Techniques in Coloproctology

, Volume 18, Issue 3, pp 265–272

Results from an American Society of Colon and Rectal Surgeons survey on the management of young-onset colorectal cancer

Authors

    • Department of Colorectal SurgeryDigestive Disease Institute, Cleveland Clinic
    • Department of Colorectal SurgeryEpworth Healthcare
    • Division of Surgical OncologyPeter MacCallum Cancer Centre
  • M. F. Kalady
    • Department of Colorectal SurgeryDigestive Disease Institute, Cleveland Clinic
    • Center for Hereditary Colorectal NeoplasiaDigestive Disease Institute, Cleveland Clinic Foundation
  • R. P. Kiran
    • Department of Colorectal SurgeryDigestive Disease Institute, Cleveland Clinic
  • J. M. Church
    • Department of Colorectal SurgeryDigestive Disease Institute, Cleveland Clinic
    • Center for Hereditary Colorectal NeoplasiaDigestive Disease Institute, Cleveland Clinic Foundation
Original Article

DOI: 10.1007/s10151-013-1052-5

Cite this article as:
Warrier, S.K., Kalady, M.F., Kiran, R.P. et al. Tech Coloproctol (2014) 18: 265. doi:10.1007/s10151-013-1052-5

Abstract

Background

Young patients with colorectal cancer (CRC) present a diagnostic and clinical challenge. The aim of our study was to survey the approaches to preoperative evaluation and clinical management of young patients with CRC by colorectal surgeons in North America.

Methods

A standard electronic survey was sent to the members of the American Society of Colon and Rectal Surgeons. The survey polled management decisions in various clinical scenarios for CRC patients less than 50 years old. Survey responses were collated and analyzed.

Results

One hundred ninety surgeons responded and 140 completed the entire survey (response rate 10 %). Eighty percent of surgeons would offer preoperative genetic testing if the patient’s family met the Amsterdam criteria compared to only 67 % if the criteria were not met. Of those offering preoperative tumor testing, 48 % test microsatellite instability, 19 % mismatch repair protein expression by immunohistochemistry, and 24 % offer both. Decisions regarding the extent of the resection for cancer were dependent on family history: Most members (86 %) would perform a segmental colectomy for CRC in a patient without family history. Eighty-four percent of respondents would offer a total abdominal colectomy if preoperative tests indicated Lynch syndrome. When questioned about MYH-associated polyposis, only 27 % recognized the appropriate diagnosis.

Conclusions

Among the American Society of Colon and Rectal Surgeons, family history influences preoperative testing and surgical management decisions. A significant portion of surgeons do not offer preoperative genetic testing, despite implications on operative management, postoperative surveillance, and screening of family members.

Keywords

Lynch syndromeColorectal cancerYoung-onsetAmerican Society of Colon and Rectal Surgeons

Introduction

Colorectal cancer (CRC) is the second leading cause of cancer deaths in the USA [1]. The majority of colorectal cancers arise in patients over 50 years of age; however, the incidence among the young adult population is increasing. Despite a decline in overall rates of CRC since 1996 [2], over the same time period, the incidence of patients with CRC aged less than 50 years old has risen [3]. Young patients with CRC pose particular challenges in diagnosis and treatment. Prevention is difficult because patients have not yet reached the age at which average risk screening is recommended. Symptoms are often attributed to benign causes, leading to delays in diagnosis. The young age at onset suggests a hereditary component to the disease, and an advanced stage at presentation carries a potential for early mortality that may have significant psychological effects and impose financial burdens on many young families.

Syndromes of inherited CRC account for up to 5 % of the overall burden of colorectal cancer [4], although when patients are diagnosed young, heredity may instead account for up to 15–25 % [5]. Only some of these patients are identified by the presence of a strong family history or intestinal polyposis. In patients with Lynch syndrome, the presence of an accelerated adenoma-to-carcinoma sequence confers a higher risk of metachronous colorectal cancer and extra colonic cancers such as endometrial, ovarian, gastric, and ureteric [6]. In order to pre-empt this increased cancer risk, surgeons may alter their surgical strategy and offer an extended prophylactic colon resection, as well as prophylactic hysterectomy and oophorectomy [7].

