results if you want a seven-day turnaround – but it's a good technique, possibly replacing several single marker assays. Also, reimbursement affects NGS uptake – molecular pathology procedures only recently began to get reimbursed in the UK. So we're introducing NGS very carefully, with extensive validation. It's important to ensure that NGS fits with laboratory workflow – the tissue preparations, the DNA extraction processes, and the libraries. More automation would help, and that's an advantage of Ion Torrent; but library preparation is still labor-intensive. It's evolving, however; template preparation and chip loading are automated, which improves data quality.

How challenging was validation?

MB: We adopted the Oncomine™ Solid Tumor DNA kit, which is a CE-IVD-marked kit, so it needs only performance verification, not full validation. Our verification procedure was based on 90 NSCLC and CRC tumor samples, previously tested with Sanger sequencing, Cobas EGFR kits and Sequenom panels. We confirmed all known gene alterations in the samples, and also identified additional driver mutations.

IC: Our validation method involved testing 155 clinical samples for variants at 87 hotspots in 22 genes relevant to lung or colorectal cancer (1). The panel was tested by seven different labs in their own clinical settings. All previously identified mutations were confirmed, and other mutations were revealed. The study was very realistic – samples included difficult 'real-life' cases. 

How do you see NGS being used in the future?

MB: In the future we will adopt more extensive NGS panels that not only include gene mutation but also gene translocation analysis in order to eliminate other time-consuming and expensive techniques, such as FISH. In the era of precision medicine, the implementation of NGS multiple biomarker tests offer a great opportunity to increase the number of therapeutic opportunities for every single patient, improving cancer patient’s care. However, we need NGS to be endorsed as a routine diagnostic assay by the regulatory agencies.

IC: Initially, NGS will be an add-on to PCR, and will be used where it's critical to identify driver genes in order to direct drug choice. Ultimately, NGS will replace the 'single biomarker per diagnostic test' approach; we've already switched to a panel-based testing approach with UK-NEQAS (UK National External Quality Assurance Scheme), which has essentially become a panel scheme. Oncologists already require mutation information for some therapies today, and NGS-sourced information will become increasingly important in guiding therapy in the future.

To learn more about Oncomine™ Solid Tumor Kits CE IVD, go to thermofisher.com/solidtumour