GASTROENTEROLOGY 2004;127:422– 427
`
`Potential Usefulness of Detecting Cyclooxygenase 2 Messenger
`RNA in Feces for Colorectal Cancer Screening
`
`SHIGERU KANAOKA,* KEN–ICHI YOSHIDA,* NAOYUKI MIURA,‡ HARUHIKO SUGIMURA,§ and
`MASAYOSHI KAJIMURA*
`*First Department of Medicine, ‡Department of Biochemistry, and §Department of Pathology, Hamamatsu University School of Medicine,
`Hamamatsu, Japan
`
`Background & Aims: Cyclooxygenase 2 (COX-2) is over-
`expressed frequently in aerodigestive tumors, especially
`in colorectal tumors. Therefore, it may be a suitable
`biomarker for colorectal cancer (CRC) screening. We
`performed a pilot study of whether detecting COX-2
`expression in fecal RNA enables us to discriminate be-
`tween patients with and without CRC. Methods: The
`study cohort included 29 patients with CRC, and 22
`control patients without neoplastic disease of the colon
`or rectum. RNA was isolated from routinely collected
`stool samples using a modified method. The expression
`levels of carcinoembryonic antigen (CEA) and COX-2
`were determined by nested reverse-transcription poly-
`merase chain reaction (RT-PCR). Results: The sensitivity
`and the specificity of fecal COX-2 assay for CRC were
`90% (95% confidence interval
`[CI], 73%–98%) and
`100% (95% CI, 85%–100%), respectively, whereas those
`of the fecal CEA assay for CRC were 100% (95% CI,
`88%–100%) and 5% (95% CI, 2%–23%), respectively.
`COX-2 messenger RNA (mRNA) was detected in 3 of 4
`patients with Dukes’ stage A, 13 of 14 patients with
`Dukes’ stage B, and 10 of 11 patients with Dukes’ stage
`C or D. COX-2 mRNA was detected in 5 of 7 patients with
`proximal cancer and in 21 of 22 patients with distal
`cancer. The COX-2 assay was superior to the CEA assay
`for detecting CRC in terms of specificity, although both
`assays had high sensitivity. Conclusions: This fecal
`COX-2 assay had high sensitivity and high specificity for
`detecting CRC. These results suggest that it would be a
`promising approach for detecting CRC, although a larger
`study is necessary to assess the sensitivity and the
`specificity.
`
`Colorectal cancer (CRC) is the second most common
`
`cause of death in the Western world, and is the most
`In the
`common fatal cancer among nonsmokers.1,2
`United States, CRC accounts for 11% of all cancers, with
`an estimated 130,200 new cases and 48,100 deaths in the
`year 2001,1 and in Japan there are 85,000 annual cancer
`registrations and 35,600 deaths caused by this disease.3
`Because a large number of patients can be treated suc-
`cessfully when metastasis does not occur,4 it is important
`
`to make an early diagnosis. Colonoscopy and sigmoidos-
`copy are highly specific and sensitive tests for neoplasia,
`but they are invasive and limited by patient compliance
`and physician availability.5 The fecal occult blood test is
`a noninvasive and simple examination that has been
`shown to reduce the incidence, morbidity, and mortality
`of CRC.6 –9
`Recently, methods for the isolation of human DNA
`directly from stool samples have allowed the analysis of
`genetic alterations associated with neoplasia.10 –15 Be-
`cause these genetic alterations are associated directly
`with the development of neoplasia, they have clear ad-
`vantages over indirect markers such as fecal occult blood.
`The disadvantage of DNA-based stool assays is the lack
`of sensitivity caused by clonal heterogeneity in CRC.16,17
`RNA-based stool assays have been reported in several
`preliminary studies,18 –21 one of which showed that
`CD44 variant expression in human feces could be de-
`tected in 68% of CRC patients by a combination of
`reverse-transcription polymerase chain reaction (RT-
`results
`and Southern hybridization.21 These
`PCR)
`prompted us to develop a more sensitive assay. The major
`disadvantage of this RNA-based stool assay is that it is
`difficult to isolate RNA without degradation and it is
`difficult to remove impurities from feces such as PCR
`inhibitors. A protocol that isolates RNA with less deg-
`radation and that can detect CRC molecular markers
`therefore is required. The ideal target molecules for an
`RNA-based stool assay are those that are expressed only
`in CRCs and not in normal mucosa. Cyclooxygenase 2
`(COX-2) gene expression is increased frequently in aero-
`digestive tumors, including esophagus, stomach, pan-
`creas, lung, and colon.22 It has been shown that COX-2
`is overexpressed in more than 80% of CRCs compared
`
`Abbreviations used in this paper: CEA, carcinoembryonic antigen; CI,
`confidence interval; COX-2, cyclooxygenase 2; CRC, colorectal cancer;
`RT-PCR, reverse-transcription polymerase chain reaction.
