Article Text

Download PDFPDF

The impact of screening and genetic registration on mortality and colorectal cancer incidence in familial adenomatous polyposis
  1. E K L Mallinson1,
  2. K F Newton2,
  3. J Bowen3,
  4. F Lalloo1,
  5. T Clancy1,
  6. J Hill2,
  7. D G R Evans1
  1. 1Department of Genetic Medicine, Manchester Academic Health Science Centre, Manchester, UK
  2. 2Department of General Surgery, Manchester Royal Infirmary, Manchester, UK
  3. 3Department of Paediatric Surgery, Royal Manchester Children's Hospital, Central Manchester University Hospitals Foundation Trust, Manchester, UK
  1. Correspondence to Katherine Newton c/o Prof. D Gareth Evans, Department of Genetic Medicine, 6th Floor, St. Mary's Hospital, Oxford Rd, Manchester M13 9LW, UK; katynewton2002{at}yahoo.co.uk

Abstract

Background Regular colonic surveillance of familial adenomatous polyposis (FAP) patients is necessary to ensure appropriate prophylactic surgery is performed before colorectal cancer (CRC) develops. Polyposis Registries have been established to coordinate screening programmes. The aim of this study was to assess the effect of screening and of the formation of the Registry on survival, incidence of CRC and age at onset of CRC, in FAP patients.

Methods Patients on the Manchester Polyposis Registry were categorised according to their mode of presentation; screening or symptomatic, and survival time from birth was calculated for each patient (n=353). The effect of the formation of the Registry was assessed by comparing survival times from birth for patients diagnosed in the 20 years before the establishment of the Registry, to patients diagnosed in the 20 years since the formation of the Registry (n=273).

Results This study demonstrated that survival was increased from 57.8 years to 70.4 years (p<0.001) by screening, and from 58.1 years to 69.6 years (p=0.007) following establishment of the Polyposis Registry. The incidence of CRC was reduced from 43.5% to 3.8% by screening, and from 28.7% to 14.0% following establishment of the Polyposis Registry. Although direct causation between improved survival and reduced CRC incidence, and establishment of the Registry cannot be proven, an association has been demonstrated. Colorectal cancer was found to develop, on average, 16 years later in the screening population.

Conclusion A regular systematic large bowel screening programme, managed by a Polyposis Registry, significantly improves the prognosis of FAP.

  • Familial adenomatous polyposis
  • screening
  • cancer genetics
  • polyposis
View Full Text

Statistics from Altmetric.com

Significance of this study

What is already known about this subject?

  • Regular screening of patients at risk of familial adenomatous polyposis (FAP), improves prognosis.

  • Polyposis Registries improve outcomes for FAP patients by improving the rate of diagnosis of FAP.

  • Polyposis Registries improve outcomes for reducing the incidence of colorectal cancer (CRC).

What are the new findings?

  • The improvement in actuarial survival of UK FAP patients due to presymptomatic screening is quantified (57.8 to 70.4 years).

  • The benefits of a Polyposis Registry in a UK population are quantified and improvement in actuarial survival demonstrated (58.1 to 69.6 years).

How might it impact on clinical practice in the foreseeable future?

  • This study supports the implementation of Polyposis Registries in areas currently lacking. The improvement in prognosis demonstrated, both in terms of CRC incidence and survival, will hopefully encourage the organisation and funding of such registries.

Introduction

Familial adenomatous polyposis (FAP) is an autosomal dominant condition, in which hundreds to thousands of precancerous adenomatous polyps develop throughout the colorectum.

The clinical diagnosis of FAP typically requires observation of diffuse adenomatous polyps in the large bowel at any age, or fewer polyps in the colorectum of a symptomatic child. Although colonic polyps may develop in the colorectum of the normal population, FAP sufferers develop multiple polyps beginning in childhood and adolescence. ‘There is an interval of approximately 10 years between the appearance of polyps and the onset of symptoms’,1 and thus large bowel endoscopic surveillance (screening) aims to monitor patients in this latent phase. Endoscopy allows quantification of polyp load and thus an assessment of colorectal cancer (CRC) risk, and appropriate planning of prophylactic surgery.

