Memorandum by the Royal College of Pathologists
1. The Royal College of Pathologists (RCPath)
is a registered charity, bound by its Royal Charter to promote
and maintain the standard of service in pathology (i. e. all of
laboratory medicine, including genetics) for the benefit of patients
and the general public. It is not a trades union and it does not
discuss the terms and conditions of employment of its members.
2. Histopathologists are concerned mainly
with the diagnosis and evaluation of disease on the basis of the
examination of diseased tissue samples, especially but not exclusively
relating to cancer. We are well aware of the potential future
impact of the detailed analysis of germline DNA on the diagnosis
of inherited disease. We are also aware of the potential for the
analysis of inherited genetic variations to provide detailed information
on a person's susceptibility to many common illnesses, as varied
as diabetes, vascular disease and cancer. These are extremely
important areas. However, our emphasis in this Submission of Evidence
is on the optimal identification and classification of established
disease processes, to facilitate better treatment and the provision
of better prognostic information.
3. Increasingly this involves aspects of
genomic medicine other than the analysis of the inherited DNA
sequence ("germline" DNA). These include testing for
acquired DNA mutations in tumours; analysis of patterns of gene
expression in diseased tissue ("transcriptomics"); and
analysis of patterns of protein production in diseased tissue
4. A separate response to this Call for
Evidence is also being made by the Joint Committee on Medical
Genetics. We support and endorse the views expressed in that submission,
which has a greater emphasis on "conventional" inherited
diseases than does this one.
5. In the diagnosis of infection, the detection
of the genome of a microbial pathogen can often produce faster,
more sensitive and more specific tests than conventional microbiological
approaches. This important area is not considered further in this
6. Diagnostic approaches bases on the analysis
of DNA (or RNA) are already an integral part of the management
of patients with some relatively rare malignancies, notably haematological
malignancies (leukaemia and lymphoma) and some sarcomas. For some
new anti-cancer therapies it is already a requirement that information
about genomic changes in the tumour must be available before the
drug can be given. Data from one haematopathology laboratory (Prof.
F. Cotter, Bart's & The London) indicate close to a threefold
increase in the use of these techniques in the last two years.
It is inevitable genomic analysis will soon be a standard requirement
for many much more common tumours. Demand is set to increase exponentially.
7. As explained above, this submission is
limited to the context of diagnostic services. In that context,
in the future these techniques:
will permit more accurate and biologically
meaningful classification of disease, including the sub-classification
of common conditions that were previously regarded as a single
condition (for example, colon cancer), thereby providing explanations
of previously unpredictable variations in response to treatment
and in prognosis;
will thereby guide more appropriate
treatment and avoid ineffective treatment, and will identify some
patients who do not need treatment;
will be an absolute requirement before
the administration of many new treatments, especially new anti-cancer
will increasingly allow the prior
prediction of severe adverse reactions to specific drugs, avoiding
the current "trial and error" approach to idiosyncratic
will permit more accurate and earlier
diagnosis using smaller and less invasive sampling methods (eg
detection of mutated DNA in urine samples to detect bladder cancer,
rather than cystoscopy and bladder biopsy); and
will facilitate new population screening
programmes to detect early disease.
8. However, all these developments will
require investment if the benefits are to be achieved, and consequently
a cost-benefit analysis is necessary before they are funded from
the public purse. Failure to do this properly will result either
in the loss of potential health benefits to the population, or
to resources being wasted.
9. The need for cost-benefit analyses apply
to large healthcare organisations such as the NHS. But it also
applies to individual members of the public, who may be persuaded
to pay for directly-marketed "genomic" investigations
without fully understanding the risks and benefits. We are concerned
about the risks to the general public posed by the unregulated
direct commercial marketing of genomic investigations. "Market
forces" can only work if the purchaser is in a position to
evaluate adequately the product which is being purchased, which
we do not believe is the case here. This area has recently been
explored by the Human Genetics Commission in their report "More
Genes Direct"(1) and, with our assistance, by the charity
"Sense About Science"(2). However, our current submission
is limited to the provision of conventional (principally NHS)
10. To assess the risks and benefits of
a new diagnostic investigation is difficult. When a potential
new avenue for disease investigation emerges from basic research,
work is typically done to measure parameters such as diagnostic
sensitivity and specificity (sometimes called "clinical validity").
International standards for the reporting of test accuracy have
been widely accepted(3).
11. Unfortunately there is rarely any evaluation
of whether a clinically valid new test actually results in any
patient benefit (ie "clinical utility"). This is an
important and complex question. There can be many reasons why
a test with excellent clinical validity may ultimately produce
no patient benefit, while conversely a relatively insensitive
or poorly-specific test might help patient care considerably if
it is relevant to a difficult clinical situation where no better
test exists. An approach to such evaluations in the context of
cancer diagnosis has been published(4) and is supported by the
National Cancer Institute of the US National Institutes of Health,
but has not been widely used.
