Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Jun 1.
Published in final edited form as: Clin Colorectal Cancer. 2019 Feb 13;18(2):102–109. doi: 10.1016/j.clcc.2019.02.003

Tumor Heterogeneity as a Predictor of Response to Neoadjuvant Chemotherapy in Locally-Advanced Rectal Cancer

Alissa Greenbaum a, David R Martin b, Thèrése Bocklage c, Ji-Hyun Lee d, Scott A Ness e,f, Ashwani Rajput a,e,*
PMCID: PMC6556409  NIHMSID: NIHMS1525835  PMID: 30935775

Abstract

Introduction/Background:

Neoadjuvant chemoradiation therapy (nCRT) is the standard of care for locally advanced adenocarcinoma of the rectum, but it is currently unknown which patients will respond. This study tested the correlation between response to nCRT and intratumoral heterogeneity, measured using next-generation sequencing assays.

Materials and Methods:

DNA was extracted from formalin-fixed, paraffin-embedded biopsy samples from a cohort of patients with locally advanced rectal adenocarcinoma (T3/4 or N1/2) who received nCRT. High read-depth sequencing of >400 cancer-relevant genes was performed using the Ion Ampliseq Comprehensive Cancer Panel™ assay. Tumor mutations and variant allele frequencies were used to calculate Mutant Allele Tumor Heterogeneity (MATH) scores as measures of intratumoral heterogeneity. Response to nCRT was pathologically scored after surgical resection.

Results:

Biopsy samples from 21 patient tumors were analyzed. 8 patients were noted to have complete response, 2 were moderate, 4 minimal and 7 demonstrated poor response. Higher MATH scores correlated with poorer response to treatment, demonstrating significantly increased tumor heterogeneity compared to those with complete response (p=0.039).

Conclusion:

The application of MATH scores as a measure of tumor heterogeneity may provide a useful biomarker for treatment response in locally advanced rectal cancer.

Keywords: Targeted sequencing, bioinformatics, personalized medicine, translational research

MICROABSTRACT

Standards therapies for locally advanced cancers include neoadjuvant chemoradiation. Currently, there is no way of knowing which patients will respond to such therapy. We analyzed 21 pre-treatment rectal cancer biopsies and report the positive correlation of the response to therapy with a quantitative Mutant-Allele Heterogeneity (MATH) Score. This NGS derived score may serve as a biomarker for response to therapy.

INTRODUCTION

Colorectal cancer (CRC) is the third most common and lethal cancer in the United States. Rectal cancers account for over one third of CRC cases, with approximately 39,000 new cases diagnosed each year (1). Treatment of locally advanced rectal cancer is different from equivalent stage colon cancer. Neoadjuvant chemoradiation therapy (nCRT) became the standard of care for locally advanced disease after a number of landmark randomized controlled trials found nCRT decreased rates of local tumor recurrence, increased rates of sphincter preservation and decreased both acute and long-term treatment toxicity (2,3,4,5). Since 2003, nCRT has been rapidly adopted and is now extensively used (6). However, patients demonstrate a spectrum of response to nCRT, ranging from complete to poor or no response to treatment. Complete response is associated with decreased local recurrence and improved overall survival (7,8,9), while poor response is associated with worse outcomes. Recent studies have explored mechanisms to explain resistance to nCRT, (10,11), but no biomarker or single somatic mutation is routinely used to predict which patients will respond to nCRT.

Intratumoral heterogeneity (ITH) has been proposed as a mechanism of resistance to both chemotherapy and radiation. Neoplastic cells are thought to proliferate from a singular common ancestor (12,13). Tumor cells can undergo additional sequential mutations, resulting in subclones within the same tumor. The presence of ITH correlates with more rapid cell proliferation, increased metastasis, resistance to therapy (14) and worse clinical outcomes. While ITH is studied in research laboratories, a reproducible method to quantify ITH in the clinical setting is not available.