`© 2004 by the American Gastroenterological Association
`0016-5085/04/$30.00
`doi:10.1053/j.gastro.2004.05.022
`
`Geneoscopy Exhibit 1061, Page 1
`
`

`

`August 2004
`
`COX-2 mRNA IN FECES FROM CANCER PATIENTS 423
`
`Table 1. Characteristics of CRC and Control Patients
`
`CRC patients
`
`Control patients
`
`n
`Sex (M/F)
`Age (yr)a
`Total RNA (␮g/g of stool)a
`
`29
`20/9
`71 (50–85)
`70 (5–414)
`
`22
`11/11
`66.5 (20–85)
`55.5 (4–379)
`
`aMedian (range).
`
`with normal colonic mucosa.23–25 Therefore, COX-2 is
`considered a good candidate gene for an RNA-based
`stool assay.
`The purpose of our study was to develop a test based
`on using fecal RNA to detect CRC. We describe here the
`modified methods used to isolate RNA quickly and with
`sufficient purity to assay by RT-PCR, allowing us to
`distinguish CRC patients from control subjects by
`COX-2 messenger RNA (mRNA) expression levels.
`
`Materials and Methods
`Patients and Samples
`All CRC patients and controls in this study were
`admitted to Hamamatsu University School of Medicine be-
`tween 1999 and 2002. We evaluated 29 patients with primary
`colorectal adenocarcinoma who were diagnosed colonoscopi-
`cally and histologically. The median age of the patients with
`CRC was 71 years (range, 50 – 85 yr), and 20 of the 29 patients
`were men (Table 1). Twenty-five of 29 patients underwent
`surgical resection of their primary tumor, 2 patients under-
`went endoscopic resection, and 2 patients had tumors that
`were considered inoperable because of metastasis to other
`organs. The tumors were classified according to Dukes’ stag-
`ing, yielding stage A (n ⫽ 4), stage B (n ⫽ 14), stage C (n ⫽
`8), and stage D (n ⫽ 3) (Table 2).
`A total of 22 control patients who had no neoplastic lesions
`colonoscopically also were included in this study (11 men, 11
`women; median age, 66.5 yr; range, 20 – 85 yr) (Table 1). The
`reasons for performing colonoscopy in the control patients
`included lower abdominal pain, anemia, constipation, or CRC
`screening, showing that the control patients in this study were
`not an average-risk population.
`Stool samples were collected before colonoscopy from all of
`the control patients and 3 of the 29 patients with CRC, and
`before surgery from the remaining 26 patients with neoplasia,
`all of whom initially were diagnosed colonoscopically in the
`outpatient unit. Their samples were collected more than 2
`weeks after colonoscopy. The samples were stored at 4°C
`immediately after collecting and transferred to a freezer set at
`⫺80°C within 6 hours. The samples were stored for up to 2
`years before isolating RNA. This study was approved by the
`institutional
`local genetic research ethics committee at
`Hamamatsu University School of Medicine. Oral and written
`informed consent was obtained from all patients.
`
`RNA Isolation From Feces
`To develop an RNA-based stool assay for the detection
`of CRC, we needed a new method to isolate fecal RNA of
`sufficient quantity, quality, and purity. The procedure we
`developed for isolating RNA from feces uses a combination of
`Isogene (Nippon Gene, Toyama, Japan) and RNeasy kit (Qia-
`gen GmbH, Hilden, Germany). Approximately 1 g of frozen
`fecal pellet was added to a sterile 5-mL tube containing 3 mL
`Isogene, and homogenized with a Handy Microhomogenizer
`(Microtech Nition, Chiba, Japan) for a few minutes. After
`homogenization, the slurry was poured into sterile 1.5-mL
`tubes, which were centrifuged at 12,000g for 5 minutes at
`4°C. The supernatant from each tube was transferred carefully
`to new sterile 1.5-mL tubes. To each tube, 0.3 mL Isogene and
`0.3 mL chloroform were added, the tubes were shaken vigor-
`ously for 30 seconds, incubated for 5 minutes at 4°C, and
`centrifuged at 12,000g for 15 minutes at 4°C.
`The aqueous phase from each tube was removed carefully
`without contamination from the interface and transferred to a
`fresh 1.5-mL tube. An equal volume of 70% ethanol was
`added, and the tubes were vortexed vigorously for 30 seconds.
