Colorectal Cancer Risk and Recommendations for Colorectal Cancer Surveillance in Adult Survivors of Childhood Cancer

Abstract

While many organizations have published guidance on the approach to colorectal cancer (CRC) screening in average-risk and certain high-risk groups, adult survivors of childhood cancer (ASCC) who have a heightened risk of CRC are rarely included as a target group for enhanced CRC surveillance. The population of ASCC continues to grow due to increasingly effective cancer therapies and improved survival. With this increased survival comes an increased risk for subsequent malignant neoplasms, including CRC. Since there is little published guidance for CRC surveillance in ASCC and limited awareness of increased CRC risk among both physicians and patients, the objectives of our paper are to review the incidence of and risk factors for colorectal neoplasia in ASCC, describe the clinical phenotypes of colorectal neoplasia in ASCC, review published surveillance strategies based on consensus-based survivorship guidelines, and outline areas for future research to optimize surveillance strategies.

Strong evidence supports current guideline recommendations for colorectal cancer (CRC) surveillance in average and many high-risk populations.^1,2 Adult survivors of childhood cancer (ASCC), defined as individuals diagnosed with cancer under the age of 19 years, are at substantially increased risk of subsequent CRC and have not been included in guidelines with rare exceptions.^3–5

There are currently over 438,000 survivors of childhood cancer in the United States.^3,4,6 Advances in cancer therapies have improved 5-year survival in patients with childhood cancer such that studies have now begun evaluating late mortality, occurring beyond the traditional 5-year benchmark.^3,7 Several prospective cohorts of ASCC aim to identify the delayed effects of cancer treatment.^8–11 The largest 4 include the North American Childhood Cancer Survivor Study (CCSS), the British Childhood Cancer Survivor Study, a collaboration between cancer registries in the Nordic countries, and the Dutch Childhood Cancer Oncology Group-Long-Term Effects After Childhood Cancer (DCOG-LATER). A description of each series is shown in Table 1.

TABLE 1 - Cohort Studies of Adult Survivors of Childhood Cancer

Cohort name	Subjects in cohort	Eligibility criteria	Year of primary malignancy diagnosis	Duration of follow-up	Number of SMN (SIR, 95% CI)	Number of deaths
North American Childhood Cancer Survivor Study (CCSS)12,13	20,483	1. Diagnosed leukemia, CNS or bone tumor, Hodgkin’s lymphoma, Wilm’s tumor, neuroblastoma, soft tissue sarcoma 2. Diagnosis and treatment at one of the 26 participating centers 3. Age ≤21 y old at primary cancer diagnosis 4. Survival of ≥ 5 y after diagnosis	1970-1986	Baseline questionnaire in 1994, with follow-ups in 2000, 2003, 2007, 2014, and 2017	802 (6.0,5.5-6.4)	2821
British Childhood Cancer Survivor Study (BCCSS)9,14	17,891	1. Age 0-14 y old at primary cancer diagnosis 2. Diagnosed in Britain 3. Survived >5 y	1940-1991	From 5 y after initial childhood cancer diagnosis until loss of follow-up, death, or at study exit date (December 31, 2006)	278 (3.9,3.6-4.2)	3121
Nordic countries Collaboration10	47,697	1. Age 0-19 y old at the time of primary cancer diagnosis 2. From cancer registries of Denmark, Finland, Iceland, Norway, and Sweden	1943-2005	From the date of diagnosis until the date of death; date of emigration; or closing date of December 31, 2002 (Norway), 2003 (Denmark, Iceland, Sweden), or 2005 (Finland)	1180 (3.3,3.1-3.5)	NA
Dutch Childhood Cancer Oncology Group-Long Term Effects After Childhood Cancer (DCOG-LATER)11	6,165	1. Treated before 18 y old 2. Treated at a Dutch pediatric oncology and stem cell transplant center	1963-2001	5 y after childhood cancer diagnosis until the date of SMN diagnosis, date of death, date of emigration, date of loss to follow-up, or end of study (January 1, 2013)	291 (5.2,4.6- 5.8)	604

CI indicates confidence interval; SIR, standardized incidence ratio; SMN, subsequent malignant neoplasm.

