NEW YORK (GenomeWeb) – A partnership between three medical and academic institutions in Heidelberg has made molecular tumor profiling available to cancer patients across Germany, impacting the treatment of many.
The initiative involves Heidelberg University Hospital's Center for Molecular Pathology, the National Center for Tumor Diseases' (NCT) Precision Oncology Program, and the German Cancer Research Institute' (DKFZ) Center for Personalized Oncology.
The collaborators are pursuing a two-pronged approach: targeted sequencing of formalin-fixed tissue with entity-specific gene panels, for routine diagnostics of primarily late-stage cancer patients, and comprehensive whole-exome or whole-genome sequencing for a subset of patients with advanced disease, a program called NCT MASTER (Molecularly Aided Stratification for Tumor Eradication).
According to Albrecht Stenzinger, head of the Center for Molecular Pathology, which is part of the Institute of Pathology Heidelberg, the two strategies complement each other and have enabled clinicians involved in routine diagnostics to join forces with those grounded in a clinical research environment. Stenzinger talked about the partnership last month at the Association for Molecular Pathology Global Congress in Berlin.
The Center for Molecular Pathology began analyzing tumors by targeted next-generation sequencing in early 2013 and has sequenced more than 10,000 FFPE samples since then from patients across Germany. At the time, Stenzinger said, no reimbursement for NGS-based tests was available from the country's public health insurance system, which factored into the center's decision to start with smaller, cancer-type-specific panels that would be cheaper to run than a single broad panel. Most samples so far have been run for diagnostic purposes, with about a fifth used to profile patient cohorts for clinical trials.
The panels, most of which were designed by the center, typically cover 40 to 60 genes or DNA regions and are available for all major cancer types. Late last summer, the center launched a program to analyze cancers of unknown primary origin, for which it uses the Oncomine Comprehensive Assay from Thermo Fisher Scientific, which covers about 150 genes and is similar to the panel used by the NCI-MATCH trial in the US. Samples are sequenced to 2,000x to 3,000x depth to allow for the analysis of small biopsies and those with low tumor cell content, in some cases as little as 10 to 15 percent, and the drop-out rate is about 3 percent. In addition, the center performs RNA sequencing to look for gene fusions. Turnaround time from receipt of an FFPE block to a clinical report is typically four to five days.
The vast majority of samples are currently profiled using sequencing platforms from Thermo Fisher Scientific — the center is equipped with two PGMs, an Ion Proton, and an Ion S5 — though it is about to bring an Illumina NextSeq online. "No matter which technology you're using, you need redundancy," Stenzinger said, so when a machine is out of service, turnaround time is not affected.
Panel sequencing is most often performed for cancer types with approved targeted therapies, such as lung cancer, melanoma, and colorectal cancer, and mostly in patients with advanced disease who are candidates for such treatments. Of the 70 or so patients with tumors of unknown primary origin who have analyzed so far, four were found to have druggable lesions, and two of them are currently receiving those drugs, he said.
Occasionally, the panels also pick up a tumor predisposition mutation, for example in a recent case of a patient with mesothelioma who had a history of colorectal and breast cancer.
Going forward, the center is considering adding gene panels that focus on certain biological themes — for example, DNA repair genes — that are involved across several tumor types.
Patients where panel sequencing does not find any druggable genetic targets might be eligible for whole-exome or whole-genome sequencing, plus RNA sequencing, under the MASTER program, which launched in Heidelberg in the summer of 2013. In early 2016, the program was opened to members of the German Cancer Consortium (DKTK), which includes 10 comprehensive cancer centers across the country, and it now has about 60 external collaborators.
According to Stefan Fröhling, an attending physician at NCT and one of the directors of the MASTER program, the idea was "to investigate whether a very broad mutational profiling approach is feasible in a clinical setting" and to evaluate its benefits. "This program is somewhere at the interface between clinical diagnostics and research. It's an effort of a research institution, such as the DKFZ, to use the technology that's available to make a difference in clinical care," he said.
The program, which is funded by the participating institutions, focuses on patients up to the age of 50 with advanced cancer of any type and no more curative options, as well as patients with very rare cancer types regardless of age. Testing is currently not reimbursed by health insurance.
The program uses fresh or cryopreserved tumor tissue whenever possible and requires a tumor cell content of at least 20 percent. Sequencing of tumor material, along with a matched normal sample, is performed by the DKFZ Genomics and Proteomics Core Facility — exome sequencing on the Illumina HiSeq 4000, to between 100x and 150x coverage, and whole-genome sequencing on the Illumina X Ten platform, to 60x coverage.