`The mixed solution (700 ␮L) was added to an RNeasy
`minispin column (Qiagen GmbH), and the columns were
`
`Table 2. Characteristics of CRC Patients
`
`Patient
`no.
`
`Age,
`yr
`
`Sex
`
`RNA,
`␮g
`
`Location
`
`Dukes’
`stage
`
`Size,
`cm
`
`CEA
`assaya
`
`COX-2
`assaya
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`29
`
`5
`F
`64
`34
`F
`77
`70
`78 M
`35
`65 M
`82
`69 M
`63
`55
`F
`89
`83 M
`62
`55
`F
`13
`70 M
`63
`73 M
`9
`64
`F
`8
`85 M
`71 M 105
`75 M 155
`50 M
`12
`74 M 110
`55 M
`31
`82 M
`9
`82 M
`13
`63 M 179
`52
`F
`117
`72 M 252
`60 M
`67
`80
`F
`195
`75 M 237
`69 M 414
`68 M 117
`76
`F
`77
`83
`F
`358
`
`A
`A
`R
`R
`C
`S
`C
`R
`R
`R
`S
`D
`R
`R
`R
`S
`R
`S
`S
`S
`A
`R
`R
`S
`C
`R
`R
`S
`A
`
`C
`B
`D
`A
`C
`B
`B
`B
`B
`A
`D
`C
`C
`C
`B
`B
`A
`C
`B
`C
`C
`B
`B
`A
`B
`B
`D
`B
`B
`
`(⫹)
`6.5
`(⫹⫹)
`7.0
`(⫹⫹)
`5.0
`(⫹)
`6.5
`(⫹⫹)
`5.0
`(⫹)
`5.2
`(⫹⫹)
`2.7
`(⫹⫹)
`2.5
`(⫹)
`5.0
`(⫹)
`1.5
`(⫹)
`3.5
`(⫹)
`3.3
`(⫹⫹)
`7.3
`(⫹⫹)
`5.5
`(⫹)
`4.2
`(⫹)
`4.8
`(⫹)
`1.2
`(⫹)
`3.2
`(⫹)
`4.0
`3.2 (⫹⫹⫹)
`(⫹)
`4.3
`(⫹)
`7.0
`(⫹⫹)
`3.6
`(⫹⫹)
`2.2
`(⫹⫹)
`4.2
`(⫹)
`5.0
`4.3 (⫹⫹⫹)
`6.5 (⫹⫹⫹)
`(⫹⫹)
`4.5
`
`(⫹)
`(⫹)
`(⫹⫹⫹)
`(⫹)
`(⫺)
`(⫹)
`(⫹)
`(⫹)
`(⫹)
`(⫹⫹)
`(⫹⫹⫹)
`(⫹)
`(⫹)
`(⫹)
`(⫹)
`(⫹)
`(⫺)
`(⫹)
`(⫹)
`(⫹⫹⫹)
`(⫹)
`(⫹⫹)
`(⫹)
`(⫹⫹)
`(⫹⫹⫹)
`(⫹⫹)
`(⫹⫹⫹)
`(⫹⫹⫹)
`(⫺)
`
`aRT-PCR was graded as negative (⫺); weakly positive (⫹); positive
`(⫹⫹); or strongly positive (⫹⫹⫹).
`A, ascending; R, rectum; C, cecum; S, sigmoid; D, descending.
`
`Geneoscopy Exhibit 1061, Page 2
`
`

`

`424 KANAOKA ET AL.
`
`GASTROENTEROLOGY Vol. 127, No. 2
`
`centrifuged at 10,000g for 15 seconds at room temperature.
`The remaining steps were performed according to the manu-
`facturer’s instructions. Total RNA concentrations were deter-
`mined by ultraviolet spectrophotometry, and the RNA sam-
`ples were stored at ⫺80°C.
