Global graying, colorectal cancer and liver metastasis: New implications for surgical management

1. Introduction 

The demographic transition seen in the last two centuries has resulted in a net increase in the number and proportion of elderly worldwide, a phenomenon referred to as “global graying” [1]. The total number of new cancer and cancer-related mortality cases is steadily increasing, despite a drop in cancer incidence and mortality rates [2]. This paradoxical phenomenon is a consequence of the effects of global graying. It is most pronounced in industrialized countries, where the elderly population is expected to continue growing from an estimated 14% in 2000 to greater than 20% by 2025 [3].

Colorectal cancer (CRC) is the third most common type of cancer and cause of cancer-related deaths in the United States in both men and women, with an estimated 145,000 new cases in 2008 [4]. Incidence increases with age, with more than 70% of affected patients over the age of 65 at time of diagnosis [2]. Approximately 50% develop liver metastasis (LM) at some point during their disease [5]. The management of CRC LM has undergone significant changes over the last decade. Optimal care for these patients is complex and poses multiple challenges in the elderly population. This review covers the effects of global graying on CRC and LM epidemiology. In addition, it highlights challenges specific to the surgical management of elderly patients and identifies areas requiring further study.

2. Population dynamics and changes in age structure 

Increased life expectancy, followed by a subsequent drop in birth and fertility rates, has led to a growing aging population, or so-called global graying [1]. The transient increase in fertility rates seen in the post-World-War II period (the generation of baby-boomers) has accentuated this trend, resulting in an increase in both the number and the proportion of elders (age ≥65 years) within the general population [6]. Furthermore, the processes leading to increased life expectancy and subsequent decreased fertility rates are unlikely to be reversed over time; hence, this trend is likely to be irreversible in the short term.

In developed countries, the median age of the population will continue to increase such that the elderly (≥65 years) will constitute greater than 20% of the population by 2025 [3], with an expected tripling in this segment within the next 50 years [7]. This demographic transition is driven by a 2.6% growth rate per year among the elderly compared with 1.3% for the general population and the proportional reduction of children in the general population. Among the elderly, the very old (i.e., those ≥80 years) are increasing at the fastest rate (3.9%), with the expectation that they will represent 20% of the total elderly population within the next 40 years [7]. In the United States the number of people ≥65 years is projected to double from 35 million in 2000 to over 70 million by 2030. Additionally, the number of people ≥80 years is expected to more than double in the same timeframe to almost 20 million; 30% of the projected elderly population by 2030 [6].

3. Implications of global graying on cancer and colorectal cancer epidemiology 

The effect of global graying on cancer is evolving [6]. Annual cancer incidence rates (per 100,000 people) increase with age, from 9.1 in those ≤50 years to over 400 in those ≥65 years [2]. Additionally, as a consequence of global graying, the total number of new cancer cases projected to 2050 is expected to double to 2.6 million [2]. Among the elderly, cancer incidence is expected to increase by 67% in the next 20 years, contributing with more than 70% of new cases by 2030 [8]. Eighty per cent of cancer-related deaths occur in those over 60 years; and cancer is now the leading cause of death among patients between 60 and 79 years, surpassing heart disease [4].

CRC is the third most common type of cancer and cause of cancer-related deaths in the United States [4]. Population-level studies have highlighted the high prevalence of CRC in the elderly. Its incidence is directly associated with increasing age, with most cases affecting the elderly; 71% of new cases occur in patients ≥65 years, and 42% occur in those ≥75 years [2]. Further, based on the Surveillance, Epidemiology and End Results (SEER) and US Census Bureau data, an estimated increase in CRC incidence of 52% is projected for the next 20 years, with over 250,000 new cases by 2030, a trend partially driven by global graying [8].

4. CRC in the elderly: implications for management 

The management and prognosis of CRC are primarily determined by stage at diagnosis. Surgery is the only curative treatment, but a strategy combining resection with specific adjuvant and neoadjuvant therapies is essential for improving outcomes in selected patients. Those with localized diseases (stages I and II) are treated with colon resection and en-bloc regional lymphadenectomy and have a 5-year overall survival close to 90%. For those presenting with regional disease (lymph-node involvement—stage III), as well as a selected group of patients with stage II disease, standard therapy includes surgery, followed by adjuvant systemic chemotherapy. This treatment strategy is associated with a 5-year overall survival over 60% [4], [9], [10]. For patients with stages II and III rectal cancer, standard treatment includes a combination of pre- or postoperative radiation therapy, chemotherapy and surgery. Their prognosis varies by pathologic stage, surgical characteristics and other tumor features [11].

