Hospital records of patients under the age of 50 years who underwent CRC operations between July 1996 and May 2012 at two institutions were reviewed (New York Presbyterian Hospital, Columbia University Campus and St. Luke’s Roosevelt Hospital, NY, United States). Specifically , the following documents where reviewed; admission notes, operative notes, discharge summaries, endoscopy records, and pathology reports. A subset of this data was also obtained from an IRB approved prospective data base of patients undergoing colorectal resection maintained by the senior author from 2006 to June 30, 2009 at New York Presbyterian Hospital and from July 1, 2009 until May 2012 at St. Luke’s Roosevelt Hospital. This prospective database also provided the information regarding CRC patients over 50 years of age (n = 392) used in this study. Additional information for this retrospective review was obtained from office charts and from telephone interviews.

The main study variables included: demographics, presenting symptoms leading to diagnosis, family history of CRC, tumor location, type of surgical resection, stage and differentiation of disease, and post operative complications. Patients with inflammatory bowel disease or known polyposis syndromes such as familial adenomatous polyposis syndrome, Gardner’s syndrome and the like were excluded from the study.

Disease stage was reported according to the TNM Classification System used by the American Joint Committee on Cancer[16,17]. “T” refers to the size or direct extent of invasion of the primary tumor; “N” refers to the degree of spread to regional lymph nodes, if any; and “M” refers to the presence of distant metastases.

Statistical methods for comparing stage and tumor distribution between the under age 50 years and the 50 and over years groups included a 2-proportion Z test.

One hundred eighty patients under 50 years of age (87 females, 93 males; range 17-49 years; mean 41.4 ± 6.2 years) underwent a CRC operation between July 1996 and May 2012 at the two institutions. In regards to the total number of patients (regardless of age) that underwent a CRC operation, complete data is only available for the period between July 2006 and May 2012; during this time period 437 CRC operations on adults were carried out of which 49 (11.2%) involved patients less than 50 years of age. When the total population of 180 patients under age 50 is considered, the distribution of CRC within age categories is as follows: under age 30 years, 8 patients (4%); age 30-39 years, 46 patients (26%); age 40-49 years, 126 (70%). Of note, 30% of the patients were younger than 40 years of age. One hundred and seventy patients (94%) reported symptoms upon presentation (Table 2).

Family history data was available for 167 patients; 13 patients (7.8%) did not know their family history. Regarding the 167 patients with family history data, 14 patients (8.4%) had a first degree family history of CRC, 20 patients (12.0%) had a second degree history, and 6 patients (3.6%) had both a first and second degree family history of CRC. Thus, a total of 12% of patients had, at least, a first degree family history of CRC. Seventy six percent were sporadic CRC cases. One patient (age 42 years) presented with synchronous primary cancers of the cecum and splenic flexure and had a family history notable for 3 first and 1 second degree relatives with CRC. The Amsterdam criteria for hereditary non-polyposis colorectal cancer (HNPCC) were met in this patient[14]. Unfortunately, none of the patients in this series were evaluated for gene mutations associated with HNPCC (i.e., hMLH1, hMSH2, etc.)[18].

The majority of tumors (77%) were located distal to the splenic flexure, with 39% involving the rectum (Table 3). Proximal (right and transverse colon) cancers were noted in only 24% of patients. In comparison, a cohort of 392 CRC patients 50 years of age and older who underwent an operation between 2006 and 2012 at the same institutions by the same surgeons, presented with more proximal tumors (age < 50 years, 24% vs age ≥ 50 years, 43%; P < 0.0001) and fewer rectal tumors (age < 50 years, 39% vs age ≥ 50 years, 27%; P = 0.002) using 2-proportion Z-test (Table 4).

A formal colorectal resection was done in the majority of patients (95%), a transanal excision of a rectal cancer in 1 patient, colocolonic bypass for unresectable Stage 4 disease in 1 patient, and proximal diversion in 6 patients with obstructing, locally invasive cancers (Table 5). Twenty three abdominoperineal resections and 3 low anterior resections with mucosectomy and subsequent coloanal anastomosis were performed. A Hartmann’s procedure was done in 4 patients with sigmoid or rectal lesions. Two patients (ages 34 and 42 years) had synchronous primary colon cancers and underwent subtotal colectomy. Twenty-four distal resection patients were temporarily diverted (16% of all patients with anastomoses). Regarding the surgical methods used in the 172 patients who underwent bowel resection, the breakdown is as follows: laparoscopic-assisted, 78 patients (45.3%); hand-assisted or hybrid laparoscopic/open technique, 29 patients (16.9%); and open methods, 65 patients (37.8%).

