N-myristoyltransferase 2–based Blood Test for the Detection of Colorectal Adenomatous Polyps and Cancer : Annals of Surgery Open

Secondary Logo

Journal Logo

Original Study

N-myristoyltransferase 2–based Blood Test for the Detection of Colorectal Adenomatous Polyps and Cancer

Rathinagopal, Tharmini LNP, MSc*,†; Bhanot, Shiv BSc; Yegrov, Sergey PhD*; Min, Jordan BSc*; Hu, Nan PhD; Fang, John MD§; Greene, Tom H. PhD; Varma Shrivastav, Shailly PhD*; Singh, Harminder MD‖,¶; Shrivastav, Anuraag PhD†,¶

Author Information
doi: 10.1097/AS9.0000000000000117

Colorectal cancer (CRC) is the second leading cause of death due to cancer leading to premature mortality worldwide.1 American Cancer Society estimates approximately 105,000 new colon cancer cases and 43,000 new rectal cancer cases, and 53,000 deaths due to CRC in 2020 in the United States of America.2

Most CRC cases arise from premalignant adenomatous polyps that grow slowly over many years before transforming into malignant colonic tumors. A 5-year survival rate of 85% to 90% is observed in patients with an early stage of disease compared with less than 12.5% in patients with advanced stages of the disease.1,3 Currently available CRC screening tests include fecal specimen based tests [guaiac fecal occult blood testing (gFOBT), fecal immunochemical test (FIT), multitargeted DNA], CT colonography, and endoscopy, including sigmoidoscopy and colonoscopy.4,5 Widely used in many countries, gFOBT has several limitations, including limited efficacy and poor acceptability in the population.6,7 The reported sensitivity of gFOBT ranges from 62% to 79% and specificity from 87% to 96% for CRC. In the United States, FIT has become the most commonly used fecal specimen based CRC screening test, which uses antibodies that react with hemoglobin protein associated with heme to identify human hemoglobin in the stool. Therefore, it is more specific than g-FOBT. The reported sensitivity of FIT with repeated testing ranges from 73% to 88% and specificity from 87% to 96% for CRC. A multitargeted stool DNA test (Cologuard Exact Sciences, Madison, WI) that identifies CRC tumor cells combined with FIT has been approved by the FDA that exhibits higher one-time sensitivity (92%) but lower specificity (84%).8 However, compliance remains a vital limitation factor with the fecal-based test. The overall compliance with the repeated recommended use of fecal-based tests in the United States and Canada varies with the target population setting and has been reported in the range of 10% to 50%.9 The importance of increased compliance rates for CRC screening using reliable blood-based methods can hardly be overemphasized. A sensitive and dependable blood-based test to screen individuals before the development of CRC could save lives.

Earlier studies have reported the presence of high N-myristoyltransferase (NMT) in the colonic tumors of rats induced by azoxymethane in comparison to normal mucosa from control rats and normal adjacent mucosa of colonic tumor-bearing rats10 and overexpression of N-myristoltransferase 2 (NMT2) in colonic polyps and cancers.11 Previously, we have demonstrated overexpression of NMT in peripheral blood mononuclear cells (PBMC) in colorectal tumor-bearing rats and further validated overexpression of NMT in PBMC of colorectal cancer patients.12 Additionally, we have identified NMT2 and methionine aminopeptidase 2 blood-based biomarkers for the detection of adenomatous polyps (AP) as well as CRC.13 We hypothesized that high levels of NMT2 in PBMCs of subjects with colorectal adenomatous polyps and cancer might serve as a molecular marker for the early detection of CRC. Therefore, our current study aims to obtain initial/preliminary estimates of the specificity and sensitivity of NMT2 levels in the PBMCs of subjects with colorectal AP or CRC. In this study, we report for the first time that NMT2 is overexpressed in the PBMC of individuals with colorectal AP and CRC. We also report a simple NMT2 based blood test that demonstrates 91% sensitivity and 81% specificity in detecting colorectal cancer adenomatous polyps and cancer.


Study Setting and Participant Recruitment

Participants were recruited from those coming for an outpatient colonoscopy at the Health Sciences Centre, affiliated with the University of Manitoba, as a convenience sample of eligible patients whenever research staff could attend. This was a small-scale cohort study to test the initial efficacy of the test. The full-scale investigation will be performed in the same population. Therefore, no formal sample size calculations were performed. All experimental procedures were performed by research personnel blinded to the colonoscopy status of the participants.

