The increase in life expectancy is a distinguishing feature of our society, resulting in higher demand for health and welfare. Colorectal cancer (CRC) is the second most common cancer for both sexes in Western countries, with more than 70 % of cases over 65 years of age [1]. This results in an increasing proportion of elderly and very elderly patients suffering from CRC perspectively candidates for surgery. Since its introduction in 1991 [2], laparoscopic colorectal surgery (LCS) has been validated by studies with level of evidence 1A, which have demonstrated its superiority over traditional surgery for the short-term outcomes and a substantial equality about the functional and oncological results [3–5]. Nevertheless, the diffusion of LCS for CRC in the elderly is still limited and published data need more evidence. The purpose of this study is to analyze the feasibility and effectiveness of LCS for the treatment of colonic and rectal cancers in very elderly, defined as people of 80 years of age or above, comparing the results with a control group consisting of patients aged between 60 and 69 years who underwent same surgery with same indications. Our hypothesis is that there are no significant differences in outcomes between cases and controls.

Materials and methods

Data on a consecutive unselected series of patients of 80 years old or greater who underwent laparoscopic surgery for malignant disease of the colon and rectum from January 2008 to December 2013 was prospectively collected and analyzed (Group A). The analysis was performed at two institutions (Division of General Surgery, Department of Surgery, Galmarini Hospital, Tradate—IT and Minimally Invasive Surgery Research Center, University of Insubria, Varese—IT), characterized by both a high specific case volume, and the perioperative management complies with international recommendations [6]. These cases were compared with a control group of patients between 60 and 69 years old, who underwent the same surgery, in the same period at the aforementioned two institutions (Group B). Procedures for benign disease and under emergency were excluded, while those in election after endoscopic placement of a colonic stent “bridge-to-surgery” were included. Laparoscopic approach is the routine indication at our institutions. Exclusion criteria to laparoscopy were just the unsuitability from an anesthesiological point of view and the dissent from the patient. We considered laparoscopic right colectomies, left colectomies and anterior rectal resections. The group “other” included laparoscopic segmental resections of both transverse colon and splenic flexure, and laparoscopic Miles’ procedures, indicated for tumors staged as T2–T4 even after neoadjuvant therapy and located within 2 cm from the anal verge. Rectal tumors in stage II or higher were candidates for long-course 5-fluorouracil-based chemoradiotherapy (CRT; 50.4 Gy in 28 fractions) with surgery scheduled 6–8 weeks after. All patients followed a low-residue and high-calorie diet intake for 5 days before the surgical procedure. The preoperative diagnostic workup included a flexible colonoscopy and a contrast-enhanced computed tomography scan of the abdomen in all patients, a water-soluble contrast enema in case of left colectomy or rectal resection. Data were stored in institutional databases prospectively updated. We analyzed demographics, surgery- and postoperative-related data, as well as the short-term outcomes for each patient. All the considered items are shown in Table 1. Comorbidity of each patient was assessed by Charlson comorbidity index [7]. No patients underwent any bowel preparation except for rectal resection that needed a mechanical bowel preparation using polyethylene glycol 70 + 70 g/2 L the day before surgery. All patients were treated with both short-term broad-spectrum intravenous antibiotics (ceftizoxime plus metronidazole) and antithrombotic prophylaxis by low molecular weight heparin according to body weight. A nasogastric tube and a urinary catheter were placed after induction of general anesthesia in all cases. All the procedures were performed by the same two teams, fully trained in both colorectal and minimally invasive surgery. Surgical instrumentation was standardized, and dissection was performed by harmonic scalpel (Harmonic ACE, Ethicon Endo-Surgery, LLC, Guaynabo, Puerto Rico). All procedures were recorded and stored in digital form.

