1. INTRODUCTION
Advances in understanding the molecular pathogenesis of cancer, along with the discovery of genes involved in oncogenesis and the evolution of next-generation sequencing have led to the development of precision medicine in cancer care. Precision medicine is a broad term for a scientific process that leads to targeted therapies based on individual patient and tumor characteristics like molecular mutations, gene expression, and other features. Precision medicine is tailor-made and targeted to the patient’s specific cancer biology. The increased understanding and rapid discovery of tumor genomic targets have made genome-driven cancer treatment the most interesting treatment approach globally.
Despite tremendous advances in cancer care, gastrointestinal (GI) cancers continue to have a poor overall prognosis with very high mortality globally. Colorectal cancer (CRC) is the third most common cancer and the third leading cause of cancer death in the United States, with an estimated 151,030 new cases and 52,580 deaths.1 Fortunately, the incidence of CRCs has decreased from 60.5 per 100,000 people in 1976 to 46.4 in 2005 and, more recently, 38.7 in 20162,3 along with a decrease in mortality.3,4 Paradoxically, while incidence among people older than 65 years decreased 3.3% annually between 2011 and 2016, the incidence increased among those younger than 65 years, with a 1% annual increase in those aged 50 to 64 years and 2% annual increase in those less than 50 years. The death rates also mirrored this age-dependent trend.3 Alarmingly, for patients 20 to 34 years of age, the estimated incidence rates for colon and rectal cancers will increase by 90.0% and 124.2%, respectively by 2030. It is more disconcerting that the cause of this trend is obscure.
Inspired by molecular targeted and immunotherapy in advanced melanoma and non-small cell lung cancers, there have been numerous studies applying precision medicine in GI malignancies, specifically CRCs. Targeted therapies, development of biomarkers, immunotherapy, and other newer treatment modalities in CRCs have lagged other malignancies but are rapidly catching up. Advances in precision medicine have also demonstrated that CRC in young patients is clinicopathologically and genetically different from CRC in older adults.5–7 In this article, we summarize the role of precision medicine in colorectal cancers with regards to screening and diagnosis, disease management, surveillance, and future perspectives. We do not intend this article to be a review of the management of CRCs.
2. SCREENING
The current stratification of CRC risk into high, moderate, average, and low is primarily based on age and family history. Screening falls into two broad categories based on available technology: structural tests (endoscopy and less commonly radiologic imaging) and stool/fecal-based tests.8 With very high sensitivity and specificity for detecting cancers and precancerous lesions, colonoscopy with tissue sampling is the gold standard for screening of CRCs. Unfortunately, the limitations and disadvantages of colonoscopies that include their invasive nature, the need for bowel prep, time involved, sedation requirements, and risks of perforation and bleeding have made it not universally acceptable. The risks of adverse events after colonoscopy increase with age as shown in a large cohort study of 53,220 Medicare patients between the ages of 66 to 95 years.9
Stool-based tests detect signs of CRC in stool specimens, traditionally occult blood, and more recently, a combination of occult blood and alterations in exfoliated DNA. Two types of Fecal Occult Blood Test types (FOBTs) are currently available: guaiac-based and immunochemical (FIT). Endoscopic surveillance detects more advanced neoplasms (like polyps) than stool testing, but at the cost of lower participation, as shown by a meta-analysis of 14 randomized controlled trials and other controlled studies.10 The COLONPREV study compared one-time colonoscopy with biennial fecal immunochemical testing (FIT) in asymptomatic adults aged 50 to 69 years, and showed that both identified similar numbers of cancers in the initial screening, but colonoscopy identified significantly more advanced and non-advanced adenomas, with more participation in the FIT arm (34.2% vs. 24.6%; P < .001).11 Stool-based tests, however, have disadvantages with sensitivity