The ethics approval for the study was obtained from the Human Research Ethics Board, University of Manitoba. Of the 100 eligible participants assessed, we recruited and included 74 (declined = 7; difficulty in blood draw = 5; other reasons such as cancelation of the procedure = 14). A total of 74 subjects were recruited prospectively (no evidence of disease [NED] with a normal gastroentroscopy = 24; adenomatous polyps [AP] = 19; non-adenomatous polyps [NAP] = 13; CRC = 14); polyps with no pathology information = 4. The NAP included hyperplastic, diminutive serrated, and inflammatory polyps. Of the total 24 subjects with NED enrolled in the study, 6 were diagnosed with ulcerative colitis, and 2 were diagnosed with Lynch syndrome placing them under high-risk patients that could potentially develop into CRC. The clinical data and the NMT2 expression are provided as supplementary material (Table S1, see https://links.lww.com/AOSO/A92).

Blood Sample Collection

The blood samples were collected before the colonoscopy from subjects following informed consent. Venous blood (10 mL) was collected in vacutainer tubes (BD Biosciences, Canada) coated with EDTA and were processed within 2 h of the collection. The mononuclear cells were isolated from peripheral blood using a Ficoll-Plaque density gradient as previously described.14 The isolated PBMCs were resuspended in RPMI culture medium, cell counts, and viability were determined using a cell counter (Bio-Rad, Canada).


Cytospin slides of PBMCs were prepared in duplicate using cytospin centrifuge (Thermo Fisher, Canada) and sent for immunohistochemistry (IHC) at the Manitoba Tumor Bank in CancerCare Manitoba, Winnipeg. The PBMC Cytospin slides were stained for NMT2 on a VENTANA (Roche Diagnostics, USA) autostainer. After staining, the slides were scored, and an “H” score was given to each sample. H-scores were derived from assessing both staining intensity (scale 0–3) and the percentage of positive cells (0–100%). These two scores were multiplied to generate an H-score of 0 to 300.15,16 The slides were scored independently by two investigators (AS, SVS). Where discordance was found, cases (n = 8) were reevaluated to reach consensus. After the samples were scored, the patient samples were unblinded to correlate the NMT2 expression with the clinical findings and medical history of the subject. Antibodies against human NMT2 were purchased from Sigma Aldrich, Canada. CD4+ and CD8+ Dynabeads coupled with antihuman CD4 and CD8 antibodies were purchased from BD Biosciences (Canada). The NMT2 antibody used for IHC was validated by peptide competition assay.

Statistical Analysis

We tested that NMT2 expression would be associated with clinical phenotypes and that this association would remain significant after accounting for clinical and demographic factors such as age and other clinical conditions. Therefore, Mann-Whitney U and Kruskal-Wallis tests were used to compare two and more participant subgroups nonparametrically. Analysis of covariance (ANCOVA) was performed after confirming the assumptions of normality and heteroscedasticity. Multivariable models were constructed using group classification as a fixed variable; covariates were chosen on the basis of significant association in univariable models (see Tables S2 and S3, see https://links.lww.com/AOSO/A93). Therefore, in the multivariable model based on four participant categories (CRC, AP, NAP, and NED) “age,” “family history of polyps/cancer,” and “diverticulosis” were included as covariates; in the multivariable model based on two participant categories (CRC vs. non-CRC) “age” and “other cancers” were included as covariates. Fold changes are ratios of medians. The receiver operating characteristics (ROC) curves analysis was used to assess the diagnostic accuracy of NMT2. The optimal cut point was decided using Youden’s index. The 95% confidence intervals (CIs) for the sensitivity and specificity at the optimal cut point was calculated using the exact Binomial method. All tests are two sided. P values <0.05 indicate statistically significant results. Statistical analyses and graphics were performed and generated using Statistical package Stata (Stata Corp., College Station, TX, USA) version 16, IBM SPSS Version 23 (New York, NY, USA), GraphPad Prism Version 6.0. (CA, USA), and R (www.r-project.org) version 3.6.1.


