bs_bs_banner
Original article
High resolution anoscopy may be useful in achieving reductions in anal cancer local disease failure rates P. GOON, PHD FRCP, Division of Virology, Department of Pathology, University of Cambridge, Cambridge, and Department of GU and HIV Medicine, Addenbrooke’s Hospital, Cambridge, V. MORRISON, RCN, Department of GU and HIV Medicine, Addenbrooke’s Hospital, Cambridge, N. FEARNHEAD, FRCS, Department of Colorectal Surgery, Addenbrooke’s Hospital, Cambridge, J. DAVIES, FRCS, Department of Colorectal Surgery, Addenbrooke’s Hospital, Cambridge, C. WILSON, FRCP, Department of Clinical Oncology, Addenbrooke’s Hospital, Cambridge, C. JEPHCOTT, FRCP, Department of Clinical Oncology, Addenbrooke’s Hospital, Cambridge, J. STERLING, PHD FRCP, Division of Virology, Department of Pathology, University of Cambridge, Cambridge, & R. CRAWFORD, MD FRCS FRCOG,
Department of Gynaecological Oncology, Addenbrooke’s Hospital, Cambridge, UK
GOON P., MORRISON V., FEARNHEAD N., DAVIES J., WILSON C., JEPHCOTT C., STERLING J. & CRAWFORD R. (2015) European Journal of Cancer Care 24, 411–416 High resolution anoscopy may be useful in achieving reductions in anal cancer local disease failure rates Anal cancer is uncommon, with an incidence rate of 0.5–1.0 per 100 000 of the population but incidence rates have been steadily increasing over the last 3 decades. Biological and epidemiological evidence have been mounting and demonstrate that anal cancer has many similarities to cervical cancer, especially in regard to its aetiology. High-resolution anoscopy (HRA) of the anal region – analogous to colposcopy of the cervix, is a technique that is not well-known in the medical and surgical fraternity. Evidence to support the use of HRA for detection and treatment in the surveillance of AIN exists and strongly suggests that it is beneficial, resulting in reduced rates of cancer progression. Pilot data from our study showed a local disease failure rate of 1.73 per 1000 patient-months compared with a published rate of 9.89 per 1000 patient-months. This demonstrates a 5.72-fold reduction in local disease failure rates of patients with T1-T3 tumours; the data therefore suggests that use of HRA for detection and treatment in surveillance of anal cancer patients will help prevent local regional relapse at the anal site. There is an urgent need for a large, randomised controlled clinical trial to definitively test this hypothesis.
Keywords: HRA, HPV, anal cancer, pre-cancers.
INTRODUCTION Anal cancer is an uncommon tumour, comprising less than 4% of large bowel cancers, with an incidence of 0.5–1.0 per 100 000 of the population in most countries in the West every year. It is clear, however, that incidence Correspondence address: Peter Goon, Division of Virology, Department of Pathology, University of Cambridge, Level 5 Laboratories Block and Clinic 1A, Box 38, Addenbrooke’s Hospital, Hills Road, Cambridge CB2 0QQ, UK (e-mail: [email protected]).
Accepted 20 November 2013 DOI: 10.1111/ecc.12168 European Journal of Cancer Care, 2015, 24, 411–416
© 2013 John Wiley & Sons Ltd
rates have been steadily increasing in the general population over the last 3 decades (Johnson et al. 2004; Renehan et al. 2005; Brewster & Bhatti 2006), but exponential rises in the HIV+ and MSM (men-who-have-sex-with-men) groups have been seen (Chaturvedi et al. 2009; Powles et al. 2009). In the UK, the ratio of women to men with anal cancer is 3:2, whereas in the USA, it is nearly equal. Similar to cervical cancer, the dominant risk factor for the development of anal cancer is now known to be infection with human papillomaviruses (HPV), specifically highrisk HPV 16 and 18. Compelling biological and epidemiological evidence now demonstrate that anal cancer has many similarities to cervical cancer. Similar detection and
GOON ET AL.