being limited by inadequate specimen collection or suboptimal processing and interpretation and the need to have a colonoscopy if a test is positive. A combination of multitarget stool DNA (mt-sDNA) and occult blood test has emerged as an option for CRC screening [Cologuard® (Exact Sciences)]. This test screens for DNA alterations associated with colorectal carcinogenesis (KRAS mutations, aberrant NDRG4, and BMP3 methylation) in tumor cells sloughed into the stool, and tests for occult blood as measured by immunoassay.12 The mt-sDNA test was more sensitive than FIT in the detection of CRC, advanced precancerous lesions, polyps with high-grade dysplasia, and sessile serrated polyp greater than 1 cm, as shown in a study that included 9989 participants at average risk for CRC, each of whom underwent FIT, mt-sDNA testing, and a colonoscopy.13 However, FIT had significantly higher specificity than mt-sDNA test among participants with non-advanced or negative findings, and many more participants were excluded because of problems with mt-sDNA testing (n=689) than because of problems with FIT (n=34).12
Developments in precision medicine have led to novel risk prediction models that include genetics and other information to further refine the risk of developing CRC. The addition of a genetic risk score (G-score) determined by 27 validated common CRC susceptibility loci identified from GWASs (genome-wide association studies) to sex, age, family history, and previous endoscopies increased discriminatory accuracy from 0.51 to 0.59 for men and 0.52 to 0.56 for women.14 Another model of CRC risk was developed by incorporating 63 CRC-associated SNPs (Single-Nucleotide Polymorphisms) to 19 lifestyle and environmental factors using data from the Colorectal Transdisciplinary Study and Genetics and Epidemiology of Colorectal Cancer Consortium. This study showed increased discriminatory accuracy of 0.53 to 0.63 for men and 0.54 to 0.62 for women.15
Liquid biopsies have a somewhat limited role in the screening of cancers to date due to the low concentration of circulating biomarkers associated with a low tumor burden.16 Nevertheless, the noninvasive nature of this approach makes it promising. Epi proColon® 2.0 assay (also called mSEPT9 assay) tests for methylated septin9 DNA in plasma17 and was FDA-approved in April of 2016 for CRC screening for those who refuse other screening modalities. FIT was compared to SEPT9 DNA methylated blood test for CRC screening and found that the specificity for CRC detection was higher for FIT (97.4% vs. 81.5%, respectively) but the sensitivity for CRC detection was not significantly different (68% vs. 73.3%, respectively).18 One study showed that more participants were willing to complete the blood test compared to FIT for CRC screening in average-risk adults.19
Direct visualization through colonoscopy, flexible sigmoidoscopy, and CT colonography (an acceptable alternative), and stool-based tests continue to be the recommended modalities of screening by USPSTF. Newer risk prediction models will play a significant role in more accurately stratifying risk in CRCs. Currently blood-based tests only have an experimental role in CRC screening, but due to being less invasive and faster, have the potential to increase screening rates.
3. DIAGNOSIS
For decades the conventional way of diagnosing cancers is through tissue biopsies that undergo histological analysis, molecular and more recently gene mutation testing.20 This method has shortcomings that include patient inconvenience and intratumoral heterogeneity, and hence impractical in certain situations.20 Liquid biopsy is an analysis of tumor-derived biomarkers from body fluids of cancer patients like blood, pleural fluid, ascitic fluid, CSF, and urine, and has emerged as an alternative to this classic approach.21 However tissue-based biopsy continues to be the gold standard of diagnosis with liquid biopsy methods playing a role in disease monitoring and mutation evaluation. Circulating tumor DNA may have a role in the diagnosis of CRC as shown by one study that showed a significantly higher and broader range for circulating tumor DNA (22–3922 ng/mL) when compared to healthy subjects (5-16 ng/mL).22 Cancer Type IDR is a new genomic test available for accurate diagnosis of metastatic cancers when there is a diagnostic ambiguity. It uses real-time RT-PCR to measure the expression of 92 genes in the tumor and compares it to a database of 50 known tumor types and subtypes, including GI cancers.23