The Mutant-Allele Tumor Heterogeneity (MATH) Score is a novel next-generation sequencing method to quantitatively measure ITH (15). The MATH score is the width of the distribution of mutant-allele fractions (MAF) among tumor loci. A locus that mutated early in the clonal evolution of the tumor will have a high MAF, while later, lower frequency loci mutations demonstrate a lower MAF (16). In head and neck squamous cell carcinoma, higher MATH scores and therefore increased ITH, correlate with shorter overall survival (OS) and worse clinical outcomes (16). We showed previously that higher MATH scores correlated with worse outcome in colon cancer (17). One previous study applied MATH scores specifically to rectal cancer, demonstrating highly variable degrees of tumor heterogeneity in 6 tumors (18). However, the correlation of MATH scores to clinical treatment or outcomes in rectal cancer is unknown. The purpose of this study was to determine if MATH scores can predict response to nCRT in locally advanced rectal adenocarcinoma.

MATERIALS AND METHODS

Twenty-three patients with locally advanced rectal cancer who underwent surgical resection of their tumors after nCRT were identified in a prospective database. Locally advanced tumors were defined as American Joint Commission on Cancer Staging (AJCC) 7th edition (19) T3 or T4 tumors, node positive (N1/N2) or T2 with possibility of achieving sphincter-sparing surgery. After Institutional Review Board approval, formalin-fixed paraffin-embedded blocks of rectal pre-treatment tumor biopsies were obtained. Patients were categorized according to nCRT response, as described by histologic neo-induced regression in the surgical pathologic specimen (complete (0), minimal (+1), moderate (+2) and poor (+3)) after surgical resection. Pre-treatment rectal biopsy samples were used for patient DNA analysis.

DNA was extracted using standard techniques. Samples were analyzed as previously described (17) using the Ion Ampliseq Comprehensive Cancer panel assay, which targets more than 400 cancer-relevant genes. Sequencing was performed on the Ion Proton instrument as described by the manufacturer, (Thermo Fisher Scientific, Waltham, MA), generating average read depths of >1200x across the targeted regions. As quality control, only samples that yielded >500x average coverage for all 4 multiplexed primer pools were used for downstream analyses. Variants were called by the Torrent Server Variant Caller (versions 4.21, 4.421, 4.607 or 5.021 with interchangeable results). The resulting Variant Caller Format (VCF) files were compared using customized bioinformatics scripts as described previously (17). Analyses were limited to single nucleotide variants (no indels) with read depth greater than 100 and at least 5 reads in each direction.

Variant Allele Frequencies (VAF) were calculated as the ratio of alternate allele observations to the read depth at each position. The MATH score is calculated as 100 x median absolute deviation (MAD) / median of the variant allele frequencies, and describes the ratio of the width of the data to the center of the distribution – effectively a score describing the spread in the data. We applied the MATH score (12) to quantify the tumor heterogeneity in our rectal tumor samples, as well as a modification to include both tumor and germline heterozygous (genotype “0/1”) variants with VAF between 0.05 and 0.75. The comparisons of MATH score between response categories were conducted using Student t-test statistics after checking of the required assumptions of the distribution and equal variance. A p-value of less than 0.05 was considered statistically significant.

RESULTS

Targeted next-generation DNA sequencing was used to identify the somatic tumor variants in a cohort of 23 retrospective rectal cancer biopsy samples that included patients whose response to nCRT treatment was judged to be “complete”, “moderate”, “minimal” or “poor” (Table 1). Two samples failed to pass quality control criteria, so results from 21 samples were used for analyses.

Table 1.

Characteristics of Rectal Cancer Patient Samples

Patient RESPONSE GENDER APC MUT KRAS MUT TP53 MUT MATH SCORE
Sample 06 Complete F yes no no 11.8
Sample 08 Complete M no no yes 17.24
Sample 09 Complete M yes no no 11.06
Sample 10 Complete M no yes yes 10.89
Sample 30 Complete F yes no no 15.64
Sample 43 Complete F no no no 12.73
Sample 45 Complete F no no no 7.88
Sample 47 Complete F no no no 14.4
Sample 16 Moderate M yes no no 6.62
Sample 17 Moderate F yes yes no 19.84
Sample 12 Minimal F no no no 8.62
Sample 13 Minimal F no no no 25.23
Sample 14 Minimal M no no yes 28.69
Sample 15 Minimal M yes yes no 15.35
Sample 20 Poor F no no yes 26.67
Sample 21 Poor F no no no 19.75
Sample 36 Poor U no no no 13.26
Sample 40 Poor M no no no 13.55
Sample 44 Poor U yes no no 27.95
Sample 46 Poor U no no no 14.31
Sample 48 Poor M yes no no 17.05
Open in a new tab