The aim of the study was to survey members of the American Society of Colorectal Surgeons (ASCRS) about their knowledge and practice in these areas. We were particularly concerned with preoperative management strategies, the extent of colectomy and postoperative surveillance.

Materials and methods

The Cleveland Clinic Institutional Review Board approved this study. Approval by the ASCRS executive council was obtained prior to distribution of the survey. A structured survey was distributed to members and post-fellowship trainees of the ASCRS. All ASCRS members are fully trained in general surgery and have undertaken a post-fellowship post in colon and rectal surgery or have a dedicated interest in colon and rectal surgery. The survey was conducted over a 6-week period with reminder e-mails sent to non-responders at the end of the fourth week.

The survey utilized scenario-based questions to investigate respondents’ clinical approach to colorectal cancer patients less than 50 years old. The survey is summarized in Table 1. A full copy of the survey is available as supplemental information [8].
Table 1

Summary of the survey distributed to the membership of the American Society of Colon and Rectal Surgeons (ASCRS)

1. Thirty-nine-year-old male with a cecal cancer, with no family history of colorectal cancer or polyps. Questions related to the preoperative, operative, and long-term surveillance strategy were asked

2. Thirty-nine-year-old male with cecal cancer, normal preoperative testing, but moderate family history. What operative and long-term surveillance strategy would you offer?

3. Thirty-nine-year-old male with a cecal cancer, with no synchronous polyps but a family history that fulfills Amsterdam Criteria for Hereditary Non Polyposis Colorectal Cancer. Questions relating to preoperative, operative, and long-term surveillance strategy (if testing was normal) were asked. In addition, the extent of colectomy and long-term surveillance frequency were asked if preoperative testing had indicated Lynch syndrome

4. Thirty-eight-year-old male with a low rectal cancer (5 cm from the anal verge) in the absence of family history and synchronous polyps. Questions related to preoperative, operative, and long- term surveillance strategy if this testing was normal were asked

5. Thirty-eight-year-old man with low rectal cancer (5 cm from the anal verge) with a family history that fulfills Amsterdam criteria. Questions were asked related to the uptake of preoperative testing. The operative and long-term surveillance strategy was asked if both testing was normal (Familial Colorectal Cancer Type X), and separately if preoperative testing indicated Lynch syndrome

6. Thirty-nine-year-old man with a young brother with colon cancer, but no other family history. The mode of inheritance was asked. At colonoscopy if 25 adenomatous polyps were found dispersed throughout the colon, what is the likely syndrome, and operative strategy that would be employed? If the polyps were hyperplastic, does this fulfill criteria for serrated polyposis syndrome and what operative strategy would be employed?

7. Thirty-nine-year-old female has finished childbearing and whose family history fulfills Amsterdam criteria. Preoperative immunohistochemistry reveals an absence of MSH6, suggestive of a MSH6 gene germline mutation. Which treatment strategy would you adopt at the time of colectomy? The answers included, a. definitive colectomy with TVUS follow-up, b. Immediate colectomy with staged hysterectomy following Family Cancer Center referral, c. Proceed with definitive colectomy and synchronous TAH/BSO, d. Immediate germline testing prior to synchronous hysterectomy at the time of colectomy

The focus of the survey was the preoperative testing threshold and strategy employed, as well as the extent of colectomy chosen. How this was influenced by age, family history, and a known familial syndrome were assessed

BSO Bilateral salphingoophorectomy, TAH Total abdominal hysterectomy, TVUS Transvaginal ultrasound

Descriptive statistics using SPSS version 19 were mean, mode, and CI where appropriate.

Definitions

In order to avoid variations in terminology, the following definitions were provided to the survey participants.

Hereditary non-polyposis colorectal cancer (HNPCC)

A patient whose family fulfilled the Amsterdam II criteria.

Amsterdam II

Three or more family members diagnosed with a HNPCC-related cancer (CRC, endometrial, ovarian, stomach, small intestine, hepatobiliary, renal pelvis, ureter), 1 of whom is a first-degree relative of the other two, 2 successive generations are affected, 1 or more of the HNPCC-related cancers is diagnosed in individuals under 50 years of age, and familial adenomatous polyposis is excluded.