Recent advances have helped identify biologically distinct molecular subtypes of CRC based on gene expression analyses. CRCs have been classified into four consensus molecular subtypes (CMS): CMS1 with good prognosis- enriched for MSI tumors with marked immune activation, CMS2- classical subtype encompassing higher CIN and strong WNT/MYC-driven tumors with epithelial characteristics, CMS3- enriched for KRAS-mutated tumors with activation of metabolic pathways, and CMS4 with poor prognosis- with mesenchymal features, high stromal content and activation of TGF-β and VEGFR pathways.24 CRCs can also be classified into intrinsic subtypes (CRIS) distinguished by specific molecular, functional and pathogenic features; (1) CRIS-A: Mucinous subtype, glycolytic metabolism, with marked MSI, mutated BRAF or KRAS; (2) CRIS-B: Active TGF-β signaling, epithelial-mesenchymal transition, bad prognosis; (3) CRIS-C: High EGFR signaling, and response to EGFR inhibitors (i.e., cetuximab); (4) CRIS-D: High WNT signaling, IGF2 gene amplification/ overexpression; and (5) CRIS-E: Paneth-like phenotype and TP53-mutated genotype.25 Analysis of these specific subtypes can predict treatment response, and improve treatment precision.26,27
Gene Sequencing- The landmark cancer genomics program, The Cancer Genome Atlas (TCGA), generated over 2.5 petabytes of genomic, epigenomic, transcriptomic, and proteomic data. In the past decade advances in next-generation sequencing (NGS) has enabled the sequencing of whole-genome, whole-exome, whole-transcriptome, and RNA, as well as the detection of large genetic aberrations.28 NGS panels or profiles contain most mutated genes and more importantly actionable genes, provide diagnostic and prognostic data, and aid in the selection of potential treatment regimens.16 NGS can detect increased tumor mutation burden (TMB- a high rate of somatic mutation) in some solid tumors which correlate with response to immunotherapy.29,30 DNA mismatch repair (MMR) deficiency and subsequent hypermutated phenotype tumors lead to microsatellite instability (MSI) which also has a good response to anti-PD1 therapy in GI cancers.31 In colorectal tumors there is a marked increase of mutations in transforming growth factor-beta (TGF-β) signaling genes and BRAF32 as well as mutations in DNA polymerase D1 (POLD1) and DNA polymerase E (POLE) genes.33 Left-sided and right-sided CRCs have distinct tumor characteristics as well as prognosis and different responses to targeted therapy. NGS has shown different mutations based on tumor side and location with 70% RAS mutations in cecal tumors but only 57% in ascending colon and 43% in hepatic flexure tumors. BRAFV600 mutations were seen in 10% of cecal, 16% of ascending colon, and 22% of hepatic flexure tumors, and PIK3CA mutations in 26% of descending colon but only 14% of sigmoid and 9% of rectosigmoid tumors.34 KDR/VEGFR2 somatic mutation is a potential genetic biomarker of patients’ responses to antiangiogenic cancer therapies.35 ALK, ROS1, and NTRK rearrangements were identified and mostly seen in older patients with right-sided tumors, locally advanced, RAS wild-type, and MSI-high cancers and has very poor outcomes.36
KRAS, BRAF, PIK3CA, TP53, CTNNB1, APC, SMAD4, and PTEN are among the most altered genes in CRC and these mutations vary among Asians and the United States population. Differences in ERBB2, APC, TP53, CDKN2A, and NRAS mutations, and genomic alterations in DNA repair genes (e.g., ATM, BLM, BRCA2, NBN, NRE11A) were noted in Japanese patients as compared to the western population with CRC.37 Currently, next-generation sequencing is mostly limited to metastatic disease but has an evolving role in early-stage disease as it can provide data to predict the risk of recurrence and enable tailored adjuvant chemotherapy to reduce this risk.38
4. DISEASE MANAGEMENT
4.1. Therapeutics