U = gender information unavailable or not provided

Since no samples of normal tissue were available for comparison, we identified tumor mutations based on their exclusion from databases of common germline polymorphisms (dbSNP 137), with variant allele frequencies below 0.50. The most common mutations that were also listed in the Catalog of Somatic Mutations in Cancer (COSMIC) [PMID 27899578] are shown in Figure 1A. Most common mutations were in APC, TP53 and KRAS genes, which is similar but not identical to the most commonly mutations observed by others (20). The differences could suggest alternative genetic or environmental mechanisms in our cohort of patients from New Mexico. Figure 1B shows a plot of the observed variant allele frequencies for each of the COSMIC mutations shown in panel A. Some mutations, (e.g. APC, KMT2C, KRAS, TP53) were observed in a broad range of allele frequencies in different tumors, suggesting that these mutations were acquired early in the evolution of the tumors. However, many of the mutations that occurred less frequently (e.g. EGFR, FBXW7, LIFR, MSH6) were only observed at low allele frequencies, suggesting that they were acquired later and were only present in a subset of tumor cells.

Figure 1. Acquired COSMIC mutations in rectal cancer samples.

Figure 1.

Open in a new tab

Frequencies of occurrence (A) and observed variant allele frequencies (B) for somatic mutations listed in the COSMIC database.

The correlation between the presence of specific mutations with either good or poor response to nCRT treatment was attempted. Figure 2 shows a heatmap indicating the presence of COSMIC mutations in all samples, which were sorted by either complete or moderate response to nCRT at left (black) and minimal or poor response at right (red). There was no correlation between response and either the presence of specific mutations or the number of acquired mutations. For example, acquired mutations in APC, KRAS or TP53 were observed in samples from both groups. Thus, there appears to be no simple way of predicting response to nCRT based solely on the use of targeted gene panel sequencing in these types of rectal cancer samples.

Figure 2. Distribution of COSMIC mutations in rectal cancer samples.

Figure 2.

Open in a new tab

The heatmap shows the presence of mutations listed in the COSMIC database in the rectal cancer samples analyzed. Samples on the left (black) showed complete or moderate response to nCRT, and the samples on the right (red) showed minimal or poor response.

We next determined whether quantitative traits in the sequencing information could be used to gain additional insight into the biology of response to nCRT. We showed previously that the MATH score as a measure of tumor heterogeneity correlated with tumor stage in colon cancer (17). We tested whether the MATH score could be used as a measure of tumor heterogeneity in rectal cancers. Figure 3 compares the distribution of variant allele frequencies from two samples, plotted as histograms in panels A and B and as plots of variant allele frequency vs. read depth in panels C and D. Sample 16, in the left panels, shows peaks of allele frequencies at 0.5 and 1.0, which are the heterozygous and homozygous variants, with very few somatic mutations at other frequencies. In contrast, sample 44 on the right shows a broad distribution of allele frequencies, and has a much higher MATH score of 27.95, compared to 6.62 for sample 16. Thus, the higher MATH score quantifies the spread in the variant allele frequencies observed in sample 44, which reflects increased tumor heterogeneity.

Figure 3. Tumor heterogeneity and MATH scores.

Figure 3.

Open in a new tab

The observed variant allele frequencies (VAF) are displayed as histograms (top, A and B) or plots of of VAF vs. read depth (bottom, C and D) for samples 16 (left, MATH = 6.62) or 44 (right, MATH = 27.95). Note the spread in the observed VAF in the sample with higher MATH score.

Next, we tested whether the MATH scores correlated with response to nCRT. As shown in Figure 4A, the samples from patients who had complete response to nCRT had significantly (P = 0.039) lower MATH scores than samples from patients with poor response to nCRT. A similar difference was observed when comparing complete response to samples in all other categories (moderate, minimal or poor, P = 0.026, panel B) or when comparing complete plus moderate response to minimal plus poor response (panel C, P = 0.02). In contrast, there was no difference in MATH scores when comparing samples that did or did not harbor mutations in APC (panel D) or other common mutations (not shown).

Figure 4. MATH scores correlate with poor response to nCRT.

Figure 4.

