The UK’s 100,000 Genomes Project reached its goal of sequencing 100,000 entire patient genomes in December 2018. Capitalizing on this success, January 2019 saw the establishment of a new, nationally commissioned Genomic Medicine Service that harnesses seven Genomic Laboratory Hubs around the country. The goal is to standardize the criteria for whole genome sequencing and targeted panel tests, simplifying patient pathways and reducing social inequalities – but is this democratization of genetic testing the best option for all institutions? We spoke to Tanya Ahmad, Consultant Medical Oncologist in London, to discover her personal outlook on in-house testing versus centralization.
I’ve been practicing as a lung cancer consultant for over nine years, a role that began just as precision medicine was beginning to flourish. I currently work across two institutions, which puts me in the interesting position of seeing things from two different perspectives. At one of my institutions, all testing is currently conducted in-house. All samples from the diagnostic teams (lab) – from CT-guided lung biopsies to EBUS/bronchoscopies – are sent to the histology lab on our campus. Once we’ve established a diagnosis of lung cancer, the sample is subject to immunohistochemistry tests and, following that, the curls of the tissue block are sent to our affiliated molecular pathology lab. Although these three steps are all carried out within one institution – i.e., “in house” with all the benefits – it is a relatively large campus and we have observed that even small changes, for example, to the location of the pathology labs, can potentially impact the delivery speed of test results.
The pathology pathway at my other institution is more complex because three geographically distinct hospital sites have merged. Biopsies performed in one location must often be transported to another for basic H&E/IHC examination – and then molecular testing is further outsourced to another central lab. Experiencing the two contrasting strategies in parallel is like having my own controlled experiment in which I can directly compare in-house and somewhat centralized testing.
It’s a combination of historical pathways, institutional politics, and the availability and quality of local support services. At the institution that conducts in-house testing, we’ve practiced molecular pathology for many years and we were routinely carrying out next-generation sequencing (NGS) before most other trusts. We also have a strong academic and research focus, so testing was already part of our standard pathway. In contrast, at my other institution, two district general hospitals merged with a larger teaching hospital. This meant that many services and departments were drastically reconfigured over the course of two years. We transformed from a place that conducted all testing in-house (before the advent of routine molecular testing for cancers), first to sending samples to another city and then back to sending them to other centers in London. Trust mergers often involve the reconfiguration of many services and pathways, a process dictated by the consolidation of expertise and various other cost implications of centralizing services.
There is a lot of heterogeneity in terms of cancer services, care pathways, and patient outcomes across the UK, part of which might be related to rapid diagnostics and patient access to clinical trials and certain drugs. The attraction of centralization is that some peripheral hospitals without adequate resources or academic expertise can access NGS, providing more detailed information about a patient’s tumor than otherwise possible. Having access to a central hub that facilitates this can improve patient care – even in the face of increased turnaround times. However, in a center that already has the facilities and expertise to carry out NGS testing and other more sophisticated genomics, centralization is unlikely to add any benefit. If you already have a system that works in-house, centralization can introduce pitfalls, such as complex lab standard operational procedures, longer turnaround times, or increased risk of losing samples in transit.
From my point of view as a medical oncologist, one of the main benefits of in-house testing is the personalized service. I know the individual multidisciplinary team (MDT) colleagues dealing with samples because I meet and talk to them every week and, if an issue arises, I can contact them directly and informally to discuss the problem. Good MDTs improve integration between pathologists, radiologists, and oncologists, all of whom might previously have worked in separate silos. There’s an interest in each other’s specialties and, as a result, a deeper understanding of the nuances with individual cases. For example, I can call my colleagues about a particular patient and explain that, although we would usually request NGS testing for EGFR, this is a very young individual who is extremely unwell. Their demographic raises strong suspicion of an EGFR mutation so, instead of waiting two weeks for the NGS results, we’d like to request a rapid EGFR test with a 48-hour turnaround time so that we can begin treatment faster. With centralized testing, each sample is anonymous – just tumor material with a serial number, perhaps lacking detailed clinical information that might help with analysis. It’s harder to have the same nuanced conversation with a central laboratory, because you don’t know the person you’re speaking to and they may not be as invested in the case as your in-house pathology colleagues. In my opinion, this is another aspect of “personalized” care.
General interaction with colleagues is another benefit; delayed results can still occur in house – but, unlike with centralization, you can easily pick up the phone and speak to the pathologist. They can then check where the sample is in the pathway and call back within minutes to confirm when the results will be available. When I’ve sent samples externally, there is a phone number to call, but the person on the other end often can’t provide much of an update and I’m left waiting for a response. The process becomes much longer and more challenging, which is why in-house testing can be beneficial for institutions whose internal labs possess all the clinical information and technical facilities to fully trust their results. It’s also important to maintain laboratory skills and academic expertise within the institution; molecular pathology is a fairly niche subject, especially as more actionable, but rare molecular targets are identified. Therefore, centralization of molecular analysis may compromise the training and experience of newer generations of pathology colleagues who won’t be exposed to it without internal facilities.
Delays are potentially clinically harmful – especially for lung cancer patients, who often present in the advanced stages of disease and with comorbidities that affect their suitability for treatment. An inadequate test result or a lost sample could be the difference between starting treatment within days or within weeks – and, because patients need to be relatively fit for certain therapies, rapid deterioration can mean they miss the opportunity for treatment entirely.
Long turnaround times are also frustrating for the patient. After being diagnosed with lung cancer, it’s often possible for them to see the oncologist on the day they receive the news – but my scope for discussing any systemic treatment with them is limited until I have all the molecular test results. Even with optimally functioning in-house testing, this can take up to two weeks (during which the patient must sit at home, knowing their diagnosis, but anxiously awaiting next steps in management). That’s extremely difficult even without delays, so any extension in turnaround time as a result of centralized testing can have a significant negative impact on patient experience and outcome.
There have been occasions when samples were lost due to the convoluted nature of centralization. There are more opportunities for errors because there are several steps in the process that are out of my institution’s control. Every step is someone else’s responsibility – and when we tried to identify issues using pathway mapping for our pathology services, it merely revealed how complex each step was. One solution we derived from this mapping was to coordinate the specimen bags by color to highlight the most urgent samples for couriers. This minor change made a noticeable improvement to turnaround times, emphasizing how much variation there can be at each stage of the process.
Another concern is ownership. Who takes responsibility for the samples once they leave the trust? If the courier gets lost or the samples are misplaced, do you wait for someone to search for them while the patient deteriorates? Or do you return to the patient, apologize, and obtain another biopsy? The latter might ensure faster results, but biopsies can be unpleasant, taxing, and sometimes inconvenient experiences – plus, there’s always an element of risk. Why should the patient have to undergo another procedure because there’s a flaw in our pathway?
From an academic perspective, I completely understand how central testing benefits certain services. However, I don’t think those advantages apply equally to all hospital trusts – and I’m not convinced that implementing “blanket” centralization was necessarily the best move in the UK, especially when we are striving to reduce cancer treatment waiting times. We’ve already found that molecular test centralization can be time-consuming with respect to both admin and turnaround times. Whether this problem improves or worsens as throughput at central lab hubs increases remains to be seen. It’s clear that institutions without an effective testing service of their own can certainly benefit from test centralization – but, for others, supporting and maintaining effective and efficient in-house testing remains the best way to optimize patient care.
Newsletters
Receive the latest analytical scientist news, personalities, education, and career development – weekly to your inbox.
References
- K Snape et al, “The new genomic medicine service and implications for patients”, Clin Med, 19, 273 (2019). PMID: 31308102.