screening techniques can therefore be used for both types of cancer and pre-cancer. HRA (high-resolution anoscopy – analogous to colposcopy) has strong potential for improving detection of significant anal neoplasias in routine clinical care. Despite HRA being accepted as the ‘gold standard’ now for diagnosis and treatment in anal intraepithelial neoplasia (AIN) management, the medical and surgical fraternity is still relatively oblivious to its existence. This was borne out recently by a review article in a major journal (Simpson & Scholefield 2011) when neither HRA nor anal cytology were mentioned or discussed. There is a need for wider recognition and debate on the implementation of newer technologies such as HRA which have strong potential for patient benefit. Colposcopy was invented by the German physician, Hans Hinselmann, in 1925. Since then, the technique has become the gold standard for detection and treatment of cervical lesions. Colposcopy was introduced into screening and treatment of pre-invasive disease prior to evidence-based medicine and there is some contention as to the value of colposcopy without multiple biopsies. The techniques involved in high resolution anoscopy (HRA) were first described in 1991 by a team in the University of California, San Francisco, where colposcopic techniques
were utilised to assess the perineum and anal canal. The term ‘HRA’ was first used in 1995 as a more accurate description of the use of a technique analogous to that in the colposcopy clinic, to visualise and characterise anal and peri-anal pre-neoplastic lesions. The technique itself is modelled on colposcopy; with the use of a binocular colposcope for magnification (see Fig. 1), use of acetic acid and Lugol’s iodine for staining, biopsy forceps, haemostatic agents such as Monsel’s solution or silver nitrate, and similarly trained staff. Over the last 15 years, expertise in HRA has gradually increased and direct anoscopic visualisation and biopsy is accepted as the most sensitive and specific method of diagnosing high-grade AIN (HG-AIN) lesions (Darragh 2008). There are now the beginnings of formal training standards for certification, accreditation or ongoing clinical quality assurance in HRA, starting with the USA. The University of San Francisco and others have established an HRA training course in conjunction with the American Society for Colposcopy and Cervical Pathology training schemes since 2005. To our knowledge, there are no similar training schemes available in Western Europe, and training and skills transfer is on an ad hoc basis. There is consequently a dearth of experts in the use of HRA and
Figure 1. HRA photos of neoplastic lesions. (A) Lesion with dense acetowhite staining – biopsy showed AIN 3 on histology. (B) Same lesion as in (A) and counterstained with Lugol’s iodine – Schiller’s test positive. (C) Dense acetowhite lesion with punctation, mosaicism, irregular vascularity, friability+++ = AIN3 on biopsy. (D) Another lesion similar to (C) but with increased mosaicism. (E) Higher magnification (×40) dense acetowhite lesion – biopsy showed superficially invasive squamous cell carcinoma. As for HG-CIN, HG-AIN is generally not palpable or visible unless acetic acid staining and magnified colposcopic visualisation is utilised. HRA, high-resolution anoscopy. HG-CIN, highgrade cervical intraepithelial neoplasia; HG-AIN, high-grade anal intraepithelial neoplasia.
412
© 2013 John Wiley & Sons Ltd
HRA may prevent anal cancer local disease failure
availability varies widely from country to country. In the UK, the authors are currently aware of only 3 centres with dedicated practitioners and routine HRA clinics that provide a service for patients and a focus for research. These are linked into the gynaecology oncological and colorectal surgical multidisciplinary teams (MDT). The shortage of skilled HRA practitioners is likely to remain a significant obstacle to any proposed anal cancer screening programme and to development of the service as an adjunct to current cervical screening. Current AIN management guidelines in the UK (Scholefield & Nugent 2011) do not recommend colorectal surgeons undertake additional HRA training, rather the establishment of tertiary referral cancer centres for MDT management of AIN and anal cancer. The similarities between colposcopy and anoscopy therefore suggest that HRA services could be embedded within Gynaecological oncological services, with the sharing of space, staff, reagents and other resources. Current practitioners in the US and Western Europe come from diverse clinical backgrounds such as primary care, dermatology, infectious diseases, genitourinary medicine, gastroenterology, oncology and surgery. Here we show data testing the hypothesis that HRA and ablation of high-grade neoplastic lesions in anal cancer follow-up will help prevent loco-regional relapse, and compare it with available published data. We then discuss the relevance of HRA with targeted biopsies plus ablative treatment with respect to two main scenarios: 1) Anal cancer recurrence; and 2) AIN detection and management.
