|Home | About | Journals | Submit | Contact Us | Français|
The proportion of cancer cases in younger patients is increasing though colorectal cancer (CRC) screening guidelines recommend starting at age 50. The national treatment patterns and outcomes of these patients are largely unknown.
This is a population-based retrospective cohort study of the nationally representative Surveillance, Epidemiology, and End Results registry for patients diagnosed with CRC from 1998-2011. Patients were categorized as being younger or older than the recommended screening age. Differences in stage at diagnosis, patterns of therapy, and disease-specific survival were compared between age groups using multinomial regression, multiple regression, cox-proportional hazards regression, and Weibull survival analysis.
Of 258,024 CRC patients, 37,847 (15%) were younger than 50. Young patients were more likely to present with regional (Relative risk ratio [RRR]: 1.3, p<0.001) or distant (RRR: 1.5, p<0.001) disease. CRC patients with distant metastasis were more likely to receive surgical therapy for their primary tumor in the younger age group (adjusted probability: 72% vs. 63%; p<0.001), and radiation therapy was more likely in younger RC patients (adjusted probability: 53% vs. 48%; p<0.001). Patients younger than screening age had better overall disease-specific survival (Hazard ratio: 0.77; p<0.001), despite a larger proportion presenting with advanced disease.
Colorectal cancer patients diagnosed before age 50 are more likely to present with advanced stage disease. However, they receive more aggressive therapy and achieve longer disease-specific survival, despite the greater proportion with advanced-stage disease. These findings suggest the need for improved risk assessment and screening decisions for younger adults.
Although colorectal cancer (CRC) is primarily thought of as a disease of older adults, this paradigm is changing. CRC is increasingly being diagnosed in young adults,1 who are too young for average-risk screening (age <50).2 In fact, it is estimated that, despite an aging population, by the year 2030, more than 1 in 10 colon cancers and nearly 1 in 4 rectal cancers will be diagnosed in individuals younger than 50 years of age.1 Current screening guidelines recommend against routine screening for “average-risk” individuals under the age of 50 years, thus younger patients are often diagnosed only after they become symptomatic. Indeed, several studies have demonstrated that young CRC patients are more likely to present with advanced disease.3–7
However, the current treatment patterns and the resultant outcomes in young CRC patients are not clear. With modern therapeutic strategies, the potential survival benefit of curative therapy for CRC in younger patients is great, but their outcomes may be worse if they are diagnosed at a later stage. The literature on this is divergent, and is limited by studies from single centers with smaller non-population-based samples,3,7,8 or older reports that do not capture modern therapeutic practices.5,6 Younger patients may also be candidates for more aggressive therapy in the absence of frailty and multiple morbidity.
In this context, we performed a population-based, nationally representative cohort study. We specifically examine the effect of having a diagnosis of CRC before the routine screening age of 50 years on: 1) disease stage at presentation; 2) treatment patterns by stage; and 3) adjusted cancer-specific survival. We hypothesize that younger patients are at higher risk of having advanced disease at diagnosis, but that they may achieve longer survival through the receipt of more aggressive therapy.
We conducted a retrospective cohort study using the Surveillance, Epidemiology, and End Results (SEER) database.9 The SEER database is the authoritative source for cancer incidence, survival, and prevalence, currently capturing 28% of the US population, and is representative of geographic, racial, and ethnic diversity. SEER collects demographic information (e.g., age, sex, and race/ ethnicity) and clinical information (e.g., primary tumor location, tumor histology, stage, treatment, and survival) from 18 cancer registries. Institutional Review Board approval is not required for publicly available data.
All patients age 20-79 years, who were diagnosed with histologically confirmed colon or rectal cancer based on Classification of Diseases for Oncology (ICD-O) site and histology codes between 1/1998 and 12/2011 were included. Patients for whom this was not their first and only malignancy were excluded.
Patients were categorized according to the age at which initiation of average-risk CRC screening is recommended into two groups – those 50 years or older and those younger than 50 years of age. This age cut-off is ubiquitously accepted as the age for which average-risk screening begins.2
Baseline clinical and demographic patient characteristics by age category were compared using chi square tests for categorical variables and two-way t-tests for continuous variables, with significance set at p<0.05.
First, we examined the association between the two age categories and the stage at presentation. We used the SEER “summary stage A” variable to stratify patients into those with: Localized disease which refers to disease confined to the colon or rectum; Regional disease which refers to contiguous and adjacent organ spread (e.g., lymph nodes, kidney, and pelvic wall); and Distant disease which refers to remote metastases. This variable is the most consistently captured variable in the SEER dataset. A multinomial logistic regression model was constructed adjusting for age, sex, race/ethnicity, marital status, tumor location, and year of diagnosis.10–12 The model yielded adjusted odds ratios (aOR) and their 95% confidence intervals (95% CI) for each covariate and its association with regional or distant disease as compared to localized disease stage at presentation.