Open in a new tab

The boxplots compare the MATH scores for patients who demonstrated complete vs poor response to nCRT (A), complete response vs. moderate, minimal or poor response (B) and complete or moderate vs. minimal or poor response (C). There was no correlation between MATH scores and the presence of mutations in the APC gene (D).

DISCUSSION

Preoperative nCRT is the standard of care for locally advanced rectal adenocarcinoma. MATH scores, a novel bioinformatics tool allowing quantitative measurement of tumor genetic heterogeneity, appears to have relevance in the study of rectal cancer. We observed higher MATH scores, and thus increased tumor genetic heterogeneity, to be associated with poorer response to nCRT. Although these results will need to be validated in a larger group of samples, it appears that measures of tumor heterogeneity such as the MATH score have potential for predicting treatment response.

Some studies examining biomarkers to predict nCRT response in rectal cancer have been promising, though others have presented contradictory results. In 2001, Ki-67 expression was found to correlate with nCRT response, with a higher Ki-67 labeling index seen in patients with complete response (10). In contrast, a study in 2008 demonstrated that lower Ki-67 expression was associated with higher level of tumor regression (11). Increased expression of epidermal growth factor receptor (EGFR) and p21 have been observed in tumors with poor nCRT response (21,22), while high thymidylate synthase levels and lower pre-treatment DNA methylation are associated with higher response rates (23). These studies suggest a number of molecular markers may predict nCRT response, though confirmatory studies are ongoing and routine use in clinical practice is currently unavailable.

While the effect of tumor heterogeneity on response to nCRT in rectal cancer is unknown, genetic heterogeneity is thought to be a major contributor to monoclonal antibody treatment resistance in CRC. A number of factors may explain how increased tumor heterogeneity contributes to treatment resistance in cancer. A tumor subclone may possess a Darwinian advantage in its ability to survive the stresses of hypoxia and cytotoxic chemotherapy (13) or harbor a subclone with a mutation that renders the tumor resistant to targeted molecular therapy. CRC is considered to be a highly heterogeneous disease, with mounting evidence revealing various genetic aberrations in both colon and rectal tumors (24).

The MATH score offers a global assessment of tumor heterogeneity, compared to single gene assessments currently used in CRC. KRAS mutations are present in nearly 40% of CRC tumors (25), and are known to drive resistance to anti-EGFR monoclonal antibody treatment. Patients with KRAS wildtype tumors should benefit from anti-EGFR therapy, though nearly 35% will not respond to treatment (25, 26). It has been hypothesized that this treatment resistance is secondary to the presence of subclones possessing KRAS mutations, many times with concurrent BRAF and PIK3CA mutations. The current standard method of tumor biopsy in both clinical and research settings cannot detect the presence or extent of such genetic heterogeneity, as only one area of the tumor is sampled. In fact, nearly one-third of CRC tumors demonstrate intratumoral heterogeneity by analyzing RAS mutations in multiple tumor areas (27). Similar rates (39%) of intertumoral heterogeneity are seen between primary tumors and either lymph node or distant metastases (27). While detection of mutations such as KRAS has helped to guide targeted molecular therapy, the extreme degree of genetic heterogeneity in CRC presents a major challenge to precision medicine (28).

Predicting which patients will respond to nCRT holds important clinical implications in rectal cancer. Complete pathologic response, defined as ypT0N0M0 after nCRT and total mesorectal excision, is seen in 15–20% of patients with rectal cancer. Complete response is associated with decreased local recurrence and improved overall survival (7,8,9). Conversely, patients with little or no response to nCRT demonstrate poorer outcomes and survival. While high ITH is known to contribute to cytotoxic and targeted molecular therapy resistance in CRC, breast, esophageal, and lung cancers (29,30), no studies to date have directly examined the relationship of rectal cancer tumor heterogeneity and response to nCRT. By stratifying the biologic profiles of locally advanced rectal cancer, providers could have the ability to discuss an individual patient’s prognosis. Recommendations could be tailored toward more aggressive adjuvant treatment or toward observation based on their predicted response to nCRT.

