Current Treatment Options in Gastroenterology

, Volume 12, Issue 3, pp 269–282

Colorectal Cancer of the Elderly


    • Division of Gastroenterology, Department of MedicineSan Francisco General Hospital and Trauma Center
  • Fernando Velayos
    • Division of Gastroenterology, Department of MedicineUniversity of California, San Francisco
Geriatrics (S Katz, Section Editor)

DOI: 10.1007/s11938-014-0021-z

Cite this article as:
Day, L.W. & Velayos, F. Curr Treat Options Gastro (2014) 12: 269. doi:10.1007/s11938-014-0021-z

Opinion statement

Colorectal cancer (CRC) disproportionately affects the elderly. Older age is a strong risk factor for both the development of precancerous adenomas and CRC, thus raising the issue of screening and surveillance in older patients. However, screening and surveillance decisions in the elderly can be complex and challenging. Elderly patients are a diverse and heterogeneous group and special considerations such as co-morbid medical conditions, functional status, and cognitive ability play a role in one’s decisions regarding the utility of screening and surveillance. Such considerations also play a role in factors related to screening modalities, such as colonoscopy, as well as CRC treatment options and regimens. This review addresses many of the unique factors associated with CRC of the elderly and critically examines many of the controversies and challenges surrounding CRC in older patients.


Colorectal cancerColonoscopyScreeningSurveillanceElderlyVery elderlyAdverse eventsEndoscopyQuality


Colorectal cancer (CRC) is the third leading cancer diagnosed and cause of cancer-related deaths in the United States. In 2013, 142,820 people will be diagnosed with and 50,830 people will die of CRC [1]. Age is an important risk factor for developing CRC [2] and the elderly are disproportionately affected by this disease, thereby necessitating the need for screening and surveillance in this group. However, screening and surveillance decisions in the elderly can be challenging, as elderly patients are a heterogeneous group with some patients having a long life expectancy while others have multiple co-morbid medical conditions and impaired functional status.

This review addresses a number of the challenges and controversies surrounding CRC and the elderly. First, we review the epidemiology and clinical presentation of CRC in the elderly and how it differs from younger patients. Second, we review the efficacy of screening modalities and examine the data regarding when not to screen a patient based on age. Moreover, we discuss colonoscopy and how various aspects of this procedure are impacted by age. Finally, we discuss CRC treatment and how age influences the efficacy, safety, and decision to treat elderly patients with CRC.


CRC is common in the elderly. The median age of CRC diagnosis in the United States is 69 years with nearly 90 % of new CRCs diagnosed in patients over the age of 50 years (Fig. 1) [1]. The incidence of CRC dramatically rises as one ages (Fig. 2). For example, the incidence of colon cancer in a person below the age of 65 years is low at 11.4/100,000 persons, but exponentially increases to 176.1/100,000 persons in people aged over 65 years. This trend is observed between both sexes and all racial backgrounds with respect to age with the same observations also noted with respect to rectal cancer incidence [1].
Fig. 1

Distribution of patients with colorectal cancer stratified by age.
Fig. 2

Incidence rates for colon and rectal cancer for all races and both sexes from 2006 to 2010 [1]. SEER = Surveillance Epidemiology and End Results.

One precursor to the development of CRC is colonic polyps [3]. Precancerous adenomatous polyps and advanced adenomatous polyps (defined as polyp size >10 mm, villous/tubulovillous histologic features, or having high-grade dysplasia) both have an increased prevalence and incidence in the elderly [46]. In fact, adenoma and advanced adenoma prevalence in persons aged 70–75 years is more than double that of persons aged 40–49 years [4, 7, 8]. In contrast, the prevalence of serrated lesions only marginally increases with age [913]. Yet, it should be noted that studies examining age and serrated lesions occurred prior to the World Health Organization’s reclassification of serrated lesions into three distinct categories (hyperplastic, sessile serrated adenoma/polyp, and traditional serrated adenoma) and the true risk of age and these individual categories may vary, but to date no study has shed light on this topic. Additionally, age plays a role with respect to a number of other factors related to polyps. Older men have a greater prevalence of adenomatous polyps than older women. Additionally, larger sized polyps and more proximal colonic polyps are found in older patients [14].

