Sage Journals: Discover world-class research

Abstract

BACKGROUND:

Detection of circulating cell-free DNA (ccfDNA) methylated in BCAT1 and IKZF1 is sensitive for detection of colorectal cancer (CRC), but it is not known if these biomarkers are present in other common adenocarcinomas.

OBJECTIVE:

Compare methylation levels of BCAT1 and IKZF1 in tissue and plasma from breast, prostate, and colorectal cancer patients.

METHODS:

Blood was collected from 290 CRC, 32 breast and 101 prostate cancer patients, and 606 cancer-free controls. Tumor and matched normal tissues were collected at surgery: 26 breast, 9 prostate and 15 CRC. DNA methylation in BCAT1 and IKZF1 was measured in blood and tissues.

RESULTS:

Either biomarker was detected in blood from 175/290 (60.3%) of CRC patients. The detection rate was higher than that measured in controls (48/606 (8.1%), OR $=$ 18.2, 95%CI: 11.1–29.0). The test positivity rates in breast and prostate cancer patients were 9.4% (3/32) and 6.9% (7/101), respectively, and not significantly different to that measured in gender-matched controls (8.0% (33/382) females (OR $=$ 0.84, 95%CI: 0.23–3.1) and 7.6% (26/318) males (OR $=$ 0.86, 95%CI: 0.65–2.1). In tumor and non-neoplastic tissues, 93.5% (14/15) of CRC tumors were methylated in BCAT1 and/or IKZF1 ( $p<$ 0.004). Only 11.5% (3/26) and 44.4% (4/9) ( $p=$ 0.083) of breast and prostate tumors were hypermethylated in these two genes.

CONCLUSIONS:

Detection of circulating DNA methylated in BCAT1 and IKZF1 is sensitive and specific for CRC but not breast or prostate cancer.

Keywords

ctDNA biomarker DNA methylation bowel cancer sensitivity specificity diagnostic accuracy

1. Introduction

Colorectal cancer (CRC) is a common disease with high mortality rates. Estimates from 2020 showed that CRC was the fourth most commonly diagnosed non-cutaneous cancer, with almost 2 million newly diagnosed cases globally and almost 1 million deaths [1]. Survival rates for CRC can be improved to over 90% if the cancer is detected at an early stage [2]. Formation of pre-cancerous lesions, such as adenomas and sessile serrated lesions, can take over 10 years to progress to malignancy after accumulation of genetic and epigenetic abnormalities [3]. Such an extended period between benign and malignant disease enables opportunity to detect and remove pre-invasive lesions before they turn cancerous. Strategies that can accurately detect CRC, particularly early-stage disease or recurrence when monitoring cases after treatment, would benefit outcomes and improve treatment.

Detection of circulating tumor DNA (ctDNA) in the blood for identifying tumor-specific changes (such as epigenetic changes and mutations) is a promising strategy to detect CRC as well as other cancer types [4, 5, 6]. We have established a sensitive DNA methylation assay that can detect circulating hypermethylated BCAT1 (branched chain amino acid transaminase 1) and IKZF1 (IKAROS family zinc finger 1) in CRC cases with a sensitivity of up to 77% [7]. When combined with the current CRC screening fecal immunochemical test (FIT), the sensitivity for detecting cancer rises to 89% [8]. We have also shown that levels of methylated BCAT1 and IKZF1 DNA in the blood increase with tumor stage and that over 95% of CRC tissues have aberrant methylation of these two genes [7, 9] providing clinical utility for the detection and monitoring of recurrence in CRC patients. Whether these biomarkers are also present in patients with other common adenocarcinomas such as breast and prostate cancer is not known.

Breast and prostate cancer are hormone dependent cancers and are common in both women and men, respectively [1], and as such may be present but unsuspected in a CRC screening population. Although originating in different organs, breast and prostate cancer share many genetic and epigenetic characteristics [10]. There are also many genetic commonalities between adenocarcinomas of the breast, prostate and colorectum, including somatic mutations in oncogenes and/or tumor suppressor genes (e.g. TP53 and BRAF) [11], but also common sites of DNA hypermethylation in key genes controlling carcinogenesis, such as the Wnt, Hippo and retinoblastoma tumor suppressor pathway [12]. Previous studies have pointed to a role of BCAT1 and IKZF1 in breast and prostate cancer progression [13, 14, 15, 16, 17, 18]. However, less is understood about the extent to which these specific epigenetic changes characteristic of CRC tissues occur in breast and prostate cancers.

To support the clinical utility and specificity of the methylated BCAT1 and IKZF1 assay for detection and management of CRC, it is important to determine whether these other common cancers also harbor these aberrantly methylated genes and release them into the circulation. Thus, we measured levels of BCAT1 and IKZF1 DNA methylation in blood and tissues from patients diagnosed with adenocarcinomas of the breast and prostate and compared these to levels observed in patients with CRC and individuals with no evidence of cancer.

