Detecting Colorectal Cancer in Stool With the Use of Multiple Genetic Targets
Detecting Colorectal Cancer in Stool With the Use of Multiple Genetic Targets
Abstract & Commentary
Synopsis: A reproducible method was developed for reliably detecting 3 tumor-associated genetic alterations in the stool of 51 colorectal cancer patients with all stages of malignancy. The results support the concept of early tumor identification by detection of molecular markers in ex vivo bodily fluids or products. This technology could potentially mitigate the need for unnecessary invasive screening tests in patients without disease or improve the chance for earlier diagnoses and cure for those with disease.
Source: Dong SM, et al. J Natl Cancer Inst. 2001;93: 858-865.
Colorectal cancer remains one of the most prevalent tumors in the Western world and certainly one of the most widely studied in Western medicine. Our understanding of this cancer has been enhanced by the discovery of sequential genetic mutations driving the process of neoplastic transformation.1,2 In the more common adenoma-to-carcinoma pathway, the initial mutation occurs in the adenomatous polyposis coli (APC) gene, associated with aberrant crypt foci and eventually polyp formation, growth, and dysplasia. Next, an oncogene such as K-RAS is activated, incurring further clonal expansion of the dysplastic lesion. Subsequently, mutation or deletion of a tumor-suppressor gene such as TP53 and/or deleted in colon cancer (DCC) occurs, leading to the requisite disruption of at least 2 growth control pathways. Malignant transformation has also been associated with germline or somatic mutations of mismatch repair genes, most notably among hereditary nonpolyposis colorectal cancer (HNPCC) patients.3 These mutations manifest as instability in microsatellite sequence loci, eg, BAT26.
Screening for any of these markers individually may be a painstaking process with sensitivities and specificities no better than the well-entrenched, fecal-occult blood screen. However, a relatively simple, highly sensitive and specific screening method for some optimal combination of these markers could be very useful indeed. In this study, Dong and colleagues devised a protocol for reproducibly isolating DNA from stool in sufficient quantity to allow detection of mutations in K-RAS, TP53, and BAT26 genes. They validated the presence of alterations by testing tumor specimens resected from the same patients.
Fifty-one patients with colonoscopy-diagnosed tumors and no history of familial adenomatous polyposis (FAP) or HNPCC were recruited for this study. One patient had Dukes’ A disease; the others had Dukes’ B (n = 17), Dukes’ C (n = 21), and Dukes’ D (n = 12). Thirteen patients had right-sided lesions and 37 had left-sided lesions. One patient had a transverse colon lesion. Stool samples and primary tumor samples were obtained from each patient and subjected to molecular analysis separately.
TP53 mutations in stool-derived DNA were detected using mismatch-ligation assays (MLAs) and confirmed with repeated testing of separate preparations. TP53 mutations in processed tumor specimens were identified after polymerase chain reaction (PCR) amplification using ligation-detection reaction (LDR) and confirmed with MLA. BAT26 deletions were assessed by minisequencing after PCR in both stool-derived and tumor-derived specimens. K-RAS gene mutation status was examined using digital PCR for stool-derived DNA. Status of the K-RAS gene in tumor-derived DNA was determined by both MLA and PCR/LDR, and confirmed with digital PCR in some specimens.
Overall, Dong et al developed methods to isolate a sufficient quantity of DNA from all 51 stool samples for evaluation of TP53 and BAT26 mutations, and 48 samples for digital PCR analysis of K-RAS. Tumor TP53 mutations were observed in 30 of 51 specimens (59%) and identical TP53 mutations were observed in each of the matched stool-derived specimens, including 1 patient with stage A disease in the right colon. No mutations were detected in stool specimens from patients who did not have mutations in the tumor specimens.
BAT26 deletions were identified in 5 of 51 stool specimens, although an exact match occurred in only 3 of the 51 corresponding tumor specimens (6%). No BAT26 deletions were seen in stool from the 46 patients without BAT26 mutations in their tumor specimens. All tumors with BAT26 deletions were located in the right colon (a frequent site of microsatellite unstable tumors).
