Skip navigation

Commissioning New Genetic Services - Inappropriate demand or unmet need?

Wed May 8th, 1996 at the Wellcome Trust, 183 Euston Road, London


Introduction: Dr Nicholas Hicks
Familial Breast Cancer: Dr Jane O'Grady
Familial Colorectal Cancer: Dr Shirley Hodgson
Venous Thrombosis: Dr David Keeling
Purchasing Genetic Services: Dr Mike Gill
Discussion
The Way Forward: Dr Muir Gray

INTRODUCTION: DR NICHOLAS HICKS

The chairman introduced the meeting by saying that today's main objective was to address the problem of whether the new genetic services were already posing an inappropriate demand or whether there was a genuine unmet need developing. We have a great opportunity for turning the rapid scientific advances accruing from the Human Genome Project into real benefits for patients but we are currently unprepared for this. "The handbook of man is in a language and syntax not clearly intelligible to public health physicians".

The new understanding of the genetic basis of disease should lead to better diagnostic and predictive tests, better lifestyle advice and new and better targeted treatments. To take advantage of these will require much education and training for both the public and for professional staff. Public health physicians will have a responsibility to understand all these issues and to advise their health authorities on the best use of resources in this area.

Today's programme was designed to start this process of informed debate by concentrating on three areas where progress in genetic research is already having an impact on service demands.

FAMILIAL BREAST CANCER DR JANE O'GRADY

Breast cancer is the commonest malignancy in women, accounting for about 27,000 new cases and 15,000 deaths per year. Only a small proportion of these cases are thought to be due to, or associated with, the inheritance of specific genetic mutations. The mutated genes involved can be roughly divided into high risk and low risk (see figs) and it is estimated that the high risk genes account for about 5% of all breast cancer cases.

Several of the genes most directly involved in breast cancer have been cloned within the last two years eg BRCA1, BRCA2 and the Ataxia-Telangiectasia (ATM) gene and there are still considerable technical problems associated with looking for mutations in these genes. Attention at present is mainly focussed on the high risk genes, BRCA1 and 2, although the importance of the lower risk ATM gene may soon be recognised.

It is possible that tumours associated with high risk genes may have a different natural history compared to sporadic breast tumours. The former tend to occur in younger women and so a higher proportion of tumours in women under 50 have an inherited component. So far there are no definitive studies on the natural history of breast cancer in women who carry mutations in known breast cancer genes and therefore the most appropriate management strategies remain uncertain.

The identification of individuals and families as inheriting an increased risk of breast cancer poses many ethical problems. The psychological, social and possibly financial and legal implications of such identifications should be considered when developing policies for genetic testing. The different possible consequences of the reactive and proactive provision of information should also be borne in mind. There is some evidence to suggest that interest in gene tests declines when subjects are required to give 'fully informed consent'. The growth of commercial organisations with the technical capacity to undertake mutation testing is also a cause for concern. There is a clear potential to do harm in this area.

The main points raised in discussion were:-

When challenged to provide advice to local health authorities on the management of familial breast cancer the meeting suggested that it was important to provide joint clinics with specialist surgeon, oncologist, nurse, etc and to provide guidelines on referral for GPs. It was also important to improve the dialogue between primary and secondary care.


FAMILIAL COLORECTAL CANCER: DR SHIRLEY HODGSON

Colorectal cancer is the second most lethal cancer in men (after lung) and the third in women (after breast and lung). There are about the same number of new colorectal cancers as there are breast cancers in the UK each year (27,000). 5 year survival of symptomatic colorectal cancer (mainly Duke's C and D) is < 40% whereas survival of screen detected tumours (mainly Duke's A) is > 80%. Thus there is stronger evidence for the effectiveness of screening and early treatment in colorectal cancer than there is for breast cancer. Colonoscopy in skilled hands detects most adenomatous polyps and early cancers, at a cost of about £250. The average cost of treatment of established colorectal cancer is about £8,500.