METHODS A prospective pilot study opened to all patients suffering anal cancer from the East of England region, attending Addenbrooke’s Hospital, Cambridge for anal cancer treatment from 2007 to 2012. Inclusion criteria include any patient with anal cancer over the age of 18 years, able to attend for hospital follow-up and treatment, able to give fully informed written consent. Exclusion criteria: Any patient unable to give fully informed written consent or outside age range, and unable to attend hospital for follow-up and treatment. Besides all routine follow-ups in Clinical Oncology, recruited patients also have additional HRA surveillance with targeted biopsies and ablation of any detected lesions for two years. The rationale for a two year HRA follow-up is that approximately 95% of loco-regional relapse occurs during the first two-three years after chemoradiotherapy (Renehan et al. 2005). All patients were © 2013 John Wiley & Sons Ltd
treated with a standard chemoradiotherapy regime according to the UKCCCR anal cancer ACT 1 trial in 1996 (1996 UKCCCR) [median dose 55 Gy, 5-fluorouracil (5-FU) and MMC]. Patients were followed up at 3 monthly intervals for the first 2 years, and then 6 monthly until 5 years. Neoplastic lesions detected were biopsied and then ablated with a CO2 laser, trichloroacetic acid, cryotherapy or surgically excised. For the purposes of analyses, local disease failure (LDF) was defined as the presence of either residual (uncontrolled disease in the perineum between 6 weeks and 3 months after the start of treatment) or recurrent (re-emergence of local disease after 3 months) (Renehan et al. 2005). All patients gave written informed consent. The study was given ethical approval by Cambridgeshire 2 REC (10/ H0308/32). A total of 23 patients accepted the opportunity to join the study, 3 self-withdrew, leaving 20 in the cohort. Local disease failure incidence rates for patients with T1-T3 tumours were calculated for this cohort, and compared with similar published data (Renehan et al. 2005) using person-time incidence calculations. Fisher’s exact test was used for categorical parametric statistical analyses.
RESULTS Twenty patients were recruited into the study, all were TNM status T1-T3 with 5 T1 tumours (25%), 11 T2 tumours (55%) and 4 T3 tumours (20%). All patients were treated with a standard chemoradiotherapy regime according to the UKCCCR anal cancer ACT 1 trial in 1996 (1996) (median dose 55 Gy, 5-FU and MMC). Median follow-up was 19 (13–64) months. Age range of patients was 42–81 years, with a median of 56.5 years. There were 16 women and 4 men, none admitted to a history of previous sexually transmitted infections, 11 were smokers (10 women and 1 man). Of the 16 women, 7 reported a history of previous cervical intraepithelial neoplasia (CIN) disease detected and treated through the national cervical smear screening programme. All patients were HIV negative. Local disease failure (LDF) rates were calculated from for T1-T3 TNM status patients who were treated with the standard chemoradiotherapy regime. There were 5 T1 tumours (25%), 11 T2 tumours (55%) and 4 T3 tumours (20%). Median follow-up was 19 (13–64) months. Age range of patients was 42–81 years, with a median of 56.5. In our study, there were 576 patient-months and 1 case of LDF. Therefore, this equates to 1.73 cases per 1000 patient-months. The data are summarised in Table 1. 413
GOON ET AL.
Table 1. Analyses of anal cancer local disease failure data T1 tumours, n (%) T2 tumours, n (%) T3 tumours, n (%) Age in years, median (range) Follow-up in months, median (range) LDF (n) Total patient-months LDF rate (per 1000 patient-months) Treatment
5 (25%) 11 (55%) 4 (20%) 56.5 (42–81) 19 (13–64) 1 576 1.98 Chemoradiotherapy
LDF, local disease failure.
