The authors have declared that no competing interests exist.
Conceived and designed the experiments: KLG. Performed the experiments: SS KLG. Analyzed the data: KLG SS IH. Contributed reagents/materials/analysis tools: KLG SS. Wrote the paper: KLG IH.
M2 pyruvate kinase (M2PK) is an oncoprotein secreted by colorectal cancers in stools. This the first report on the accuracy of a rapid stool test in the detection of colorectal cancer (CRC).
To determine the sensitivity, specificity and positive and negative predictive value of a rapid, point of care stool test M2 PK- the M2PK Quick.
Consecutive cases of endoscopically diagnosed and histological proven CRC were recruited. Stools were collected by patients and tested with the immunochromatographic M2PK Quick Test (Schebo Biotech AC, Giessen, Germany). Controls were consecutively chosen from patients without any significant colorectal or gastrointestinal disease undergoing colonoscopy. CRC was staged according to the AJCC staging manual (7th Edition) and location of tumor defined as proximal or distal.
The sensitivity, specificity, positive predictive value, negative predictive value and overall accuracy were: 93%, 97.5%, 94.9%, 96.5% and 96.0% respectively. The positive predictive value for proximal tumors was significantly lower compared to distal tumors. No differences were seen between the different stages of the tumor.
The M2-PK Quick, rapid, point-of-care test is a highly accurate test in the detection of CRC. It is easy and convenient to perform and a useful diagnostic test for the detection of CRC in a clinical practice setting.
Stool tests that have been widely used in the detectionand screening of colorectal cancer (CRC) are based on the testing for occult blood in stools. The guaic based tests have been in use for alongtime but suffers from the drawback of a high false positive resultdue to the inherent nature of the test which depends on the oxidative capacity of the guaicsubstrate. In general the guaic based fecal occult tests have limitedsensitivity and specificity. In one study,the sensitivity of guaiac based test (HemoccultSensa) was 79.4% and the specificity was 86.7% [
M2-pyruvate kinase (M2-PK) is an isoenzyme of pyruvate kinase, which is a key enzyme in glycolysis where it catalyzes the conversion of phosphoenolpyruvate to pyruvate. This isoenzyme is generally a highly active, tetrameric form. However in tumor tissue, on exposure to oncoproteins, M2-PK is converted into a less active dimeric form, a change which is necessary for tumor metabolism [
In CRC,tumor M2-PK is shed in the colonic lumen andtherefore can be detected in the stools of patients. This forms the basis of the M2-PK stool test. The use of M2-PK test stool was first reported by Hardt et al in 2003 whenthe authors reported quantification of M2 PK levels in stools and suggested that it may be useful in CRC screening [
Recently a newer version of the M2-PK test- the rapid office based qualitative test–M2PK Quick was introduced for clinical use. This is the first study to validate the M2PK Quick for the detection of CRC.The aim of this study is to determine the diagnostic accuracy of the M2-PK Quick stool test in the detection of CRC.
This is a case control study that was carried out at the University of Malaya Medical Centre (UMMC) fromJanuary 2013 to December 2013. Approval for the study was specifically obtained from the Institutional Review Board of the hospital: the ethics committee of the UMMC and the study was carried out in accordance to ICH-GCP guidelines. A patient information leaflet and a written consent form, which is standard policy, were approved by the committee. The study aims and procedures were explained in detailed to all participants. The patient information leaflet and written consent were given to all participants. All written consent forms obtained were kept in the case record files of each patient.
Consecutive patients with CRC histologically confirmed on colonoscopy and biopsy were recruited for the study. The casesconsisted of patients undergoing CRC screening(asymptomatic) or patients who presented with symptoms and signs suggestive of CRC (tenesmus, alteration in bowel habit, PR bleed and weight loss). Controls consisted of consecutive patients who had undergone screening colonoscopy with no evidence of CRC, adenomasnor dysplasia and wereotherwise healthy. Control subjects who had any clinical suspicion of any GI tumor were excluded.The recruitment ratio of cases and controls was 1: 2. The diagnosis of CRC was confirmed on biopsies and resected surgical samples sent for histopathologicalexamination.CRC patients were subcategorized according to tumor location and staging. For staging,the American Joint Committee on Cancer (AJCC)classification based on CT scan and on operative findings was used. Tumor location was categorized as either proximal (cecum,ascendingcolon,hepaticflexure,transverse colon and splenic flexure) or distal (descending colon,sigmoidcolon,rectum and anal canal).
The stool testing for M2PK was carried out by a single trained personnel who was blinded to the results of the colonoscopy.Stools were collected either before the colonoscopy and bowel preparation was carried out or at least one week after the colonoscopy. “Fresh” stool samples were tested as soon as collected. If testing were not possible, stool samples were kept at 40 C in the refrigerator for no longer than 24 hours.