The MATH score originally developed in 2013 and studied in head and neck squamous cell carcinoma (12), now presents a novel potential biomarker in CRC. We previously demonstrated that higher MATH scores are associated with more advanced stages of disease and increased metastatic capability (17). One previous study to date has specifically applied MATH scores to rectal cancer (18). NGS of 2–3 separate areas of 6 rectal tumors found that while all patients demonstrated heterogeneity, the degree varied significantly. Our study is consistent with these findings, and with the added correlation of tumor heterogeneity to treatment response. The MATH score offers a promising method to identify high risk patients at high risk of treatment failure, both in practice and in clinical trial settings. It should be mentioned however, that while computation of MATH score may allow for patient stratification and prediction of nCRT response, it does not address the complex issue of how to improve treatment for patients with increased tumor heterogeneity.

Limitations of this study include the small sample size of 21 rectal tumors. However, this is the largest analysis employing MATH scores in rectal cancer to date, and the comparison of MATH scores by nCRT response categories was striking enough to yield significant results. Second, the rectal cancer samples and MATH scores were retrospectively obtained. A prospectively designed study must be conducted to further investigate MATH score as a clinical biomarker. Third, DNA sampling and analysis was performed from one area of the tumor. While future studies will sample multiple tumor areas, whole tumor sequencing is not yet possible. Future studies in rectal cancer should also examine MATH scores in relation to oncologic outcome data including overall survival, disease-free survival and local recurrence.

Tumor heterogeneity presents a challenge to the developing field of precision oncology. In this pilot study, we applied a novel bioinformatics approach, the MATH score, to quantitatively measure tumor genetic heterogeneity in locally advanced rectal cancer. Increased MATH scores were associated with poorer response to nCRT. The calculation of MATH scores presents a potential biomarker to predict nCRT response in rectal cancer in both clinical and research settings.

CONCLUSIONS

The Mutant-Allele Heterogeneity (MATH) Score is a quantitative measure of tumor heterogeneity. This technique does not rely on a single gene mutation or a unique signaling pathway. The MATH score uses the entire 400+ gene data set to look at the shape of the data. This paper demonstrates that MATH scores generated from pre-treatment rectal cancer biopsies correlate with response to neoadjuvant chemoradiation therapy. Thus, knowing a patient’s MATH score may allow for customizing treatment protocols.

CLIINICAL PRACTICE POINTS

Rectal Cancer remains a leading cause of cancer related morbidity and mortality. Locally advanced tumors are standardly treated with chemotherapy and radiation prior to surgical resection. This approach leads to decreased local recurrence rates, peri-operative complications and possibly increases the rate of sphincter preservation. Currently, it is not known which patients will have a favorable response to chemoradiation. By using Next Generation Sequencing techniques, this paper demonstrates that the quantitative MATH score, a reflection of tumor heterogeneity, correlates with the response to neoadjuvant chemoradiation treatment. Our data demonstrates that the higher MATH scores correlate with a poorer response to neoadjuvant chemoradiation. This is important information to know as patients with less favorable response fare worse compared to patients who have a good response to such treatments. Knowing a patient’s MATH score and thus the degree of heterogeneity, may allow for a more practical and economic surveillance strategy and possibly better personally designed treatment algorithms.

Synopsis:

MATH score is a quantitative assessment of tumor heterogeneity. This score is shown to correlate and predict a patient’s response to neoadjuvant chemotherapy in the treatment of locally advanced rectal cancer.

ACKNOWLEDGEMENTS

The authors acknowledge outstanding technical support from Jennifer Woods, Maggie Cyphery, Jamie Padilla and Kathryn Brayer. We thank Patricia J. Young for assistance in manuscript preparation. Some experiments used the facilities or services provided by the Analytical and Translational Genomics Shared Resource, which is supported by the State of New Mexico and the UNM Comprehensive Cancer Center P30CA118100.

FUNDING

This work is supported by: S.A. Ness: NIH grants 5R01CA170250, 5R01DE023222, 5P30CA118100.