Although 40% of CRC cases are localized, 36% and 19% of patients present with regional and metastatic disease, respectively. One-third of CRCs are located in the rectum, with the remaining two-thirds being distributed throughout the colon [4]. Based on this distribution and outlined management recommendations, a multidisciplinary approach for treatment is essential and carries improved outcomes [9], [10], [11]. Adequate evidence is necessary to guide indications, timing, and treatment duration; and this is particularly relevant to the elderly, in whom CRC has a similar anatomic distribution, generally more advanced stage at presentation and worse overall outcomes [12], [13], [14].

Increasing age has been repeatedly demonstrated to be a poor prognostic factor in CRC [12], [13], [15], [16], [17], [18], [19]. Studies evaluating long-term (oncologic) and short-term (perioperative) outcomes in younger and older adults have confirmed this finding [12], [13], [20], [21], [22]. Although these differences could be attributed in part to cancer biology and physiologic function specific to the elderly [23], [24], [25], explicit process-of-care characteristics have been shown to drive these outcome differences. Specifically, treatment disparities related to diagnosis, surgical care, and adjuvant and neoadjuvant therapies have been identified when comparing younger and older patients [12], [13], [20], [26], [27]. Inaccurate perceptions of poor outcomes, based on age alone, have been postulated as critical in the medical decision-making process leading to these differences [26], [27], [28], [29].

Specific disparities in diagnosis and surgical care in elderly CRC patients are well recognized. In a systematic review with a population exceeding 35,000 patients, suboptimal management in the elderly was reported at various steps during the process of care, including later diagnosis resulting in more advanced disease at presentation; lower rate of surgical treatment; and lower rate of curative resections [12]. Similar care patterns have been identified in elderly patients with rectal cancer, including less cancer-directed therapy and lower rates of radical surgery [13]. These differences in surgical care are in part due to concerns regarding physical and mental frailty and increased co-morbidities in elderly patients [28]. However, chronological age alone is not a reliable predictor of postoperative outcome [30]. Surgical care to the elderly can be delivered within safe standards when following outlined guidelines [31], [32]. Further, when adjusting for treatment disparities, short- and long-term outcomes in well-selected elderly patients are equivalent to those seen in the younger population [12], [33]. The critical question is which patients are likely to benefit from these interventions. Preoperative geriatric assessment can help stratify elderly patients, using premorbid health and functional condition as a basis [34], [35], [36], [37], [38]. The preoperative assessment of cancer in the elderly (PACE) [39] is a multidimensional assessment tool used to evaluate overall health and function in elderly cancer patients. A recent multi-institutional cohort study validated its ability to identify elderly cancer patients at increased risk of significant complications after surgical treatment [40]. Specifically, elderly cancer patients with preoperative instrumental activities of daily living score <8, abnormal performance status (defined as >1), or a moderate/severe brief fatigue inventory (defined as 4–10) are associated with an increased risk of postoperative complications [40]. On the basis of these data and the above considerations, adequately selected elderly CRC patients should be offered surgical treatment; and age alone should not be a factor driving eligibility for this curative strategy [41].

Despite a National Institutes of Health Consensus Conference recommending the use of adjuvant systemic chemotherapy for stage III colon and rectal cancer [42], adjuvant and neoadjuvant therapies are underused in eligible elderly patients [26]. A population-level study of over 85,000 stage III colon cancer patients revealed a decline in the use of adjuvant chemotherapy associated with increasing age [43]. Only 39% of patients >80 years received chemotherapy, despite equivalent overall survival in those receiving postoperative chemotherapy when compared with the younger population [43]. Concerns regarding increased toxicity and questioned efficacy of systemic therapy in the elderly can partially explain chemotherapy underutilization. However, data derived from pooled analysis and population-based studies support the use of adjuvant systemic therapy in the elderly (primarily those <75 years) [43], [44], [45]; and clinical trials have shown a similar toxicity profile to that in younger patients, despite specific management modifications [46], [47], [48]. Thus, adjuvant systemic chemotherapy is recommended for well-selected elderly patients [41].

A decline in appropriate use of chemotherapy and radiation treatment for rectal cancer among the elderly has also been noted. Among 21,390 patients with rectal cancer in the SEER database, a 36% reduction in radiation therapy was associated with increasing age [13]. When adherence to recommended guidelines for the management of CRC in the Veterans Affairs population was examined, advancing age was independently associated with lower rate of recommended therapy for rectal cancer [26]. This study highlights the importance of physicians’ perceptions regarding medical fitness and candidacy for adjuvant therapy in the decision-making process for use of chemotherapy and found this to be the most common reason for deviations from recommended management strategies in elderly patients [26].