ASCC are at higher risk of mortality due to the recurrence of their primary malignancy within the first decade after diagnosis.⁷ However, in later decades, ~20 to 30 years from diagnosis, mortality from subsequent malignant neoplasm exceeds the risk of death from primary tumor recurrence.⁷ Subsequent malignancies account for 15% to 20% of new cancer diagnoses in cancer registries^10,11,14 and are the second leading cause of mortality in ASCC.⁷ In the CCSS cohort, the cumulative incidence of a subsequent cancer was 20.5% (95% confidence interval (CI): 19.1–21.8%) 30 years after childhood cancer diagnosis.¹⁵ Compared with the general population, ASCC are at an increased risk of subsequent malignancies, with an overall standardized incidence ratio (SIR) of 3.3 to 6.4 compared with the general population.^8,10,16

COLORECTAL CARCINOMA AS A SUBSEQUENT MALIGNANT NEOPLASM

Multiple studies have reported an increased risk of CRC in ASCC.^14,17,18 Based on 2014 to 2018 trends, the age-adjusted estimated incidence of CRC in the general US population in 2021 is 37.85 per 100,000 men and women.¹⁹ Two prospective cohort studies report that CRC occurs in ASCC at a much higher incidence than in the general population, with SIR ranging from 3.3 to 10.9.^20,21 The highest incidence of CRC was reported in a study of adult survivors of Hodgkin Lymphoma with an SIR of 36.4.¹⁷

CRC not only occurs more frequently in ASCC, but it also develops at a younger age than in the general population. The median age for CRC diagnosis in the general population is 66 years.²² In a study of 19 ASCC, the median age at the time of CRC diagnosis was 33.3 years old (range 11.2–47.1).²⁰ Similarly, in a CCSS cohort study of subsequent gastrointestinal cancers where 53% (24/45) were CRC, the median age at the time of gastrointestinal malignancy diagnosis was 33.5 years (range: 9.7–44.8), with 57.8% of patients less than 35 years of age at the time of cancer diagnosis.¹⁸

RISK FACTORS FOR COLORECTAL CARCINOMA IN ASCC

Known cancer-related risks for subsequent CRC include radiotherapy, chemotherapy type (alkylating agents and platinum drugs), primary malignancy type, and genetic predisposition syndromes.^14,18,23 Other risk factors, such as diet, lifestyle, and environmental exposures, remain to be explored.²³

1. Radiation

Radiotherapy is a well-established risk factor for subsequent malignancy and is dose-dependent and location-dependent.^{18,20,24–26} Patients who receive abdominal radiotherapy as part of primary cancer therapy are at 4- to 11-times higher risk of developing CRC, and the risk increases significantly in ASCC greater than 40 years old.^17,18,20,27 The cumulative incidence of CRC by age 50 in patients who were treated with direct abdominopelvic radiotherapy was 1.4% (95% CI: 0.7–2.6%), which is similar to patients who have at least 2 first-degree relatives affected by CRC (1.2%).¹⁴

The risk of CRC in ASCC has been shown to increase with higher doses of radiotherapy. A French study of 36 CRC cases in ASCC demonstrated that in patients who received radiotherapy doses of 20 to 29.99 versus ≥30 Gy, the risk of CRC increased from 8.9-fold to 19.3-fold, respectively.²⁸ In another study of 19 ASCC with invasive adenocarcinoma of the colon or rectum treated with radiotherapy, the risk of metachronous CRC increased by 70% (odds ratio (OR) 1.7, 95% CI: 1.2–2.5) with every 10-Gy increase in radiotherapy dose to the colon.²⁰

CRC develops in body areas in or near prior radiotherapy exposure.¹² In a study of 19 ASCC with CRC, ASCC were 15.4 times (95% CI: 2.0–699.9) more likely to develop cancer in the segment exposed to radiotherapy compared with controls who were matched for age at diagnosis of primary malignancy and did not receive radiotherapy. Increased colonic exposure to radiotherapy was associated with an increased risk of developing CRC (OR 1.8, 95% CI: 1.2–3.2).²⁰ Further studies to assess the risk of CRC in patients exposed to contemporary radiotherapy techniques are needed to accurately quantify risk in ASCC.