Clinical Performance of NMT2 in the Detection of Colorectal Adenomatous Polyps and Cancer

Cases were segregated into four clinically relevant groups, no evidence of disease (NED), nonadenomatous polyps (NAP), adenomatous (AP), and colorectal cancer (CRC). Table 1 lists the median H scores and clinicopathologic characteristics of NED, NAP, AP, and CRC groups. The subjects with CRC were significantly older and more often had a history of non-CRC cancers (Table 1). Subjects with NED and NAP displayed lower expression of NMT2 compared with subjects with AP and CRC (Fig. 1A). The subjects with colorectal AP had significantly higher H score compared with subjects with NED and NAP (Fig. 1B). Additionally, a multivariate ANCOVA analysis was performed using H-scores of all four groups (Table S2, see https://links.lww.com/AOSO/A93). Our results indicate that the H-scores were significantly higher for the AP and CRC groups (P < 0.001) compared with the group with NAP and NED (Table S2, see https://links.lww.com/AOSO/A93). Whereas, NAP was not significantly different (P = 0.773) compared with NED (Table S2, see https://links.lww.com/AOSO/A93). Since the expression of NMT2 in PBMC was higher and comparable in both CRC patients and subjects with colorectal AP (Fig. 1A), we combined them in one group, “AP+CRC” (n = 33). Low H scores were obtained in subjects with NED except for 4 cases and subjects with NAP except for 3 patients (Fig. 1B). Therefore, subjects from both groups were combined under a single “NED+NAP” group (n = 37). The demarcation in the expression of NMT2 in AP+CRC versus NED+AP groups resulted in further refinement of the ability of NMT2 expression in differentiating colorectal AP and cancer from NED and NAP. There were four cases of polyps with no pathology information and thus were excluded from the analyses (n = 70). In subjects with NED, NMT2 expression in PBMC ranged from negative to weak positivity except for 4 cases. The median H score for NED was 60 [IQR 22.5–87.5] (Table 1). In cases where the polyps were characterized as NAP, the median H score was 90 [IQR 12.5–135] (Table 1). In contrast, CRC patients and subjects with colorectal AP displayed intense NMT2 staining and a high percentage of positive cells. The median H scores for colorectal AP and cancer were 150 [IQR 140–180] and 180 [IQR 150–270], respectively (Table 1). In the univariate analysis, compared with the NED, subjects with CRC and AP had significantly elevated NMT2 H scores (median values 180 and 150, respectively, P < 0.001 for both). The median H-score of subjects with NAP was 1.25 fold higher but was not significant than the subjects with NED, while significantly lower than that seen in subjects with CRC and AP, 2.2-fold (P = 0.002), and 1.9-fold (P = 0.017), respectively. In a multivariate analysis, these differences remained significant (Table S2, see https://links.lww.com/AOSO/A93).

TABLE 1. - Median H Scores and Clinicopathologic Characteristics of Subjects (n = 70) Grouped Under CRC and AP, NAP, and NED
Participant Characteristic CRC (n = 14) AP (n = 19) NAP (n = 13) NED (n = 24) P
Median H score (IQR) 180 (150–270) 150 (140–180) 90 (12.5–135) 60 (22.5–87.5) <0.001
Median age (IQR) 68 (61–73) 66 (60.0–71.0) 62 (52.3–69.3) 53.5 (32.5–64.5) 0.002
Men, n (%) 5 (35.7) 11 (57.9) 9 (69.2) 8 (33.3) 0.111
Women, n (%) 9 (64.3) 8 (42.1) 4 (30.7) 16 (66.7)
Other non-CRC cancers present, n (%)* 3 (21,4) 0 (0) 1 (7.6) 0 (0) 0.028
Family history of polyps or cancer, n (%) 8 (53.3) 4 (21.1) 6 (50.0) 5 (20.8) 0.056
Diverticulosis 1 (7.1) 7 (36.8) 8 (61.5) 3 (12.5) 0.003
Hemorrhoids 3 (21.4) 8 (42.1) 5 (38.5) 5 (27.8) 0.003
Inflammatory bowel conditions, n (%), of which: 2 (13.3) 3 (15.8) 2 (16.7) 5 (20.8) 0.890
 UC/CD 1 (6.7) 1 (5.3) 2 (16.7) 3 (12.5)
 Other 1 (6.7) 2 (10.5) 0 (0) 2 (8.3)
Lynch syndrome 0 (0) 0 (0) 0 (0) 2 (8.3) 0.267
Total participants included in the analysis were n = 70.
*Four individuals had tumors present in the breast and liver, pancreas, liver and lungs, prostate, kidney.
†Ileocecal valve ulcer, patchy inflammation in rectum, mild-moderate colitis.
For n = 4, non-CRC participants with missing pathology data specifying polyp status.
AP indicates adenomatous polyps; CD, Crohn’s disease; CRC, colorectal cancer; IQR, interquartile range; NAP, nonadenomatous polyps; NED, no evidence of disease; UC, ulcerative colitis.