12. This lack of evidence on patient outcomes
makes undertaking risk benefit analysis very difficult, resulting
in reliance on "expert opinion" or a failure to make
any decision at all.
13. We believe that this lack of research
on clinical utility is driven by several factors, including the
organisational difficulty of conducting this type of research;
its relative lack of "prestige" amongst the scientific
community; and a traditional reluctance of the major grant-giving
bodies to fund "mundane" research into such practical
matters, research which is not directed towards improving our
understanding of the underlying disease processes. We also suspect
that the distinction between clinical validity and clinical utility,
explained above, is not widely understood.
14. The current ethical and regulatory framework
for such patient-based research has largely evolved with therapeutic
research on living patients in mind. The result is a system which
is perhaps excessively cumbersome and bureaucratic for research
which can involve little or no risk for patients, as the research
materials are typically anonymnised samples of surplus biological
material linked to relevant clinical data.
15. Industry contributes very little to
this type of research. We suspect this is partly because there
is no regulatory requirement to undertake it, so companies find
aggressive marketing to be more commercially effective than research;
but it is also true that there are much lower profit margins in
diagnostics than in pharmaceuticals, so the funds available for
such research are quite limited.
16. Consequently we recommend that incentives
and funding should be established to encourage high quality research
to evaluate whether and to what extent the many new genomic-based
diagnostic approaches actually improve patient outcome.
17. We are concerned that the approach to
evaluation and implementation of new diagnostic tests is fragmented,
incomplete and ineffective. The introduction of new drugs for
NHS use is overseen by a sophisticated regulatory system; there
is no such framework for new diagnostic tests.
18. The MHRA is concerned to assess the
safety of new "in vitro diagnostic devices", but this
is largely limited to ensuring compliance with EU regulations.
It does not address clinical validity or utility as defined above.
It has been suggested that this is analogous to insisting that
a drug company ensures that a new drug is chemically pure and
does not poison patients, without any requirement to demonstrate
that it can actually produce any benefit for patients.
19. NICE and NHS QIS (Quality Improvement
Scotland) have a responsibility to evaluate new diagnostic modalities
in addition to their better-known role in relation to therapeutics,
but in practice they have evaluated only a very small number of
innovations in laboratory diagnostics.
20. The Department of Health (England) has
identified several other organisations that have a responsibility
in this area, including the National Horizon Scanning Centre,
the Health Technology Assessment programme and the Centre for
Evidence-Based Purchasing. However, we suggest that the existence
of many bodies with overlapping responsibilities demonstrates
that the work is not well coordinated. Furthermore, all these
agencies are selective in the topics they will address, and many
new innovations are not covered by the remit of any of them. In
practice, evaluations from these bodies have had very little impact
on laboratory practice in the UK, and we see no evidence to suggest
that this situation will change in the near future.
21. We therefore recommend the establishment
of a new UK-wide body, or reorganisation of one of the existing
bodies, to provide a single comprehensive function for horizon-scanning
and evaluation of new laboratory investigations. Such body should
also advise on areas where research is needed to resolve the uncertainties
explained above. It should work towards the development of an
authoritative guide to the appropriate use of all laboratory investigations,
in a way analogous to the service currently provided for drugs
by the "British National Formulary".
22. This area was the subject of a recent
conference of diagnostic specialists(5). The list of recommendations
of that conference is attached as Appendix A. Very similar conclusions
were independently recently reached by the Science Council in
an almost simultaneous report(6). We are not yet aware of any
formal Government response to either of these reports.
23. In the absence of a national system
of evaluation, discussed above, the implementation of new diagnostic
techniques is left to local evaluation and decision-making.
24. In the NHS, the provision of laboratory
diagnostics is usually based on a fixed annual budget, which may
(or may not) be sensitive to major changes in demand, but which
typically does not include efficient mechanisms for funding new
25. Many of the new developments in genomics
are dependent on investment in costly machines and extensive training.
The rapid rate of progress in technology means that almost as
soon as new equipment is acquired it is necessary to start planning
for its next replacement. This is a difficult concept for many
NHS managers to accept.
26. To deal with the data handling, storage
and security, it will also be necessary for NHS IT schemes to
invest in sufficient server capacity (as well as staff), PCs,
equipment interfaces, and sufficiently sophisticated Laboratory
Information Management Systems (LIMS).
27. Hence investment is an obvious requirement
if the potential of genomic medicine is to be translated into
improved diagnostic services for patients. But the availability
of the necessary investment and facilities is being curtailed
by the "silo" structure of funding in the NHS, and by
a historical under-resourcing of pathology services. These problems
are clearly identified and analysed in Lord Carter's recent report
on the provision of pathology services in England (7), but solutions
to these problems have not yet been implemented.
28. Current Department of Health policy
is to devolve decision-making to a local level as far as is possible.