2. Chemotherapy

Alkylating agents such as procarbazine and platinum are risk factors for CRC, possibly due to direct mucosa exposure to chemotherapy, although the exact mechanism is unknown.¹⁸ In a nested case-control study of 19 ASCC with CRC, univariate analysis found that alkylating agent exposure was a significant independent risk factor for CRC (OR 8.8, 95% CI: 1.2–405.4).²⁰ Another multivariate analysis evaluated a Dutch cohort of 55 HL survivors with subsequent CRC. Patients who had received infra-diaphragmatic radiotherapy and a procarbazine dose >4.2 g/m² were at very high risk of CRC development (hazard ratio (HR) 6.8, 95% CI: 3.0–15.6) compared with those who had received neither of these treatments.²⁹ In patients who had only received procarbazine, the SIR for CRC increased as the procarbazine dose increased (SIR 1.7 [95% CI: 0.4–5.1) vs. SIR 6.1 [95% CI: 2.9–11.2] for doses ≤4.2 vs. >8.4 g/m², respectively). As these patients received outdated treatment regimens, studies with modern regimens are needed to quantify risk in patients who receive chemotherapy.

3. Type of Primary Malignancy

Survivors of Wilms tumor, central nervous system tumors, and Hodgkin Lymphoma are recognized to be at high risk for subsequent CRC, which may be related to primary treatment strategies. In the CCSS cohort study where 53% (24/45) of subsequent gastrointestinal cancers were CRC, patients who were survivors of Wilms tumor, central nervous system tumors, and Hodgkin Lymphoma had the highest SIR (SIR 15.5, 95% CI: 5.0–47.9; SIR 6.8, 95% CI: 2.6–18.2; SIR 5.7, 95% CI: 3.0–11.0, respectively).¹⁸ Similarly, a study of 71 patients from the CCSS cohort demonstrated that Wilms tumor and Hodgkin Lymphoma survivors had the highest risk of CRC (SIR 25.4, 95% CI: 6.3–101.4, and SIR 12.3, 95% CI: 5.5–27.4).¹² Potential interactions between primary malignancy and treatment regimen could not be assessed due to sample size.

4. Genetic Predisposition Syndromes

Genetic predisposition and family factors are likely significant contributions to CRC development. More research is needed to understand the contribution of CRC risk due to germline pathogenic variants associated with hereditary CRC syndromes, and pathogenic variants causing extra-colonic childhood cancer syndromes and CRC. The most common hereditary syndromes associated with young onset CRC include Lynch syndrome and the constitutional mismatch repair deficiency syndrome, which are due to mono-allelic and bi-allelic DNA mismatch repair gene pathogenic variants, respectively, familial adenomatous polyposis, which is due to the APC mutation, and Li-Fraumeni syndrome, caused by a TP53 gene pathogenic variant.³⁰ Possible germline pathogenic variants related to childhood cancer and subsequent malignancy include RB1, associated with retinoblastoma and subsequent sarcoma, and TP53, seen in Li-Fraumeni syndrome but also in subsequent leukemia.³¹ The interaction of multiple low-penetrance and intermediate-penetrance genetic variants, treatment exposures, and environmental factors likely results in subsequent malignancy development.³²

An observational prospective single-center study integrated germline and somatic clinical sequencing in the management of 102 pediatric patients with refractory or recurrent solid or hematological cancer. Forty-six percent (42/102) of pediatric patients with recurrent tumors had clinically actionable findings, defined as any genomic finding that changed clinical management by providing a targetable molecular mutation (eg, BRAF mutation resulting in BRAF inhibitor initiation, or NF1 mutation resulting in MEK inhibitor initiation), change in cancer diagnosis, or the initiation of patient or family counseling due to potential risk of other cancers. Of those with clinically actionable findings, the researchers were able to act on 25% of these patients and their families, with 10% demonstrating clinical response.³³ Further studies to elucidate the interaction between CRC-specific germline pathogenic variants, somatic pathogenic variants, and childhood cancer risk are needed, as these may be clinically actionable findings that impact cancer treatment.