2. Materials and methods

2.1 Experimental subjects

Individuals with confirmed diagnosis of prostate cancer, breast cancer and CRC, and who attended clinic appointments at one of three South Australian public hospitals (Repatriation General Hospital, Flin- ders Medical Centre, and Noarlunga Health Service) between 2011 and 2020 were assessed for presence of methylated BCAT1 and IKZF1 in blood. Where possible, and providing that neoadjuvant therapy had not been given prior to surgery, tumor tissue and adjacent normal tissue samples were collected at the time of cancer resection or retrieved from archival samples obtained through the Flinders Tissue Biobank (Bedford Park, South Australia) for assessment of hypermethylation in these two genes. Inclusion criteria comprised patients aged 18 years and older with known diagnosis of primary invasive adenocarcinoma (i.e. not stage 0) of any pathological stage, of the breast, prostate, or colorectum. Cases were excluded from the study if they had already commenced any type of treatment, if an insufficient blood volume was collected for assay, or if they had a known active cancer in any other organ. Controls recruited from the same hospital sites between 2011 and 2014 were males and females aged over 18 years who had a blood sample collected prior to colonoscopy, with no known history of cancer diagnosis and who were negative for precancerous lesions (adenoma or sessile serrated lesions) and CRC at colonoscopy. Informed written consent was obtained prior to sample collection.

Experiments were undertaken with the understanding and written consent of each subject, and that the study conforms with The Code of Ethics of the World Medical Association (Declaration of Helsinki), printed in the British Medical Journal (18 July 1964). Approval was granted by the Southern Adelaide Clinical Human Research Ethics Committee (human ethics committee approvals 162.16, 134.045). The study was prospectively registered through the Australian New Zealand Clinical Trials Registry (registration numbers: ANZCTR 12616001138471 and 12611000318987). All authors had access to the study data and had reviewed and approved the final manuscript.

2.2 Clinical procedures

Pathology and clinical data from cases where blood was collected for methylation analysis were recorded to determine any associations with blood test positivity. TNM and AJCC staging [19, 20, 21] were assessed based on findings at surgery or at MRI if no surgery was performed. Staging by MRI in prostate and rectal cancer patients not undergoing surgery is a widely accepted clinical practice [22, 23], and demonstrates good accuracy for diagnosis of T stage [24, 25]. For colorectal tissue sampling, normal biopsies were collected at a minimum distance of 20 mm adjacent to the tumor (range 20–350 mm). For prostate tissue samples, 5 mm biopsy punch samples were collected from each quadrant of intact whole resected prostate glands after radical prostatectomy, with half of the biopsy sent for pathology diagnosis and the other half banked for DNA extraction. For breast tissues, whole resected breast samples were sectioned and tumor tissue identified by macroscopic examination by a pathologist. Normal breast tissue was taken from the quadrant with no visible macroscopically confirmed tumor tissue for DNA extraction. For prostate cancer, indicators of disease severity including Gleason score of the tumor and blood PSA were also collected. For breast cancer, tumor sub-type (invasive ductal carcinoma, lobular or mixed ductal/lobular), histology grade, and tumor size (at surgical resection or by imaging) were recorded, and hormone receptor status for ER, PR and HER2 were assessed by standard clinical immunohistochemical staining.

2.3 Blood DNA methylation testing

Approximately 18 mL of blood was collected via venipuncture (K3-EDTA Vacuette tubes, Greiner Bio-One, Frickenhausen, Germany), followed by plasma isolation, which was stored at $-$ 80 ${}^{\circ}$ C until analysis of circulating cell free (ccfDNA) methylated in BCAT1 and IKZF1 at Clinical Genomics (North Ryde, NSW). Detailed procedures for assaying DNA isolated from plasma for methylated BCAT1 and IKZF1 have been previously described [8, 26]. Briefly, ccfDNA was extracted from 5 mL of plasma using the QIAsymphony Circulating Nucleic Acid kit (Qiagen, Hilden, Germany) and bisulfite-converted using the Epitect Fast 96 Bisulfite Conversion kit (Qiagen GmbH) on a QIAcube HT liquid handler (Qiagen). Triplicate wells of bisulfite-converted ccfDNA were analyzed using QuantiTect Multiplex PCR NoROX mastermix (Qiagen) and oligonucleotides [26]. Standard curves were generated for each methylated gene (BCAT1, IKZF1) and reference gene (ACTB) for quantification based on serial dilution of fully methylated DNA (Zymo, USA). The 3-plex qPCR assay was performed on an LC480 II (Roche Diagnostics) with cycle threshold (Ct) values calculated using the second derivative algorithm provided with the LC480 software. The assay result was qualitatively called “positive” if any PCR replicate was positive for amplification of methylated BCAT1 or IKZF1 within 50 amplification cycles.