Genetic analysis is not recommended until the child is able to understand the consequences of the result, which is generally agreed to be around the age of 11 years.2 This is also the time at which genetic diagnosis would alter the management of the child. Some individuals present symptomatically either due to new mutations or if they are not known to a screening programme, and are subsequently diagnosed with FAP. These individuals have a worse prognosis.3 Colonic surveillance is performed annually if polyps are found, but may be reduced to biennially if no polyps are present. In typical FAP, it is recommended that this commences at age 10–12.4 5 There is increasing evidence supporting the existence of an attenuated FAP phenotype determined by genotype. These patients tend to have fewer, more proximal polyps and develop CRC 10–15 years later than those with typical FAP.6 7 It is thought that biennial colonoscopy commencing at age 18–20 years is appropriate for these patients.4 This obviously avoids unnecessary procedures in a proportion of the paediatric FAP population.

Regular screening and follow-up is accepted as vital in maintaining the health of FAP patients. A central organisation directing and coordinating such events should be beneficial. In 1925, Lockhart-Mummery demonstrated that presymptomatic examination of at-risk family members, and prophylactic colectomy reduced the incidence of CRC and hence improved survival of FAP patients. Following this, the St Mark's Hospital Polyposis Registry was established as the first Polyposis Registry in the world.8 Polyposis Registries have now been established across the world, both regionally and nationally. However, despite the British Society of Gastroenterologists (grade B) recommendation, published in 2002, that all FAP families should be notified to a Regional Registry,5 there are currently only two formal Polyposis Registries in the UK (Manchester Regional Genetics Centre and the North West Thames Regional Genetics Service: St Mark's Polyposis Registry). The Manchester Regional Genetics Centre serves a population of 4.7 million. The St Mark's Polyposis Registry serves a regional population of 3.5 million, although as this Registry receives many referrals from other centres, both in London and nationwide, the exact population served is difficult to define. One other centre (Oxford Regional Genetics Centre; serves a population of 3.6 million) has a more informal local database, which essentially functions as a Registry. Patients are regularly followed up and screening outcomes are documented. There are 23 regional genetics centres in the UK.

The aims of a Polyposis Registry include: registration of polyposis patients and their family members, counselling and genetic testing, initiation and coordination of screening of family members at risk, distribution of information and advances in scientific knowledge and research.9 A further benefit of an established register is the maintenance of post surgical follow-up, which is of particular importance after ileo-rectal anastomosis surgery, to ensure monitoring of the retained rectum.10 The continuity of care provided and annual contact maintained with consistent departmental staff including genetic counsellors also benefits the psychological and social aspects of patients' health.

Several European studies have demonstrated the effect a Registry has on the FAP populations it monitors. An improvement in survival, an increase in detection rate of FAP and a reduction in the incidence of CRC have been demonstrated.4 11 12 Two UK groups have reported on the outcomes of regional registries. Rhodes et al reported on the early results of the Northern Region Polyposis Registry in 1991. They found that their Registry increased the identification of gene carriers and hence screening of at-risk relatives. Mean age of diagnosis was reduced from 36.6 years to 24.7 years. At that time the Registry had only been established for 36 months.13 Morton et al reported similar results following the establishment of the West Midlands Polyposis Register. Median age of diagnosis was reduced from 32 years to 23 years (p=0.0004), and incidence of CRC was reduced from 35% to 14% (p<0.05).14 No UK group has reported on the improvement on survival or age of onset of CRC attributable to a Polyposis Registry.

The aim of this study was to assess the effect of mode of initial presentation and of a UK Regional Polyposis Registry on survival, incidence of CRC, and age of onset of CRC.

Methods

This study represents a part retrospective, part prospective longitudinal study of FAP patients listed on the Manchester Polyposis Registry. Patient details are contained upon a Filemaker Pro database. Familial adenomatous polyposis patients, from families ascertained through the genetic register, were confirmed from medical records or cancer registration and entered on the database along with all available information on endoscopy screening. Dates of FAP diagnosis and dates of cancer diagnosis were obtained, along with dates of death or last follow-up, from the North West Cancer Intelligence Service (NWCIS), family genetic records and also from the NHS Spine Portal. Patients were entered onto the Filemaker Pro database if they were found to have over 100 colonic adenomas and/or an adenomatous polyposis coli (APC) mutation. Thus, the database includes some families who were latterly found to have MutYH associated polyposis.

Only patients living within the strict regional boundaries of the Genetic Registry and NWCIS were included in the study and the database was interrogated to retrieve the necessary data.