Second, we examined the association between the two age categories and the receipt of cancer-directed surgery or radiation using logistic regression models, which adjusted for the variables in the previous model in addition to the stage at presentation. The β coefficients of the covariates generated by the multivariable model allow for the calculation of an adjusted rate for the receipt of cancer-directed therapy for each age category. This is achieved by fixing the value of the exposure variable for each patient while keeping all other covariate values as observed.
Lastly, we examined the association between the two age groups and 5-year cancer-specific survival. We used multivariable Cox proportional hazards models to obtain adjusted hazard ratios (aHR) and their 95% CI's, while accounting for sex, race/ethnicity, marital status, tumor location, stage, receipt of cancer-directed surgery or radiation, and year of diagnosis. We also used Weibull survival analyses to obtain adjusted cancer-specific survival estimates stratified by the stage at presentation for each age category, while accounting for the above listed variables in addition to the interaction between age and stage at diagnosis.
All statistical tests were two-sided and p-values<0.05 were considered significant. Statistical analyses were conducted using STATA special edition (version 13.1, StataCorp, College Station, TX).
Of 258,024 patients with histologically confirmed colon or rectal cancer from 1998 to 2011, 37,847 (14.7%) patients were too young for average-risk screening at a mean age of 42.5 ±6.0 years. There was a nearly 1:1 male-to-female ratio between the two age groups. Compared to their older counterparts, these patients were more likely to be of African American race (14.8% vs. 12%) or American Indian/AK native or Asian/Pacific Islander (10.6% vs. 8.5%; all p<0.001). Younger patients were also less likely to have right-sided colon cancer (20% vs. 31.1%; p<0.001) and generally more likely to have had rectal cancer (31.2% vs. 22.4%) as noted in Table 1.
After adjusting for sex, race, marital status, tumor location and year of diagnosis, younger patients were significantly more likely to present with regional disease than localized disease (RRR=1.37, p<0.001) and even more likely to present with distant disease than localized disease (RRR=1.58, p<0.001) compared to their older counterparts (Table 2). Of note, patients with rectal cancer were significantly less likely to present with advanced stages, compared to patients with colon cancer. The multinomial regression model allows for the calculation of adjusted proportions of stages at presentation by age category and Figure 1 portrays theses differences in stage at diagnosis between the two age groups (all p<0.001).
Table 3 shows the adjusted rates of receiving cancer-directed therapy by stage for both age categories and accounting for sex, race/ethnicity, marital status, tumor location, and year of diagnosis. While the majority of patients receive surgery for their primary tumor, younger patients with distant disease are more likely to receive cancer-directed surgery on their primary (70.8% vs 66.6%, p<0.001) than their older counterparts. The differences are more pronounced for receiving cancer-directed radiation where young patients with rectal cancer are more likely to receive radiation therapy for all stages (all p<0.001).
Overall crude 5-year survival for young patients and all stages combined was 67.7% vs. 66.8% for patients ≥ 50 years (p=0.008). However, notable differences are seen after comparing survival stage-for-stage and adjusting for sex, race, marital status, tumor location, stage, receipt of cancer-directed surgery or radiation, and year of diagnosis. The stage-for-stage effect of being too young for average-risk screening on adjusted 5-year cancer-specific survival is shown in Figure 2. These results show that adjusted cancer-specific survival is better in younger patients, across all stage groups. The 5-year cancer-specific survival for patients too young for average-risk screening was 95.1% vs. 91.9% for patients 50 and older (p<0.001) for localized disease, 76% vs. 70.3% (p<0.001) for regional disease, and 21.3% vs. 14.1% (p<0.001) for distant disease, respectively.
In this nationally representative study, we found differences in CRC outcomes and treatment patterns in patients too young for average-risk screening compared to patients 50 years and older. Importantly, we found that nearly 1 in 7 CRC patients were younger than 50. Our study confirms that these patients were more likely to present with advanced stage disease, for which they received more aggressive cancer-directed therapy, and that they achieved longer cancer-specific survival. These results forecast the imminent demand for unique survivorship care in these young, aggressively treated patients.
In the present study young patients were more likely to present with advanced stage disease, which may largely be explained by the fact that current screening guidelines do not recommend routine screening for this seemingly low-risk population.2 These results mirror what we are seeing in our clinical practice: an increasing number of very young patients with colorectal cancer, often symptomatic and with a delay in diagnosis. This emphasizes the need for close attention to young patients who present with hematochezia, anemia or a change in bowel habit, and raises the importance of public and medical community awareness of the increasing prevalence of CRC in younger patients.1,4,7 Exactly when young patients with gastrointestinal symptoms should undergo colonoscopy remains to be decided.