DISCUSSION Anal cancer recurrence and management Chemoradiotherapy is now the definitive first-line treatment for anal cancer since the publication of the results from the UKCCCR anal cancer ACT 1 trial in 1996 (UKCCCR Anal Cancer Trial Working 1996). ACT 2 showed that the local control rates for all stages of anal cancer were high – approximately 95% (James & Glynn-Jones 2009). However, it is estimated that 20–25% will suffer local disease relapse, predominantly during the first 3 years post diagnosis (Clark et al. 2004; Renehan et al. 2005). The definitive treatment for local disease relapse is salvage surgery, in most cases radical abdomino-perineal resection (APR). Morbidity is high. Known risk factors for local disease relapse are age >75 years, immunosuppression, large original tumour size and inability to complete primary chemoradiation. It follows that detection of early local relapsed disease may offer the opportunity for a second chance for a cure. There are some data to support this hypothesis; as patients selected for salvage surgery have 40–60% 5-year survival rates, compared with approximately 5% at 3 years for those not selected for surgery (Akbari et al. 2004; Renehan et al. 2005). Due to the high post-operative morbidity and a recognised perioperative mortality of
Original article
High resolution anoscopy may be useful in achieving reductions in anal cancer local disease failure rates P. GOON, PHD FRCP, Division of Virology, Department of Pathology, University of Cambridge, Cambridge, and Department of GU and HIV Medicine, Addenbrooke’s Hospital, Cambridge, V. MORRISON, RCN, Department of GU and HIV Medicine, Addenbrooke’s Hospital, Cambridge, N. FEARNHEAD, FRCS, Department of Colorectal Surgery, Addenbrooke’s Hospital, Cambridge, J. DAVIES, FRCS, Department of Colorectal Surgery, Addenbrooke’s Hospital, Cambridge, C. WILSON, FRCP, Department of Clinical Oncology, Addenbrooke’s Hospital, Cambridge, C. JEPHCOTT, FRCP, Department of Clinical Oncology, Addenbrooke’s Hospital, Cambridge, J. STERLING, PHD FRCP, Division of Virology, Department of Pathology, University of Cambridge, Cambridge, & R. CRAWFORD, MD FRCS FRCOG,
Department of Gynaecological Oncology, Addenbrooke’s Hospital, Cambridge, UK
GOON P., MORRISON V., FEARNHEAD N., DAVIES J., WILSON C., JEPHCOTT C., STERLING J. & CRAWFORD R. (2015) European Journal of Cancer Care 24, 411–416 High resolution anoscopy may be useful in achieving reductions in anal cancer local disease failure rates Anal cancer is uncommon, with an incidence rate of 0.5–1.0 per 100 000 of the population but incidence rates have been steadily increasing over the last 3 decades. Biological and epidemiological evidence have been mounting and demonstrate that anal cancer has many similarities to cervical cancer, especially in regard to its aetiology. High-resolution anoscopy (HRA) of the anal region – analogous to colposcopy of the cervix, is a technique that is not well-known in the medical and surgical fraternity. Evidence to support the use of HRA for detection and treatment in the surveillance of AIN exists and strongly suggests that it is beneficial, resulting in reduced rates of cancer progression. Pilot data from our study showed a local disease failure rate of 1.73 per 1000 patient-months compared with a published rate of 9.89 per 1000 patient-months. This demonstrates a 5.72-fold reduction in local disease failure rates of patients with T1-T3 tumours; the data therefore suggests that use of HRA for detection and treatment in surveillance of anal cancer patients will help prevent local regional relapse at the anal site. There is an urgent need for a large, randomised controlled clinical trial to definitively test this hypothesis.
Keywords: HRA, HPV, anal cancer, pre-cancers.
INTRODUCTION Anal cancer is an uncommon tumour, comprising less than 4% of large bowel cancers, with an incidence of 0.5–1.0 per 100 000 of the population in most countries in the West every year. It is clear, however, that incidence Correspondence address: Peter Goon, Division of Virology, Department of Pathology, University of Cambridge, Level 5 Laboratories Block and Clinic 1A, Box 38, Addenbrooke’s Hospital, Hills Road, Cambridge CB2 0QQ, UK (e-mail: [email protected]).
Accepted 20 November 2013 DOI: 10.1111/ecc.12168 European Journal of Cancer Care, 2015, 24, 411–416
© 2013 John Wiley & Sons Ltd
rates have been steadily increasing in the general population over the last 3 decades (Johnson et al. 2004; Renehan et al. 2005; Brewster & Bhatti 2006), but exponential rises in the HIV+ and MSM (men-who-have-sex-with-men) groups have been seen (Chaturvedi et al. 2009; Powles et al. 2009). In the UK, the ratio of women to men with anal cancer is 3:2, whereas in the USA, it is nearly equal. Similar to cervical cancer, the dominant risk factor for the development of anal cancer is now known to be infection with human papillomaviruses (HPV), specifically highrisk HPV 16 and 18. Compelling biological and epidemiological evidence now demonstrate that anal cancer has many similarities to cervical cancer. Similar detection and
GOON ET AL.