The lifetime risk of dying of colorectal cancer for the general population is usually quotedas about 1 in 50, (although figures of 1in 23 for males and 1 in 33 for females have been quoted for Scotland). The risk rises to about 1 in 17 for individuals who have one first degree relative with colorectal cancer, 1 in 10 if diagnosed by age 45, and 1 in 6 with two first degree relatives affected. Like breast cancer the risk rises with the number of relatives and a younger age of onset of disease. About 5% of cases of colorectal cancer are thought to have a strong inherited predisposition; less than 1% are FAP (familial adenomatous polyposis) and most of the rest are HNPCC (heriditary non-polyposis colon cancer). Both conditions are conferred by rare autosomal dominantly inherited genes. HNPCC tumours are more likely to be proximal, on the R side of the colon and multiple and the average age of onset is 45.

The 'Amsterdam Criteria' are generally accepted as providing a basis for uniformity in the diagnosis of HNPCC. They are:-

Four DNA repair genes have been implicated in HNPCC. Most mutations are found in the hMSH2 or the hMLH1 genes and these tend to cause protein truncation which makes them easier to detect. It is also possible that the detection of the rer +ve phenotype (replication error prone) will provide a screening test for individuals most likely to carry these mutations. It should be noted that unaffected individuals cannot currently be tested for APC or HNPCC mutations unless a mutation has been detected in one of their affected relatives.

There has been much debate about the relative merits of different screening protocols and their effectiveness and cost-benefit profiles.

The major questions are:-


There is a move towards a consensus advising that screening should only be offered routinely to individuals with a risk of 1 in 10 or greater. It has been estimated that this could save about 1,600 lives /year in the UK. However to make rational, evidence-based decisions we need to have better studies on the genetic epidemiology and natural history of the 'inherited' versus the 'sporadic' disease and on the effectiveness of screening and treatment.

There is little doubt that a network of coordinated family cancer clinics with nationally agreed protocols would improve early detection and hence treatment and outcomes for colorectal cancer. It has been claimed that such a network would require, per district , the equivalent of 0.25 WTE nurse counsellor: 0.1 WTE consultant geneticist; 0.25 WTE molecular geneticist and secretarial support. (A working group has been set up under the chairmanship of Professor Peter Harper to report and make recommendations to the CMO on these issues later this year.)

DISCUSSION

The chairman asked whether the problems we face in screening and treating colorectal cancer are significantly different from those in breast cancer. It is interesting that much of the drive for improved services in breast cancer comes from the public and media, whereas there is relatively little interest in colorectal cancer. Here the drive for progress is almost entirely from within the profession.

The pre-malignant polyp, which is potentially detectable before colorectal cancer develops does provide an opportunity for early detection and treatment that is not apparently available in breast cancer. It was noted in passing, that the predictive value of rectal bleeding and occult blood tests is poor. It was agreed that there are many uncertainties shared by breast and colorectal cancers which need to be resolved before rational decisions can be reached about the level of services that we should recommend to purchasers and providers. A simple message applicable to both conditions is that early onset cancers are likely to have a different aetiology and require different approaches. But since the numbers of such cases are small the provision of new services for them should not pose a large problem.

VENOUS THROMBOSIS: DR DAVID KEELING

DVT is responsible for about 3 deaths per 100,000 population per year. There are 2 homeostatic systems to counterbalance coagulation, the anti-thrombin serine proteases (AT) and the activated protein C system (APC) which cleaves coagulation co-factors.

There are now 4 recognised inherited defects affecting these systems which may lead to an increased tendency to thrombosis:-


The first factor V mutations were discovered only about 3 years ago as a consequence of investigations of patients with an increased resistance to protein C activation. All the mutations detected to date involve the same DNA change, namely a G to A substitution resulting in an arginine to glutamine change at amino acid 506 in the Factor V protein. This DNA change can be simply and reliably detected with a Mnl I restriction enzyme cleavage test. Moreover the test is cheap, about £25.

However APC resistance itself can be detected even more cheaply, indeed a simple biochemical clotting test costs only about £2 and a recently introduced improved version of the test which has a potential negative predictive value for the factor V mutation approaching 99% costs only about £4. This test should substantially reduce the number of DNA tests that can be justified.