Table 1 Data analyzed
Full size table

Surgical technique

All the procedures were performed by a 4-port standardized technique with an approach from medial to lateral. Laparoscopic right colectomy has provided a “high” vascular ligation and an extracorporeal ileocolic anastomosis until 2009–2010, then intracorporeal that is a totally laparoscopic right colectomy. For laparoscopic left colectomy and anterior rectal resection, there was a radical ligation of the inferior mesenteric pedicle with preservation of the hypogastric nervous plexus in both sexes. Mobilization of the left colonic flexure was selective in left colectomy, standardized in rectal resection. In case of anterior rectal resection, a total mesorectal excision (TME) was always performed. The anastomosis was transanal and mechanical according to Knight–Griffen technique anyway. In case of cancer of the middle and lower rectum, a loop ileostomy was performed at the end of the procedure. We did not use drains routinely. Postoperative medical and nursing care was standardized. The clinical course was documented for each patient. All patients were mobilized early with removal of the urinary catheter. The nasogastric tube was removed after the first flatus. Criteria for the discharge included the absence of symptoms, tolerance of a minimum of 3 meals without restrictions and passage of stool. All adverse events that occurred within 30 days after surgery were considered complications. Complications were classified using the Clavien–Dindo classification system [8]. The term anastomotic leakage defines all conditions with clinical or radiological anastomotic dehiscence, with or without the need for surgical revision. The proximal and distal clearances from tumor are expressed in centimeters. Regarding rectal surgery, the distal and circumferential resection margins (CRM) are defined positive, that is not adequate, if the clearance is less than 1 cm and 2 mm, respectively. The assessment of mesorectal integrity is expressed as satisfactory (intact mesorectum with defects <5 mm, smooth CRM), partly satisfactory (moderate irregularities, partial conization, irregular CRM) and unsatisfactory (highly incomplete mesorectum, defects reaching the muscle layer, highly irregular CRM) [9]. Short-term follow-up was conducted at 30 days after surgery.

Statistical analysis

Continuous variables were expressed as median and interquartile range (IQR) and then analyzed by nonparametric Mann–Whitney U test. Categorical variables were expressed as a percentage and analyzed by Fisher’s exact or Chi-square test where appropriate. P values <0.05 were considered significant. Statistical analysis was performed using XLSTAT version 2015.1.01 (Addinsoft, New York, USA).

Ethics

The approval of the institutional review committee was not required because the data of the present study were collected in the course of routine clinical practice and, therefore, are considered valid the informed consent signed by each patient for any surgery or other procedure. The study protocol conforms to the ethical guidelines contained in the “World Medical Association Declaration of Helsinki—Ethical Principles for Medical Research Involving Human Subjects” adopted by the 18th WMA General Assembly, Helsinki, Finland, June 1964, then revised in Tokyo 2004 [10].

Results

Colon cancer

Ninety-six patients over 80 years old (Group A) who underwent laparoscopic surgery for colonic cancer were compared with 220 patients aged between 60 and 69 (Group B). Demographics and disease-related data are summarized in Table 2. There were no statistically significant differences by gender, body mass index and previous abdominal surgery between Group A and Group B. Type of surgical procedures is also comparable. As expected, Group A showed a CCI (7 [IQR 5–9] vs 4 [IQR 3–5]; P 0.001) and an ASA score significantly higher than Group B. Regarding the stage of disease, T in situ is more represented in the cohort between 60 and 69 years (0 vs 6.4 %; P 0.011). Table 3 shows data related to surgery and short-term outcomes. Mean operative time (180 [IQR 150–200] vs 180 [IQR 150–210] min; NS), estimated blood loss (50 [IQR 25–75] vs 50 [IQR 25–120] mL; NS) and conversion rate (1.0 vs 0 %; NS) are statistically comparable between the two groups. Timing of first stool is quantitatively less for Group A (3 [IQR 2–3.25] vs 3 [IQR 2–5] days; P 0.080), although not statistically significant. Anastomotic leakage (1.0 vs 0.9 %; NS), postoperative hospital stay (7 [IQR 6–8] vs 7 [IQR 6–8] days; NS) and the rate of readmission within 30 days after surgery (1.0 vs 0.45 %; NS) are statistically similar. The proximal (7 [IQR 5–10] vs 7 [IQR 5–9] cm; NS) and distal resection margins (9 [IQR 6–12] vs 10 [IQR 8–12] cm; NS), as well as the number of lymph nodes harvested (21 [IQR 15–27] vs 20 [IQR 14–31]; NS) are substantially overlapping between Group A and Group B. Complications according to Clavien–Dindo classification system are summarized in Table 4, showing no significant differences.