Testing for the M2-PK protein in the stools of patients was carried out by using the rapid, point of care,M2-PKQuickTest(ScheBoBiotech AG, Giessen, Germany).Thetest was carried out according to the manufacturer’s instructions. The interpretation based on the appearance of bands in the test (T) and control (C) areas of the test strip when a solubilized stool sample is placed. The M2PK Quick test is an immunochromatographic rapid test. When the M2PK protein is present in stools, it reacts with a monoclonal antibody bound to gold particles and forms a gold-labelled complex, called the antibody-M2PK complex. This complex is then migrates along the membrane and reached the Test line (T) which has second monoclonal antibody will bind to this antibody-M2PK complex, and subsequently develop a pink colour line at T.
The study was reported according to Standards for the Reporting of Diagnostic accuracy studies guidelines.
Statistical analysis was done by using SPSS version 16 for Windows (SPSS Inc, Chicago, Illinois). For the baseline demography, Student t test and Chi-square testing were used where applicable. The sensitivity, specificity and positive and negative predictive values of the rapid serology test were calculated against the gold standard for diagnosis of CRC based on histological confirmation. 95% confidence intervals (C.I.) were calculated for proportions of these values.
One hundred and fifteen patients were diagnosed to have CRCduring the period of the study. However stool samplescollected appropriately were obtained from only 100 patients and tested for the M2-PK protein. Likewise,stool samples were requested from 213 subjects with normal colonoscopy but only 200 cases were appropriately collected and tested for the M2-PK protein (
The baseline demography of the cancer and control cases are summarized in
Cancer | Controls | P value | |
---|---|---|---|
65.31±13.16 | 60.31 ± 11.78 | ||
|
58 (58%) | 100 (50%) | |
|
42(42%) | 100 (50%) | |
|
25 (25%) | 100(50%) | |
|
62 (62%) | 60(30%) | |
|
13 (13%) | 40(20%) |
The sensitivity, specificity, positive predictive value, negative predictive value and overall accuracy of the M2-PK test were as follows; 93%, 97.5%, 94.9%,96.5% and 96.0% respectively (
M2PK positive | M2PK negative | Sensitivity % (95% CI) | Specificity % (95% CI) | PPV % (95% CI) | NPV % (95% CI) | Accuracy % (95% CI) | |
---|---|---|---|---|---|---|---|
93 | 7 | 93 (86.3–96.6) | 97.5 (94.3–98.9) | 94.9 (88.6–97.8) | 96.5 (93.0–98.3) | 96.0 (93.1–97.7) | |
5 | 195 |
The diagnostic accuracy of the tests according to tumor location. Out of the 100 patients, 87 were distal tumors and 13 were proximal. The sensitivity, specificity, positive predictive value and negative predictive value of the M2-PK test according to tumor location is summarized in
M2PK positive | M2PK negative | Sensitivity, %(95% CI) | Specificity, %(95% CI) | PPV, %(95% CI) | NPV, %(95% CI) | Accuracy, %(95% CI) | |
---|---|---|---|---|---|---|---|
12 | 2 | 85.7(60.1–96.0) | 97.5(94.3–98.9) | 70.6(46.9–86.7) | 99.0(96.4–99.7) | 96.7(93.4–98.4) | |
81 | 5 | 94.2(87.1–97.5) | 97.5(94.3–98.9) | 94.2(87.1–97.5) | 97.5(94.3–98.9) | 96.2(93.7–98.1) | |
5 | 195 |
The sensitivity, specificity, positive predictive value and negative predictive value of the M2-PK test according to tumor stagingis summarized in
AJCC Stage | M2PK positive | M2PK negative | Sensitivity, %(95% CI) | Specificity, %(95% CI) | PPV, %(95% CI) | NPV, %(95% CI) |
---|---|---|---|---|---|---|
16 | 1 | 94.1(73.0–99.0) | 97.5(94.3–98.9) | 76.2(54.9–89.4) | 99.5(97.2–99.9) | |
16 | 2 | 88.9(67.2–96.9) | 97.5 (94.3–98.9) | 76.2(54.9–89.4) | 99.0(96.4–99.7) | |
27 | 1 | 96.4(82.2–99.4) | 97.5(94.3–98.9) | 84.4(68.3–93.1) | 99.5(97.2–99.9) | |
28 | 3 | 90.3(75.1–96.7) | 97.5(94.3–98.9) | 84.8(69.1–93.4) | 98.5(95.6–99.5) | |
5 | 195 |
The M2-PK fecal test is a novel test in the detection of CRC. As proliferatingtumour cells shed M2-PK readily from the surface of the tumor, this test would allow us to reliably detect cancers from the gastrointestinal tract.