The recurrence of adenomas, advanced adenomas, and serrated lesions is less influenced by age (Table 1). Age does not appear to impact the recurrence of adenomas after an initial diagnosis during colonoscopy [1517]. Other factors such as index polyp size (polyp > 1 centimeter) [16], number of polyps at index colonoscopy [18], or initial incomplete polyp resection [19] are more strongly associated with recurrence. With respect to advanced adenomas, similar predictors of recurrence are present as those found for adenomas, but age may play a slightly greater role [20, 21]. In addition to polyp characteristics (polyp size, number of polyps, and histology), a number of colonoscopy-related characteristics are associated with detecting advanced colorectal neoplasia at surveillance colonoscopy [18]. Specifically, insufficient bowel preparation and incomplete examination both predict the recurrence of advanced neoplasia detected on future surveillance colonoscopies. Few data exist on the topic of serrated lesion recurrence and age, but limited studies suggest that age does not have an impact on recurrence [22] and consensus guidelines on surveillance intervals for serrated lesions focus more on the size, number, and location of the index serrated lesion(s) [23].
Table 1

Factors that increase the recurrence of adenomas, advanced adenomas, and colorectal cancer


 Index polyp size (polyp >1 cm)

 Number of index polyps

 Incomplete polypectomy

Advanced adenomas

 Number of index adenomas

 Index polyp size (polyp >1 cm)

 Villous histology on pathology

 Insufficient bowel preparation

 Incomplete examination (unable to reach farther than the distal colon)

Colorectal cancer

 Family history of colorectal cancer

 Presence of extra-colonic malignancy

 Detection of synchronous lesions

 Coexisting adenomas

 Perforation at time of diagnosis


Along the same lines, while CRC incidence increases with age, the recurrence of it is not influenced by age. Instead a number of other factors, such as family history, findings on index colonoscopy [24, 25], and presenting symptoms [26], have a stronger association (Table 1).

Given that age does not strongly influence the recurrence of either colonic polyps or CRC, surveillance guidelines have not been tailored by age.


CRC presentation is similar in younger and older patients although more proximal cancer is detected in older patients [6]. Likewise, older patients may be less likely to present asymptomatically [27, 28]. Elderly patients with CRC may have symptoms that include occult blood loss, rectal bleeding, change in stool caliber, weight loss, or they may have signs of bowel obstruction or perforation. While no one presenting symptom predominates in elderly patients, it should be recognized that elderly patients may have a more subtle presentation of CRC such as vague abdominal pain or a new microcytic anemia. Such symptoms cannot be attributed to other etiologies and deserve a thorough cancer evaluation in the elderly.

Screening modalities for colorectal cancer (CRC)

CRC screening can detect both precancerous polyps and CRC, which can reduce both the incidence and deaths related to CRC [2939]. While no one screening method is preferred in the United States, a number of consensus documents offer several screening recommendations including those from the US Preventative Services Task Force [40], the American Cancer Society, the US Multi-Society Taskforce with the American College of Radiology [41], and the American College of Gastroenterology [42]. Screening tests available and recommended include examining stool for occult blood, radiologic and endoscopic methods (colonoscopy and flexible sigmoidoscopy). To date, no one screening test has proven superior for CRC screening in the elderly.

Colonoscopy in elderly patients

Rates of CRC screening in the United States have been steadily rising over the last 20 years, in large part to the increases in the use of colonoscopy for screening [43]. Furthermore, lower endoscopy, specifically colonoscopy, has been shown to significantly reduce both distal and proximal colorectal in older individuals [44]. As more elderly patients undergo CRC screening using colonoscopy, one has to consider how specific factors related to this procedure and age may interact. In particular, earlier recognition by the endoscopist of multiple factors such as a patient’s co-morbid medical conditions, cognitive function, mobility, and polypharmacy needs to occur prior to performing endoscopy in elderly patients.

Several reviews [4547] and a consensus guideline from the American Society of Gastrointestinal Endoscopy [48] have addressed multiple issues with regard to the elderly and endoscopy. Very few changes are recommended for elderly patients with respect to the pre-procedure process and sedation. However, the endoscopist should be aware of two additional recommendations to consider when performing endoscopy on an elderly patient. First, it is recommended that providers assess for an elderly patient’s cognitive ability and capacity to understand the procedure and fully document functional status and depression screening during the pre-procedure assessment [48, 49]. Second, providers should use fewer sedative medications at lower doses and medications should be infused at slower rates when sedating an elderly patient [48].

Adverse events

The occurrence of adverse events during a colonoscopy and how age may modify this risk are important considerations for the elderly patient. Age does not play a role in minor adverse events [50]. However, major adverse events such as perforation, bleeding, and cardiopulmonary adverse events are all affected by age, though the individual risk varies and can be influenced by additional factors. Of all adverse events associated with colonoscopy, the greatest risk associated with age is perforation. Patients aged >65 years have a 30 % higher risk than younger patients undergoing colonoscopy and a 14-fold higher risk than patients of the same age who do not undergo the procedure, with the risk even higher in patients over 80 years [51]. Furthermore, older patients with more co-morbid medical conditions have a greater risk of experiencing an adverse event if the colonoscopy is performed with general anesthesia assistance [52].