screening techniques can therefore be used for both types of cancer and pre-cancer. HRA (high-resolution anoscopy – analogous to colposcopy) has strong potential for improving detection of significant anal neoplasias in routine clinical care. Despite HRA being accepted as the ‘gold standard’ now for diagnosis and treatment in anal intraepithelial neoplasia (AIN) management, the medical and surgical fraternity is still relatively oblivious to its existence. This was borne out recently by a review article in a major journal (Simpson & Scholefield 2011) when neither HRA nor anal cytology were mentioned or discussed. There is a need for wider recognition and debate on the implementation of newer technologies such as HRA which have strong potential for patient benefit. Colposcopy was invented by the German physician, Hans Hinselmann, in 1925. Since then, the technique has become the gold standard for detection and treatment of cervical lesions. Colposcopy was introduced into screening and treatment of pre-invasive disease prior to evidence-based medicine and there is some contention as to the value of colposcopy without multiple biopsies. The techniques involved in high resolution anoscopy (HRA) were first described in 1991 by a team in the University of California, San Francisco, where colposcopic techniques
were utilised to assess the perineum and anal canal. The term ‘HRA’ was first used in 1995 as a more accurate description of the use of a technique analogous to that in the colposcopy clinic, to visualise and characterise anal and peri-anal pre-neoplastic lesions. The technique itself is modelled on colposcopy; with the use of a binocular colposcope for magnification (see Fig. 1), use of acetic acid and Lugol’s iodine for staining, biopsy forceps, haemostatic agents such as Monsel’s solution or silver nitrate, and similarly trained staff. Over the last 15 years, expertise in HRA has gradually increased and direct anoscopic visualisation and biopsy is accepted as the most sensitive and specific method of diagnosing high-grade AIN (HG-AIN) lesions (Darragh 2008). There are now the beginnings of formal training standards for certification, accreditation or ongoing clinical quality assurance in HRA, starting with the USA. The University of San Francisco and others have established an HRA training course in conjunction with the American Society for Colposcopy and Cervical Pathology training schemes since 2005. To our knowledge, there are no similar training schemes available in Western Europe, and training and skills transfer is on an ad hoc basis. There is consequently a dearth of experts in the use of HRA and
Figure 1. HRA photos of neoplastic lesions. (A) Lesion with dense acetowhite staining – biopsy showed AIN 3 on histology. (B) Same lesion as in (A) and counterstained with Lugol’s iodine – Schiller’s test positive. (C) Dense acetowhite lesion with punctation, mosaicism, irregular vascularity, friability+++ = AIN3 on biopsy. (D) Another lesion similar to (C) but with increased mosaicism. (E) Higher magnification (×40) dense acetowhite lesion – biopsy showed superficially invasive squamous cell carcinoma. As for HG-CIN, HG-AIN is generally not palpable or visible unless acetic acid staining and magnified colposcopic visualisation is utilised. HRA, high-resolution anoscopy. HG-CIN, highgrade cervical intraepithelial neoplasia; HG-AIN, high-grade anal intraepithelial neoplasia.
412
© 2013 John Wiley & Sons Ltd
HRA may prevent anal cancer local disease failure
availability varies widely from country to country. In the UK, the authors are currently aware of only 3 centres with dedicated practitioners and routine HRA clinics that provide a service for patients and a focus for research. These are linked into the gynaecology oncological and colorectal surgical multidisciplinary teams (MDT). The shortage of skilled HRA practitioners is likely to remain a significant obstacle to any proposed anal cancer screening programme and to development of the service as an adjunct to current cervical screening. Current AIN management guidelines in the UK (Scholefield & Nugent 2011) do not recommend colorectal surgeons undertake additional HRA training, rather the establishment of tertiary referral cancer centres for MDT management of AIN and anal cancer. The similarities between colposcopy and anoscopy therefore suggest that HRA services could be embedded within Gynaecological oncological services, with the sharing of space, staff, reagents and other resources. Current practitioners in the US and Western Europe come from diverse clinical backgrounds such as primary care, dermatology, infectious diseases, genitourinary medicine, gastroenterology, oncology and surgery. Here we show data testing the hypothesis that HRA and ablation of high-grade neoplastic lesions in anal cancer follow-up will help prevent loco-regional relapse, and compare it with available published data. We then discuss the relevance of HRA with targeted biopsies plus ablative treatment with respect to two main scenarios: 1) Anal cancer recurrence; and 2) AIN detection and management.