Conventional 5-FU based chemotherapy alone was the backbone of CRC treatment for decades and it had unpredictable and varied responses. Targeted therapy was developed and added to conventional chemotherapy for treating CRCs and has dramatically improved outcomes. Bevacizumab (a VEGF Inhibitor), cetuximab, and panitumumab, (EGFR inhibitors) added to a 5-FU based chemotherapy upfront is the backbone for treatment of advanced, and unresectable CRCs currently. Cetuximab and panitumumab have been shown to benefit only left-sided tumors with no KRAS/NRAS/BRAF mutations in the first-line setting. For HER2-amplified and RAS and BRAF WT CRCs, trastuzumab + [pertuzumab39,40 or lapatinib41]) or fam-trastuzumab deruxtecan-nxki42 is the frontline treatment for patients not appropriate for intensive therapy and as the second line for all patients. Cetuximab or panitumumab is also recommended as a single agent in the first-line setting for KRAS/NRAS/BRAF WT and left-sided tumors in patients with poor general health. For BRAF V600E patients, the combination of encorafenib and cetuximab or panitumumab can be used in the second-line setting. The addition of MEK inhibitor, binimetinib, to cetuximab and encorafenib was studied in the randomized, phase III BEACON trial and showed no improvement in OS or ORR.43,44 For tumors positive for the rare NTRK gene fusion entrectinib or larotrectinib are targeted agents. TMB is a potential biomarker for response to immunotherapy and pembrolizumab has been FDA-approved for patients with unresectable or metastatic, TMB-high (TMB-H, 10 or more mutations/megabase) solid tumors that have progressed following prior treatment and have no satisfactory alternative treatment options.45 This was based on results from the phase 2, KEYNOTE-158 study46 in which none of the 796 patients who were evaluated for efficacy had CRCs. However, pembrolizumab significantly prolonged progression-free survival than chemotherapy in first-line therapy for MSI-H–dMMR metastatic CRCs.47 NCCN does not currently recommend TMB biomarker testing for CRCs unless measured as part of a clinical trial. In stage II and some stage III CRCs, MSI-H and/or dMMR are predictive biomarkers that predict lack of benefit for adjuvant 5-FU-based chemotherapy.48–51 Adjuvant chemotherapy is not recommended for low-risk MSI-H/dMMR stage II CRC patients due to the mostly good prognosis of these patients and a lack of treatment benefit.48
Liquid biopsies post initial diagnosis from tissue have led to improved therapeutic interventions and hence improved patient outcomes and overall survival.52 It can be used for detection of MRD (minimal residual disease), drug selection including sensitivity, monitoring recurrence, and response to targeted agents.53 Circulating tumor cells (CTCs) are tumor cells that circulate, either as a single cell or in clusters, to metastatic sites after they are detached from primary or metastatic lesions.16,54 Experiments on the establishment of colon CTC cultures and permanent cell lines may provide genetic and epigenetic information on tumor biology, and probably assess sensitivity to anticancer drugs.55 Circulating tumor DNA (ctDNA) analysis has the potential to guide specific systemic chemotherapy and targeted agents. Emerging RAS mutations while on treatment with anti-EGFR antibody reveal resistance in patients with metastatic CRC (mCRC) and hence serially monitoring for RAS mutations in peripheral blood of RAS wild-type patients may be imperative.56 Mutant KRAS DNA was detected in the peripheral blood of patients who were initially wildtype while receiving monotherapy with panitumumab.57 OncoBEAM RAS CRC Assay identifies the (cell-free) cfDNA of the most frequent KRAS and NRAS mutations by using BEAMing technology and thus can replace tissue testing for RAS mutation.58 In anti-EGFR treatment-refractory patients, detecting treatment-induced genetic alterations by sequencing ctDNA could identify biomarkers for treatment screening.59 A single-arm phase II trial (eSCOUT - Erbitux Study of CPT 11, Oxaliplatin, UFT oral Targeted-therapy) in patients with previously untreated KRAS wild-type advanced CRC showed that stratification of patients by CTC count identified those who might benefit the most from an intensive four-drug regimen (irinotecan, oxaliplatin, and tegafur-uracil with leucovorin and cetuximab), avoiding high-toxicity regimens in low-CTC groups.60
4.2. Monitoring and surveillance