NMT2 levels are augmented in PBMC obtained from the peripheral blood samples of subjects with colorectal adenomatous polyps and cancers. A, PBMC cytospin slides were stained for NMT2 using polyclonal anti-NMT2 antibody by IHC. NMT2 expression in PBMC from subjects with NED, subjects with NAP, subjects with colorectal AP and CRC. B, H scores of NMT2 expression in PBMC of NED (n = 24), NAP (n = 13), AP (n = 19), and CRC (n = 14). Bars of the histogram represent the median “H” scores and interquartile ranges. There is no significant difference between NED and NAP. C, ROC curve of CRC + AP vs. NED + NAP. AP indicates adenomatous polyps; CRC, colorectal cancer; IHC, immunohistochemistry; NAP, non-adenomatous polyps; NED, no evidence of disease; NMT2, N-myristoyltransferase 2; PBMC, peripheral blood mononuclear cells; ROC, receiver operating characteristics.

Since the H scores were not significantly different between subjects with CRC and AP (Fig. 1B), we pooled these two groups for ROC curve analyses. ROC curve is the plot of sensitivity versus 1-specificity. We performed a non-parametric ROC analysis to assess the overall performance of NMT2 through the area under the curve (AUC), which can be interpreted as the percent of agreement between the test score and the true diagnostic status. We observed that CRC+AP and NED+NAP were separated in terms of the H score of the NMT2 except for 7 subjects (Table S1, see https://links.lww.com/AOSO/A92). The ROC curve showed a high AUC (0.85 (95% CI: 0.75–0.95). Our results indicate a positive correlation between age and NMT2 levels. Because of this confounding effect of age, we constructed a model taking into account age and depicted this model as a ROC curve (Fig. 1C). We also performed ROC analysis for CRC+AP versus NED with and without age-adjusted (Fig. 1D). Furthermore, the positive predictive value (PPV) and negative predictive values (NPV) were calculated at each cutoff point for NMT2 expression (Table 2). At the optimal cutoff, which maximizes the average of the sensitivity and specificity, the overall probability of correct classification was optimized. Various cutoffs for H scores with respect to sensitivity and specificity are provided in Table 2. An H score of 120 represented an optimal sensitivity (91.15%; 95% CI: 84.49–97.80) and specificity (80.56%; 95% CI: 71.28–89.83) in identifying both AP and CRC compared with NED and NAP. Interestingly, H scores increased across the groups, with the lowest score among individuals with no neoplastic findings and highest among those with colorectal cancer (Fig. 1A, B).

TABLE 2. - Sensitivity and Specificity of NMT2 in Detecting CRC/adenomatous Polyps
Cut Point Sensitivity (%) Specificity (%) Correctly Classified (%) PPV (%) NPV (%)
(≥0) 100.00 0.00 48.57 49.27 NA
(≥2) 100.00 5.56 51.43 50.70 100.00
(≥5) 100.00 11.11 54.29 52.21 100.00
(≥20) 100.00 16.67 57.14 53.82 100.00
(≥30) 100.00 27.78 62.86 57.35 100.00
(≥40) 100.00 30.56 64.29 58.31 100.00
(≥50) 100.00 33.33 65.71 59.30 100.00
(≥60) 100.00 41.67 70.00 62.48 100.00
(≥70) 100.00 52.78 75.71 67.29 100.00
(≥80) 100.00 58.33 78.57 69.98 100.00
(≥90) 100.00 66.67 82.86 74.45 100.00
(≥100) 91.18 77.78 84.29 79.94 92.11
(≥120) 91.18 80.56 85.71 82.00 92.11
(≥140) 76.47 80.56 78.57 79.25 81.40
(≥150) 64.71 80.56 72.86 76.38 74.47
(≥160) 50.00 83.33 67.14 74.44 67.31
(≥180) 44.12 83.33 64.29 71.99 64.82
(≥210) 29.41 86.11 58.57 67.28 59.33
(≥240) 20.59 94.44 58.57 78.25 56.46
(≥270) 11.76 97.22 55.71 80.42 53.85
(>270) 0.00 100.00 51.43 NA 50.73
The table provides the sensitivity, specificity, PPV, and NPV at each NMT2 cutoff. At the optimal cutoff (shown in bold), which maximizes the average of the sensitivity and specificity, the overall probability of correct classification was optimized as well.
CRC indicates colorectal cancer; NMT2, N-myristoyltransferase 2; NPV, negative predictive value; PPV, positive predictive value.