`
`RT–PCR
`Complementary DNA (cDNA) was synthesized using
`ReverScript II (Wako Chemical, Osaka, Japan) with 1 ␮g fecal
`RNA and 250 ng random hexamers according to the manu-
`facturer’s instructions, and amplified using nested PCR. The
`cycling conditions were as follows: carcinoembryonic antigen
`(CEA), 95°C for 5 minutes, followed by 20 cycles at 95°C for
`1 minute and 72°C for 2 minutes; COX-2, 95°C for 5 min-
`utes, followed by 20 cycles at 95°C for 1 minute, 56°C for 1
`minute, and 72°C for 1 minute. The nested PCR reactions
`were as follows: 95°C for 5 minutes, followed by 25 cycles at
`95°C for 1 minute, 69°C was used for CEA, and 56°C was used
`for COX-2 for 1 minute, and 72°C for 1 minute. CEA primers
`were as described previously26 and COX-2 primers were de-
`signed according to published sequence information.27 The
`primers used were as follows: CEA A primer: 5⬘–TCTG-
`GAACTTCTCCTGGTCTCTCAGCTGG-3⬘; CEA B primer:
`5⬘–TGTAGCTGTTGCAAATGCTTTAAGGAAGAAGC-3⬘;
`CEA C primer: 5⬘–GGGCCACTGCTGGCATCATGATTG-
`3⬘; COX-2 A primer: 5⬘–CTGAAACCCACTCCAAACA-
`CAG-3⬘; COX-2 B primer: 5⬘–ATAGGAGAGGTTAGAG-
`AAGGCT-3⬘; COX-2 C primer: 5⬘–GCACTACATACTTAC-
`CCACTTCAA-3⬘.
`Primers A and B were used for the first round of PCR, and
`primers C and B were used for the second round. To distin-
`guish from contaminated genomic DNA, we selected both
`forward and reverse primers at different exons. The 131-bp
`CEA PCR product and the 178-bp COX-2 PCR product were
`identified by electrophoresis of 10 ␮L through 4% NuSieve
`3:1 agarose (BioWhittaker Molecular Applications, Rockland,
`ME) in Tris-acetate-ethylenediaminetetraacetic acid buffer and
`ethidium bromide staining. Negative controls for the RT-PCR
`consisted of either a reverse-transcribed sample without total
`RNA or PCR mixture only. To ensure reproducibility of
`results, all samples were reverse transcribed and amplified in
`triplicate. In addition, the fidelity of both the 131-bp CEA
`PCR product and the 178-bp COX-2 PCR product from the
`stool samples was confirmed by DNA sequencing.
`
`Statistical Analysis
`Sensitivity and specificity were estimated relative to
`the results of colonoscopy in the usual manner; 95% confidence
`intervals (CIs) for these estimated parameters were based on
`the exact binominal distribution. Statistical significance was
`determined by the Fisher exact test, the Mann–Whitney test,
`the Kruskal–Wallis test, and Spearman coefficient by rank
`test. P values less than 0.05 were interpreted as statistically
`significant. All reported P values are evaluated by 2 sides.
`
`Table 3. Comparison of COX-2 Assay With CEA Assay
`
`COX-2 assay
`
`95% CI
`
`CEA assay
`
`95% CI
`
`90% (26/29) 73%–98% 100% (29/29) 88%–100%
`Sensitivity
`Specificity 100% (22/22) 85%–100%
`5% (1/22)
`2%–23%
`
`Results
`RNA Isolation
`We modified the method of Chomczynski and
`Sacchi28 by using both Isogene and RNeasy kit. After
`purification, the RNA was analyzed to detect CEA and
`COX-2 expression. The median yield of RNA was 70 ␮g
`(range, 5– 414 ␮g) per gram of stool from cancer patients
`and 55.5 ␮g (range, 4 –379 ␮g) per gram of stool from
`control patients, respectively (Tables 1 and 2). There was
`no significant difference in RNA yield between the 2
`groups (P ⬎ 0.05, by the Mann–Whitney test), and the
`RNA isolation took only 1 hour using our method.
`
`Nested RT-PCR to Detect CEA
`CEA mRNA was detected in stool samples from
`all cancer patients (100%; 95% CI, 88%–100%) and in
`all but one control patient (95%; 95% CI, 77%–98%)
`(Tables 2 and 3). Representative results are shown in
`Figure 1. There was no significant difference on CEA
`expression levels in feces between CRC patients and
`control patients (P ⬎ 0.05, by the Mann–Whitney test).
`Further, the factors such as Dukes’ stage, location, or size
`of the tumor had no influence on CEA expression levels
`in feces (P ⬎ 0.05, by the Kruskal–Wallis test). This
`revealed that CEA is not suitable for use in an RNA-
`based stool assay to detect CRC, however, detecting CEA
`mRNA could prove the isolation of RNA was sufficient
`to assay RT-PCR.
`
`Nested RT-PCR to Detect COX-2
`We performed RNA-based stool assays unblinded
`to the clinical data. Because COX-2 expression has been
`found in nearly 20% of normal mucosa,25 we prelimi-
`narily needed to prepare various quantities of cDNA
`made from control patients for the first round of PCRs.