The arguments for this policy have been well presented and are
understood. But in the present context of genome-based diagnostics,
to fund any new development at local level demands that someone
locally has the time and energy to draw up a well-argued business
plan. This is difficult and time-consuming, for reasons given
29. To replicate such a time-consuming bidding
and evaluation process in every hospital is an extremely inefficient
use of staff time and will generate inequalities in provision
across the UK (ie "Postcode diagnostics").
30. However, the more likely outcome is
that local proposals to implement new developments are not made
at all. This is because of the difficulty of the process and the
lack of evidence of patient benefit explained above; because NHS
staff rarely have the time to undertake this sort of work; and
because the size of the investment needed is widely perceived
as being unlikely to be available within the "silo"
budgets of an individual Trust.
31. The logical route to implementation
is for local hospital units to co-operate and form collaborative
networks, much as was recommended by Lord Carter (7).
32. Unfortunately, the Department of Health
(England)'s current policy of "contestability" and the
establishment of independent Foundation Trusts have encouraged
a climate of competition between NHS Trusts, making such collaboration
difficult or impossible.
33. We therefore suggest that the efficient
implementation of new genomic-based diagnostic approaches will
demand central direction and funding for the establishment of
specialist laboratories to provide these services, on a model
similar to that of the current National Commissioning Group (formerly
NSCAG). This nationwide approach has already been adopted in the
comparatively small arena of "conventional" inherited
disease, as is explained in the separate Submission of Evidence
by the Joint Committee on Medical Genetics; but this system is
far too small to address the issues raised by the involvement
of genome-based diagnostics in much commoner diseases.
34. We suggest that if this model is adopted
there will remain a need for local expertise in the requesting
and interpretation of new investigations, but that the technical
provision of the tests will be undertaken, as Lord Carter recommended,
on a network basis.
35. An alternative would be to leave provision
to the private sector. But the need for individual NHS units to
evaluate what services should be purchased would remain; so most
of the problems outlined above would not be resolved. In the absence
of central guidance, commercial provision would raise the possibility
of inappropriate local decisions being made on the basis of commercial
pressure rather than on the basis of scientific evidence. The
provision of individual investigations by a commercial organisation
would presumably be limited to tests that are commercially attractive,
thereby potentially disadvantaging patients with rare diseases.
36. In order to take advantage of future
developments in Genomic and Molecular medicine it will be necessary
to ensure that current medical practitioners as well as the next
generation are educated in this subject. The RCPath is exploring
how genomic and molecular pathology might be brought into the
curricula for trainee pathologists and Clinical Scientists. The
intention is to have core level of understanding for all pathologists
and more advanced training and curricula for specialists; but
providing this training on such a large scale is proving difficult.
37. There is a need to train staff who specialise
in medical bioinformatics, to support those staff carrying out
testing and those staff receiving results.
38. There is also a need to improve training
in molecular genetics in the undergraduate medical curriculum.
That area is the responsibility of the General Medical Council.
39. In the UK a mere five individuals are
qualified in the application of genomics to "acquired"
disease, in distinction to diseases that are conventionally regarded
as "inherited". Only one of these is in NHS employment
as a Genetic Pathologist (in Cardiff), the rest being employed
in senior academic posts or in one case as a Clinical Geneticist.
Of these five individuals, four are cancer geneticists.
40. There are nominally just two Genetic
Pathology Specialist Registrar posts in the UK. In the absence
of any consultant posts to absorb trainees the RCPath has recently
had to conclude that the specialty should be suspended in terms
of training. This is surely a bizarre development, driven by the
reality of short-term economics rather than any logical assessment
of future need.
1. More Genes Direct. A report on developments
in the availability, marketing and regulation of genetic tests
supplied directly to the public. Human Genetics Commission: 2007,
2. Making sense of testing: A guide to why scans
and health tests for well people aren't always a good idea. Sense
About Science: 2008. At: http://www.rcpath.org/resources/pdf/Making_Sense_of_Testing08.pdf
3. Bossuyt PM, Reitsma JB, Bruns DE, et al. Towards
complete and accurate reporting of studies of diagnostic accuracy:
the STARD initiative. Standards for Reporting of Diagnostic Accuracy.
Clin Chem 2003; 49 (1): 1.
4. McShane LM, Altman DG, Sauerbrei W, Taube
SE, Gion M, Clark GM. Reporting recommendations for tumor marker
prognostic studies. J Clin Oncol 2005; 23 (36): 9067.
5. The evaluation of diagnostic laboratory tests
and complex biomarkers. Summary of a Diagnostic Summit, 14-15
January 2008. Royal College of Pathologists and the PHG Foundation:
2008. At: http://www.rcpath.org/resources/pdf/DiagnosticSummitFinalreport08.pdf
6. Integration and Implementation of Diagnostic
Technologies in Healthcare. A report from the Science Council's
Science in Health Group. The Science Council: 2008. At: http://www.sciencecouncil.org/documents/diagnostics_execsummary.pdf
7. Report of the review of NHS pathology services
in England (chaired by Lord Carter of Coles). Commissioned by
the Department of Health, England: 2006. At: http://www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPolicyAndGuidance/DH_4137606