Clinical Performance of NMT2 in the Identification of CRC Subjects

As a secondary analysis, we evaluated the performance of NMT2 in detecting CRC by setting our primary endpoint to assess the level of NMT2 expression in PBMCs of subjects with CRC (n = 14) compared with subjects without CRC (n = 60). Table 3 lists the median H score and clinicopathologic characteristics of these two groups (CRC vs. non-CRC). The subjects with CRC were significantly older and more often had a history of non-CRC cancers. In the univariate analysis, subjects with CRC had significantly elevated expression of NMT2 as determined by high H scores (median of 180) compared with subjects without CRC (median 90), Fig. 2A, B, Table 3 (P < 0.001). The high-risk patients with inflamed bowel had a median H score of 85, whereas two Lynch Syndrome patients had 105 (two subjects one had an H score of 0 and the other had an H score of 210).

TABLE 3. - Median H Scores and Clinicopathologic Characteristics of Participants (n = 74) Segregated into Two Groups CRC and Non-CRC
Subject Characteristics CRC (n = 14) Non-CRC (n = 60) P
Median H Score (IQR) 180 (150–270) 90 (50–150) <0.001
Median age (IQR) 68 (61–73) 60 (49–67) 0.011
Men, n (%) 5 (35.7) 29 (48.3) 0.393
Women n (%) 9 (64.3) 31 (51.7)
Other cancers of non-CRC origin, n (%)* 3 (21.4) 1 (1.6) 0.003
Family history of polyps or cancer, n (%) 8 (53.3) 17 (28.8) 0.040
Diverticulosis, n (%) 1 (7.1) 18 (30) 0.078
Hemorrhoids, n (%) 3 (21.4) 20 (33.3) 0.386
Inflammatory bowel conditions, n (%), of which: 2 (13.3) 10 (16.9) 0.980
 UC/CD 1 (6.7) 6 (10.2)
 Other 1 (6.7) 4 (6.8)
Lynch syndrome, n (%) 0 (0) 2 (3.4) 0.489
*Four individuals had tumors present in the breast and liver, pancreas, liver and lungs, prostate, kidney.
†Ileocecal valve ulcer, patchy inflammation in rectum, mild-moderate colitis.
CD indicates Crohn’s disease; CRC, colorectal cancer; IQR, interquartile range; UC, ulcerative colitis.

Differential expression of NMT2 as determined by IHC in PBMC of CRC patients compared with subjects with no CRC. (A). PBMC cytospin slides were stained for NMT2 using polyclonal anti-NMT2 antibody by IHC. NMT2 expression in PBMC from subjects with NED, subjects with NAP, subjects with colorectal AP and CRC. (B). Histogram showing the “H” score from subjects with CRC (n = 14) and non-CRC (n = 60) subjects, bars of the histogram represent the median “H” scores and the interquartile ranges. (C). ROC curve analysis was performed to determine the performance of NMT2 in differentiating CRC from subjects without CRC. When adjusted for age, the accuracy of NMT2 in differentiating CRC is 86%. AP indicates adenomatous polyps; CRC, colorectal cancer; IHC, immunohistochemistry; NAP, non-adenomatous polyps; NED, no evidence of disease; NMT2, N-myristoyltransferase 2; PBMC, peripheral blood mononuclear cells; ROC, receiver operating characteristics.

In a multivariate analysis accounting for the effects of age and history of non-CRC cancers, this difference remained significant (P < 0.001) (Table S3, see https://links.lww.com/AOSO/A93). The ROC curve was constructed based on these data and had an area under the curve (AUC) of 0.83, which increased upon adjusting for age and history of non-CRC cancers to 0.86 (Fig. 2B).