K-RAS gene alterations, the best studied in stool due to a low number of codon sites, were assessed in 50 tumor specimens and 48 stool specimens due to technical limitations. K-RAS mutations were present in 17 of the 47 tumor specimens that had a corresponding stool specimen. K-RAS alterations were revealed in 8 of the 17 stool specimens corresponding to these K-RAS-positive tumors for an overall sensitivity of 17% (8/48). All 8 primary tumors with K-RAS alterations were located in the left colon. None of the stool specimens from patients with K-RAS mutation-negative tumors had evidence of K-RAS mutations.
Taken together, evaluation of stool-derived DNA for presence of any of these 3 genetic alterations revealed 71% sensitivity (36/51) and 100% specificity. Sensitivity increased to 92% (36/39) when only cases with mutation-positive tumors were considered. Dong et al note that this represents progress in the search for reliable, noninvasive tests for colorectal cancer, although specificity needs to be further examined in early-stage and asymptomatic patients.
Comment by Arden Morris, MD
Genetic alterations leading to identifiable tumor markers have provided powerful new tools for detection and characterization of tumors. Identification of these tumor markers from sloughed cells in cancer patients may be an excellent source for detecting abnormalities. Despite technically challenging contaminants, stool is a particularly useful medium for development of such tests, given its reflection of mucosal cells from the entire organ and the well-documented sequence of genetic changes in colorectal cancer. Furthermore, the prevalence of this malignancy both assures availability of tissue for study and compels ongoing investigation into early diagnosis as the best hope for cure.
In this study, Dong et al have undertaken 2 tasks. First, they sought to generate an efficient and reliable method for extracting DNA from the stool of colorectal cancer patients. Their ability to examine nearly all specimens (51/51 for TP53 and BAT26; 48/51 for K-RAS) attests to their success in this endeavor. Second, they developed and validated assays to detect a limited number of common cancer-associated genetic alterations. This latter goal was slightly compromised by a lower sensitivity in detecting K-RAS (17% overall, 47% in those cases with positive tumors), which decreased overall sensitivity for the combined genetic markers. The presence of K-RAS on only 3 codons has made it the most frequently studied single genetic marker in stool.2,4,5 However, targeting K-RAS is limited by its presence in only about 50% of colorectal cancers, ultimately halving the sensitivity of even the most precise assay. Moreover, K-RAS is also found in normal mucosa and nonmalignant lesions, potentially leading to false-positive results.6 Additionally, although the test for BAT26 deletion had better accuracy (60%), the low prevalence of BAT26 deletions decreased sensitivity for this marker. The development of a simple and reliable assay for detection of APC, one of the most commonly mutated genes in colorectal cancer, could substantially improve the effectiveness of this multitargeted panel.
In summary, Dong et al have presented a protocol that may provide a template for future reliable analysis of stool for colorectal cancer-associated genetic mutations. The reliability of this multitargeted test was noteworthy in tumors of all stages and in both right- and left-colon lesions. With modification for better overall sensitivity and trials incorporating asymptomatic patients, we may anticipate the availability of a noninvasive tool for identifying early stage patients in the clinical setting.
References
1. Fearon ER, Vogelstein B. Cell. 1990;61(5):759-767.
2. Jen J, et al. Cancer Res. 1994;54(21):5523-5526.
3. Kinzler KW, Vogelstein B. Cell. 1996;87(2):159-170.
4. Sidransky D, et al. Science. 1992;256(5053):102-105.
5. Minamoto T, et al. Cancer Detect Prev. 2000;24(1): 1-12.
6. Ahlquist DA. BMJ. 2000;321(7256):254-255.
Subscribe Now for Access
You have reached your article limit for the month. We hope you found our articles both enjoyable and insightful. For information on new subscriptions, product trials, alternative billing arrangements or group and site discounts please call 800-688-2421. We look forward to having you as a long-term member of the Relias Media community.