Two types of fecal M2-PK test are available: A quantitative test, ELISAbased laboratory test and aqualitativepoint of care immunochromatographic test. An office-based, point of care test for stool detection is not as practical as a similar test for blood samples as result as will not be immediately available. Stools will usually have to be collected at a separate occasion and usually at home. However,this test allows the busy doctor in clinical practice to perform the test in his own clinic with results being available almost as soon as the stool sample is submitted.On other hand, a quantitative test requires the stool sampleto be sent to a well fairly equipped laboratory with an ELISA reader and withtrained personnel, requiring several hours before results can be obtained.
The M2PK quantitative test has been widely tested and have shown a sensitivity ranging from 73–97% and specificity from 78.6–100.0% [
Fecal Test | Sensitivity, % | Specificity, % | |
---|---|---|---|
Hardt11 | M2-PK | 73.3 | 77.8 |
Naumann10 | M2-PK | 85.2 | 65.3 |
gFOBT | 63 | 86.7 | |
Vogel12 | M2-PK | 77.3 | 71.8 |
gFOBT | 27.3 | 89.1 | |
iFOBT | 90.9 | 93.8 | |
Tonus14 | M2-PK | 77.8 | 92.9 |
Shastri13 | M2-PK | 81.1 | 71.1 |
gFOBT | 36.5 | 92.2 | |
Koss16 | M2-PK | 90.6 | 92.3 |
Mulder15 | M2-PK | 84.6 | 90.5 |
iFOBT | 92.3 | 96.8 | |
Shastri17 | M2-PK | 78.2 | 73.8 |
iFOBT | 70.9 | 96.3 | |
This study | M2-PK (Quick Test) |
93 | 97.5 |
*Case-control validation study
Our study has also shown that the test is highly accurate regardless of the stage of the tumor although this is in contrast to other studies. For example,Shastri and colleagues found a significantly higher proportion of patients with Dukes C and D cancers (89.5%) vs. Dukes A and B cancers (63.9%) having a positive quantitative M2-PK test [
How has the M2PK test performed in the detection of CRC? In our case control study we have included patients with overt CRC and did not have patients withvery early CRC confined to the mucosa. Cases with colorectal adenomas were not included in the study. In a fairly large screening study in a German population, using the laboratory based EISA assay and looking specifically at colorectal adenomas, stool M2-PK showed a low sensitivity of only 22% and 23% for advanced and non-advancedadenomas(early and intermediate)respectively although a specificity of 82% was reported [
The other point to consider was that both symptomatic and asymptomatic patients were recruited into the study who were consecutively found to have CRC. It is not clear whether or not the M2-PK test will have similar diagnostic accuracy in a purely asymptomatic population as part of a CRC screening program as one could argue that symptomatic patients require a colonoscopy anyway. However, in a developing country such as Malaysia where endoscopic services remain limited, the M2-PK test may allow risk stratification in terms of need and timing for colonoscopy.
Another confounding factor in clinical practice would be the presence of concomitant inflammation in the colon. Positive tests have also been reported in inflammatory bowel disease which constitutes a “false positive” test. In a small group of patients, Mulder etal showed that in IBD patients, the M2-PK test was positive in 15 of 19 (78.9%) [
The most widely used stool test for screening for CRC remain the fecal occult based tests and there have been studies comparing the accuracy of M2-PK compared to these tests. When compared to guaic based FOBT for the detection of CR adenomas, a meta-analysis by Tonus et al amalgamating the results of 4 studies showed a significant better detection rate with the M2PK test for adenomas both less and morethan1cm in size [
However as the main utility of stool based testing is for screening in an asymptomatic population, a future study should be planned to look at the sensitivity, specificity and overall accuracy of the rapid point ofcare M2-PK test in a screening population and compare it to the rapid point of care (qualitative) FIT. Another possibility would be the use of both tests in combination although the cost effectiveness of this strategy needs to be determined. A study by Leen at al [
The M2-PKQuick, rapidpoint-of-care test which has been shown to be a highly accurate test in the detection of CRC. It is easy and convenient to perform in a doctor’s office. It is certainly a useful diagnostic test for the detection of CRCin a clinical practice setting.
(DOCX)
Alere provided financial support and supply of the test kits. Special thanks to Mr Joachim Hevler, International Medical Director, Alerefor technical support and general advice. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. We wish to acknowledge the help of DrAusama of Colorectal Surgery with help in collection of cases and Ms SashikalaMayakrisnan (research assistant) for data collection and stool testing.