Since we observed a high percentage of positive staining corresponding to NMT2 in the PBMC of subjects with CRC, we further fractionated CD4+ T cells and CD8+ T cells. The CD4+ and CD8+ T cells were stained for NMT2 by IHC. Weak staining for NMT2 in CD4+ T and CD8+ T cells from subjects with NED was observed. In comparison, NMT2 was found to be overexpressed in both CD4+ and CD8+ T cells from CRC patients relative to cells from NED (Fig. 3). However, CD8+ T cells from subjects with CRC displayed the highest levels of NMT2 expression (Fig. 3).

Determination of NMT2 expression in CD4+ T cells and CD8+ T cells by IHC. PBMC were separated using Ficoll Paque followed by the isolation of CD4+ T cells and CD8+ T cells, cytospin slides were prepared and stained for NMT2 using polyclonal anti-NMT2 antibody by IHC. Representative images for NMT2 expression in PBMC of subjects with NED and CRC are shown. CRC, colorectal cancer; IHC, immunohistochemistry; NED, no evidence of disease; NMT2, N-myristoyltransferase 2; PBMC, peripheral blood mononuclear cells.


In the present study, we have identified higher expression of a protein-based biomarker, NMT2, in PBMCs from subjects with colorectal adenomatous polyps and cancer compared with subjects with nonadenomatous polyps and no evidence of disease. When expression of NMT2 was determined in fractionated T cells, we observed the highest expression in CD8+ T cells. Our results suggest that NMT2 could be used as a novel blood-based biomarker to screen for CRC and premalignant lesions. Although the mechanism of increased expression of NMT2 in PBMC is unknown and is currently being investigated by us, it is consistent with the increased expression and activation of c-Src in PBMC of CRC patients.17 The activation of c-Src requires myristoylation, and the activated c-Src has been demonstrated to be a surrogate marker in PBMC for the CRC progression and treatment response to dasatinib.17 In CRC, increased expression of c-Src correlates with malignant potential and metastasis with increasing expression in premalignant lesions and adenomas but with the highest expression in malignant polyps.18–20 These observations indicate that the increased expression of NMT2 in PBMC correlates with c-Src.

CRC detection has been carried out using circulating tumor cells, nucleic acids, and proteins that can be detected in blood serum or plasma or PBMCs of cancer patients. Several protein-based biomarkers like carcinoembryonic antigen, carbohydrate antigen 19-9, tissue polypeptide antigen, and tissue polypeptide specific antigen have been used to detect CRC; however, they have proven to be unreliable.21,22

The “cancer immunogram” has become a cornerstone in the prognosis of cancer, of which T cells play a vital role in tumor immunology.23 There are many biomarkers that are specifically of prognostic relevance in CRC. Significantly, the role of the microenvironment in CRC revealed CD8+ T cells infiltration leads to a favorable prognosis.24 NMT1 and NMT2 were found to be critical for the development of T cells. The conditional knockout (KO) of NMT1 or NMT2 in T cells revealed that NMT1 is essential for the development of T cells. NMT1 abnormal T cell development. NMT1 KO or double KO of NMT1 and NMT2 mainly affected the development of T cells and the impairment in TCR-mediated T cell activation, which was attributed to the mislocalization of Lck. NMT2 KO affected the selection of the double-positive T-cell stage.25 CD8+ T cell differentiation and dysfunction are associated with tumorigenesis, and one of the reasons is the exhaustion of CD8+ T cells.26 Our results suggest that the overexpression of NMT2 during the early stages of colorectal neoplasms could affect T cell function via impaired selection through double-positive stages.25 Further research is warranted to understand the implication of NMT2 overexpression in T cells’ functions.

Magnuson et al have demonstrated that NMT activity is elevated in colonic epithelial neoplasms, which appears at an early stage in animal models of colonic carcinogenesis.10 Additionally, a recent study has indicated that NMT2 expression is elevated in the early stages of CRC with high NMT2 expression in polyps compared with normal colonic tissue.11 The reason for overexpression of the protein and elevated activity of NMT in CRC is not well understood but is consistent with the greater demands of myristoylation of oncoproteins during tumorigenesis for activation or altered localization.

Other blood-based biomarkers used to identify individuals with CRC include monitoring circulating DNA from apoptotic cells or cells undergoing necrosis or live cells. Circulating DNA apart from cancers can also be found under inflammation and autoimmune disorders in serum or plasma, making them unreliable.

