Personalized medicine in colorectal cancer
History and background of personalised medicine in CRC
Medicine for a long time has been practiced under 'one-size-fits-all' paradigm, though throughout centuries there has been some kind of personalised medicine depending on the age or body mass index or gender… [1]; but not until recently it hasn’t been more evident that each individual have widely different and unique characteristics at the molecular, physiological, environmental exposure and behavioural levels, things like DNA sequencing have further provided this diversity and variation in the disease process and progression, so this things should be investigated considered in order to determine the fittest therapy for that individual[2].
Despite being the fourth most common type of cancer, Colorectal cancer hasn’t seen much improvement treatment wise, though in the last decade there has been some improvement but the situation was almost static in a way that, the 5-year survival probability since 1996 was stagnant at a rate of 60–65% till 2014[3], And since its discovery, 5-fluorouracil (5-FU) has been the most effective treatment and usually combined with some other agents[4], Nevertheless, the response to this agent is not that great with 80% of the patients being unresponsive to 5-FU[5]. So in order to spare the patient from the side effects of these agents, there as been an increasing focus on determining which patient group will respond to which agents and vise versa, hence personalizing the medicine. One of the earliest works on this was a study in 1998, whose results contradicted prior studies, in this study (which was done in vivo), it claimed that decreased levels of thymidine phosphorylase (TP) were associated with better response to 5-FU in contrast to prior studies (which were done in vitro)[6]. Another study by Scherf et al in 2000, showed that increased level of the rate limiting enzyme in metabolism of 5-FU, Dihydropyrimidine dehydrogenase (DPYD, encoded by DPYD) is negatively correlated with 5-FU response as the exposure time of the tissues to 5-FU will be decreased[7]. in mid 2003, where thioredoxin-1 was found to be a prognostic factor of poor outcome hence starting a quest for finding thioredoxin-1 inhibitor to be used as a treatment for those patient groups[8]. In that same year a retrospective study was done assessing, Thymidylate synthase (TS), Ki-67, and p53 to determine their prognostic value, and was found that though they bear a substantial prognostic value but they weren’t any different (for better or worse) in response to adjuvant chemotherapy[9]. A research conducted in Albert Einstein Cancer Center, was yet another step towards further personalizing medicine and using of developed techniques to determine those factors, presenting that overexpression of proto-oncogene c-Myc directly proportional to the response of camptothecin-induced apoptosis[10]. Looking at another way of personalized medicine, was to evaluate the adverse effects of a chemotherapy in an individual and come up with investigations to be done beforehand to asses risk/benefit ration; Van Bebber in 2006 investigated a way to assess the extent of irinotecan (Which is a chemotherapy used for advanced CRC) adverse effects, through UDP glucuronosyltransferase (UGT1A1) genotyping, later (in July 2005), it became US FDA approved, stating that UGT1A1 homozygous genotype should be considered for a lower initial starting dose of irinotecan[11]. Since then personalised medicine has come a long way and it has never looked more promising, and we will be presenting the current and emerging techniques to further personalise medicine.
Brief introduction of Heterogeneity in CRC
Colorectal cancer is a one of the most heterogenous diseases, this heterogeneity can present as either intertumoral, that being variability between patients with the same histological classification or intra-tumoral heterogeneity that can demonstrate as having heterogenous cell populations inside the same tumor that may demonstrate different gene expression or different genetic mutations[12]. Intra-tumor heterogeneity can be further classified as spatial and temporal heterogeneity, Spatial heterogeneity stating the difference in genetics between cells in the same tumor or genetic heterogeneity between the primary and metastatic tumor, and Temporal heterogeneity describes the change in genetic composition of tumor over the time[13]. Using single-cell RNA sequencing (scRNA-seq) Weier Dai in his study found 5 clusters; cluster one being mostly immune cells, cluster two contained genes responsible for major histocompatibility complex, cluster three of cells harbored genes responsible for stabilizing the cell, energy transportation and cell regulation (such as TSPAN6, PFDN4,andTIMM13), cluster four comprised of cells breakdown of extracellular matrix and remodeling of tissues and cluster five had genes that are expressed in cancer cells like WFDC2[14] Figure (1). This intra-tumor categorization could shed light on better prognostic predictors and even more personalized medicine [15]. One origin of the Intertumoral heterogeneity is the genetic alteration that give rise to the tumor, we broadly have three mechanisms leading to CRC, chromosomal instability (CIN), microsatellite instability (MSI) and CpG island methylator phenotype (CIMP), with that being said most tumors have different composition of different genetic variation, in addition tumor microenvironment also plays a role in heterogeneity through its effect on tumor spread and growth[16]. With regard of RAS proto-oncogenes family (KRAS, NRAS, and HRAS), and their downstream effect on epidermal growth factor receptor (EGFR), studies show only 40-45% of the CRC demonstrate alterations in KRAS KRAS proto-oncogene, observed mainly in codons 12 (G12D, 13%; G12V, 9%) and 13 (G13D, 8%) of exon 2[17], NRAS mutations alongside KRAS mutations were present in 1–3% of all CRC and no reports on HRAS mutations, keeping in mind that mutations in either KRAS or NRAS will lead to malignant transformation of the CRC through activation of EGFR[13]. Presently whether there is KRAS mutation or not determine the usage of anti-EGFR agents for CRC[18], that’s because KRAS and NRAS mutations are downstream of the EGFR pathway[13]. Differences in genotype and phenotype also are present in association to the site of the tumor (Right-sided or left-sided) owing to the fact that different genetic alterations result in the localization of the tumor[19]. The left sided tumors show CIN and also activation of the EGFR pathway, in contrast right-sided were observed to poses BRAF mutations and MSI to a higher degree and tends to develop in certain groups with genetic predisposition (e.g., Lynch syndrome)[20].