`We assayed the first round of PCR using 3 different
`amounts of cDNA, the 75%, the 45%, and the 15% part
`of cDNA synthesized from 1 ␮g of fecal RNA. When the
`75% quantity of cDNA was used for the first round of
`PCR, COX-2 mRNA was detected in 2 of 22 control
`patients (data not shown). However, when assayed using
`less cDNA, COX-2 mRNA was detected in none of the
`control patients. So we decided to use the 45% quantity
`of cDNA synthesized from 1 ␮g of fecal RNA for the
`first round of PCR. COX-2 mRNA was detected in 26 of
`
`Geneoscopy Exhibit 1061, Page 3
`
`

`

`August 2004
`
`COX-2 mRNA IN FECES FROM CANCER PATIENTS 425
`
`Discussion
`The data presented herein show the potential use
`of RT-PCR to detect COX-2 mRNA in feces from CRC
`patients using the appropriate primers and assay condi-
`tions. In the past decade, a number of studies have
`reported neoplasm-specific DNA changes in feces from
`patients with CRC or large adenomas.10 –15 The DNA-
`based stool assays used in these studies typically have
`analyzed mutations on either a single gene or more than
`2 genes. Because the gene mutations were specific to
`neoplasms with an exemption such as K-ras, these assays
`were virtually 100% specific. Indeed, a multitarget
`DNA-based assay recently has been reported to yield
`91% sensitivity and 100% specificity, when analyzed in
`22 patients with CRC and 28 subjects with endoscopi-
`cally normal colons. After recovery of human DNA from
`stool using a sequence-specific hybrid capture technique,
`assay components targeted point mutations at any of 14
`mutational hot spots on p53, and APC genes, mutations
`on BAT-26, microsatellite instability marker, and highly
`amplifiable DNA.12
`It is known that neoplasms continuously exfoliate
`luxuriant populations of viable colonocytes, unlike the
`sparse and largely apoptotic cells shed from normal mu-
`cosa.29 Whole colonocytes have been recovered from stool
`samples,30 and isolated fecal colonocytes can be incorpo-
`rated into assay systems using immunocytochemical
`methods or RT-PCR assays.21,31 However, cytolytic fac-
`tors in stools may compromise the stability of sloughed
`colonocytes.32 To address these factors, we devised a
`method of isolating RNA from feces by homogenizing a
`frozen fecal pellet in the presence of guanidine salt,
`thereby inactivating both cytolytic factors and RNase in
`stools and removing PCR inhibitors. As a result, RNA
`could be isolated more efficiently (in a less degraded
`condition) and more rapidly (approximately 1 h), allow-
`ing specific RT-PCR assays for both CEA and COX-2 to
`be performed. A technique to recover colonocytes from
`stool with an average yield of more than 106 cells per
`gram of stool and with the viability rate as high as 80%
`has been reported previously.30 Theoretically, 106 viable
`cells could produce approximately 10 ␮g of total RNA.
`It also was shown that 5–30 ␮g of RNA per gram of
`stool from cancer patients and approximately 5 ␮g of
`RNA per gram of control stool were obtained.20 Our
`yields of fecal RNA from both cancer and control pa-
`tients were much higher than in this previous report, and
`there was no significant difference between cancer pa-
`tients and control patients in terms of yield (70 ␮g, 55.5
`␮g/g of stool, respectively). Therefore, most of the RNA
`isolated using our method originated from intestinal
`
`Figure 1. Nested RT-PCR for CEA and COX-2. (A) Lanes 1–7 show CEA
`nested RT-PCR results from patients with CRC and lanes 8 –13 show
`those from control patients. All lanes show a PCR product of the
`expected length of 131 bp. (B) The same samples were analyzed for
`COX-2 mRNA. Lanes 2–7 show a PCR product of the expected length
`of 178 bp. Lane 1 and all lanes from control patients (lanes 8 –13)
`show negative results. Lanes 1–7 corresponds to patient numbers 5,
`7, 11, and 22–25 in Table 2, respectively. M, 100-bp ladder size
`marker; NC, negative control.
`
`the 29 stool samples from patients with CRC (90%; 95%
`CI, 73%–98%). None of the 22 control samples showed
`positive results (0%; 95% CI, 0%–15%; P ⬍ 0.0001, by
`the Fisher exact test) (Tables 2 and 3). The representative
`results are shown in Figure 1.