Reliable noninvasive screening of CRC is essential to enable the early detection of CRC. Cologuard increased the compliance for CRC screening since it is less invasive than a colonoscopy but has better sensitivity and specificity than FIT, however as demonstrated by the Septin-9 study, the compliance increases to 83% if the test offered is blood based.27 Therefore, the compliance will increase with a simple NMT2-based blood test and the ease of its inclusion in the annual physical. Our study indicates that NMT2 shows high sensitivity (91%) and specificity (81%) in PBMCs of subjects with colorectal adenomatous polyps and cancer. Most importantly, NMT2 based blood test may become valuable in triaging patients for colonoscopy. The Cologuard has a sensitivity of 42% in detecting advanced adenomas compared with 91% for NMT2.28

The levels of NMT2 protein in PBMC appear to gradually increase along with the natural progression of polyp development from nonadenomatous to adenomatous to CRC (Fig. 1). A significant proportion of CRC patients in our study reported having a family history of polyps or cancer (Tables 1 and 3). This is not surprising since cancer has a strong genetic component. It is plausible that enhanced NMT2 expression is also a genetic trait, which should be explored in future studies. Our findings should be interpreted in the light of limitations, including we did not run other tests such as FIT and EpiProcolon alongside.

Since age strongly correlates with CRC prevalence, it is a confounder that is often challenging to overcome in small cohort studies of cancer biomarkers. Age was indeed a strong correlate of CRC in our study. Interestingly, age did not substantially affect the performance of NMT2 as a biomarker when it was included in the multivariable model; on the contrary, the AUC of NMT2 H-score was slightly enhanced.

The IHC based tests encounter discordance in results.29 We are currently planning a study with a larger sample size that will also assess other screening tests alongside NMT2 expression. We plan to include the FACS based analysis of NMT2 expression and compare it with IHC results. Based on the results of the larger trial, moving forward, an alternate FACS based test could be developed similar to staining of PBMC cells for other indications such as HIV. We have already developed a FACS protocol for staining NMT2 in PBMC (Plesniarski et al, unpublished).


Given the high mortality due to CRC worldwide, a simple screening test that could be adopted with ease would significantly reduce deaths due to CRC. Our preliminary data suggest that the sensitivity of NMT2 in detecting adenomatous polyps is high. The results from this study indicate that NMT2 is a unique marker in peripheral blood for detecting CRC at an early adenoma stage and can improve compliance to CRC screening while preventing CRC dramatically.


S.V.S. acknowledges the support from CSB Manitoba, Canada. The authors thank Dr. S. V. Good for her help with some of the statistical analyses.