Lynch syndrome

A patient with a proven pathogenic mutation in a DNA mismatch repair gene (MLH1, MSH2, MSH6 or PMS2).

Familial colorectal cancer type X

A patients who fulfills Amsterdam I/II criteria, but who has a microsatellite stable cancer [9].

In addition, the following definitions are applied for the purposes of discussion.

Limited colectomy

A standard oncological resection with or without anastomosis is performed to remove the primary malignancy.

Extended colectomy

A limited colectomy plus an additional resection of normal colon or rectum is performed to reduce the risk of future CRC.

Results

One hundred ninety surgeons responded (10 % response rate), and 140 completed the entire survey. Ninety-one percent of respondents indicated an interest in familial CRC. Fifty-one percent had access to hereditary screening clinics.

Preoperative testing

In a patient diagnosed with CRC before the age of 50 and without a family history of CRC, 59 respondents (33.1 %) would proceed to definitive surgery without preoperative DNA testing of tumors. Forty-five (25.3 %) would seek microsatellite instability (MSI) testing on endoscopic biopsies, 18 (10.1 %) would request evaluation for expression of DNA mismatch repair (MMR) protein expression by immunohistochemistry (IHC), and 24 respondents (13.5 %) would obtain both tests.

Comment

Patients who develop colorectal cancer under age 50 should be referred for genetic counseling. If this is not available, a careful family history should be taken. This involves at least 3 generations, noting diagnosis of cancer and causes of death. A family history consistent with HNPCC (Amsterdam criteria) will miss 50 % of MSI tumors, and therefore, the microsatellite status of the tumor should be tested routinely on young CRC patients. Testing for MSI or with IHC on tumor biopsies may delay surgery but can alter surgical planning. Testing for a germline mutation is, in most cases, too time-consuming in the preoperative setting to influence surgical strategy [10].

If a patient under 50 years of age diagnosed with colorectal cancer is from a HNPCC family, only 34 respondents (20.4 %) would immediately proceed to definitive surgery. Sixteen (9.6 %) would test for MSI, 12 (7.2 %) would test for MMR expression, and 24 (14.4 %) would perform both. Only 24 surgeons (14.4 %) would offer germline testing as the initial test. Fifty-seven surgeons (34.1 %) would refer the patient to a genetic counselor.

Comment

Referral for genetic counseling and testing for a germline mutation should be routine. MSI testing can be done quickly, however, and would expedite the diagnosis of Lynch syndrome if the tumor were MSI-high. IHC is a valid alternative [11].

Extent of colectomy

For a right-sided colon cancer in a young patient without a family history of CRC, the majority of respondents (86.0 %) would offer a limited colectomy. If there were a family history of CRC (not meeting the Amsterdam criteria), 76 % would offer a segmental colectomy. If the patient’s family fulfilled the Amsterdam criteria, but preoperative germline testing was normal, then 44.7 % of respondents would still offer a total colectomy and ileorectal anastomosis. If preoperative testing suggests Lynch syndrome, 84.9 % of respondents would recommend a total colectomy and ileorectal anastomosis. In the same situation, 6.3 % would offer a restorative proctocolectomy. The results are summarized in Table 2.
Table 2

Surgical strategy in patients under 50 with colon cancer, depending on family history

 

No family history (%)

Moderate family historya (%)

Familial colorectal cancer type X (%)

Lynch syndrome (%)

RHC

153 (86)

128 (76.6)

68 (42.8)

9 (5.7)

ERC

14 (7.9)

14 (8.4)

18 (11.3)

5 (3.1)

TAC and IRA

10 (5.6)

25 (15)

71 (44.7)

135 (84.9)

RPC and IPAA

1 (0.6)

0 (0)

2 (1.3)

10 (6.3)

Total

178 (100)

167 (100)

159 (100)

159 (100)

RHC Right hemicolectomy, ERC Extended right hemicolectomy, TAC Total Abdominal colectomy, IRA Ileorectal anastomosis, RPC Restorative proctocolectomy, IPAA Ileoanal pouch anal anastomosis

aModerate Family History—Clinical scenario with one first-degree relative (father) with CRC aged 60