ABBREVIATIONS

nCRT

neoadjuvant chemoradiation therapy

CRC

Colorectal Cancer

ITH

intratumoral heterogeneity

MATH

Mutant-Allele Tumor Heterogeneity

MAF

Mutant-Allele Fraction

OS

Overall Survival

VCF

Variant Caller Format

VAF

Variant Allele Frequencies

NGS

Next Generation Sequencing

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

REFERENCES

  • 1).Siegel RL, Miller KD, Jemal A. Cancer statistics, 2016. CA Cancer J Clin 2016; 66:7–30. [DOI] [PubMed] [Google Scholar]
  • 2).Sauer R, Becker H, Hohenberger W, Rodel C, Wittekind C, Fietkau R, et al. Preoperative versus postoperative chemoradiotherapy for rectal cancer. N Engl J Med 2004; 351: 1731–1740. [DOI] [PubMed] [Google Scholar]
  • 3).Kapiteijn E, Marijnen CA, Nagtegaal ID, et al. Preoperative radiotherapy combined with total mesorectal excision for resectable rectal cancer. N Engl J Med 2001;345:638–46. [DOI] [PubMed] [Google Scholar]
  • 4).Folkesson J, Birgisson H, Pahlman L, Cedermark B, Glimelius B, Gunnarsson U. Swedish Rectal Cancer Trial: long lasting benefits from radiotherapy on survival and local recurrence rate. J Clin Oncol 2005;23:5644–5650. [DOI] [PubMed] [Google Scholar]
  • 5).Improved survival with preoperative radiotherapy in resectable rectal cancer. Swedish Rectal Cancer Trial. N Engl J Med 1997;336:980–987. [DOI] [PubMed] [Google Scholar]
  • 6).Mak RH, McCarthy EP, Das P, Hong TS, Mamon HJ, Hoffman KE. Adoption of preoperative radiation therapy for rectal cancer from 2000 to 2006: a Surveillance, Epidemiology, and End Results patterns-of-care study. Int J Radiat Oncol Biol Phys 2011; 80(4):978–984. [DOI] [PubMed] [Google Scholar]
  • 7).Martin ST. Systematic review and meta-analysis of outcomes following pathological complete response to neoadjuvant chemoradiotherapy for rectal cancer. Br J Surg 2012; 99:918–928. [DOI] [PubMed] [Google Scholar]
  • 8).Maas M, Nelemans PJ, Valentini V, Das P, Rodel C, Kuo LJ, et al. Long-term outcome in patients with a pathological complete response after chemoradiation for rectal cancer: a pooled analysis of individual patient data. Lancet Oncol 2010; 11:835–844. [DOI] [PubMed] [Google Scholar]
  • 9).Capirici C, Valentini V, Cionini L, De Paoli A, Rodel C, Glynne-Jones R, Coco C, et al. Prognostic value of pathologic complete response after neoadjuvant therapy in locally advanced rectal cáncer: long-term analysis of 566 ypCR patients. Int J Radiat Oncol Biol Phys 2008. 72(1):99–107. [DOI] [PubMed] [Google Scholar]
  • 10).Kim NK, Park JK, Lee KY, Yang WI, Yun SH, Sung J, et al. p53, BCL-2, and Ki-67 expression according to tumor response after concurrent chemoradiotherapy for advanced rectal cancer. Ann Surg Oncol 2001;8:418–24. [DOI] [PubMed] [Google Scholar]
  • 11).Jakob C, Liersch T, Meyer W, Becker H, Baretton GB, Aust DE. Predictive value of Ki67 and p53 in locally advanced rectal cancer: correlation with thymidylate synthase and histopathological tumor regression after neoadjuvant 5-FU-based chemoradiotherapy. World J Gastroenterol 2008;14:1060–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12).Mroz EA, Tward AD, Pickering CR, Myers JN, Ferris RL, Rocco JW. High intratumor genetic heterogeneity is related to worse outcome in patients with head and neck squamous cell carcinoma. Cancer 2013;19(16):3034–3042. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13).Seoane J, De Mattos-Arruda L. The challenge of intratumour heterogeneity in precisión medicine. J Intern Med 2014;276(1):41–51. [DOI] [PubMed] [Google Scholar]