Patients were grouped according to their mode of initial presentation; whether they had presented at presymptomatic screening or presented symptomatically. For ease of reading, these groups will be referred to as the screening group and the symptomatic group respectively. The symptomatic group were all entered into the screening programme following initial presentation. Patients were excluded from the study if their mode of presentation was unknown and also if no date of last follow-up could be established. Survival times were then calculated for each patient and the incidence of CRC determined. The time until onset of CRC was also calculated. Statistical analyses were then completed using Kaplan-Meier curves.

For the second part of the study patients were selected according to whether they were diagnosed in the 20 years before the formation of the Registry (1969–1988) or after its formation (1989–2009). All de novo cases were eliminated from the study as the establishment of the Registry would not have affected these patients. Patients were again excluded if they had no known date of last follow-up or an unknown mode of presentation. Survival times were calculated, along with the time till onset of CRC and CRC incidence.

In both parts of the study, survival times were calculated using the patient's date of birth and date of death. If the patient was still alive at the time of the study then the date of last follow-up was used instead of date of death, and the survival time was predicted using Kaplan-Meier curves. The time until onset of CRC was calculated using the patients' date of birth and date of diagnosis of cancer. If the patient had not been diagnosed with CRC at the time of the study then the date of last follow-up was used instead of the date of diagnosis, and a cancer-free time was predicted using Kaplan-Meier curves. The incidence of CRC was also determined.

SPSS statistical software, version 16.0 was used to complete the analyses. Kaplan-Meier was used to estimate survival rates and the log rank test was used to analyse differences between the groups. A p value of <0.05 was considered to be statistically significant.

Results

A total of 377 patients registered on the Manchester Polyposis Registry live within the regional geographic and NWCIS boundaries. To assess the effect of presymptomatic screening upon the expected survival of FAP patients, a sample of 353 eligible patients was used. One hundred and eighty five (52.4%) of these were diagnosed by screening, whereas the remaining 168 patients (47.6%) presented symptomatically. The male to female ratios for the two groups were 95:90 and 97:71 respectively (figure 1).

Figure 1

Screening and symptomatic populations.

At the time of the study, the mortality of the screening population was 5.9% (11 patients), compared with a mortality of 56.0% (94 patients) in the symptomatic population (p<0.001).

The mean survival age of the screening population was 70.4 years (95% CI 64.4 to 76.3), compared to 57.8 years (95% CI 54.6 to 61.1) for the symptomatic population. The cumulative life expectancy of the screening population was significantly better than that of the symptomatic population (p<0.001) (figure 2).

Figure 2

Kaplan-Meier survival curve to show the cumulative survival of familial adenomatous polyposis (FAP) patients dependent on the screening population and the symptomatic population (χ2=14.98, 1 df, p≤0.001, log rank test).

At the time of this study, only seven of the patients in the screening population had developed CRC (3.8%), compared to 73 patients in the symptomatic population (43.5%) (p<0.001).

In this study, the mean age at onset of CRC for the screening population was 73.6 years (95% CI 67.8 to 79.4), whereas the symptomatic population developed CRC at 57.5 years (95% CI 54.3 to 60.7, p<0.0001) (figure 3).

Figure 3

Kaplan-Meier survival curve to show the cumulative probability of familial adenomatous polyposis (FAP) patients developing colorectal cancer (CRC), of the screening population and the symptomatic population (χ2=18.73, 1 df, p≤0.001, log rank test).

To assess the effect of the establishment of the Manchester Polyposis Registry on expected survival, a sample of 273 eligible FAP patients was used. Eighty-seven patients (31.9%) were diagnosed with FAP in the 20 years prior to the formation of the Manchester Polyposis Registry in 1989. The remaining 186 FAP patients (68.1%), were diagnosed in the 20 years after the Registry was established. The male to female ratios for the two groups were 45:42 and 99:87 respectively (figure 4).

Figure 4

Pre-Registry and post-Registry populations.

The rate of diagnosis of FAP has more than doubled since the Registry was formed. On average, pre-Registry patients were diagnosed at a rate of 4.4 per annum, whereas post-Registry diagnoses (excluding de novo cases) increased to 9.3 patients per annum (table 1).

Table 1

Age at diagnosis and death and cause of death in individuals with inherited FAP over two time spans

There were 140 patients detected at screening in the post-Registry population. Four of these patients died (3.0%). None of these deaths were CRC=related; although three of the four deaths were related to desmoids (table 1).