Comment

While the top priority of the resection is to cure the presenting cancer, the extent of surgery can vary according to the risk of metachronous cancers and the likely effects of the resection on bowel function, lifestyle, and quality of life. Young patients can tolerate an extended resection and have longer for metachronous lesions to develop. If their family history is Amsterdam-positive, they should be considered for an extended resection [12]. Alternatively, a limited colectomy with long-term yearly or 2-year colonoscopic surveillance can be offered following discussions with the patient. Testing the tumor preoperatively for MSI or MMR gene expression (IHC) makes decision making easier, as an extended resection can be offered for patients with MMR-deficient tumors (MSI-high, lack of expression of an MMR gene).

Long-term colonoscopic postoperative surveillance

In young patients without a family history of CRC, after resection for CRC, 85 surgeons (47.0 %) would offer annual postoperative colonoscopic surveillance. Forty-one respondents (23.0 %) offered endoscopic surveillance every 2 years, while 31 surgeons (17.4 %) offered surveillance every 3 years.

For patients with familial cancer type X, 110 respondents (69.2 %) would offer annual surveillance, 21 respondents (13.2 %) would offer every 2 years, and 16 respondents (10.1 %) would offer surveillance every 3 years. If the affected patient had Lynch syndrome, 116 surgeons (74.8 %) would offer annual endoscopic surveillance, 20 surgeons (12.9 %) would offer endoscopic surveillance every 2 years, and 7 surgeons (4.5 %) would offer endoscopic surveillance every 3 years.

Table 3 shows the long-term surveillance intervals stratified according to family history and germline status in young-onset CRC patients.
Table 3

Long-term surveillance intervals in patients under 50 with colon cancer following a limited colectomy, depending on family history or germline status

 

No family history (%)

Familial colorectal cancer type X (%)

Lynch syndrome (%)

6 monthly

5 (3)

8 (5)

11 (7.1)

Annual

79 (47.3)

110 (69.2)

116 (74.8)

2 yearly

43 (25.7)

21 (13.2)

20 (12.9)

3 yearly

28 (16.8)

16 (10.1)

7 (4.5)

5 yearly

12 (7.2)

4 (2.5)

1 (0.6)

Total

167 (100)

159 (100)

155 (100)

Comment

Resection for colorectal cancer is usually followed a year later by a baseline surveillance colonoscopy. After that, regular surveillance begins. The extremes of surveillance intervals in patients who have had colorectal cancer are every year for Lynch syndrome, to every 5 years for patients with sporadic colorectal cancer and no family history. The exact interval in any particular patient is determined by what is found and by the strength of the family history. Amsterdam-positive, mutation-negative patients should have yearly examinations as they might have Lynch syndrome. If their tumors are microsatellite stable, surveillance can be relaxed to every 2 years as they are type X [13].

Modality of postoperative surveillance

For surveillance of a young patient with Lynch syndrome after standard resection of CRC, 125 surgeons (80.6 %) would perform standard colonoscopy, 13 surgeons (8.4 %) would perform chromo-endoscopy, and 17 surgeons (11 %) would use narrow band imaging colonoscopy.

Comment

Careful standard colonoscopy in a clean colon should protect against death from CRC.

Surgical management after segmental resection and a postoperative diagnosis of Lynch syndrome

When presented with the situation of a young patient who has undergone a segmental colectomy for CRC and is diagnosed postoperatively with Lynch syndrome, 103 respondents (66.5 %) stated they would offer a prophylactic completion colectomy.

Comment

The two options are completion colectomy and annual colonoscopy. In a compliant patient with an easily examined colon and a careful colonoscopist, surveillance is reasonable. There is limited evidence to support a subsequent prophylactic colectomy in this setting, and benefits must be balanced with the risk of an additional major surgery. In a non-compliant patient, or those with a colon that is difficult to examine, completion colectomy could be considered.

Low rectal cancer in a young patient

When presented with a case of low rectal cancer in a patient less than 50 years old with no family history, 59 (39.3 %) surgeons would proceed to definitive surgery without preoperative testing. Twenty-six (17.3 %) surgeons would order MSI testing, 13 (8.7 %) MMR protein expression testing, and 20 (13.3 %) would order both tests. Without a family history of CRC, 143 (95.3 %) surgeons would offer an ultra-low anterior resection for low rectal cancer in a young patient. In a patient with a low rectal cancer and confirmed Lynch syndrome, the majority of surgeons (69.1 %) would perform a restorative proctocolectomy.