Table 2 Demographics and disease-related data (median [IQR]—#, %)
Full size table
Table 3 Short-term outcomes (median [IQR]—#, %)
Full size table
Table 4 Complications according to Clavien–Dindo classification system
Full size table

Rectal cancer

We have analyzed 33 patients aged 80 years old or more (Group A) and 82 controls between 60 and 69 (Group B). Table 2 shows demographics and disease-related data. There are no statistical differences, except for ASA score and CCI (7 [IQR 5–8] vs 4 [IQR 3–5]; P 0.001), significantly higher above 80 years. Group A and Group B show a significant number of T3 (60.6 vs 22 %, P 0.007) and T2 tumors (18.2 vs 39.0 %, P 0.031), respectively. Surgery-related data and short-term outcomes are shown in Table 3. Operative time (180 [IQR 160–210] vs 180 [IQR 165–240] min; NS), estimated blood loss (50 [IQR 0–150] vs 50 [IQR 25–108.7] mL; NS), conversion rate (3.0 vs 2.4 %; NS), timing of first stool (3 [IQR 2–4] vs 3 [IQR 2–5] days; NS), hospital stay (8 [IQR 8–9] vs 8 [IQR 7–9] days; NS) and 30-day readmission rate (6.1 vs 1.2 %; NS) do not differ statistically. The rate of both anastomotic leakage (6.1 vs 2.4 %; NS) and reoperation (0 vs 1.2 %; NS) is well comparable too. About the oncological proficiency, the positivity rate of both the distal margin (3.0 vs 2.4 %; NS) and the CRM (6.1 vs 4.9 %; NS), as well as the number of lymph nodes harvested (18 [IQR 12–25] vs 20.5 [IQR 14–24.7]; NS) is comparable between the two cohorts. The anatomopathological evaluation of the mesorectal specimen does not show any findings classified as unsatisfactory in all cases analyzed. The classification of complications is comparable between the two groups, as shown in Table 4.

Discussion

Colorectal cancer (CRC) is the third most common cancer in men (746,000 cases/year, 10.0 % of the total) and the second in women (614,000 cases/year, 9.2 % of the total) worldwide [11]. Statistically, about 4.7 % of the population will have a diagnosis of CRC in lifetime. For example, a man now in his sixties has an estimated risk of developing CRC in the next 10 years of 1.26 %, a septuagenarian of 1.96 % and so on [12].