5. Other Factors

Other risk factors for CRC in ASCC include advancing age and higher rates of obesity due to hypothalamic/pituitary radiotherapy ≥20 Gy, lower socioeconomic status, and poor physical function.^34,35 In the general population, obesity is associated with a 30% to 70% increased risk of CRC in average-risk men ³⁶ and may likely play a role in CRC in ASCC.

CLINICAL PHENOTYPE OF COLORECTAL NEOPLASIA IN ASCC

The development of sporadic CRC occurs by the adenoma-carcinoma sequence, where somatic pathogenic variants lead to adenoma formation, with advanced adenomas (AAs) as the bridge to CRC.² AAs are defined as the presence of ≥3 adenomas, polyps with size ≥10 mm, tubulovillous or villous histology, or polyps with high-grade dysplasia. Understanding how prior malignancy and cancer treatment impacts these premalignant stages is imperative for the prevention and early detection of CRC.

Studies assessing colorectal neoplasia in ASCC report an AA detection rate range of 0.6% to 28% at a median age of 35 to 45 (Table 2), compared with a median AA detection rate of 8.9% (IQR 6.3%–12.4%) in average-risk patients with a mean age of 56.8.⁴⁴

TABLE 2 - Colonoscopy Findings in Adult Survivors of Childhood Cancer

Reference year Country	Study design	Population of interest and inclusion criteria	Control population	Primary outcome	Indication for colonoscopy	Median or mean age at the time of colonoscopy	Proportion of patients with following colonoscopy findings
Adult Survivors of Childhood Cancer
Daly37 2016 Canada	Prospective Single-arm	1. 54 ASCC* treated with ≥25 Gy RT† to abdomen, pelvis, or spine, or ≥12 Gy of RT to total body 2. RT ≥10 y prior to study entry 2. Current age 35-49	None	Proportion of patients with adenomatous polyps	Surveillance	Median age: 45 y	NAA‡: 44% (24/54) AA§: 28% (15/54) CRC∥: 0%
Glockenberg38 2017 USA	Retrospective	1. 183 ASCC diagnosed at mean age 16.6 y 2. Colonoscopy performed between 2005 and 2015	None	Prevalence of colorectal neoplasia	Diagnostic (n=118), screening (n=53), surveillance (n=12)	Median age: 34.9 y (range 6.6- 65.9)	NAA: 23% (43/183) AA: 16% (29/183) CRC: 3.8% (7/183)
Au39 2018 Canada	Retrospective Case-Control study	1. 29 ASCC treated with abdomen/pelvic/spine or total body RT >10 Gy 2. First radiation treatment 10 y before first colonoscopy 3. Age <50 at the time of the first colonoscopy	1. Average-risk patients aged 17-49 (n=58) 2. Average-risk patients aged 50-75 (n=58)	Prevalence of adenomas	Surveillance	35 y old	ASCC: NAA: 38% (11/29) AA: 21% (6/29) CRC: 0% Average-risk patients aged 17-49 NAA: 17% (13/58) AA: 6.9% (4/58) CRC: 0% Average-risk patients aged 50-75 NAA: 28% (16/58) AA: 3.4% (2/58) CRC: 0%
Rigter40 2019 Netherlands	Multicenter Cohort	1. 101 HL¶ survivors diagnosed at mean age 25 y old treated with any of the following: 1. Abdominal RT of at least para-aortic and iliac fields 2. A cumulative procarbazine dose ≥ 2.8 g/m2 3. Abdominal RT (any field or fields) and chemotherapy (any regimen)	1426 asymptomatic average-risk individuals agreed 50 to 75 y old	Prevalence of colorectal neoplasia	Surveillance	Mean age: 51 y old	Findings in HL survivors: NAA: 40.6% (55/101) AA: 13.9% (14/101) CRC: 0% SPS#: 5.9% (6/101) Average-risk patients: NAA: 2.0% (29/1426) AA: 0.6% (9/1426) CRC: 0.6% (9/1426) SPS: 0%
Therapy-Associated Polyposis
Yurgelun41 2014 USA	Case Series	5 patients treated with alkylating chemotherapy and abdominal RT	None	Initial colonoscopy findings	Patient 1: diagnostic; patient 2-5: surveillance	Mean age: 24.8 y (range 13-35)	Patient 1: 9 adenomas, 5 inflammatory polyps Patient 2: >30-40 SSPs** Patient 3: ~40 adenomas Patient 4: ~50 polyps (mix of adenomas, inflammatory polyps, and mixed hyperplastic polyps) Patient 5: “many” SSPs
Rigter42 2016 Netherlands	Case Series	3 patients treated with chemotherapy and radiation	None	Colonoscopy findings	Patient 1-3: diagnostic	Median age: 62 y (range 36-70)	Pt 1:>30 adenomatous polyps Pt 2: 21 tubular adenomas, rectal adenocarcinoma Pt 3: 36 adenomas
Biller43 2020 USA	Multicenter cohort	1. 34 patients diagnosed ≤30 y old or between ages 31-45, if the first polyp identified ≥10 y after initial cancer treatment 2. Polyposis†† without known genetic predisposition 2. Prior exposure to chemotherapy and/or RT	None	Colonoscopy findings	Surveillance	Median age: 49 y old	NAA: 100% (34/34) AA: N/A‡‡ SPS: 29% (10/34) CRC: 29% (10/34)