5. CRC liver metastasis and the elderly: time to change the surgical paradigm? 

It is estimated that 30–50% of patients diagnosed with CRC will develop LM, commonly presenting as isolated liver disease in 76% of cases [5], [49], [50]. Synchronous LM occurs in 40–50% of patients, and the remaining present as metachronous lesions within the first 2 years post-CRC diagnosis [5], [50], [51]. A French registry noted a temporal increase in synchronous LM, from 12% to 17% over a 20-year period, possibly an effect of population growth and global graying [50]. However, the effect of global graying on CRC LM epidemiology has not been formally studied. Nonetheless, based on the increasing number of new CRC cases and the growth of the elderly population, the number of LM cases in older patients is expected to increase in coming years. The current management of CRC LM is complex and rapidly evolving and carries multiple challenges when caring for the elderly population.

The management strategy for CRC LM is critical to improve overall survival and maximize cure rates. Best supportive care and fluorouracil-based systemic chemotherapy are associated with 5-year overall survival <5% and median survival of 6–12 months [49], [50], [51]. Newer chemotherapy regimens with better cytotoxic agents, used alone or in combination with biologic targeted therapy, improve median overall survival up to 24 months; but 5-year overall survival remains below 10% [5], [10], [41], [51]. Liver resection offers the only curative option and is feasible in 10–20% of patients [49], [50], [52]. Postoperative mortality following liver resection is below 5% [51], [53], [54], and 5-year overall survival for these patients is over 30% [51], [54]. Nonetheless, surgical therapy is seldom offered to elderly patients with LM. Provider perception of high postoperative mortality and lack of oncologic benefit may contribute to this pattern [55].

Age is often used as a criterion when evaluating candidates for liver resection; among patients in a French cancer registry, age >75 years was the strongest predictor of symptomatic treatment used alone and was an independent predictor for reduced use of chemotherapy and surgical treatment [49]. Although the use of surgery as a major treatment strategy has increased over time, younger age (<75 years) is also independently associated with a higher likelihood of undergoing curative surgery (liver resection) for both synchronous and metachronous disease [50]. Not surprisingly, elderly patients are poorly represented in studies evaluating liver resection for CRC LM [55], limiting widespread availability of data supporting resection for this population. However, single-institution retrospective studies have reported acceptable, although slightly higher, postoperative mortality rates (≤8%) in well-selected elderly patients following liver resection [56], [57], [58], [59], [60], [61], [62], [63], [64], [65], [66], [67], [68]. Similarly, a population study based on SEER-Medicare data, including more than 800 patients over 65 years of age, demonstrated a 30-day mortality rate of 4.3% [51]. Further, when comparing long-term outcomes between those with and without liver resection, this study confirmed the significant oncologic benefit of hepatectomy in the elderly, with a 5-year overall survival of 32%, compared with 10.5% in those not undergoing hepatectomy [51]. Table 1 lists selected studies evaluating postoperative and oncologic outcomes in elderly patients undergoing hepatectomy for CRC LM specifically [51], [56], [58], [59], [60], [61], [62], [63], [64], [65], [67], [68]. Although a few studies report postoperative and long-term outcomes as being slightly better in the younger population, all report an acceptable mortality rate and a significantly increased overall survival compared with historic controls not receiving surgical treatment (5-year overall survival <5%). Based on these findings, liver resection is recommended for elderly patients when deemed safe and is expected to improve overall survival [41]. However, the specific group of patients that can benefit from this “aggressive” strategy still needs to be better defined.

Table 1. Selected studies assessing outcomes in elderly patients undergoing liver resection for colorectal liver metastasis.

	Author (year)	Patient population	N	POP complications (%)	POP mortality (%)	Overall survival
						Median (months)	3 years (%)	5 years (%)
	Zieren [65]	>70 years	18	28	6	18	38	25
		27–78 years	90	26	3	27	45	32
	Brunken [58]	>70 years	25	28	4	–	–	47
		<70 years	141	26	2	–	–	36
	Brand [56]	≥70 years	41	39	7.3	–	29	16
		<70 years	126	30.9	2.4	–	42	21
	Zacharias [64]	>70 years, first resection	56	41	0	33	44	21
		>70 years, repeat resection	16	38	7	17	25	0
	Nagano [62]	≥70 years	62	19.7	0	47	47	34
		<70 years	150	23.3	0.49	–	63	53
	Turrini [63]	≥70 years	15	33	0	22	–	–
	Figueras [60]	≥70 years	160	41	8	–	48	3
		<70 years	488	34	3	–	62	45
	Mazzoni [61]	≥70 years	53	20.7	5.7	28	–	30
		<70 years	144	14.6	2.1	31.5	–	38
	Cummings [51]	≥65 years	833	32	4.3	45	–	33
	Mann [67]	≥70 years	49	30	4	–	38	31
		<70 years	142	19	3	–	54	43
	de Liguori Carino [59]	≥70 years	178	38.5	4.9	–	43	32
	Nojiri [68]	≥75 years	64	29.6	0	–		33
		<75 years	337	23.4	0.5	–		48