Present evidence suggests that overall about 20% of all patients with DVT have factor V mutations. This increases to about 50% in patients with a family history of thrombosis and as many as 60% of pregnant patients with thrombosis.

The relative risk of DVT for heterozygous carriers of the mutation is about 8 but for homozygous carriers this increase to about 80 fold.

For all women taking oral contraceptives the relative risk is about 4 but for women with a factor V mutation taking the pill the RR increases to about 35.

Not surprisingly important and urgent questions arise from this data:-


Having decided whom to screen the next dilemma is whom to treat, and how, and for how long.

Dr Keeling emphasized the importance of having a wide debate on these issues and the key role that public health physicians must play in determining policy. He himself is rather conservative, in favour of limiting screening at present to high risk groups and those in clinical trials. He expressed some concerns that unless we develop and implement clear policies in the near future practice will, to some extent, be driven by litigation or the fear of it.

Dr Sally Davies also addressed the issues surrounding thrombophilia screening. She thought that a case could be made for advising women with a family history of thrombosis who carried the factor V mutation to avoid the use of oestrogens . Similarly pregnant women with a history of thrombosis who carry the mutation should be considered for anticoagulant treatment, especially for 6 weeks post partum. It was emphasized that this treatment did not endanger breast-fed neonates.

Dr Davies referred to recent work on risk factors for myocardial infarction. This suggested that levels of factors VII and VIII and polymorphisms in platelet glycoprotein IIb could increase risk. Allele A2, present in about 25% of the population may increase relative risk by 2-6 fold.

In discussion it was emphasized that there is no evidence that factor V mutations are risk factors for arterial disease. Conversely although smoking increases the risk of arterial disease greatly it appears to be neutral for venous thrombosis.

The problems of presenting risk information to patients was highlighted. The differences between absolute risk, relative risk and NNTs were often confusing to patients and there is clearly an urgent need to improve transfer of information so that individuals can make genuinely 'informed choices'.

The chairman commented that since the issues arising from the role of factor V in venous thrombosis seem to be more clearly defined, ie a simple mutation with simple, reliable and cheap tests, and a better appreciation of the underlying pathobiology, than in the cases of the cancer genes that the meeting has considered, we ought be be able to arrive at rational policies more easily.

PURCHASING GENETIC SERVICES DR MIKE GILL

The evolution of regional specialties has an interesting history, driven by the size and needs of populations, the academic contribution and innovation to the subject and the central support available. Medical genetics has been one of the best examples of such services. But regional organisation does not fit easily with the new funding arrangements which are essentially locally based.

The scope of potential new genetic services is rapidly widening and the situation is made worse by the considerable confusion engendered by the many different uses of the word 'genetic'.

Commissioners of genetic services need to be explicit in the services they wish to purchase, but it is very difficult for a regional service to operate satisfactorily on the basis of multiple, often inconsistent, local contracts. A number of regions are therefore attempting to operate consortia where a professional advisory group reports to the relevant DHAs on behalf of the providers.

Key items on their agenda are:-


It must be appreciated that the develoment and introduction of new genetic services is the same as, and therefore competing with, the introduction of all other new medical services

Discussion

In the new system it is proving difficult to move large amounts of funds to develop new services. It is much easier to support incremental drift and to deal with individual or small groups of patients. This is not a satisfactory mechanism of achieving a more advanced service and emphasises the importance of collaborative working. Lead purchasing or 'fiat from above' are alternatives. However experience to date has shown that it is not easy to set up effective consortia.

A new problem arises from the fact that the 'new genetics' is in areas of existing clinical activities and it is not yet clear how commissioning boundaries should be drawn. A view was expressed that medical genetics should not be purchased as a separate service but embedded within the appropriate existing clinical services. Others thought that combined services would be better than embedded.

It is generally assumed that primary care will become progressively more involved in genetics issues, since these will inevitably concern more and more individuals but there is some anxiety that the necessary skills may not yet be widely available.


THE WAY FORWARD:

Dr Muir Gray set out an R&D perspective on general issues in this field. The appreciation of quality and strength of evidence should always be kept in mind.

The chairman thanked all the speakers and discussants and drew the meeting to a close by reminding the participants of three key questions:-