`Although CRC patients were predominately men
`(69%), positive results were obtained similarly in both
`18 of 20 male patients and 8 of 9 female patients (Tables
`1 and 2). When patients were assigned according to
`Dukes’ classification, positive results were obtained in 3
`of the 4 patients with stage A cancer, 13 of the 14
`patients with stage B cancer, 7 of the 8 patients with
`stage C cancer, and all 3 of the patients with stage D
`cancer. Positive results were obtained in 4 of 5 patients
`with small-size tumors (1.2–3 cm), 16 of 18 patients
`with medium-size tumors (3.1– 6 cm), and in all 6
`patients with large-size tumors (⬎6 cm). It is notewor-
`thy that 5 of 7 patients with cancer proximal to the
`splenic flexure had positive results, as did 21 of 22
`patients with more distal cancer. There was no correla-
`tion between CEA expression levels and COX-2 expres-
`sion levels in feces from CRC patients (P ⬎ 0.05, Spear-
`man coefficient by rank test) or between COX-2
`expression levels in feces from CRC patients and Dukes’
`stage, location, or size of the tumor (P ⬎ 0.05, by the
`Kruskal-Wallis test). There was no significant difference
`in COX-2 expression levels in feces between men and
`women (P ⬎ 0.05, by the Mann–Whitney test). In
`conclusion, the COX-2 assay was superior to the CEA
`assay in terms of specificity (100% vs. 5%), although
`both assays had high sensitivity (90% vs. 100%)
`(Table 3).
`
`Geneoscopy Exhibit 1061, Page 4
`
`

`

`426 KANAOKA ET AL.
`
`GASTROENTEROLOGY Vol. 127, No. 2
`
`flora, with the proportion of RNA from colonic epithe-
`lium being considered only a small part. Consequently,
`we needed to amplify both CEA and COX-2 mRNA by
`nested PCR instead of standard PCR methods.
`Serum CEA protein currently is used clinically to
`monitor the management of colorectal carcinoma. The
`gene for CEA is one of the most widely expressed genes
`in cancer cells.33 It is expressed in 95% of colorectal,
`gastric, and pancreatic cancers, similarly being present in
`columnar epithelial cells such as normal colon and stom-
`ach. Expression of CEA mRNA in normal surface epi-
`thelium is almost similar to that of epithelial cells of
`colorectal carcinoma patients.34 We hypothesized that
`expression levels of fecal CEA mRNA of CRC patients
`would be higher than that of control patients. Contrary
`to expectation, there was no significant difference in
`expression levels between the 2 groups, indicating that
`the mRNA extracted from stool samples using our
`method was sufficient for RT-PCR but that the CEA
`molecule is not suitable for RNA-based stool assays to
`detect CRC.
`COX-2 gene expression is up-regulated in most colo-
`rectal carcinomas compared with surrounding normal
`mucosa by approximately 50-fold.23 Immunohistochem-
`ical analysis detected COX-2 in cancer cells, inflamma-
`tory mononuclear cells, vascular endothelial cells, and
`fibroblasts.25 Furthermore, exfoliation of colonocytes and
`leukocytes are quantitatively greater from CRC epithelia
`than from normal mucosa.29 We selected COX-2 as a
`molecular marker for detecting CRC by our method, and
`succeeded in detecting fecal COX-2 mRNA from the
`CRC patients with high sensitivity. To our surprise, this
`assay had 90% sensitivity and 100% specificity, however,
`this extraordinary specificity came from only 22 control
`patients. We will need to examine a larger number of
`control patients for estimating specificity of this assay.
`We found no significant difference between levels of
`COX-2 mRNA and Dukes’ stage, location, or size of the
`tumors. Therefore, it is conceivable that recovery of fecal
`colonocytes from CRC patients, even those with proxi-
`mal colon cancer, would be sufficient to detect fecal
`COX-2 mRNA from small tumors. We performed a
`COX-2 assay on 26 cancer patients using stool samples
`obtained more than 2 weeks after colonoscopy. It is not
`certain whether colonoscopy together with forceps biopsy
`can influence the effect of COX-2 expression. Further
`evaluation is needed to clarify this issue.
`In summary, it is possible to detect COX-2 mRNA in
`feces from patients irrespective of the clinical stage of
`CRC. Moreover, our method has certain advantages com-
`pared with DNA-based stool assays; approximately 1 g of
`
`fecal pellet is sufficient for the assay and more sensitive in
`terms of analyzing a single molecule. It currently re-
`mains unclear, however, that the fecal COX-2 assay is a
`useful screening test for CRC in the clinical setting. A
`larger study remains to be performed. First, to assess the
`sensitivity for a much broader spectrum of tumors, in-
`cluding early Dukes’ A cancer as well as premalignant
`adenoma; second, to assess the specificity for a much
`broader spectrum of controls, in which possible cause of
`the false-positive results should be clarified; and, finally,
`to determine whether the fecal COX-2 assay is as sensi-
`tive and specific as the fecal occult blood test in average-
`risk persons. When these issues are solved, our approach
`would be attractive for CRC screening.