Bowel preparation

Bowel preparation is a significant issue to consider in patients undergoing colonoscopy as poorer bowel preparations can lead to missed polyps and cancers. Two agents are available (polyethylene glycol electrolyte lavage solution (PEG) and oral sodium phosphate (OSP)). A number of adverse events are observed in elderly patients taking PEG and OSPs (Table 2). PEG has a much better safety profile compared with OSPs, with a major concern that elderly patients taking OSPs are at significantly greater risk of having electrolyte disturbances [53, 54] and developing acute kidney injury [14, 55]. It is for this reason that OSPs are no longer recommended in the elderly and that PEG is the preferred bowel preparation agent.
Table 2

Adverse events associated with bowel preparation in elderly patients

Polyethylene glycol (%)

Oral sodium phosphate (%)

Dizziness (48)

Hyperphosphatemia (58.1100)

Fecal incontinence (27–39)

Fecal incontinence (2355)

Abdominal pain (7–23)

Elevated creatinine/renal injury (55.2)

Nausea (2–17.5)

Hypocalcemia (5.158)

Insomnia (13)

Hypokalemia (5.456)

Fatigue (12.7)

Abdominal pain (1132)

Headache (7.9)

Nausea (936)

Hypokalemia (2.9–20.5)

Insomnia (15)

Dysnatremia [hyponatremia/hypernatremia] (4.1)

Dizziness (355)

Emesis (3.2)

Emesis (47)

Aspiration pneumonia (<1)

Hypotension (4)

Pancreatitis (<1)


Ischemic colitis (<1)


Non-adherence to PEG-based regimens in the elderly is reported to be as high as 32 % [56]. While spilt PEG dosing regimens have been shown to be equally effective [57] versus standard one-time dosing, this approach has not been studied in the elderly. Of critical importance is that elderly patients remain adequately hydrated when taking PEG [58].

Poor bowel preparations in the elderly undergoing colonoscopy are in the range of 4–57 % [28, 5967] with bowel preparation being more difficult to achieve in very elderly patients (aged >80 years) regardless if they are compliant with or of the type of bowel preparation used [61, 65]. Poorer bowel preparations observed in the elderly can occur for a variety of reasons including altered gastrointestinal motility, increased rates of medication-related constipation, previous surgeries, decreased understanding of bowel preparation instructions, greater burden of co-morbid medical conditions, and/or functional limitations.

Completion of colonoscopy

Key to a high-quality colonoscopy is the successful completion of it (i.e., intubation of the cecum). Endoscopists subjectively judge a colonoscopy to be more difficult in an elderly patient [68], but completion rates vary from 78 to 86 % in the elderly and from 52 to 95 % [59, 60, 69] in the very elderly [28, 6163, 6567]. Age may be an independent risk factor for lower completion rates, but it is apparent that other factors such as poor bowel preparation and a patient’s underlying disease process play a stronger role [59, 70].

Decisions regarding not to screen for CRC in the elderly

One of the issues that have to be addressed when discussing the elderly, CRC, and screening is the role of co-morbid medical conditions. Elderly patients have a greater number and severity of co-morbid medical conditions with over one-quarter of patients aged >65 years having more than five co-morbid medical conditions [71]. Such an increase in co-morbid medical conditions can play a role in developing CRC as well as reducing the benefit from screening and treatment. It is clear that the benefit of screening is reduced with increasing disease burden. For example, Ko et al. showed that the greatest number needed to screen to prevent a CRC death was in older more ill patients and that screening-related adverse events were also greater than the benefit in this same group of patients [72]. Additionally, adverse events related to colonoscopy (including screening) increase as the number of co-morbid medical conditions also increases [73]. Furthermore, patients with greater co-morbid medical conditions have lower survival rates after an initial diagnosis of CRC [7476], poorer survival after chemotherapy [76, 77], and more prolonged hospitalizations as a consequence of their CRC [78].

While detecting CRC earlier is beneficial, the question that remains is: does screening extend life in older patients who may have a shorter life expectancy? This raises the question at what point does CRC screening cease to provide an important potential extension in life expectancy and therefore not be offered? Many have attempted to answer this question addressing it from two perspectives: 1) gain in life expectancy through screening and 2) at which age does screening not provide additional benefit to the patient.