Comment

The top priority of treatment is cure for the rectal cancer. This usually involves an anterior proctosigmoidectomy. While the significant risk of metachronous advanced colonic neoplasia in HNPCC patients and Lynch syndrome patients has been highlighted by recent papers by Kalady and Win et al. [14, 15], this is probably not worth the risks of complications and disturbed bowel function that go with proctocolectomy and ileal pouch anal anastomosis, especially with patients around 50 years of age or older [16].

Recognition of MYH-associated polyposis syndrome

Eighty-one surgeons (56.3 %) were able to correctly identify an autosomal recessive mode of CRC inheritance based on family history. Despite this, only 39 surgeons (27.1 %) were able to identify MYH-associated polyposis in a scenario of attenuated polyposis and in a family with an autosomal recessive inheritance pattern. The majority (83.7 %) would offer an abdominal colectomy with ileorectal anastomosis.

Comment

Surgery can be planned according to the number and distribution of polyps at the time of diagnosis. If there are more than 100 synchronous adenomas, or if the adenomas cannot be controlled endoscopically, colectomy and ileorectal anastomosis are the best options.

Serrated polyposis syndrome

Eighty-five surgeons (60.3 %) were able to recognize and name serrated polyposis syndrome.

Attitudes toward synchronous prophylactic hysterectomy

In a scenario of a young woman who has completed her family and who has a right-sided MSH6-deficient colon cancer, 75 surgeons (53.6 %) would proceed to definitive colon resection and synchronous total abdominal hysterectomy with bilateral salphingooophorectomy.

Comment

There are no data to show that surveillance for uterine or ovarian cancer in Lynch syndrome is effective. The best way to prevent cancer is by hysterectomy and oophorectomy. MSH6 was chosen because of the particularly high risk of endometrial cancer associated with loss of this gene [17].

Discussion

Less than 10 % of colorectal cancers are diagnosed in patients younger than 50 years of age [18]. Young age of onset suggests an inherited tendency to cancer. This combination is enough for patients with CRC diagnosed under the age of 50 to be viewed with suspicion. Potential syndromes include those featuring polyposis (familial adenomatous polyposis (FAP), MYH-associated polyposis, serrated polyposis, juvenile polyposis, Peutz–Jeghers polyposis) and those not featuring polyposis (Lynch syndrome and familial colorectal cancer type X). Polyposis syndromes are often obvious because multiple polyps are seen on colonoscopy, but this is not always the case. Sometimes MYH-associated polyposis does not present with unusually large numbers of polyps, and sometimes f FAP patients also present with few adenomas [19]. An algorithm for the investigation into and management of a young patient with CRC is presented in Fig. 1. While there are no 100 % “right” answers, the survey revealed a variety of practice patterns, including a significant deviation from ASCRS Practice Parameters recommendations [20, 21].
https://static-content.springer.com/image/art%3A10.1007%2Fs10151-013-1052-5/MediaObjects/10151_2013_1052_Fig1_HTML.gif
Fig. 1

Algorithm for investigation into and management of a young patient with colorectal cancer. *If an abnormal MSI or IHC testing, then BRAF testing or methylation testing can be used to direct treatment. A BRAF V600 E mutation would suggest the cancer is occurring due to a sporadic mutation and not in a Lynch syndrome patient and therefore a segmental colectomy can be performed. If BRAF is wild type (WT) in a patient with a HNPCC family, this would indicate the patient has Lynch Syndrome. If BRAF is WT in a patient without a family history then immediate germline testing can be instituted prior to colectomy

The study shows that while colorectal surgeons in North America are willing to consider preoperative evaluation in a young patient, one-third of respondents will not order any molecular or genetic testing. The result is concerning given that 90 % of respondents indicated an interest in hereditary CRC, and is likely to be larger with an increasing response rate. A more extensive family history, specifically in a family that meets Amsterdam criteria, triggers an increase in the percentage of respondents who will pursue preoperative evaluation [22, 23]. Among those that utilize preoperative testing, there is no uniform approach as MSI and MMR protein testing were selected in equal numbers.