Despite the inception of the Registry, a number of individuals (46 people) still presented symptomatically. Twenty-five of these were new patients from families with a significant history but as yet undiagnosed, seven were relatives of a new patient (proband) who had had a recent cancer or polyps, one patient was known to the Registry but was too young for screening when she had symptoms, and three patients had been called for screening but had repeatedly failed to attend. Four patients were new diagnoses and were latterly found to have MutYH polyposis. The remainder were from families who were known to either this or another Registry but had either not been screened due to a breakdown in family communication or had been lost to follow-up after a change of address. One patient was found to have been diagnosed elsewhere on screening and had been lost to follow-up after a change of address. He presented to the present Registry with symptoms. Thirteen CRC-related deaths occurred in this post-Registry group, all these patients presented symptomatically (table 1). Only two deaths occurred in individuals known to the Registry. These individuals repeatedly failed to respond to calls for screening. Four deaths were due to failure of the spouse of an affected person who had died (prior to formation of the Registry) to inform their offspring of their risk (their offspring were <10 years of age at time of their affected parent's death). The other seven deaths occurred in individuals from families who were not known to the Registry at the time of their symptomatic diagnosis.

Since the inception of the Registry, 95 FAP patients have been diagnosed presymptomatically by DNA testing, 64 occurring at age 12 years or less (table 1). No DNA testing was possible before the Registry, which is reflected in the higher median age of diagnosis in the pre-Registry screened group (table 1).

The mean survival age in the post-Registry group (69.6 years, 95% CI 64.0 to 75.2) is 11.5 years greater than the pre-Registry group (58.1 years, 95% CI 53.2 to 63.0, p=0.007) (figure 5).

Figure 5

Kaplan-Meier survival curve to show the cumulative survival of familial adenomatous polyposis (FAP) patients who were diagnosed in the 20 years before the formation of the Polyposis Registry and those patients diagnosed in the 20 years since the formation of the Registry (χ2=7.36, 1 df, p=0.007, log rank test).

At the time of this study, 28.7% of the pre-Registry population (25 patients; screen detected n=2, symptomatic n=23) had been diagnosed with CRC compared to only 14.0% of the post-Registry population (26 patients; screen detected n=4, symptomatic n=22). This difference in CRC incidence, however, may be attributable to the difference in age of the two populations rather than the formation of the Registry.

The mean age at onset of CRC in the study, was 61.3 years (95% CI 57.2 to 65.3) for the post-Registry population and 61.9 years for the pre-Registry population (95% CI 56.6 to 67.2, p=0.778). Thus, there is no significant difference in the age at onset of CRC between the pre-Registry and post-Registry FAP populations.

Discussion

This study demonstrates an improvement in survival of FAP patients as a result of screening and the establishment of the Manchester Polyposis Registry. There have been a limited number of published studies demonstrating this and none have used actuarial survival from birth as the main outcome. No UK group has previously published regional survival data.

This study demonstrates that when screened before the onset of symptoms, FAP patients survive more than 12 years longer than when presenting symptomatically. This supports previous studies. Bulow et al calculated that 94% of their call-up patients (patients undergoing presymptomatic screening) survived 10 years post diagnosis in comparison to only 41% of probands (patients diagnosed with symptoms).10 A further smaller study has confirmed these findings.12 Heiskanen et al also observed a statistically significant improved survival in screen detected patients following colectomy.15

The present study calculated overall survival from date of birth rather than from date of diagnosis. It is recognised that FAP patients detected on screening tend to be younger at diagnosis than those patients detected symptomatically.7 10 11 Thus, it was acknowledged that lead time bias may occur if date of diagnosis was used in the survival calculation. Previous studies calculated the survival of both symptomatic and screening populations, using the date of diagnosis of polyposis. Therefore, direct comparison between the present results and others cannot be made. This study, however, confirms that the presymptomatic screening of FAP patients improves the long-term survival as suggested by previous studies.11 14 15 Indeed, actuarial survival from birth is a more accurate assessment of life expectancy and the authors believe this method of analysis is more appropriate.