`
`References
`1. Greenlee RT, Hill-Harmon MB, Murray T, Thun M. Cancer statis-
`tics, 2001. CA Cancer J Clin 2001;51:15–36.
`2. Garfinkel L, Mushinski M. U.S. cancer incidence, mortality and
`survival: 1973–1996. Stat Bull Metrop Insur Co 1999;80:23–32.
`3. The Editorial Board of Cancer Statistics in Japan. Cancer Statis-
`tics in Japan-2001, Tokyo: Foundation for Promotion of Cancer
`Research, 2001.
`4. Eisenberg B, Decosse JJ, Harford F, Michalek J. Carcinoma of the
`colon and rectum: the natural history reviewed in 1704 patients.
`Cancer 1982;49:1131–1134.
`5. Frazier AL, Colditz GA, Fuchs CS, Kuntz KM. Cost-effectiveness of
`screening for colorectal cancer in the general population. JAMA
`2000;284:1954 –1961.
`6. Mandel JS, Bond JH, Church TR, Snover DC, Bradley GM, Schu-
`man LM, Ederer F. Reducing mortality from colorectal cancer by
`screening for fecal occult blood. Minnesota Colon Cancer Control
`Study. N Engl J Med 1993;328:1365–1371.
`7. Kronborg O, Fenger C, Olsen J, Jorgensen OD, Sondergaard O.
`Randomised study of screening for colorectal cancer with faecal-
`occult-blood test. Lancet 1996;348:1467–1471.
`8. Hardcastle JD, Chamberlain JO, Robinson MH, Moss SM, Amar
`SS, Balfour TW, James PD, Mangham CM. Randomised con-
`trolled trial of faecal-occult-blood screening for colorectal cancer.
`Lancet 1996;348:1472–1477.
`9. Saito H. Screening for colorectal cancer by immunochemical fecal
`occult blood testing. Jpn J Cancer Res 1996;87:1011–1024.
`10. Sidransky D, Tokino T, Hamilton SR, Kinzler KW, Levin B, Frost P,
`Vogelstein B. Identification of ras oncogene mutations in the
`stool of patients with curable colorectal tumors. Science 1992;
`256:102–105.
`11. Eguchi S, Kohara N, Komuta K, Kanematsu T. Mutations of the
`p53 gene in the stool of patients with resectable colorectal
`cancer. Cancer 1996;77:1707–1710.
`12. Ahlquist DA, Skoletsky JE, Boynton KA, Harrington JJ, Mahoney
`DW, Pierceall WE, Thibodeau SN, Shuber AP. Colorectal cancer
`screening by detection of altered human DNA in stool: feasibility
`of a multitarget assay panel. Gastroenterology 2000;119:1219 –
`1227.
`13. Dong SM, Traverso G, Johnson C, Geng L, Favis R, Boynton K,
`Hibi K, Goodman SN, D’Allessio M, Paty P, Hamilton SR, Sidran-
`sky D, Barany F, Levin B, Shuber A, Kinzler KW, Vogelstein B, Jen
`J. Detecting colorectal cancer in stool with the use of multiple
`genetic targets. J Natl Cancer Inst 2001;93:858 – 865.
`14. Traverso G, Shuber A, Levin B, Johnson C, Olsson L, Schoetz DJ
`Jr, Hamilton SR, Boynton K, Kinzler KW, Vogelstein B. Detection
`
`Geneoscopy Exhibit 1061, Page 5
`
`

`

`August 2004
`
`COX-2 mRNA IN FECES FROM CANCER PATIENTS 427
`
`of APC mutations in fecal DNA from patients with colorectal
`tumors. N Engl J Med 2002;346:311–320.
`15. Traverso G, Shuber A, Olsson L, Levin B, Johnson C, Hamilton
`SR, Boynton K, Kinzler KW, Vogelstein B. Detection of proximal
`colorectal cancers through analysis of faecal DNA. Lancet 2002;
`359:403– 404.
`16. Fearon ER, Vogelstein B. A genetic model for colorectal tumori-
`genesis. Cell 1990;61:759 –767.
`17. Boland CR, Sato J, Saito K, Carethers JM, Marra G, Laghi L,
`Chauhan DP. Genetic instability and chromosomal aberrations in
`colorectal cancer: a review of the current models. Cancer Detect
`Prev 1998;22:377–382.
`18. Davidson LA, Jiang YH, Lupton JR, Chapkin RS. Noninvasive
`detection of putative biomarkers for colon cancer using fecal
`messenger RNA. Cancer Epidemiol Biomarkers Prev 1995;4:
`643– 647.