The conventional method of disease surveillance and treatment response in CRCs is through CT imaging and serial CEA monitoring. Precision medicine has an evolving role in disease management and surveillance in the metastatic as well as early-stage setting. In patients with early-stage surgically treatable cancers, liquid biopsy has helped in the detection of MRD. Detection of postoperative circulating tumor DNA indicates the presence of occult tumor cells. In a study of 65 CRC patients who underwent surgery, tumor-specific DNA mutations like somatic KRAS and BRAF in the peripheral blood were more frequently detected in patients who had R2 resection (macroscopic residual disease).61 Presence of this mutation in the peripheral blood may be a good estimate of surgical clearance of the disease.61 Postoperative detection and serial monitoring of ctDNA in CRC can be used to monitor for residual disease, inform the response to intervention, and predict relapse risk with a high probability.62
Results from large observational trials, such as TRACC63 (NCT04050345) and ADNCirc (NCT02813928) will help establish the standards for ctDNA as a marker for MRD and also show that ctDNA clearance is a surrogate marker for survival. The IMPROVE-IT2 (NCT04084249) trial is testing the hypothesis that ctDNA guided post-operative surveillance combining ctDNA and radiological assessments could result in earlier detection of recurrent disease and identify more patients eligible for curative treatment. Ongoing trials of stage II CRCs, like the COBRA study (NCT0406810), the CIRCULATE trial (NCT04120701), and the DYNAMIC-II study (ACTRN12615000381583), are testing if ctDNA will enable the identification of patients with a high risk of disease recurrence who might benefit from adjuvant chemotherapy. While the DYNAMIC-III study (ACTRN12617001566325) is evaluating the clinical utility of chemotherapy de-escalation or escalation as informed by ctDNA status in stage III colon cancer, the BESPOKE Study (NCT04761783) is examining the impact of ctDNA on treatment decisions on tumor assessment timepoints after initiation of immunotherapy. Several other trials are ongoing to establish that ctDNA is an acceptable surrogate marker of survival and adjuvant therapy effectiveness (Table 1).64
Gene assays also have a role in disease surveillance. In an observational study, a novel 2-gene (methylated BCAT1 and IKZF1 DNA) blood test was more sensitive for recurrence than CEA, with twice the odds of recurrence with a positive BCAT1/IKZF1 result compared to a positive CEA result.68 Multigene assays like MSK-IMPACT, NCC Oncopanel, Todai OncoPanel, Oncomine Dx Target test, Foundation OneCDx, and CANCERPLEX can estimate the risk of relapse after definitive surgery.16 In metastatic CRC, detection of KRAS, GNAS, and SMAD4 mutations in peripheral blood were significantly associated with lung metastasis.69
ctDNA sequencing may improve clinical trial enrollment rates as well. Japanese researchers compared clinical trial enrollment using ctDNA sequencing in 1,687 advanced gastrointestinal (GI) cancer patients in the trial SCRUM-Japan GOZILA (no. UMIN000016343) to enrollment using tumor tissue sequencing in the same centers and network (GI-SCREEN, 5,621 patients), and found that genotyping of ctDNA significantly shortened the screening duration (11 versus 33 days) and improved the trial enrollment rate (9.5% versus 4.1%) without compromising treatment efficacy compared to tissue genotyping.70
6. FUTURE DIRECTIONS
Precision medicine is in its nascent stage and with advances in molecular imaging, tumor gene sequencing, and advanced bioinformatics capability, it will expand in the immediate future. Organoids are complex three-dimensional (3D) multicellular stem cell-derived constructs generated in-vitro and mimic their corresponding organ in vivo.71 Patient-Derived Organoids (PDOs) can effectively recapitulate tumor-specific characteristics and may play a valuable role in precision oncology.72 Scientists created a CRISPR-mediated knock-out panel of all RASGAPs (RAS GTPase-activating proteins) in CRC PDOs and found that only loss of NF1, and no other RASGAPs, results in enhanced RAS-ERK signal amplification and improved tolerance towards limited EGF stimulation. This suggests that NF1-deficient CRCs are likely not to show resistance to anti-EGFR monotherapy and can be a biomarker for progression.73