Population aging is a question of primary interest in our society. In Italy, life expectancy at birth in 2011 was 79.4 and 84.5 years, respectively, for males and females, while life expectancy at 65 years old, which is the number of years that a person of that age has still to live on average, is 18.4 for men and 21.9 for women, with no significant geographical differences [13]. Considering that colorectal tumors reach their peak of incidence between the seventh and eighth decade of life, it can be inferred how the amount of elderly patients affected by CRC will be constantly increasing. Surgery plays a pivotal role in the treatment of CRC, although age of patients represents an independent indicator of postoperative morbidity and mortality according to some authors [1, 14]. Surgical treatment is therefore a challenge for this group of patients, due to their frailty caused by reduced physiological reserves and frequent comorbidities [15, 16]. Laparoscopic colorectal surgery (LCS) has shown, with levels of evidence 1A, short-term outcomes better than traditional techniques, due to a substantial reduction in surgical trauma and therefore in the immunological postoperative reaction of the patient. This comes from less handling and exposure of the viscera, together with reduced intraoperative blood loss [17]. The expression of the DR receptor of human leukocyte antigen (HLA-DR) on monocytes remains higher after laparoscopic than conventional colorectal surgery [18], and the seriated measurements of C-reactive protein and interleukin-6 are lower in the 48 h following a laparoscopic procedure, overall indicating a better metabolic response to surgical stress. The long-term analysis after LCS has then highlighted oncological outcomes substantially comparable to open surgery, giving a definitive validation for the minimally invasive treatment of cancers of the colon and, more recently, of the rectum [19–23]. Despite these outcomes, LCS is still much less widespread in the elderly than in younger, because of some considerations limiting the indications. For example, it has been suggested that laparoscopic surgery layout could affect, for extended operating times, a reduction in cardiac output and an increased risk of postoperative atelectasis. But it has been shown that a prolonged pneumoperitoneum or reverse Trendelenburg position of the patient does not cause significant intra- or postoperative problems in patients over 70 years old. Already 10 years ago the guidelines from the European Association for Endoscopic Surgery (EAES) highlighted how, in patients with a reduced cardio-pulmonary function, laparoscopic surgery is not contraindicated, recommending however to maintain levels of pneumoperitoneum not more than 12 mmHg (Grade of Recommendation B) [24]. The percentage of elderly patients included in the trials published on LCS is limited. Out of 52 study protocols available today, 44 % exclude the elderly simply because of age or comorbidities. In 86 % of trials, the mean age of patients enrolled is below 70 years of age without any specification of concomitant comorbidities [25]. There are some limitations related to the age of patients even in clinical practice. A Dutch population-based analysis over 33,000 patients showed that the rate of LCS for cancer still registers a statistically significant difference in distribution between patients under and over 75 years of age (49.2 vs 40.3 %) [26]. Hence, we need to clarify the effectiveness of LCS in the elderly population, whose results are shown in Tables 5 and 6, respectively, as results of LCS in elderly versus younger and outcomes of LCS versus open surgery in elderly patients. The age limit to indicate the elderly patient is difficult to define worldwide, and it often shows some differences even in developed countries. Conventionally, recent literature has indicated 70 years as the cutoff to determine this limit, although a shared definition is not yet reached. Most of the authors retrospectively compared open and laparoscopic colorectal surgery over a limit of age varying between 65 and 85 years. However, this method of comparison could highlight a selection bias, when the criteria for the choice of a type of surgery over another are not well specified [27], or when both previous abdominal surgery and locally advanced diseases constitute criteria for the exclusion of patients from the laparoscopic cohort [28]. Few papers analyzed the outcomes of LCS in the elderly compared to younger patients. However, in the study of Schwandner et al., the two cohorts were not well comparable in terms of surgical procedures [29], while Reissman et al. used an inadequate aging cutoff fixed at 60 years [30]. The paper by Yamamoto et al. was methodologically correct, although the small sample size has weakened the statistical significance [31]. Those by Chautard et al. [32] and Keller et al. [33], with a cutoff of 70 years, have shown outcomes substantially comparable between young and elderly patients undergoing LCS, although the second one included patients with benign disease for the 60 % of young patients and 40 % of the elderly. The research from Jeong et al., with an age limit of 75 years, showed a significantly higher rate of complications in elderly patients (12 vs 6.2 %) and therefore concluded that age represented an independent predictor of postoperative morbidity [34]. Papers concerning very elderly patients over 80 years of age are rare. Steward et al. and Lian et al. on the one hand, Mukai et al. and Nakamura et al. on the other, compared LCS and open surgery with a cutoff fixed at 80 and 85 years, respectively, highlighting significant results in terms of length of hospital stay and morbi-mortality in favor of minimally invasive surgery [35–38]. In this study, we tried to rigorously define the methodology of selection for both cases and controls. We chose octogenarians cases who, for uniformity of characteristics, better represented the object of an analysis about the effectiveness of LCS in the elderly. Then, we considered a group of controls limited in age between 60 and 69 years, as a very significant decade for the epidemiological impact of CRC, representing a cohort of controls neither too young nor too close to the group of cases, so as to minimize the risk of bias for the analysis. The cohorts analyzed were found to be uniform regarding both the surgical treatment and the perioperative management. All the procedures were performed by minimally invasive technique, which is of choice for us, regardless of limits of inclusion related to comorbidities or disease’s stage. As expected, the groups differed regarding both the preoperative assessment by the America Society of Anesthesiologists score and the index of comorbidities. This latter aspect has been evaluated by Charlson comorbidity index through 19 tracer conditions corresponding to different scores, complemented by an additional set of values for decades of age. This score is derived from internal medicine but represents a useful tool to weigh the comorbidity of compared groups and to estimate the postoperative morbidity and mortality. In fact, the surgical risk is directly proportional to the comorbidities of patients, and, as noted in the original work by Charlson, a score greater than 3 doubles 1-year mortality risk [7].

Table 5 Laparoscopic colorectal surgery in the elderly compared to younger patients
Full size table
Table 6 Laparoscopic versus open colorectal surgery in the elderly
Full size table

Referring to the stage of disease, colonic cancers showed a significantly greater rate of in situ tumors in Group B (0 vs 6.4 %). Cancers of rectum instead showed a higher number of T2 tumors in the arm between 60 and 69 years (18.2 vs 39.0 %), while Group A showed a higher proportion of T3 tumors (60.6 vs 26.8 %), probably in relation to the reduced rate of neoadjuvant therapy in this cohort (15.2 vs 20.2 %). Colonic and rectal cancers show anatomical and pathophysiological similarities; however, there are some differences in epidemiological, clinical and therapeutic matters that have suggested a differentiated analysis. Surgical treatment of rectal cancer seems to be more challenging, especially in the presence of patients with narrow pelvis or bulking tumors. A proper surgical technique is directly related to both functional and oncological outcomes. Recognition and respect of the hypogastric plexus and its branches allow the maintenance of a quality of life even in the elderly, so also it is well known how a proficient surgery gets good long-term outcomes in terms of local and systemic recurrences as well as disease-free and overall survival [39]. Therefore, a great care has been taken in the analysis of the specimens.