`19. Davidson LA, Aymond CM, Jiang YH, Turner ND, Lupton JR,
`Chapkin RS. Non-invasive detection of fecal protein kinase C
`betaII and zeta messenger RNA: putative biomarkers for colon
`cancer. Carcinogenesis 1998;19:253–257.
`20. Alexander RJ, Raicht RF. Purification of total RNA from human
`stool samples. Dig Dis Sci 1998;43:2652–2658.
`21. Yamao T, Matsumura Y, Shimada Y, Moriya Y, Sugihara K, Akasu
`T, Fujita S, Kakizoe T. Abnormal expression of CD44 variants in
`the exfoliated cells in the feces of patients with colorectal can-
`cer. Gastroenterology 1998;114:1196 –1205.
`22. Mohan S, Epstein JB. Carcinogenesis and cyclooxygenase: the
`potential role of COX-2 inhibition in upper aerodigestive tract
`cancer. Oral Oncol 2003;39:537–546.
`23. Eberhart CE, Coffey RJ, Radhika A, Giardiello FM, Ferrenbach S,
`DuBois RN. Up-regulation of cyclooxygenase 2 gene expression in
`human colorectal adenomas and adenocarcinomas. Gastroenter-
`ology 1994;107:1183–1188.
`24. Sano H, Kawahito Y, Wilder RL, Hashiramoto A, Mukai S, Asai K,
`Kimura S, Kato H, Kondo M, Hla T. Expression of cyclooxygen-
`ase-1 and -2 in human colorectal cancer. Cancer Res 1995;55:
`3785–3789.
`25. Hao X, Bishop AE, Wallace M, Wang H, Willcocks TC, Maclouf J,
`Polak JM, Knight S, Talbot IC. Early expression of cyclo-oxygen-
`ase-2 during sporadic colorectal carcinogenesis. J Pathol 1999;
`187:295–301.
`
`26. Gerhard M, Juhl H, Kalthoff H, Schreiber HW, Wagener C, Neu-
`maier M. Specific detection of carcinoembryonic antigen-express-
`ing tumor cells in bone marrow aspirates by polymerase chain
`reaction. J Clin Oncol 1994;12:725–729.
`27. Kosaka T, Miyata A, Ihara H, Hara S, Sugimoto T, Takeda O,
`Takahashi E, Tanabe T. Characterization of the human gene
`(PTGS2) encoding prostaglandin-endoperoxide synthase 2. Eur
`J Biochem 1994;221:889 – 897.
`28. Chomczynski P, Sacchi N. Single-step method of RNA isolation by
`acid guanidinium thiocyanate-phenol-chloroform extraction. Anal
`Biochem 1987;162:156 –159.
`29. Ahlquist DA, Harrington JJ, Burgart LJ, Roche PC. Morphometric
`analysis of the “mucocellular layer” overlying colorectal cancer
`and normal mucosa: relevance to exfoliation and stool screening.
`Hum Pathol 2000;31:51–57.
`30. Albaugh GP, Iyengar V, Lohani A, Malayeri M, Bala S, Nair PP.
`Isolation of exfoliated colonic epithelial cells, a novel, non-inva-
`sive approach to the study of cellular markers. Int J Cancer
`1992;52:347–350.
`31. Chung K, Mullick T, Motevalli S, Vinayek S, Tantry U, Nair P, Dutta
`S. Evaluation of sensitivity and specificity of CD44 expression in
`fecal colonocytes as a noninvasive marker for detecting colonic
`polyps and cancer. Gastroenterology 1998;114:A580.
`32. Harrington J, Limburg P, Roche P, Klee G, Simko V, Ahlquist D.
`Interference of fecal colonocyte recovery by endogenous cytoly-
`sis. Gastroenterology 1998;114:A607.
`33. Hammarstrom S. The carcinoembryonic antigen (CEA) family:
`structures, suggested functions and expression in normal and
`malignant tissues. Semin Cancer Biol 1999;9:67– 81.
`34. Jothy S, Yuan SY, Shirota K. Transcription of carcinoembryonic
`antigen in normal colon and colon carcinoma. In situ hybridization
`study and implication for a new in vivo functional model. Am J
`Pathol 1993;143:250 –257.
`
`Received September 18, 2003. Accepted April 29, 2004.
`Address requests for reprints to: Shigeru Kanaoka, M.D., First De-
`partment of Medicine, Hamamatsu University School of Medicine,
`1-20-1 Hanndayama, Hamamatsu 431-3192, Japan. e-mail: kanaoka@
`hama-med.ac.jp; fax: (81) 53-434-9447.
`The authors thank Dr. N. Shirai for statistical evaluation.
`
`Geneoscopy Exhibit 1061, Page 6
`
`