METHODS A prospective pilot study opened to all patients suffering anal cancer from the East of England region, attending Addenbrooke’s Hospital, Cambridge for anal cancer treatment from 2007 to 2012. Inclusion criteria include any patient with anal cancer over the age of 18 years, able to attend for hospital follow-up and treatment, able to give fully informed written consent. Exclusion criteria: Any patient unable to give fully informed written consent or outside age range, and unable to attend hospital for follow-up and treatment. Besides all routine follow-ups in Clinical Oncology, recruited patients also have additional HRA surveillance with targeted biopsies and ablation of any detected lesions for two years. The rationale for a two year HRA follow-up is that approximately 95% of loco-regional relapse occurs during the first two-three years after chemoradiotherapy (Renehan et al. 2005). All patients were © 2013 John Wiley & Sons Ltd
treated with a standard chemoradiotherapy regime according to the UKCCCR anal cancer ACT 1 trial in 1996 (1996 UKCCCR) [median dose 55 Gy, 5-fluorouracil (5-FU) and MMC]. Patients were followed up at 3 monthly intervals for the first 2 years, and then 6 monthly until 5 years. Neoplastic lesions detected were biopsied and then ablated with a CO2 laser, trichloroacetic acid, cryotherapy or surgically excised. For the purposes of analyses, local disease failure (LDF) was defined as the presence of either residual (uncontrolled disease in the perineum between 6 weeks and 3 months after the start of treatment) or recurrent (re-emergence of local disease after 3 months) (Renehan et al. 2005). All patients gave written informed consent. The study was given ethical approval by Cambridgeshire 2 REC (10/ H0308/32). A total of 23 patients accepted the opportunity to join the study, 3 self-withdrew, leaving 20 in the cohort. Local disease failure incidence rates for patients with T1-T3 tumours were calculated for this cohort, and compared with similar published data (Renehan et al. 2005) using person-time incidence calculations. Fisher’s exact test was used for categorical parametric statistical analyses.
RESULTS Twenty patients were recruited into the study, all were TNM status T1-T3 with 5 T1 tumours (25%), 11 T2 tumours (55%) and 4 T3 tumours (20%). All patients were treated with a standard chemoradiotherapy regime according to the UKCCCR anal cancer ACT 1 trial in 1996 (1996) (median dose 55 Gy, 5-FU and MMC). Median follow-up was 19 (13–64) months. Age range of patients was 42–81 years, with a median of 56.5 years. There were 16 women and 4 men, none admitted to a history of previous sexually transmitted infections, 11 were smokers (10 women and 1 man). Of the 16 women, 7 reported a history of previous cervical intraepithelial neoplasia (CIN) disease detected and treated through the national cervical smear screening programme. All patients were HIV negative. Local disease failure (LDF) rates were calculated from for T1-T3 TNM status patients who were treated with the standard chemoradiotherapy regime. There were 5 T1 tumours (25%), 11 T2 tumours (55%) and 4 T3 tumours (20%). Median follow-up was 19 (13–64) months. Age range of patients was 42–81 years, with a median of 56.5. In our study, there were 576 patient-months and 1 case of LDF. Therefore, this equates to 1.73 cases per 1000 patient-months. The data are summarised in Table 1. 413
GOON ET AL.
Table 1. Analyses of anal cancer local disease failure data T1 tumours, n (%) T2 tumours, n (%) T3 tumours, n (%) Age in years, median (range) Follow-up in months, median (range) LDF (n) Total patient-months LDF rate (per 1000 patient-months) Treatment
5 (25%) 11 (55%) 4 (20%) 56.5 (42–81) 19 (13–64) 1 576 1.98 Chemoradiotherapy
LDF, local disease failure.
DISCUSSION Anal cancer recurrence and management Chemoradiotherapy is now the definitive first-line treatment for anal cancer since the publication of the results from the UKCCCR anal cancer ACT 1 trial in 1996 (UKCCCR Anal Cancer Trial Working 1996). ACT 2 showed that the local control rates for all stages of anal cancer were high – approximately 95% (James & Glynn-Jones 2009). However, it is estimated that 20–25% will suffer local disease relapse, predominantly during the first 3 years post diagnosis (Clark et al. 2004; Renehan et al. 2005). The definitive treatment for local disease relapse is salvage surgery, in most cases radical abdomino-perineal resection (APR). Morbidity is high. Known risk factors for local disease relapse are age >75 years, immunosuppression, large original tumour size and inability to complete primary chemoradiation. It follows that detection of early local relapsed disease may offer the opportunity for a second chance for a cure. There are some data to support this hypothesis; as patients selected for salvage surgery have 40–60% 5-year survival rates, compared with approximately 5% at 3 years for those not selected for surgery (Akbari et al. 2004; Renehan et al. 2005). Due to the high post-operative morbidity and a recognised perioperative mortality of
