ePrivacy and GPDR Cookie Consent by Cookie Consent

Episode 2: Large panels in clinical routine? A reality check from a molecular pathologist - Prof. Hummel

Mélanie Moxhet

Dec. 1, 2020


Prof. Hummel, Charité Hospital (Berlin)

After talking about large panel testing in molecular oncology in the first episode of your ONCOmmunity podcast, let’s take a look behind the scenes of a molecular pathology laboratory.

Professor Michael Hummel is head of the molecular pathology department at the Charité Hospital in Berlin. His pathology laboratory performs comprehensive molecular genomic testing on cancer patients in everyday clinical practice. In this episode, Prof. Hummel will share the experience acquired in using NGS in pathology over the last six years and his insights on the future developments in the field of molecular genomic tumour analysis.

In the area of approved biomarkers and therapies, it is very clearly defined what we have to depict in our reports. It becomes more difficult with broad genomic testing that goes beyond the approved target structures. The availability of according therapeutic agents that match the respective molecular changes of the patient is often a major challenge”, he mentions about the clinical utility of large panels.


0:06 Welcome to “ONCOmmunity – the molecular oncology podcast” by OncoDNA. After talking about large panel testing in molecular oncology in a previous podcast, today we are going to take a look behind the scenes of a molecular pathology laboratory.

Professor Michael Hummel, Head of Molecular Pathology at the Charité Hospital in Berlin is here with us today. Welcome.

Professor Hummel

00:36 Welcome everyone. I’m happy to have the opportunity to share the experience we have acquired in using NGS in pathology over the last six years. And I’m looking forward to discussing this with you today.


00:51 In your pathology laboratory in Berlin, you perform comprehensive genome testing on cancer patients in everyday clinical practice. Could you please tell us which panels you use and how such tests can be carried out?

Professor Hummel

01:07 Yes, as mentioned, we already have several years of experience with gene panels. Of course, we started at a relatively small scale in 2014 with very small panels, and also the number of cases was significantly different compared to today.

01:24 We now perform a wide range of different gene panels for various diagnostic questions – for myeloid diseases, for lymphoma diseases, for various tumours, and especially for lung cancer. We are one of the centres within the national network for lung diseases, nNGM, and we offer a variety of solutions up to our customised large gene panels with three megabases and more than 600 genes. This helps us to find the right answers for various diagnostic questions, to offer various test applications for the different diagnostic requests and we are therefore positioned quite broadly.


02:11 Do you get enough DNA from tissue biopsies of solid tumours?

Professor Hummel

02:18 Well, that can definitely be a challenge, because obviously in the interest of the patient, the tissue samples should be as small as possible. We are talking about fine needle biopsies which contain a few cells only from which the DNA can be extracted. But we are quite successful. The laboratory staff has acquired a lot of experience over many years, so that in the vast majority of cases we can obtain sufficient DNA – nor should we forget the world of RNA – in order to ultimately make a reliable statement about the mutations or the transcriptome of the relevant patient.


03:03 So, how fast and how feasible are the sequencing workflows?

Professor Hummel

03:10 Good question. The workflows that we use are based on Thermo Fischer technology and are partly automated, meaning, we set up robotics-based processes for the library preparation and some other steps which clearly facilitates the work of the technical staff and also allows us to handle the high sample throughput in the first place. Therefore, the actual workflow with the Thermo Fischer platforms is quite well described.

We can expect a significantly higher level of automation in molecular diagnostics in the future.

03:50 How fast are we? So, the system runs for about one day. The sample preparations take another half a day. We need about a day and a half from DNA to getting the results, the raw data, and then of course we have to interpret the data that we have collected by drawing up a report in the end. Here, the most important results are summarised for the enquiring physicians, to then support their clinical decisions. Those are our processes.

04:20 I would say that the majority of cases are processed within five days.


04:34 What is the actual outlay for you in terms of time and materials as a laboratory? Is the whole thing cost-efficient?

Professor Hummel

4:43 Those are good questions. The question of cost-efficiency: You have to consider that there are various accounting systems in Germany. For inpatients in a hospital like the Charité in Berlin there is actually no additional reimbursement for molecular diagnostics from the health insurers, which means that everything is billed via the treatment costs. These DRG codes are relatively old and usually do not include molecular diagnostics. The hospital has to do a mixed cost calculation to prevent shortfalls.

05:15 For external submissions from the outpatient sector, we have the so-called EBM catalogue, which enables us to bill according to the corresponding services. The situation is of course more favourable here, as we are able to cover the costs of the patient cases that we analyse to a reasonable extent.