For colonic cancer surgery, through the assessment of both the number of lymph nodes harvested and the clearance of resection margins, which should, respectively, be at least of 14 nodes and 2 cm [6].

Regarding rectal cancer, as well as through the analysis of the items mentioned above, with a minimum cutoff of 1 cm for the resection margin of lower rectum cancers, in addition we checked the integrity of mesorectum excised using the methodology proposed by the Belgian group PROCARE [40], and we evaluated the circumferential resection margin (CRM). CRM is an accurate predictor of local recurrence, with a prognostic value independent of TNM classification. Our analysis has provided a positive CRM cutoff less than 2 mm, which, if negative, should reduce the risk of local recurrence (5.8 % compared to 16.0 % in the case of positive 2-mm CRM) [41]. The same limit was applied for the COLOR II study, showing a positive CRM in 9.9 % of cases [42], while other trials have defined as positive a CRM less than 1 mm, resulting in 15.1 % in the MRC-CLASICC [43] and in 3.5 % in COREAN trial [44]. The comparison of these data with the results reported by us (6.1 vs 4.9 %; NS) emphasized the effectiveness of rectal surgical resections for both cases and controls. Groups A and B showed no statistically significant differences for either colonic or rectal cancer in terms of operative time (180 [IQR 150–200] vs 180 [IQR 150–210] min; NS—180 [IQR 160–210] vs 180 [IQR 165–240] min; NS), intraoperative blood loss (50 [IQR 25–75] vs 50 [IQR 25–120] mL; NS—50 [IQR 0–150] vs 50 [IQR 25–108.7] mL; NS) and conversion rate (2.1 vs 2.7 %; NS—3.0 vs 2.4 %; NS). Short-term outcomes of LCS are well known and supported by level 1A evidence [45–47]. This study showed that these results can be achieved even in very elderly patients over 80 years of age. Timing of first stool (3 [IQR 2–3.25] vs 3 [IQR 2–5] days; NS—3 [IQR 2–4] vs 3 [IQR 2–5] days; NS), length of hospital stay (7 [IQR 6–8] vs 7 [IQR 6–8] days; NS—8 [IQR 8–9] vs 8 [IQR 7–9] days; NS) and 30-day readmission rate (1.0 vs 0.45 %; NS—6.1 vs 1.2 %; NS) revealed no significant differences between age groups, similar to literature data. Even the postoperative morbidity and mortality rate were statistically comparable between elderly and 60- to 69-year-old patients, for both colonic (47.9 vs 43.6 %; NS) and rectal (63.6 vs 52.4 %; NS) cancers. The high rate of complications was substantially due to the choice of the Clavien–Dindo classification system [10]. This tool stratifies complications in five grades of severity, depending on the type of treatment. This minimizes the risk of bias and allows a rigorous method of comparison between the series in the literature, otherwise difficult using the classical categorization in minor and major complications. Grades I and II include only minor deviations from the normal postoperative course, which could be solved by drug therapy, blood transfusions, total parenteral nutrition and placement of simple devices such as a urinary catheter and nasogastric tube. Grades III and IV instead collect severe complications, which require a surgical, endoscopic or radiological interventional treatment, rather than ICU management. Almost all of the postoperative complications of our series fell in Clavien–Dindo grades I and II, with only a significant difference consisting in a higher rate of blood transfusions in the arm over 80 years underwent laparoscopic colonic surgery, likely due to an increased comorbidity rate.

Conclusions

Laparoscopic surgery is effective and safe for the treatment of colonic and rectal cancer even in very elderly over 80 years old. Short-term outcomes and results in terms of oncological proficiency are similar to those from the control group between 60 and 69 years old. Age does not appear to be a risk factor or a limitation for laparoscopic colorectal surgery, and indeed, very elderly patients may benefit from a minimally invasive approach.