Figure 1 Intra-tumor heterogeneity; Different cell populations within one tumor
Omics technologies for personalised medicine
As a part of “The Cancer Genome Atlas” project, first large scale and inclusive profiling for Colorectal cancer genomics was done in 2012, it showed that 16% were hypermutated and part excluding these hypermutated tumors, the rest had similar genomic alteration, apart from the well established gene mutations (i.e. APC, TP53, SMAD4, PIK3CA and KRAS) some new ones were spotted like SOX9, FAM123B and ARID1A also fusion of NAV2 and WNT pathway member TCF7L1[21]. Ironically the more mutation load a CRC has the more chance its has in the survival of the patient[22], owing to that a tumor with more mutations will produce more neoantigens that can be recognized and targeted by tumor-infiltrating lymphocytes also more memory T-cells(Figure-1)[23], though through selective pressure, tumors has evolved mechanisms to evade immune response, through many mechanisms (i.e. down regulation of MHC-1 or T-cell through inhibition immune checkpoint molecules or tumor suppression via tryptophan metabolism modulation)[24], making these types vulnerable to immune checkpoint inhibitors[25]. Regarding Metabolomics, when comparing normal colorectal mucosa with colorectal cancer tissues, Chan et al found that there were lower amounts of arachidonic acid, malate and fumarate, saturated and unsaturated fatty acids and lipids in the CRC tissue and higher levels of phosphocholine, phosphatidylcholine, phosphatidylethanolamine, lactate and glycine in the CRC, complying that generally CRC possessed higher activity and had more quantities of substrates needed for glycolysis and hypoxia, nucleotide biosynthesis, lipid and steroid metabolism, and inflammation.[26] While Warburg’s effect was thought to imply that high glycolytic rate in presence of oxygen meant a defective mitochondrial system but recent works has shed a light on this and states that actually the TCA cycle intermediates are needed for anabolism and also a study found that knocking out components of Electron transport chain in Kras-mutant mouse lung adenocarcinoma cells suppresses tumour growth while the gaining of functional mitochondrial DNA from nearby stromal cells in vivo would restore tumour growth[27]; recently there has been a promising development in what’s known as Metabolite imaging, which constitutes visualization and snapshot of metabolites in a tissue be either in vivo or in vitro, this using nuclear magnetic resonance (NMR), magnetic resonance spectroscopy (MRS), positron emission tomography (PET), matrix assisted laser desorption/ionization (MALDI)MS, secondary ion MS (SIMS) or desorption electro spray ionization MS (DESIMS) techniques; with MRS and PET being the promising ones as they are noninvasive and also development in MRS has made possible of detecting metabolites in an amount of as low as 20; these all allow us for a better diagnosis and metabolic phenotyping and hence a more customized therapies[28]. In 2014 study, it was found that CRC roughly had five distinct major proteomic subtypes (A,B,C,D and E); among several distinctive features of these subtypes was one was that almost all hypermutated and MSI-high tumors were included in subtypes B and C, and that the The Cancer Genome Atlas Program(TCGA) CpG island methylator phenotype-high (CIMP-H) methylation subtype was exclusively associated with proteomic subtype B, in contrast subtype C was significantly associated with a non-CIMP subtype, subtype B also showed lack of TP53 mutations and chromosome 18q loss, which concludes strong association between subtype B and MSI-High and CIMP, while other subtypes were related to Chromosomal instability (CIN), another well established genetic property of CRC[29], among those subtype E showed a much higher association with TP53 mutations and 18q loss, genomic properties that are regularly associated with CIN tumors[30].
Figure 2 Relation of mutation load with immune response
References
1. Elemento, O., The future of precision medicine: towards a more predictive personalized medicine. Emerg Top Life Sci, 2020. 4(2): p. 175-177.
2. Goetz, L.H. and N.J. Schork, Personalized medicine: motivation, challenges, and progress. Fertil Steril, 2018. 109(6): p. 952-963.
3. Patel, J.N., M.K. Fong, and M. Jagosky, Colorectal Cancer Biomarkers in the Era of Personalized Medicine. J Pers Med, 2019. 9(1).
4. Heidelberger, C., et al., Fluorinated pyrimidines, a new class of tumour-inhibitory compounds. Nature, 1957. 179(4561): p. 663-6.