On the first question, modeling studies have examined the impact of various colorectal screening methods on life expectancy at different age groups [7981]. Younger patients had a much more significant decrease in life expectancy than did elderly patients after a diagnosis of CRC. Simultaneously, there was a nearly 75 % reduction in the benefit of screening for elderly patients when compared with younger patients regardless of which screening modality was used. Similar results have been noted in clinical studies addressing this question. The benefit of screening, in particular colonoscopy, after an initial negative screening test is greatly reduced in elderly patients as older patients will succumb to other illnesses (cardiovascular, pulmonary, extra-colonic malignancies) rather than CRC itself [74]. Along the same lines, in elderly patients who have had a positive fecal occult blood test, nearly half of the patients who did not undergo colonoscopy died from other causes. Of those who underwent colonoscopy, 10 % experienced an adverse event and 86 % of patients with worse life expectancy did not benefit from screening [82].

With regard to the second question, multiple studies have examined various ages at which CRC screening should be discontinued. A number of ages at which to stop screening have been advocated. Some have shown that decreasing the screening age from 85 to 75 years yields small reductions in life-years gained and uses fewer resources, and stopping screening at age 75 years provides almost the same benefit as stopping at age 85 years [83]. Others have demonstrated diminishing returns for days of life lost after the age of 70 for any form of screening [84] and that after the age of 60 years the percentage of life-years saved declines precipitously after a single colonoscopy [80]. Additionally, there does appear to be an age in which some patients achieve no benefit from screening. Men aged >85 years and women aged >90 years do not achieve any benefit from CRC screening regardless of modality [72], and patients over 80 years have a shortened life expectancy (median survival of <5 years) after a diagnosis of CRC regardless of co-morbid medical conditions or functional status [74]. Unfortunately, there is very little guidance on when to stop CRC screening from national medical societies (Table 3).
Table 3

Colorectal cancer screening guidelines and the elderly



US Multi-Society Task Force and the American Cancer Society

In those with a prior polyp: discontinuation of surveillance colonoscopy should be considered in persons with serious co-morbidities and with less than 10 years of life expectancy

American Gastroenterological Association

No comment on when to stop screening

Comment on need for shared decision making and an individualized approach

American Geriatrics Society

Not recommended in those unlikely to live more than 5 years or who have significant co-morbidity that would preclude treatment

British Society of Gastroenterology

FOBT every 2 years offered to all persons aged 5069 years (depending on location) with current plans to extend to age 75 years in most areas

Kaiser Permanent Care Management Institute

Discontinuation of screening is generally recommended at age 75 years, provided that there is a history of routine screening

Discontinuation is recommended at age 80 years for those with no history of routine screening. The decision to discontinue screening should be based on physician judgment, patient preference, the increased risk of complications in older adults, and existing co-morbidities

FOBT fecal occult blood test

Thus, health, life expectancy, and functional status as well as age should all play a role when considering the option to discontinue screening. However, this practice is quite variable. Some providers take an individualized approach with their elderly patients and CRC screening with more physicians choosing to engage healthier elderly patients [85]. Whereas a larger body of evidence suggests that others perform inappropriate CRC screening in patients with severe co-morbid medical conditions and low life expectancies [86, 87]. Such disparities in practice highlight the need for greater education among providers on CRC screening in the elderly.

Special population screening: inflammatory bowel disease

The risk of CRC increases with age, but whether the risk is higher in patients with inflammatory bowel disease (IBD) compared with an age-matched cohort without IBD has not been well studied. A recent systematic review highlighted an important variation in the risk reported in studies [88]. Notably, a recent prospective cohort from France demonstrated no increased risk of developing CRC among patients with elderly-onset IBD (Crohn disease (CD), standardized incidence ratio (SIR) = 1.2, 95 % confidence interval [CI] 0.5–2,4; ulcerative colitis (UC), SIR = 1.0, 95 % CI 0.5–1.9) [89]. This variation highlights the importance of clearly defining whether an elderly patient with IBD has had long-standing disease or late-onset disease, where the risk of CRC in the latter case would be expected to be similar to the background population given the more recent onset of inflammation.

With regard to performance of the colonoscopy itself, the miss rate appears to be increased in elderly IBD patients compared with age-matched controls without IBD. A study using the Surveillance, Epidemiology and End-Results Medicare-linked database identified patients who were aged 67 years or older at colonoscopy during 1998–2005 and those who were subsequently diagnosed with CRC within 36 months. Compared with older non-IBD patients, early/missed CRC was three times higher for IBD patients (CD, odds ratio [OR] = 3.1, 95 % CI 2.2–4.2; UC, OR = 3.1, 95 % CI 2.4–3.8) [90].