When considering liver resection in the elderly, balancing postoperative risks and oncologic benefits is a step that must be included in the decision-making process. Guidelines to help with this assessment need to be developed. Currently, postoperative risks may be estimated by preoperative health assessment, using tools such as the comprehensive geriatric assessment (CGA) or the preoperative assessment of cancer in the elderly (PACE) tool [34], [39], [41], [69]. However, these tools are not validated for liver resection; and selection of patients based on perioperative risk remains an imprecise task, which relies on individual subjective assessments. Data guiding the individual risk of postoperative morbidity and mortality after liver resection in the elderly are limited. The use of minimally invasive techniques reduces the surgical insult and can help overcome some postoperative complications. Nevertheless, systematic approaches to better define patients more likely to withstand and hence benefit from open and laparoscopic approaches for liver resection are required.

Similarly, oncologic outcomes after hepatectomy have not been well studied in the elderly. Additionally, only a few studies have examined disease-free survival in the elderly, a particularly relevant outcome considering the competing causes of mortality in this segment of the population [68]. Current approaches estimate overall survival using clinical and tumor-related variables [70]. Despite limitations from this approach, it provides the best currently available data to help identify patients more likely to benefit from liver resection. Recently, a score based on five such predictors has been developed for elderly patients (Table 2) [61]. The presence of three or more of these factors was associated with a median survival of 23 months or less, compared with as high as 46 months for those with lower scores [55], [61]. Based on these data, the benefit of liver resection for elderly patients with three or more predictors is limited; and other less aggressive but potentially curative strategies, such as ablation, should be considered for this group of patients until better predictive tools become available [55], [61].

Table 2. Clinical risk score to predict overall survival after liver resection in elderly patients treated for CRC liver metastasis [61].

	Variables	Points
	Node-positive primary tumor	1
	Disease-free interval<12 months	1
	More than one metastatic lesion	1
	Tumor size>5cm	1
	CEA>100ng/mL	1

	Score	Predicted median overall survival (months)
	0–2	30–46
	3–4	20–23
	5	12

Additional challenges arise when considering the evolving management of LM. Eligibility criteria for liver resection of CRC LM have been redefined [5], [71], [72], allowing expanded indications and improving mortality rates to below 3% [73]. Previously, indications for liver resection were based on tumor-related factors (i.e., size, number of lesions, etc.), but current criteria are driven by future liver remnant characteristics; liver resection is generally indicated, regardless of tumor characteristics, as long as a negative-margin resection is feasible while preserving sufficient liver to support homeostatic functions after hepatectomy [5], [71], [72], [74]. The necessary future liver remnant volume is defined based on preoperative liver function [72]. However these data are not validated in elderly patients. Liver physiology changes with increasing age, resulting in decreased hepatic flow and functional liver-mass loss [23]. These physiologic changes often present with preserved baseline liver function that is typically impaired when challenged (e.g., after liver resection) [23]. Based on this, extensive liver resections in elderly patients can be associated with increased risk of postoperative liver insufficiency, which may result in higher delayed (>30 days) postoperative mortality [75], [76]. To decrease risks associated with these physiologic changes, operative technique with new and emerging surgical instruments must focus on minimizing blood loss, unnecessary transfusions, and ischemic time while performing parenchyma-preserving operations whenever feasible [55].