Surgical decision making was altered by the suspicion or presence of Lynch syndrome in this study. In a young patient with presumed sporadic young-onset colorectal cancer, 93 % of respondents would perform a limited colectomy, but this was altered by the diagnosis of Lynch syndrome with 84.9 % proposing a total abdominal colectomy and ileorectal anastomosis. A small proportion of the members would offer a restorative proctocolectomy and ileoanal pouch reconstruction in this population group; however, the authors consider this approach excessive. While such a strategy would remove all the at-risk colonic mucosa, this fear remains theoretical in patients with Lynch syndrome. In a large multicentre study of Lynch syndrome patients conducted by Parry and colleagues, no metachronous rectal cancers were observed over 414 person years of follow-up in 50 patients following total colectomy with ileorectal anastomosis [24].

In a young patient with a low rectal cancer, the majority of respondents were willing to offer a sphincter-preserving procedure. Interestingly, if the patient had HNPCC but tested germline negative, then 47 surgeons (31.5 %) would offer a restorative proctocolectomy (RPC) compared with 103 surgeons (69.1 %) if the patient had a confirmed mutation. Kalady et al. reported on 31 primary HNPCC rectal cancers with appropriate endoscopic follow-up. The authors found 42 high-risk adenomas in 9 patients (29 %) and 5 metachronous colon cancers, at a median of 6 years following initial proctectomy [14]. These findings are supported by a recent publication by Win and colleagues who calculated the cumulative risk of colon cancer to be 69 % at 30 years following index proctectomy in 79 Lynch syndrome carriers [15]. Again the additional functional compromise and risks of surgery may outweigh the benefits of surgery. Certainly where restoration is not possible, proctocolectomy and end ileostomy should be considered for a young Lynch syndrome patient.

The majority of respondents (87.9 %) were willing to offer risk-reducing gynecological surgery at the time of colonic resection in a female beyond childbearing age if the patient had a MSH6-deficient tumor. This approach is supported by retrospective data, indicating that prophylactic hysterectomy can prevent endometrial cancer and ovarian cancer [25]. However, our survey results indicate that while a majority of respondents were willing to address this at the time of definitive colectomy, when questioned on preoperative testing in young patients with and without a strong family history, 33 and 20 % of respondents, respectively, would progress straight to definitive colectomy and therefore miss the ability to preselect such patients. These results further emphasize the importance of preoperative testing and family cancer referrals when a diagnosis of CRC is made as clearly surgical planning is altered.

MYH was poorly identified by respondents in the survey. Affected patients mostly express an attenuated colonic polyposis phenotype [26], and unlike other hereditary colorectal cancers, the syndrome is inherited in an autosomal recessive pattern. This trend was also noted in a survey of the Australasian colorectal surgical society [27]. However, despite not recognizing the syndrome, most surgeons would offer a total abdominal colectomy given the phenotype presented. Thirteen percent would over-treat with a proctocolectomy perhaps extrapolating from FAP.

Despite the distribution through the ASCRS, the response rate was low (190 respondents out of 1,900 members). This may be due to the complexity of the survey, and the nature and relative rarity of its subject matter. These respondents are likely the most informed about hereditary CRC, but the low uptake may indicate a lack of interest in familial CRC on the part of non-responders. On the contrary, the majority of respondents indicated an interest in familial CRC. Hence, with the wide spectrum of practice demonstrated, this would likely be wider with a higher response rate. Many clinical scenarios are managed quite adequately by one of a number of options, and the option of choice is a matter of opinion. Furthermore, each patient and each family is subtly different from the others. It is impossible to include such subtleties in a survey questionnaire such this.

Conclusions

Colorectal surgeons with an interest in hereditary colorectal cancer are inclined to perform preoperative molecular and genetic testing in young colorectal cancer patients to diagnose a hereditary syndrome and will alter management based on the results. Even in this motivated group, however, there is considerable variety of practice patterns and some deviation from guidelines.

Conflict of interest

None.

Copyright information

© Springer-Verlag Italia 2013