Gut microbiota affects immunity, metabolism, and tissue development74–76 and has a promising role in GI cancers as shown by the association of Fusobacterium and response to chemotherapy.77,78 A large population of F. nucleatum was shown to be associated with a high risk of CRC recurrence and may serve as a prognostic marker.79 The gut microbiome enhances the effect of immunotherapy80–85 but can also be associated with the adverse effects of immunotherapy, especially colitis.86,87 The impact of the gut microbiome on immunotherapy depends on the bacterial strains in the gut.88 Researchers at MD Anderson found that responders to immunotherapy in Melanoma have a higher abundance of Faecalibacterium and Ruminococcaceae bacteria than nonresponders.83 Prospective studies should explore the relationship between the gut microbiome and immunotherapy in CRC patients. Future immunotherapy approaches will be focused on creating a functional anti-tumor immune system by using different immune cells like NK and T cells, and various immune mechanisms like targeting specific surface antigens, T-cell chimeric antigen receptor technology (CAR-T), dendritic cell vaccine, or immune checkpoint inhibitors.89
Artificial intelligence and the capability to analyze big data will enhance our knowledge about tumor genomics, immunological landscape, molecular characteristics, and expedite the development of targeted drugs, and associated predictive and prognostic biomarkers. Comprehensive genomic profiling by DNA whole-exome sequencing, RNA whole transcriptome sequencing, and immunohistochemistry can be generated, and using highly advanced bioinformatics and machine learning, we can identify unique signatures of individual cancers at the molecular level. This signature can predict benefits from a specific therapy that is unique and tailored for the individual. Using CRC patients’ gene expression data of HTSeq-FPKM and matching data from The Cancer Genome Atlas (TCGA) datasets, researchers identified nine new biomarkers that are independent prognostic indices for CRC patients and can aid in accurate survival evaluation of CRC patients.90
Barriers like the costs of these sophisticated tests, access, inadequacies in healthcare institutions, and socioeconomic barriers among minority patients91 are major limiting factors and will exaggerate the current inequality in cancer care. The scarcity of resources in the developing world92 and a lack of genomic and molecular data will limit the benefit of precision medicine to the developed world. Awareness of these barriers by physicians, policymakers, and scientists will lead to a multi-level effort to address this and may help attenuate some of the disparities. Researchers and scientists should include patients from Low and Middle-Income Countries (LMICs) in international studies. A global effort, funded by high-income countries or international philanthropic organizations, is needed to ensure representation from LMICs and identify ways of implementing cost-effective approaches to precision medicine. Mandating that polyomic data is deposited in international databases will allow access to researchers in LMICs and the worldwide scientific community. Through high-throughput drug screening on organoids and other models, re-classification and categorizing diseases based on their molecular signature, with next-generation sequencing, proteomic analysis, active efforts to reduce costs, and with improved access, we can expand the armory of precision medicine.
7. CONCLUSION
Precision medicine has evolved tremendously over the past few decades and continues to enhance the accuracy of cancer diagnosis and develop a newer and more precise classification of this deadly disease. It has changed the landscape of CRC disease management and has an evolving role in screening and initial diagnosis. Identifying unique characteristics and the development of a molecular signature of tumors that can be targeted by the least toxic and most precise therapeutic agents has become the standard of care. An increase in the incidence of young CRCs is an area of concern and future research is needed to be done to clearly identify the characteristics of young colorectal cancers. Improving access to precision medicine among the developing world should be made a priority.
Conflict of Interest
None
Funding information
N/A
Ethical statements
N/A
Acknowledgments
N/A
Author contribution
i. Janeesh Sekkath Veedu, Jill Kolesar: conception and design
ii. Janeesh Sekkath Veedu, Chaitanya Iragavarapu: data collection and assembly and manuscript writing.
iii. Jill Kolesar, Chaitanya Iragavarapu: data analysis, manuscript writing.
All authors have approved the manuscript.