5. Mariadason, J.M., D. Arango, and L.H. Augenlicht, Customizing chemotherapy for colon cancer: the potential of gene expression profiling. Drug Resist Updat, 2004. 7(3): p. 209-18.
6. Metzger, R., et al., High basal level gene expression of thymidine phosphorylase (plateletderived endothelial cell growth factor) in colorectal tumors is associated with nonresponse to 5-fluorouracil. Clin Cancer Res, 1998. 4(10): p. 2371-6.
7. Scherf, U., et al., A gene expression database for the molecular pharmacology of cancer. Nat Genet, 2000. 24(3): p. 236-44.
9. Seitz, D.E., Personalized medicine: the search for prognostic and predictive factors in colorectal cancer. J Lab Clin Med, 2003. 142(1): p. 5-6.
10. Allegra, C.J., et al., Prognostic value of thymidylate synthase, Ki-67, and p53 in patients with Dukes' B and C colon cancer: a National Cancer Institute-National Surgical 11. Adjuvant Breast and Bowel Project collaborative study. J Clin Oncol, 2003. 21(2): p. 241-50.
10. Arango, D., et al., c-Myc overexpression sensitises colon cancer cells to camptothecininduced apoptosis. Br J Cancer, 2003. 89(9): p. 1757-65.
11. Van Bebber, S.L., et al., Novel personalized medicine technology: UGT1A1 testing for irinotecan as a case study. Per Med, 2006. 3(4): p. 415-419.
12. Bedard, P.L., et al., Tumour heterogeneity in the clinic. Nature, 2013. 501(7467): p. 355-64.
13. Sagaert, X., A. Vanstapel, and S. Verbeek, Tumor Heterogeneity in Colorectal Cancer:
What Do We Know So Far? Pathobiology, 2018. 85(1-2): p. 72-84.
14. Dai, W., et al., Single-cell transcriptional profiling reveals the heterogenicity in colorectal cancer. Medicine (Baltimore), 2019. 98(34): p. e16916.
15. Li, H., et al., Reference component analysis of single-cell transcriptomes elucidates cellular heterogeneity in human colorectal tumors. Nat Genet, 2017. 49(5): p. 708-718.
16. Zhai, Z., et al., Colorectal cancer heterogeneity and targeted therapy: Clinical implications, challenges and solutions for treatment resistance. Semin Cell Dev Biol, 2017. 64: p. 107-115.
17. Song, M., et al., PTEN loss increases PD-L1 protein expression and affects the correlation between PD-L1 expression and clinical parameters in colorectal cancer. PLoS One, 2013. 8(6): p. e65821.
18. Punt, C.J., M. Koopman, and L. Vermeulen, From tumour heterogeneity to advances in precision treatment of colorectal cancer. Nat Rev Clin Oncol, 2017. 14(4): p. 235-246.
19. Tejpar, S., et al., Prognostic and Predictive Relevance of Primary Tumor Location in Patients With RAS Wild-Type Metastatic Colorectal Cancer: Retrospective Analyses of the CRYSTAL and FIRE-3 Trials. JAMA Oncol, 2017. 3(2): p. 194-201.
20. Molinari, C., et al., Heterogeneity in Colorectal Cancer: A Challenge for Personalized Medicine? Int J Mol Sci, 2018. 19(12).
21. Muzny, D.M., et al., Comprehensive molecular characterization of human colon and rectal cancer. Nature, 2012. 487(7407): p. 330-337.
22. Giannakis, M., et al., Genomic Correlates of Immune-Cell Infiltrates in Colorectal Carcinoma. Cell Rep, 2016. 15(4): p. 857-865.
23. Kloor, M., S. Michel, and M. von Knebel Doeberitz, Immune evasion of microsatellite unstable colorectal cancers. Int J Cancer, 2010. 127(5): p. 1001-10.
24. Kather, J.N., N. Halama, and D. Jaeger, Genomics and emerging biomarkers for immunotherapy of colorectal cancer. Semin Cancer Biol, 2018. 52(Pt 2): p. 189-197.
25. Le, D.T., et al., PD-1 Blockade in Tumors with Mismatch-Repair Deficiency. New England Journal of Medicine, 2015. 372(26): p. 2509-2520.
26. Chan, E.C., et al., Metabolic profiling of human colorectal cancer using high-resolution magic angle spinning nuclear magnetic resonance (HR-MAS NMR) spectroscopy and gas chromatography mass spectrometry (GC/MS). J Proteome Res, 2009. 8(1): p. 352-61.
27. DeBerardinis, R.J. and N.S. Chandel, We need to talk about the Warburg effect. Nat Metab, 2020. 2(2): p. 127-129.
28. Wishart, D.S., Emerging applications of metabolomics in drug discovery and precision medicine. Nat Rev Drug Discov, 2016. 15(7): p. 473-84.
29. Zhang, B., et al., Proteogenomic characterization of human colon and rectal cancer.
Nature, 2014. 513(7518): p. 382-7.
30. Fearon, E.R., Molecular genetics of colorectal cancer. Annu Rev Pathol, 2011. 6: p. 479-507.