^*ASCC = adult survivor of childhood cancer.

^†RT = radiotherapy.

^‡NAA = non-advanced adenoma: 1-2 adenomas <10 mm in size that do not meet criteria for AA.

^§AA = advanced adenoma: ≥3 adenomas, ≥10 mm in size, tubulovillous or villous histology, or high-grade dysplasia.

^∥CRC = colorectal cancer.

^¶HL = Hodgkin Lymphoma.

^#SPS = serrated polyposis syndrome.

^**SSPs = sessile serrated polyps.

^††Polyposis = cumulative lifetime incidence ≥10 gastrointestinal polyps.

^‡‡Patients were not stratified into non-advanced versus advanced adenoma.

One of the first prospective studies that supported the need for earlier colonoscopy surveillance in ASCC was a multicenter US study that evaluated colonoscopies performed at a median age of 45 years (range 36–49 y) in 54 ASCC who received abdominopelvic radiotherapy. It revealed a 27.7% prevalence of adenomatous polyps, and 15/54 patients had tubular, tubulovillous, or serrated adenomas.³⁷ This study suggested that patients who are ASCC have a preclinical phase where potentially precancerous polyps can be detected.

In a subsequent Canadian study, the adenoma detection rate was compared between 29 ASCC who received radiotherapy, 58 average-risk controls (gender-matched and age-matched), and 58 average-risk patients 50 to 75 years of age.³⁹ In the ASCC population, the adenoma detection rate was higher (58.6%) than both the control (17.2%, P=<0.001) and average-risk groups (27.6%, P=0.009). AAs were detected in 20.7% of ASCC patients compared with 6.9% of controls and 3.4% in the average-risk population.³⁹

Advanced neoplasia (AA and CRC) has also been studied in Hodgkin Lymphoma survivors. A multicenter European study compared colonoscopy findings in 101 Hodgkin Lymphoma survivors treated with abdominal radiotherapy and/or procarbazine to 1426 asymptomatic patients aged 50 to 75 years undergoing screening colonoscopy.⁴⁰ Younger Hodgkin Lymphoma survivors (age 40–49) had a higher prevalence of advanced neoplasia compared with older controls (age 50–59) (19% vs. 9%, P=0.045). Patients who received both radiotherapy and procarbazine also had a higher prevalence of advanced neoplasia than the control group (31% vs. 12%, P=0.001). Furthermore, 6 patients were diagnosed with serrated polyposis syndrome based on World Health Organization’s 2010 Criteria.⁴⁵

Therapy-associated Polyposis

Therapy-associated polyposis is a poorly understood phenomenon due to its rarity but is currently defined as polyposis that occurs in ASCC in the absence of identifiable germline or familial susceptibility.⁴³ The phenotypic spectrum of therapy-associated polyposis remains to be clearly defined, as some studies have included patients who were diagnosed with primary malignancies up to age 35 years rather than age 19.