  • 14).Rocco JW. Mutant allele tumor heterogeneity (MATH) and head and neck squamous cell carcinoma. Head Neck Pathol 2015;9(1):1–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15).Mroz EA, Rocco JW. MATH, a novel measure of intratumor genetic heterogeneity, is high in poor-outcome classes of head and neck squamous cell carcinoma. Oral Oncol 2013; 49:211–215. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16).Mroz EA, Tward AD, Hammon RJ, Ren Y, Rocco JW. Intra-tumor genetic heterogeneity and mortality in head and neck cancer: analysis of data from the Cancer Genome Atlas. PLoS Med 2015; 12(2):31001786. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17).Rajput A, Bocklage T, Greenbaum A, Lee JH, Ness SA. Mutant-allele tumor heterogeneity scores correlate with risk of metastases in colon cancer. Clin Colorectal Cancer 2017; 16(3):e165–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18).Hardiman KM, Ulintz PJ, Kuick RD, Hovelson DH, Gates CM, Bhasi A, et al. Intratumor genetic heterogeneity in rectal cancer. Lab Invest 2016;96(1):4–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19).Edge SB, Byrd DR, Compton CC, Fritz AG, Greene FL, Trotti A, editors. AJCC cancer staging manual (7th ed). New York, NY: Springer; 2010 [Google Scholar]
  • 20).Crumley SM, Pepper KL, Phan AT, Olsen RJ, Schwartz MR, Portier BP. Next-generation sequencing of matched primary and metastatic rectal adenocarcinomas demonstrates minimal mutation gain and concordance to colonic adenocarcinomas. Arch Pathol Lab Med 2016, 140(6):529–535. [DOI] [PubMed] [Google Scholar]
  • 21).Giralt J, Eraso A, Armengol M, Rossello J, Majo J, Ares C, et al. Epidermal growth factor receptor is a predictor of tumor response in locally advanced rectal cancer patients treated with preoperative radiotherapy. Int J Radiat Oncol Biol Phys 2002;54:1460–5. [DOI] [PubMed] [Google Scholar]
  • 22).Reerink O, Karrenbeld A, Plukker JT, Verschueren RC, Szabo BG, Hospers GA, et al. Molecular prognostic factors in locally irresectable rectal cancer treated preoperatively by chemo-radiotherapy. Anticancer Res 2004;24:1217–21. [PubMed] [Google Scholar]
  • 23).Tsang JS, Vencken S, Sharaf O, Leen E, Kay EW, McNamara DA, et al. Global DNA methylation is altered by neoadjuvant chemoradiotherapy in rectal cáncer and may predict response to treatment- A pilot study. Eur J Surg Oncol 2014;40(11)1459–1466. [DOI] [PubMed] [Google Scholar]
  • 24).Punt CJ, Koopman M, Vermeulen L. From tumour heterogeneity to advances in precision treatment of colorectal cancer. Nat Rev Clin Oncol 2017;14(4):235–246. [DOI] [PubMed] [Google Scholar]
  • 25).Sorich MJ, Wiese MD, Rowland A, Kichenadasse G, McKinnon RA, Karapetis CS. Extended RAS mutations and anti-EGFR monoclonal antibody survival benefit in metastatic colorectal cancer: a meta-analysis of randomized controlled trials. Ann Oncol 2015; 1:13–21. [DOI] [PubMed] [Google Scholar]
  • 26).Douillard JY, Siena S, Cassidy J, et al. Final results from PRIME: randomized phase III study of panitumumab with FOLFOX4 for first-line treatment of metastatic colorectal cancer. Ann Oncol 2014; 25:1346–1355. [DOI] [PubMed] [Google Scholar]
  • 27).Jeantet M, Tougeron D, Tachon G, Cortes U, Arachmbaut C, Fromont G, et al. High intra- and inter-tumoral heterogeneity of RAS mutations in colorectal cancer. Int J Mol Sci 2016; 17(12). Pii:E2015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28).Russo M, Siravegna G, Blaszkowsky LS, Corti G, Crisafulli G, Ahronian LG, Mussolin B, et al. Tumor heterogeneity and lesion-specific response to targeted therapy in colorectal cancer. Cancer Discov 2016; 6(2):147–153. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29).Obulkasim A, Yistra B, van Essen EF, Benner C, Stenning S, Langley R, et al. Reduced genomic tumor heterogeneity after neoadjuvant chemotherapy is related to favorable outcome in patients with esophageal adenocarcinoma. Oncotarget 2016; 7(28):44084–44095. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30).Murtaza M, Dawson S-J, Tsui DWY, Gale D, Forshew T, Piskorz AM, et al. Noninvasive analysis of acquired resistance to cancer therapy by sequencing of plasma DNA. Nature. 2014; 497:108–112. [DOI] [PubMed] [Google Scholar]

RESOURCES