1. Xi Y, Xu P. Global colorectal cancer burden in 2020 and projections to 2040. Transl Oncol. 2021;14:101174.
2. Siegel RL, Miller KD, Fuchs HE, et al. Cancer Statistics, 2021. CA Cancer J Clin. 2021;71:7–33.
3. Coppedè F, Lopomo A, Spisni R, et al. Genetic and epigenetic biomarkers for diagnosis, prognosis and treatment of colorectal cancer. World J Gastroenterol. 2014;20:943–956.
4. Colorectal Cancer Screening. Recommendation statement from the Canadian Task Force on Preventive Health Care. CMAJ. 2001;165:206–208.
5. Winawer S, Fletcher R, Rex D, et al.; Gastrointestinal Consortium Panel. Colorectal cancer screening and surveillance: clinical guidelines and rationale-update based on new evidence. Gastroenterology. 2003;124:544–560.
6. Moayyedi P. Colorectal cancer screening lacks evidence of benefit. Cleve Clin J Med. 2007;74:545, 549–550, 552 passim.
7. Nicholson FB, Barro JL, Atkin W, et al. Review article: population screening for colorectal cancer. Aliment Pharmacol Ther. 2005;22:1069–1077.
8. Imperiale TF, Ransohoff DF, Itzkowitz SH, et al. Multitarget stool DNA testing for colorectal-cancer screening. N Engl J Med. 2014;370:1287–1297.
9. Lee JK, Resi V, Liu S, et al. Improving fecal occult blood testing compliance using a mailed educational reminder. J Gen Intern Med. 2009;11:1192–1197.
10. Magnuson BA, Raju RV, Moyana TN, et al. Increased N-myristoyltransferase activity observed in rat and human colonic tumors. J Natl Cancer Inst. 1995;87:1630–1635.
11. Selvakumar P, Smith-Windsor E, Bonham K, et al. N-myristoyltransferase 2 expression in human colon cancer: cross-talk between the calpain and caspase system. FEBS Lett. 2006;580:2021–2026.
12. Shrivastav A, Varma S, Saxena A, et al. N-myristoyltransferase: a potential novel diagnostic marker for colon cancer. J Transl Med. 2007;5:58.
13. Varma Shrivastav S, Shrivastav A. N-Myristoyltransferase (NMT)1, NMT2 and methionine aminopeptidase 2 overexpression in peripheral blood and peripheral blood mononuclear cells is a marker for adenomatous polyps and early detection of colorectal cancer. USPTO 2017. International Publication Number WO 2017/190241 Al.
14. Fuss IJ, Kanof ME, Smith PD, et al. Isolation of whole mononuclear cells from peripheral blood and cord blood. Curr Protoc Immunol. 2009;Chapter 7:Unit7.1.
15. Shrivastav A, Bruce M, Jaksic D, et al. The mechanistic target for rapamycin pathway is related to the phosphorylation score for estrogen receptor-α in human breast tumors in vivo. Breast Cancer Res. 2014;16:R49.
16. Hirsch FR, Varella-Garcia M, Bunn PA Jr, et al. Epidermal growth factor receptor in non-small-cell lung carcinomas: correlation between gene copy number and protein expression and impact on prognosis. J Clin Oncol. 2003;21:3798–3807.
17. Serrels A, Macpherson IR, Evans TR, et al. Identification of potential biomarkers for measuring inhibition of Src kinase activity in colon cancer cells following treatment with dasatinib. Mol Cancer Ther. 2006;5:3014–3022.
18. Bolen JB, Veillette A, Schwartz AM, et al. Activation of pp60c-src protein kinase activity in human colon carcinoma. Proc Natl Acad Sci U S A. 1987;84:2251–2255.
19. Cartwright CA, Kamps MP, Meisler AI, et al. pp60c-src activation in human colon carcinoma. J Clin Invest. 1989;83:2025–2033.
20. Cartwright CA, Meisler AI, Eckhart W. Activation of the pp60c-src protein kinase is an early event in colonic carcinogenesis. Proc Natl Acad Sci U S A. 1990;87:558–562.
21. Gao Y, Wang J, Zhou Y, et al. Evaluation of serum CEA, CA19-9, CA72-4, CA125 and ferritin as diagnostic markers and factors of clinical parameters for colorectal cancer. Sci Rep. 2018;8:2732.
22. Nicolini A, Ferrari P, Duffy MJ, et al. Intensive risk-adjusted follow-up with the CEA, TPA, CA19.9, and CA72.4 tumor marker panel and abdominal ultrasonography to diagnose operable colorectal cancer recurrences: effect on survival. Arch Surg. 2010;145:1177–1183.
23. Blank CU, Haanen JB, Ribas A, et al. CANCER IMMUNOLOGY. The “cancer immunogram.” Science. 2016;352:658–660.
24. Lalos A, Tülek A, Tosti N, et al. Prognostic significance of CD8+ T-cells density in stage III colorectal cancer depends on SDF-1 expression. Sci Rep. 2021;11:775.
25. Rampoldi F, Bonrouhi M, Boehm ME, et al. Immunosuppression and Aberrant T cell development in the absence of N-myristoylation. J Immunol. 2015;195:4228–4243.
26. Collier JL, Weiss SA, Pauken KE, et al. Not-so-opposite ends of the spectrum: CD8+ T cell dysfunction across chronic infection, cancer and autoimmunity. Nat Immunol. 2021;22:809–819.
27. Adler A, Geiger S, Keil A, et al. Improving compliance to colorectal cancer screening using blood and stool based tests in patients refusing screening colonoscopy in Germany. BMC Gastroenterol. 2014;14:183.
28. Issa IA, Noureddine M. Colorectal cancer screening: an updated review of the available options. World J Gastroenterol. 2017;23:5086–5096.
29. Hoda RS, Brogi E, D’Alfonso TM, et al. Interobserver variation of PD-L1 SP142 immunohistochemistry interpretation in breast carcinoma: a study of 79 cases using whole slide imaging. Arch Pathol Lab Med. 2021;145:1132–1137.

adenomatous polyps; colorectal cancer; N-myristoltransferase 2; PBMC

Supplemental Digital Content

Copyright © 2022 The Author(s). Published by Wolters Kluwer Health, Inc.