According to the TNM system for cancer staging by the American Joint Committee on Cancer[16,17], 37 patients (21%) had Stage 1 disease, 47 patients (26%) had Stage 2 disease, 70 patients (39%) had Stage 3 disease, and 26 patients (14%) had Stage 4 disease (Table 4). Three patients who underwent neoadjuvant chemoradiation for T3 rectal lesions based on preoperative endorectal ultrasound imaging had no residual disease on final pathology at the time of colorectal resection and were considered as Stage 2 lesions. Likewise, 3 patients who had sessile polyp cancers removed colonoscopically (invasion into submucosa noted on pathology) who underwent formal resection that revealed no residual cancer or involved lymph nodes on pathologic evaluation were classified as having Stage 1 cancers. Twenty patients with Stage 4 disease had known hepatic involvement and 3 patients had peritoneal carcinomatosis diagnosed at laparotomy.

Thirty-five patients (19%) underwent neoadjuvant chemoradiation, 3 patients (1.7%) underwent neoadjuvant chemotherapy alone, 10 patients (5.6%) underwent adjuvant chemoradiation, 56 patients (31%) underwent adjuvant chemotherapy, and 1 patient underwent adjuvant radiation alone for bony metastases. Of note, compared to the cohort of 392 CRC patients 50 years of age or older who underwent an operation between 2006 and 2012, patients under age 50 more often presented with Stages 3 and 4 disease (age < 50 years, 53% vs age ≥ 50 years, 41%; P = 0.003 using 2-proportion Z-test) (Table 4).

Moderately or well differentiated cancers were noted in 124 patients (69%) whereas poorly differentiated cancers were found in 22 patients (12%). Of those with poorly differentiated histology, 67 percent presented with advanced Stage CRC (Stage 3 or 4). Thirty-four patients (19% of total) had mucinous adenocarcinomas of which 62% had advanced stage CRC.

Regarding postoperative complications, there was 1 anastomotic leak and 4 intra abdominal/pelvic abscesses (reoperation in 1 patient, percutaneous drainage in 3 patients). Other postoperative complications included: ileus, 6 patients; small bowel obstruction, 6 patients (all required reoperation); wound infection, 7 patients; wound dehiscence, 2 patients (reoperation × 2); urinary retention, 3 patients; portal vein thrombosis, 1 patient; C. difficile colitis, 1 patient (treated with antibiotics); and incisional hernia, 1 patient (surgically repaired). There were no deaths perioperatively. Twenty-two patients (13%) developed recurrence and/or progression of disease to date. Three patients (ages 42, 42, and 49 years) went on to develop metachronous primary colon cancers within 3-4 years of their initial resection.

There are currently no screening guidelines in place by the American Society of Colon and Rectal Surgeons (ASCRS) or the American Cancer Society (ACS) pertaining specifically to young patients who present with symptoms that could signify an underlying neoplasm. The ASCRS[25] does recommend, in general, that anyone “with symptoms or signs that suggest the presence of CRC or polyps who fall outside the domain of screening should be offered an appropriate diagnostic evaluation.” However, it appears that in the United States too few practitioners are following this recommendation when confronted with young patients who report bleeding or other symptoms.

Questions regarding cancer surveillance and screening for patients and their families commonly arise when caring for young CRC patients. The following guidelines from the ASCRS and the ACS have evolved over the last 3 decades; although there are some minor differences they are quite similar. People with a first-degree relative (parent, sibling, or child) who has had CRC or adenomatous polyp(s) are advised to have screening colonoscopy starting at age 40 or ten years younger than the age, at diagnosis, of the youngest family member with CRC (whichever comes first) with repeated evaluation every 5 years. Individuals with two second degree relatives (grandparent, aunt, or uncle) with CRC or polyps are advised to be screened as average risk patients (see below), but beginning at age 40 years. Individuals with a single second or third degree relative (cousin, great-grandparent) with CRC or polyp(s) are advised to follow average risk screening guidelines. Individuals in HNPCC kindreds are advised to begin colonoscopic evaluation starting at 20-25 years of age or 10 years before the age of diagnosis in the youngest CRC patient in the immediate family, whichever comes first[25].

In average risk patients the ASCRS[25] recommends that routine screening commence at age 50 years. As per the ASCRS, acceptable screening strategies for average risk patients include: (1) flexible sigmoidoscopy every 5 years; (2) a double contrast barium enema every 5-10 years; or (3) a colonoscopy every 10 years. The ACS recommendations are almost identical to those of the ASCRS except that they include virtual colonoscopy every 5 years as an acceptable screening option. In regards to average risk patients under age 50 years without family history, the ASCRS[25] advises that they commence annual digital rectal examination and fecal occult blood testing at age 40 years; the ACS makes no recommendations for this group.