In this FAP population, the incidence of CRC was reduced by presymptomatic screening. Only 3.8% of the population screened developed CRC, compared to 43.5% of patients within the population presenting with symptoms. All other studies agree that fewer patients diagnosed on screening subsequently develop CRC.10 11 16 17 Morton et al investigated patients notified to the West Midlands Polyposis Registry and found the rate of CRC to be 6% in the screen detected population and 64% in the symptomatic population (p<0.001).14 Rhodes et al found a similar 10-fold reduction among patients notified to the Northern region Polyposis Register.13 In the present study, the overall incidence of CRC was studied rather than CRC incidence at the point of FAP diagnosis. This allows better assessment of whether presymptomatic screening actually reduced the incidence of CRC overall.

A statistically significant difference in age at onset of CRC between those detected on screening and those who presented symptomatically was also demonstrated by this study. When comparing the mean age at onset of CRC for these two groups, a difference of more than 16 years was observed.

This study also demonstrates an improvement in survival of FAP patients since the establishment of the Manchester Polyposis Registry in 1989. This improvement in survival is attributable to the reduction in incidence of CRC. No CRC deaths have occurred in screened individuals since the inception of the Registry. The improved survival in the study was calculated to be an increase of more than 11 years. This apparent effect upon patient survival is almost certainly attributable to the improvement in organisation and coordination of patient screening since the Registry's formation. Although some of the improvements in outcomes may have resulted from better endoscopy recall systems utilised by many endoscopy units, these systems cannot fulfil all the functions of a Registry. Bulow and colleagues demonstrated that screening became a much more common practice after the establishment of the register in Denmark in 1971.10 A greater proportion of patients in the post-Registry population in the present study were screened prior to the onset of symptoms (75.3%), compared to the pre-Registry population (46.0%). Three deaths in young FAP patients (1.6% of all post-Registry patients) have occurred due to desmoids since the Registry was established. This is in agreement with the Danish series (2%),3 but is much lower than the incidence in the Dutch series (7%)11 and the Cleveland series (11%).18 Timing of surgery needs to be carefully considered, particularly in those with mutations in the 3′ region of the APC gene who frequently have attenuated polyp disease.19

Of the 23 Regional Genetics Centres in the UK, only two have a formal FAP Registry used to coordinate screening. Several centres have either previously had Registries or have attempted to establish Registries but have encountered problems with administrative and managerial support, and funding. It appears that current practice is for the diagnosis to be made by the Clinical Genetics departments. Recommendations for screening are then made to the local Colorectal surgeons or Gastroenterologists. The responsibility for coordinating timely screening lies with these clinicians rather than being organised through a formal Registry.

Previous studies have recorded improvements in the prognosis of FAP patients since the establishment of polyposis registries in Denmark17 and in the Balearic Islands.12 Bulow et al compared the cumulative 10-year survival for FAP patients over two consecutive periods of time. They attributed the increase in survival, from 49% in the 1950–1969 period, to 75% in the 1970–1992 period, to the formation of the Danish Polyposis Register and also to the implementation of a screening programme.10 Bulow et al subsequently published further results of the Danish National Registry and found a significantly improved cumulative life expectancy since the inception of the Registry (p<0.0001).3

The present study indicates that the establishment of the Manchester Polyposis Registry has improved the detection rate of FAP, as more affected individuals have been diagnosed. The dramatic increase in rate of diagnosis, (from 4.4 per annum (pre-Registry) to 9.3 per annum (post-Registry)) is in agreement with previous studies.10 11 13 16

There are currently no available published contemporary data regarding prognosis in patients with FAP from regions currently lacking a Polyposis Registry. A study investigating the prognosis for patients with FAP in regions without a Registry would be important to assess whether there are other ways to manage systematic bowel screening. In the meantime, the present study provides further evidence that FAP registries improve survival, delay the onset of, and reduce the overall incidence of, CRC in FAP.

Conclusion

This study has demonstrated that the expected survival of FAP patients can be improved by presymptomatic large bowel surveillance of at-risk individuals, and by instituting a Registry to coordinate the screening programme. The implementation of both a Registry and screening has resulted in a reduction in incidence of CRC, which is reflected in improved patient survival.

Therefore, the findings from this study support the establishment of Polyposis Registries in areas lacking such an organisation.

References

View Abstract

Footnotes

  • Funding DGR Evans, J Hill, F Lalloo, J Bowen and T Clancy are supported by the Manchester NIHR Biomedical Research Centre. Other funders: NIH.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles

  • Digest
    Emad El-Omar Severine Vermeire Alexander Gerbes