The first US case study of 5 patients with therapy-associated polyposis was reported in 2014. One patient who was treated for metastatic adrenal neuroblastoma at 9 months old underwent an endoscopy/colonoscopy for hematochezia at age 24. He was found to have predominantly upper gastrointestinal inflammatory polyps, along with a total of 9 cumulative colorectal adenomas and 5 colorectal inflammatory polyps. The other 4 patients diagnosed with Hodgkin Lymphoma at ages 15 to 21 had an average of >30 polyps on the first screening colonoscopy (mean age 43.5, range 33–51 y old).⁴¹ Three out of 5 patients were also diagnosed with thyroid, ductal breast cancer, and retro-peritoneal or extra-abdominal desmoid tumors. All had a history of alkylating agent treatment or radiotherapy to either the abdominal or para-aortic lymph nodes. While all patients underwent comprehensive germline testing for APC and MYH pathogenic variants, it is possible that these subjects may have had a germline pathogenic variant in a rarer polyposis gene.⁴⁶

A second Dutch case series described 3 patients with adenomatous polyposis who did not have APC or MYH gene mutations. The first patient was treated for testicular cancer diagnosed at age 31, and 39 years later was found to have ≥30 colonic adenomas. The second patient was treated for nephroblastoma diagnosed at age 10, and on surveillance, endoscopies were found to have 21 colonic adenomas total. The third patient was treated for Hodgkin Lymphoma at age 20, and at age 62 underwent subtotal colectomy for 36 detected colonic adenomas. All patients received radiation and chemotherapy, and 2 were also found to have duodenal adenomas.⁴²

A multicenter US study identified 34 patients with therapy-associated polyposis, of which 22 (65%) did not have pathogenic variants on multigene panel testing. Of all, 10/34 (29%) only had single-gene testing, and 2/34 did not get tested. Patients had a median lifetime aggregate of 32 polyps (IQR 16–52), diagnosed at a median age of 49 years old (IQR 37–54). Of the 21 patients who received radiotherapy, 5 (24%) had polyps detected before the age they would have been recommended for screening by Children’s Oncology Group-Long Term Follow-up (COG-LTFU) guidelines. Interestingly, 23/34 (74%) of therapy-associated polyposis cases had features consistent with an inherited gastrointestinal cancer syndrome. For example, 18 patients had a colorectal adenomatous polyp burden that had clinical features of familial adenomatous polyposis.⁴³ These studies further support that earlier and more aggressive CRC surveillance than endorsed for the average population is needed for ASCC.

STRATEGIES FOR RISK REDUCTION OF CRC IN ASCC

Current CRC Surveillance Recommendations

The National Comprehensive Childhood Network (NCCN) and COG-LTFU Guidelines for ASCC recommend CRC surveillance to start at an earlier age and occur with increased frequency, stratified by treatment type, age, family history, and comorbidities that inherently increase CRC risk such as ulcerative colitis (Table 3). Furthermore, if Lynch syndrome or another hereditary CRC syndrome is suspected, earlier screening is recommended according to published guidelines for hereditary risk.