The stage-specific treatment patterns for younger patients are noteworthy. Receipt of cancer-directed surgery for localized and regional CRC was similar between the two age groups, but a higher rate of surgery on the primary tumor was noted for younger patients with distant CRC. Recently, Hu and colleagues13 demonstrated that utilization of primary tumor resection in CRC patients with distant disease is decreasing with the advent of effective systemic chemotherapy. The present study demonstrates the potential over-use of primary tumor resection specifically in younger patients. This may be due to more frequent symptomatic or emergent presentations, and it is also plausible that younger CRC patients and their surgeons have increased willingness to pursue more aggressive forms of therapy than older patients,3 especially that their ability to safely tolerate and complete treatments is likely higher. This phenomenon is also evident for receipt of radiation therapy, which was notably more frequently utilized in younger patients with RC across all stages. This finding is complimentary to a different study by Kneuertz and colleagues14 that demonstrate more intense use of multi-agent chemotherapy in young colon cancer patients. Although SEER does not have data on chemotherapy use, the present study adds to the growing literature on the patterns of therapy of CRC and that younger patients are indeed treated more aggressively.
Despite presenting with more advanced disease, we find that younger patients do not suffer inferior survival estimates compared to their older counterparts. On the contrary, five-year survival was greater for younger patients, both stage-for-stage and overall, perhaps associated with the more aggressive approach to therapy for advanced stage disease. However, the paradoxically superior survival of the young colorectal cancer patients was not due to improved survival only for advanced stage patients, but rather was a phenomenon present across stages.
This study has several limitations. First, we cannot ascertain causation from this observational study, as overall survival and treatment choice is subject to considerable selection bias that cannot be entirely mitigated with risk adjustment. It is also possible that a higher proportion of young CRC patients have a hereditary cancer syndrome, and CRC associated with such syndromes may behave less aggressively.15,16 We were unable to evaluate this point specifically as cancer genetics and family history are significant risk factors for CRC,17 but are not available in SEER. Finally, although there were more African Americans in the younger age group, the magnitude of this difference cannot in and of itself explain the marked differences in incidence and survival, knowing that African Americans are at an inherently increased risk for CRC.18 Notwithstanding these limitations, the findings are informative and relevant as the national patterns of care and cancer outcomes are largely unknown for this growing patient population.
From a policy standpoint, these findings suggest the need to improve current medical practice, and to study the possibility of changing screening practices. In the realm of current practice improvement, this study is a wake-up call to the medical community regarding the increasing prevalence of CRC in young patients. We must avoid the currently common delay in diagnosis for young patients with red-flag symptoms of CRC such as a change in bowel habit or hematochezia not typical of a benign anorectal source. In addition, more attention to patients' family history, as the most identifiable clinical risk factor, would reveal many non-average-risk patients not currently undergoing earlier age screening, for whom it would be recommended.
Unfortunately, it is not clear from this study or other published studies how to target “average-risk” patients for whom earlier-age CRC screening would be appropriate. Future clinical and basic science research will be required to identify characteristics of early-onset CRC patients without a family history that might be useful for risk-stratification. We can imagine a future in which more universal genetic testing for cancer and other disease risk factors would allow for such risk stratification. For average-risk screening, the increasing rate of early-onset CRC would suggest that we should consider low-cost/low-risk screening with a fecal immunochemical test kits or fecal DNA tests for the entire population at an earlier age (e.g., 40 or 45 years). The cost-effectiveness of such a strategy could be formally tested in future studies.
In conclusion, in this nationally representative study nearly 1 in 7 CRC patients were diagnosed before the age of average risk screening. A diagnosis of CRC in this population is associated with a higher likelihood for advanced stage disease, more aggressive cancer-directed therapy, and a slightly better cancer-specific prognosis suggesting some compensation for later diagnosis. These results suggest the need for future research on the cost-effectiveness of earlier-age screening policies for CRC with low-risk/low-cost fecal testing, and further research on survivorship care in these young, aggressively treated patients.
Colorectal cancer patients diagnosed before age 50 are more likely to present with advanced stage disease. However, they receive more aggressive therapy, and achieve longer disease-specific survival, suggesting some compensation for later diagnosis.
Funding/Support: ZMA is supported by AHRQ T32 HS000053-23. SLW is supported by AHRQ 1K08 HS20937-01 and American Cancer Society RSG-12-269-01-CPHPS. SER is supported by the American Society of Colon and Rectal Surgeons Research Foundation. SH is supported by NIH/NCI 1K07 CA163665-22 and the American Society of Colon and Rectal Surgeons Research Foundation.
Role of the Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Disclosures: The authors have no disclosures to make.
Previous presentation: This work was presented in part as a poster presentation at the American Society of Colon and Rectal Surgeons annual meeting on June 1, 2015, Boston MA.
Author Contributions: Drs. Abdelsattar and Hendren had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.Conception & design: Abdelsattar and Hendren
Data Acquisition: Abdelsattar
Analysis and interpretation: All authors
Drafting the manuscript: All authors
Critical revision: All authors
Final Approval: All authors.
Conflict of Interest Disclosure: Nothing to disclose