The assays, which have a turnaround time that ranges between four and 10 weeks, also involve the Center for Molecular Pathology, which has a board-certified pathologist evaluate each tumor sample prior to sequencing for tumor cellularity and type, and which validates each potentially targetable alteration with an orthogonal method such as Sanger sequencing.
Between 10 to 20 percent of samples cannot be sequenced successfully, often because of insufficient tissue quantity or quality, which is one reason the program prefers fresh tissue.
The assays' relatively long turnaround time means doctors need to plan well ahead. "We need to be very careful in selecting the timepoint when we initiate the analysis," Fröhling said. "In my personal experience, it works quite well if the clinician thinks about enrolling the patient while the patient is receiving, say, the second-to-last standard treatment."
Results for each patient are discussed by a multidisciplinary molecular tumor board that convenes weekly and involves oncologists, pathologists, medical geneticists, biologists, and bioinformaticians from several institutions, some participating via video conference. "This has been a fairly interesting experience for us, and there is a lot of added value doing this together with other academic centers," Fröhling said.
As of March, approximately 600 patients in the program had received results, among them about 60 to 70 from whole-genome sequencing, which was added late last year and is currently only available to patients in the Heidelberg area.
In approximately 75 percent of cases, the analysis resulted in some type of clinical recommendation, Fröhling said, classified according to the level of evidence associated with the genetic alteration found. For example, there could be a link to an approved targeted treatment, treatment response in a clinical trial, or case reports of successful treatments in the literature.
For 40 percent of patients who received a recommendation, doctors acted upon that advice, a rate that continues to increase, he said. In many cases, this means patients received the recommended treatment on an off-label basis. "Whenever possible, we recommend that patients go into a molecularly stratified clinical trial," Fröhling said. "In contrast to the situation in the US, these trial opportunities are rare in Germany, so the majority of patients gets treatment as part of compassionate or off-label use of targeted drugs."
The program organizers are currently building a clinical trial portfolio, he said, and have started writing trial protocols for molecularly stratified patient cohorts, with plans to launch the first trials soon.
In about 10 percent of cases, the test uncovers a germline alteration that results in genetic counseling of patients and their families, he added.
A formal analysis of what molecular alterations were detected by exome or genome sequencing that would have been missed by a more targeted assay will be conducted in the future. "It's a question we get all the time, also for economic reasons," Fröhling said.
Data from the MASTER program is shared across the local partner institutions, as well as within the German Cancer Consortium, and data sharing was initially difficult to implement, he said.
The program has so far been quite well received among patients, which Fröhling said can be a double-edged sword because it creates high expectations. Many patients entering the program are well informed about their disease and travel considerable distances to receive the test. "As you can imagine, for patients who are in a desperate situation, this creates a lot of hope," he said. "[We need to] make sure that patients and physicians have a realistic view of what we do and what we can achieve with this program."
Attitudes from oncologists have varied, with some being excited about the program and others being more skeptical, he added.
Going forward, both Stenzinger and Fröhling believe that molecular tumor profiling needs to look beyond next-gen sequencing. In the late 1990s, Stenzinger recalled, when RNA-based gene expression profiling emerged, "everyone thought this would be the Holy Grail and we would not need any other approach." A decade later, people realized that it did not provide the full story, and with NGS, it will likely be the same. "That's why I think we need to be quite technology-agnostic."
"If you think cancer precision medicine really through to the end, where one would leave histology at the end of the day, then you would really need a comprehensive profiling approach that involves many different biological layers, like RNA, DNA, epigenetics, [and] proteomics in terms of maybe the phosphoproteome. And that would probably more accurately reflect tumor biology than testing does now," he said.
His center already complements targeted sequencing with other tests, such as PD-L1 immunohistochemistry to predict response to immune checkpoint blockers, and is working on new sequencing-based assays to help predict checkpoint inhibitor response.
It is also working on integrating new technologies. For example, the Institute of Pathology recently acquired a MALDI mass spec imaging device for analyzing proteins in single FFPE slices, which he said has "great potential" for complementing DNA-based analyses.
Fröhling said the MASTER program is in the process of incorporating array-based DNA methylation profiling, targeted proteomic analyses, and radiogenomic approaches, "and we are constantly exploring ways in which additional tools and technologies can be used for stratification and individualization of therapy."