TABLE 3 - Colorectal Carcinoma Surveillance Recommendations for Adult Survivors of Childhood Cancer

Organization	Highest risk parameters	Surveillance recommendations
NCCN47	Radiation to abdomen, flank, pelvis, or total body irradiation ≥20 Gy	Colonoscopy every 5 y starting at age 30 or 5 y after radiation, whichever occurs later
Children’s Oncology Group Long-Term Follow-Up Guidelines for ASCC (COG-LTFU), 2018 version48	Radiation to abdomen, pelvis, flank, or spine, or total body irradiation* Familial adenomatous polyposis (FAP) Hereditary Nonpolyposis Colon Cancer (HNPCC) Lynch syndrome Inflammatory bowel disease (IBD) Personal history of GI malignancy, adenomatous polyps, or hepatoblastoma Family history of colorectal cancer or polyps in first-degree relative	For patients with radiation history, gold standard: Colonoscopy starting at age 30 or 5 y after radiation, whichever occurs later Other options (also starting at the above time point): multi-target stool DNA test every three years Select surveillance type based on informed decision-making between patient and provider For patients at high risk due to personal or family history or hereditary colorectal cancer predisposition syndrome: screening should be performed based on current guidelines for their specific history or hereditary syndrome

*Additional factors to consider that may increase risk:

Patient factors: current age ≥45.

Cancer/treatment factors: Radiation dose ≥20 Gy, combination with chemotherapy (especially alkylators).

Comorbidities: obesity.

Health behaviors: high fat/low fiber diet.

GAPS IN AWARENESS OF CRC RISK

Areas for Further Investigation

Due to the greatly improved mortality in childhood cancer patients, the population of ASCC is rapidly expanding. Multivariate models that simultaneously assess radiotherapy dosage, chemotherapy type, age at cancer diagnosis, presence of familial predisposition syndromes, and other risk factors will help improve risk stratification in ASCC and guide CRC surveillance. Studies assessing the accuracy of noninvasive CRC screening methods such as fecal immunochemical tests and multi-target stool DNA in ASCC are imperative. Contemporary studies of high-quality colonoscopies in ASCC are needed to ascertain true non-AA versus AA detection rates, and details on the biology, tumor characteristics, stage, and survival of CRC in ASCC.

Strategies to increase awareness of the need for CRC surveillance among ASCC and their providers must be continued. Between 2003 and 2016, multiple initiatives were implemented to improve cancer surveillance awareness among ASCC and physicians. Subsequent COG-LTFU surveillance guideline adherence increased from 14.3% to 41.0%. While substantial improvement has been made, adherence to COG-LTFU is still suboptimal.⁴⁹

Current Efforts

Efforts are underway to address these gaps in our knowledge. A multicenter international study is assessing primary childhood cancer therapy-induced adverse sequelae to better understand their mechanism of development, clinical trajectory, and severity (PAR-21-329).⁵⁰ The International Guideline Harmonization Group for Late Effects of Childhood Cancer is developing recommendations for CRC surveillance in ASCC.⁵¹ The Pediatric Normal Tissue Effects in Clinic, a collaborative research group, is assessing the effects of radiotherapy dose/volume and late therapy consequences. This group will soon be providing data on CRC after pediatric radiotherapy.⁵² Finally, Activating cancer Survivors and their Primary care Providers to Increase Colorectal cancer Screening (ASPIRES), is testing a mobile health intervention to improve CRC surveillance among ASCC.⁵³

CONCLUSION

Available data demonstrate an increased risk of colorectal neoplasia, including AA and CRC, in ASCC. However, study design limitations, small cohort sizes, and evolving cancer treatment strategies beg for larger and more contemporary studies.

CRC surveillance guidelines must also be revised. Since CRC develops at an earlier age in ASCC, it is reasonable that surveillance be initiated at age 30 or 5 years after radiation therapy, whichever occurs first. Treatment regimens, including alkylating agents and primary malignancies such as Hodgkin Lymphoma and Wilms tumor should be included as risk factors for CRC. ASCC with possible germline pathogenic variants should also be discussed, as studies have suggested that patients with recurrent malignancy may have clinically actionable findings that alter treatment and/or prognosis.

Both ASCC and primary care physicians have reported decreased awareness of the risk and surveillance recommendations for CRC in ASCC. Although numerous efforts are underway, further initiatives are needed to increase awareness and delineate effective CRC risk mitigation strategies.