05:34 And then we also have privately insured patients who order a certain analysis. A cost-effective and cost-covering analysis is always possible in such cases.

05:46 Last but not least, something that is becoming increasingly important is patients who are examined under a so-called special contract or selective contract with health insurers. Bills are settled directly with the health insurer and can, and this is really important, also cover inpatients.


06:05 You are talking about the patients. For which cancer patients do you use these comprehensive molecular genomic tests as a standard diagnostic tool?

Professor Hummel

6:18 When talking about the standard diagnostics, certainly we must pay attention to the approval conditions. There are biomarkers, molecular biomarkers, that are associated with an approved, licensed therapy, and, in practical terms, that’s the basic information we have to provide. These examinations are defined by the therapy approval and the licence covers a few tumour entities, such as lung, colon, skin cancer and some more.

06:46 There is currently only one biomarker that can be used across virtually all tumour entities. These are the NTRK fusions. The official approval states no correlation with a specific tumour entity, so the situation can be summarised as follows: So far there are relatively few tumour entities treated based on NTRK, but the focus lies on the main tumour entities.

07:10 We will soon be adding more tumour entities like prostate cancer, presumably breast cancer, where the analysis can also be carried out on a molecular basis.


07:28 The oncologists send their requests for laboratory tests, genomic tests, to you. Do you give them any support? Do you provide training for oncologists?

Professor Hummel

7:42 Yes, surely, we provide assistance, especially for those with less experience, who also send us their test requests. We are back to the subject of enquiry. If we look at an enquiry and think: Hmm, maybe we could address this or that question better by running a different test, or maybe the request is a bit vague, then we are already involved in this first step, i.e. when the enquiry is defined, and we provide practical advice.

08:13 The whole topic becomes particularly important with regards to the results: How do we deal with that? As pathologists, we are of course not the treating physicians, but the diagnosing is institution, and for large gene panels, i.e. those which go beyond approved biomarkers, we and many other institutions organise what are known as molecular tumour conferences or molecular tumour boards. Here, the results of our genomic tests are discussed interdisciplinarily and in the context of clinical information, thus, selecting the best therapeutic option and recommendation for the patient.


08:58 How relevant are such comprehensive molecular genomic tumour analyses in today’s clinical routine?

Professor Hummel

09:08 Here too, as mentioned for the accounting regulations, approved biomarkers have to be considered. Of course, those results from molecular diagnostics are crucial in determining further therapeutic decisions. That’s why there are these approved biomarkers and therapeutic procedures. This group is clear, as the decisions are completely predefined.

09:38 It is more difficult, or let’s say less precise, when using broad genomic testing that goes beyond the approved biomarkers. In this case, the benefit for the patient often depends on the availability of appropriate drugs that are suitable for the particular molecular changes of the relevant patient. They are not always available.

10:06 In addition, there are of course advanced clinical trials or clinical studies that are actively recruiting, and the latter in particular is the decisive, major part for me, where our patients with large panel testing have a chance to be enrolled into such clinical trials, where new forms of therapy are being established and developed for the benefit of the patient.

10:29 And the identification of such patients for clinical trials is, I believe, a very, very important task, especially for molecular pathology departments at university hospitals, in order to identify the patients who qualify for inclusion in a clinical trial.


10:50 “Clinical trials” is one area. But how do you see the future and further developments in the field of molecular genomic tumour analysis?

Professor Hummel

11:03 Well, I have no crystal ball, but looking at the developments over the last six years - that’s how long we have been practising molecular diagnostics with Next Generation Sequencing - and what has been achieved during this time, it is not rocket science to predict that developments will basically continue.

11:22 Sooner or later we will also have to be able to offer very broad molecular procedures for a larger number of patients, and we will be working more and more in a pan-cancer fashion.

11:42 In addition, we will have to deal with further approvals, which will of course expand the portfolio as such, and the obvious question here is what will happen with the liquid biopsies in the future?

11:59 It started some time ago with one specific biomarker, the T790M EGF receptor mutation in lung cancer. The molecular pathology departments perform this test, but liquid biopsies are not requested very often.

12:16 We will soon be able to detect PIK3CA mutations in breast cancer using liquid biopsy. It is developing much slower than originally expected, and we will see what the future holds. In the end, liquid biopsies may not be key to find all crucial therapeutic biomarkers but may rather become an important application for monitoring the recurrence of cancer diseases.


12:42 Thank you, Professor Hummel, for your insights and explanations.

In the next episode of “ONCOmmunity – the molecular oncology podcast” by OncoDNA, we will discuss this topic from the perspective of molecular oncologists