Is endoscopic screening indicated for asymptomatic patients under 50 years of age? The rising incidence of CRC in this group and the tendency towards advanced Stage at presentation would argue in favor of such programs. Yet, the cost of screening colonoscopy programs would be very high and given the current economic climate and the fact that the incidence of CRC is still considerably lower in this group (vs the over age 50 population) the initiation of such programs is highly unlikely. However, perhaps a case can be made for a single screening sigmoidoscopy at age 40. In the present series a screening sigmoidoscopy would most likely have revealed a significant number of the neoplasms. As per the ACS, sigmoidoscopy identifies 70%-80% of individuals with advanced lesions and is associated with a 60%-80% reduction in CRC mortality for the area of the colon within its reach. Furthermore, in a recent multi-center randomized trial a single screening sigmoidoscopy carried out between the ages of 55 and 64 reduced CRC incidence by 33% and mortality by 43%[26]. Sigmoidoscopy, although invasive, is a safe procedure with a very low rate of perforation that is well tolerated without sedation. Although the chances for successful initiation of a flexible sigmoidoscopy screening program for young patients are small, the authors believe that our current dismal record with young CRC patients justifies the effort.

The lack of genetic testing for HNPCC in this study population is a clear weakness of this study. Similarly, the lack of long term cancer recurrence and survival results is also a major shortcoming. Ideally, genetic categorization of the tumors via microarray would be done which would permit a detailed analysis and comparisons.

In conclusion, this review corroborates recent national data regarding the rising incidence of CRC in the under age 50 population and the fact that a greater percentage of younger patients present with Stages 3 and 4 disease when compared to the entire CRC population. The vast majority of the young patients in this study had sporadic CRC; only 12% had a first degree family history of CRC and only 1 patient met the clinical criteria for HNPCC. The histologic breakdown of the tumors was similar to that reported in the general population of CRC patients.

Although it is impossible to confirm, it is the impression of the authors that in many patients there was a delay, either by the patient, physician, or both, in carrying out the appropriate diagnostic evaluation. The widely held belief that CRC occurs in patients age 50 and older likely contributes to this mind set. Clearly, at the very least, young symptomatic patients with rectal bleeding, a change in bowel habits, or abdominal pain should be promptly evaluated, preferably with a full colonoscopy. Finally, the medical and surgical community need to consider the concept of some type of screening program for young patients, perhaps flexible sigmoidoscopy, beginning at age 40 years. The goal is to diagnose CRC in this population at an earlier stage so that the recurrence rates and mortality can be reduced. The medical community and the public must be made aware of the fact that CRC occurs in patients under age 50 years with regularity and that thorough evaluation is indicated for symptomatic patients regardless of their age.

While the incidence of colorectal cancer (CRC) is declining in the overall population, it is on the rise among individuals under 50 years of age. Many patients and practioners, alike, believe most cases of early-age onset CRC are attributed to a family history of CRC; however, a growing number of small studies now show that the majority of these young patients have no family history.

A growing number of small studies are being published that examine the rising incidence of CRC among patients under age 50 years. It is becoming increasingly evident that the majority of these patients do not have a family history of CRC; however, it is unclear what predisposition these patients have to develop CRC at an earlier age. The research hotspot may be to perform genetic categorization of these tumors via microarray analysis to determine how they differ from sporadic CRC in older patients.

Previous studies in the literature that have examined the incidence of early-age onset CRC often include patients with hereditary, familial, and sporadic CRC’s (see below). In the study, the authors excluded patients with a known genetic CRC predisposition or history of inflammatory bowel disease. The vast majority of patients under age 50 years in the study had no reported family history of CRC and did not meet criteria by family history for hereditary nonpolyposis colorecal cancer (HNPCC). Therefore, the authors believe the majority of patients in the study had early-age onset “sporadic” CRC similar to the general population.

The study corroborates the findings of several other studies evaluating early-age onset CRC in that the majority of patients have no family history of CRC and most tumors were found in the distal colon and rectum. This contradicts the believe of many practioners that early-age onset CRC’s are more often located in the proximal colon and are attributable to inherited predispositions (i.e., hereditary non-polyposis colorectal cancer). The next logical step is to investigate what triggers the development of a “sporadic” CRC at a younger age of onset among some individuals.

Familial CRC is defined as CRC that presents 10-20 years earlier than the general population with no clear inheritance pattern. It is presumed that lower-penetrance susceptibility genes may play a role. Hereditary CRC accounts for approximately 5%-10% of patients diagnosed with CRC. Examples include familial adenomatous polyposis syndrome and HNPCC (Lynch Syndrome), which show autosomal dominant inherited germline mutations. Tumors are often located in the proximal colon and can be synchronous or metachronous in nature. Sporadic CRC occurs in patients without identifiable genetic predispositions or a reported family history of CRC.

This is a descriptive study in which the authors investigate the incidence of early-age onset CRC among two institutions. The results corroborate a growing body of literature that now suggests that most CRC’s, regardless of age of onset, occur without a particular genetic or familial predisposition.