Oncosurgical strategies have improved long-term outcomes in patients with CRC LM [5], [72], [74]. Five-year overall survival reported in recent series reaches 58% [73], [77]. More effective perioperative chemotherapy, use of preoperative portal vein embolization (PVE) to induce and assess liver hypertrophy prior to resection, and more aggressive and innovative operations (e.g., staged and repeat hepatectomies, combined liver and colon resection, ablation) are some of these new strategies [72], [74], [78]. The role, indications and complications of these strategies are not yet well defined; however, their use in well-selected individuals is supported by available data. Adjuvant chemotherapy after liver resection is associated with improved disease-free and overall survival [79], [80]. The efficacy and safety profile of postoperative chemotherapy in the elderly are well recognized, and its use is recommended, although with special caution and dose modifications when using specific agents (i.e., bevacizumab and capecitabine) [28]. The use of preoperative chemotherapy is associated with improved long-term outcomes in selected patients [81], [82] and can additionally be used for “downstaging” of unresectable disease, allowing for subsequent liver resection in other well-selected individuals [28], [83], [84], [85], [86], [87]. Elderly patients are well represented in some studies supporting this strategy, and its use is recommended at this time, based on current existing data [81], [82]. However, preoperative chemotherapy is associated with different patterns of liver injury and increased postoperative complications, based on the specific regimen used [88], [89], [90]. Additionally, overall toxicity can limit the ability to proceed with surgery, based on changes in performance status. These side-effects have not been explicitly evaluated in the elderly and require further assessment. The use of PVE to induce liver hypertrophy is a safe strategy that allows resection while minimizing postoperative liver insufficiency, when used following appropriate indications [72], [91]. Liver regeneration is impaired and/or delayed in patients with diabetes mellitus or patients with cirrhotic livers [92]. Similarly, a substantial number of pre-clinical studies suggest a reduced and/or delayed liver regeneration in older patients [93]. However, it is hard to predict in which patients the regenerative process is severely impaired. When used following standard recommendations, PVE has the additional advantage of helping select older patients with adequate regenerative response, and hence more likely to withstand major liver resections [72]. Based on this, and experience in patients with impaired regeneration due to cirrhosis, PVE should also be considered in older patients who are otherwise candidates for major liver resection (≥lobectomy), to assess the degree of hypertrophy and the post-PVE FLR volume, and using standard accepted benchmarks for these measures, consider hepatectomy on an individual basis [92], [94]. Finally, the effects of more recent and novel surgical strategies (e.g., staged and repeat hepatectomies, combined liver and colon resection, and ablation) have not been appropriately assessed in older patients and should be considered only after individual, case-by-case evaluation. Nonetheless, when the risk–benefit ratio of resection is not favorable—such as in the oldest group patients with multiple co-morbidities or in those with a clinical risk score ≥3, as previously discussed—less aggressive and newer surgical strategies including ablation appear to be adequate alternatives, although further research to characterize outcomes in this setting is required.

Lastly, health-related quality-of-life outcomes are not commonly considered when managing patients with CRC LM. Specific measures to assess health-related quality-of-life in this group of patients are available. They have been validated [95], [96] and have recently been shown to play a prognostic role in long-term outcome [97]. These data must be incorporated in future strategies aimed at developing individualized guidelines to help with the medical decision-making process in elderly patients with CRC LM. Further, implementation of these different therapies is best accomplished within the context of an appropriate multidisciplinary approach. Research evaluating factors related to the implementation and maintenance of this approach needs to be developed.

6. Conclusions 

Global graying is evolving and has significant effects on epidemiology and management of cancer. The implications for CRC are considerable. Treatment disparities among young and elderly patients are partially driven by providers’ inaccurate perceptions regarding risks and benefits of treatment with standard-of-care strategies. As a result of global graying, CRC cases are projected to double in the next two decades; and additional challenges in the process of care for elderly patients are expected to occur. Systematic analysis of existing data, new clinical assessment tools, and guidelines for management of elderly CRC patients support the use of aggressive surgical and adjuvant therapies for well-selected patients and should help in narrowing the existing gap between the younger and older populations.

Data regarding the effect of global graying on epidemiology, management and outcomes of CRC LM in the elderly are lacking. Although the use of hepatectomy in well-selected patients is supported by current available data, a more systematic approach to evaluating factors such as safety, toxicity and outcomes of current and emerging liver-directed therapies in the elderly is yet to be developed. In the future, evaluation of factors related to the implementation of these strategies is needed as well. A substantial increase in the number of new CRC LM in the elderly is likely to occur; and, with surgery as the only curative approach, evidence-based recommendations using higher-level data are required to guide care for this segment of the population. Further, quality-of-life measures need to be studied for elderly patients undergoing liver resection. This can help in the decision-making process when considering individualized, patient-centered treatment of such complex conditions in this high-risk population.

Conflict of interest statement 

The authors have no conflict of interest to declare.

Reviewers 

Dr. Raymond S. Yeung, University of Washington, Department of Surgery, 1959 Northeast Pacific, Seattle, WA 98195, United States.

Professor Riccardo Audisio, Consultant Surgeon; Surgical Oncologist, Whiston Hospital, Department of Surgery, Warrington Road, Prescot, Merseyside L35 5DR, United Kingdom.