Although there are no specific guidelines for CRC screening in elderly patients with IBD, the available data suggest that early missed CRC is higher than in an age-matched cohort without IBD, and that elderly patients with long-standing disease are likely the higher risk group than those with elderly-onset IBD. There is active discussion regarding whether the elderly with IBD should be placed in a special category of risk.


CRC is more frequently complicated in the elderly at diagnosis [91] with the need for more emergency surgery performed in this group [92]. Treatment for CRC (surgical and medical) in the elderly differs in comparison with younger patients. Elderly patients with CRC less frequently are discussed at multidisciplinary cancer team meetings [92, 93], undergo less surgery [92, 93], have less adjuvant chemotherapy prescribed [91, 93, 94], and undergo less palliative therapy [91]. These data do not suggest this difference is appropriate or inappropriate, but rather that age impacts treatment for CRC.

Age appears to be a risk factor in the outcome of CRC surgery with older patients experiencing higher 6-month mortality rates [95]. Much debate has centered on open versus laparoscopic colectomy as the best surgical approach for elderly patients, with a number of recent studies shedding more light on this subject. Laparoscopic CRC surgery in the elderly appears to be an effective procedure compared with an open approach with equal cure rates between the two, but less adverse events [96], less blood loss [9698], and shorter postoperative hospitalizations [9698] noted with laparoscopic colectomy.

With respect to the use of chemotherapy, reviews have addressed specific treatment regimens to consider in an elderly patient [99]. Chemotherapy data on CRC and the elderly are limited as advanced age is an exclusion factor in many clinical trials [99102]. Many oncologists reduce the dose of chemotherapy for perceived risks of impaired liver or renal function in older patients [103]. Nevertheless, most chemotherapeutic agents are safe in the elderly and can achieve the same therapeutic benefits of reduced recurrence and mortality as observed in younger patients, especially in the adjuvant setting [104106]. Given these data, most advocate that age alone should not factor into one’s decision to treat CRC [102, 104, 107]. In most instances, co-morbid medical conditions play a larger role in the decision to proceed with chemotherapy, as these have a more pronounced impact on life expectancy after CRC diagnosis and treatment [76].

Tools, such as the Comprehensive Geriatric Assessment, have been developed to help evaluate elderly patients being considered for CRC treatment (for both surgery and chemotherapy) and their use has proven effective in reducing adverse outcomes among the elderly [108, 109]. While elderly patients represent over 50 % of metastatic CRC, their gain in survival over the last 25 years has been significantly less compared with younger patients, highlighting the need for better palliative therapies in this group [110]. Oncologists advocate a multidisciplinary approach in treating elderly patients with CRC. A patient’s functional status, co-morbid medical conditions, goals, preferences, and cancer stage all have to be factored into the decision of determining what regimen (if any) to use when treating the elderly patient with CRC [99, 102].


A disproportionate number of CRC diagnoses and deaths occur in elderly patients. Advancing age is an independent risk factor associated with both CRC and adenomas, whereas the detection of recurrent CRC and adenomas after a screening colonoscopy is unaffected by age. While a number of CRC screening modalities are available for elderly patients, it is important to remember that at specific ages the risks and increased resources may outweigh the benefit of screening in some elderly patients. Additionally, older patients with a greater number and severity of co-morbid medical conditions derive less benefit from screening. Subsequently, controversy exists over when to discontinue CRC screening and surveillance in the elderly. Yet, most agree that the decision to proceed with screening and surveillance requires an individualized assessment of the elderly patient that takes into account the risks and benefits and balances this with the patients’ health, functional status, and preferences. Some aspects of screening, such as with colonoscopy, need to be factored into this decision-making process. Older patients have a higher risk of adverse events during a colonoscopy, poorer bowel preparations, and possibly lower successful completion rates. Last, a variety of CRC treatments are available for elderly patients and, just as with decisions about when to discontinue screening, a patient’s functional status, co-morbid medical conditions, and preferences have to be factored into the decision-making process when determining which treatment regimen to offer.

Compliance with Ethics Guidelines

Conflict of Interest

Lukejohn W. Day and Fernando Velayos declare that they have no conflicts of interest.

Human and Animal Rights and Informed Consent

This article does not contain any studies with human or animal subjects performed by any of the authors.

Copyright information

© Springer Science+Business Media, LLC 2014