2.4 Tissue DNA methylation testing

Tumor and matched normal adjacent tissues (pro- state $N=$ 9, breast $N=$ 26, and colorectal $N=$ 15) were excised at the time of surgery and stored in RNAlater (Ambion Inc) solution at 4 ${}^{\circ}$ C overnight and transferred to $-$ 80 ${}^{\circ}$ C the following day, or snap frozen in liquid nitrogen, and analyzed at the Flinders Centre for Innovation in Cancer. DNA was extracted from all tumor and adjacent non-tumor tissues using DNeasy Blood & Tissue Kit (Qiagen) and DNA was used for bisulfite conversion using the EpiTect Fast Bisulfite Conversion kit (Qiagen) following the manufacturer’s instructions. A single-plex ACTB PCR assay was used to determine the DNA yield, whereafter the levels of methylated BCAT1 and IKZF1 were determined in 5 ng of bisulfite converted DNA using the QIAGEN Rotor gene, as previously described [27]. Levels of BCAT1 and IKZF1 were expressed as percentage of methylation normalized to ACTB concentration. Samples with BCAT1/IKZF1 methylation levels of 10% or more were deemed positive [28, 29, 30].

Table 1
Demographics and blood BCAT1 and IKZF1 positivity rates by cancer status

Cancer status	Breast	Prostate	Colorectal	Controls*
Cases	32	101	290	606
Gender
Female	32	0	112	359
Male	0	101	178	247
Mean age (SEM)
Female	64.9 (2.1)	–	65.8 (1.20)	60.4 (0.53)
Male	–	69.3 (0.78)	67.1 (0.88)	63.6 (0.50)
BCAT1/IKZF1 assay positivity (%)
BCAT1 and/or IKZF1	3 (9.4)	7 (6.9)	175 (60.3)	49 (8.1)
BCAT1	2 (6.3)	7 (6.9)	146 (50.3)	36 (5.9)
IKZF1	1 (3.1)	4 (3.7)	126 (43.5)	15 (2.5)

*Individuals undergoing colonoscopy with no history of cancer diagnosis. SEM $=$ Standard error of the mean.

2.5 Statistical analysis

Data are presented as percentages or as mean $\pm$ SEM for quantitative variables. For the blood methylation test results, age-adjusted ordinal logistic regression was used for CRC to compare test positivity rates for cancer cases versus non-cancer controls (limiting analysis of the control set to females for breast cancer, and males for prostate cancer). Age-adjusted logistic regression was used to compare test positivity rates between CRC cancers versus breast and prostate cancers. Age-adjusted Poisson regression was used for breast and prostate cancer cases to compare test positivity rates for cancer cases versus non-cancer controls, Age-adjusted Poisson regression was used for breast and prostate cancer cases to compare test positivity rates for cancer cases versus non-cancer controls. Due to low positive cases numbers and distribution of the data logistic regression could not be used. Multivariable logistic regression was used to explore relationships between test positivity and clinicopathological findings with age as a confounding factor. Odds ratio (OR) and relative risk (RR) were used for logistic and Poisson regression tests, respectively, in conjunction with 95% confidence intervals (95% CI) to designate the magnitude and direction of the outcomes. Mean percent (%) methylation $\pm$ SEM for each patient was determined using the ratio of BCAT1 and IKZF1 (pg of DNA per well) to ACTB (pg of DNA per well). Comparisons of % methylation between control, breast, prostate and CRC cases who returned a positive blood test were compared using Kruskal-Wallis equality-of-populations rank test.

For tissue methylation analysis, paired t-tests were used to compare % methylation $\pm$ SEM of cancer and matched normal tissues. McNemar’s tests were used to compare overall test positivity (positive for BCAT1 and/or IKZF1) in cancer versus normal tissues, using $\geqslant$ 10% DNA methylation normalized to ACTB as the cut-off for test positivity.

All data were analyzed using statistical package StataMP 16.0 ${}^{\@setsize{\scriptsize}{8pt}{\viipt}{\@viipt}\textregistered}$ StataCorp LLC (Texas, USA) and differences were considered significant at $p<$ 0.05.

3. Results

3.1 Experimental subjects

Overall, there were 32 breast, 101 prostate, and 290 CRC patients and 606 non-cancer controls who met inclusion criteria. Case data are outlined in Table 1. Female CRC patients and female non-cancer controls were generally younger than their male counterparts, and prostate cancer patients were the oldest cohort with a mean age of 69.3 years $\pm$ 0.78 standard error of the mean (SEM).

Figure 1.

Blood test positivity rates and concentration of circulating methylated BCAT1 and IKZF1. (A) Blood test positivity rates ( $\pm$ 95% confidence interval) for breast, prostate and colorectal cancer versus non-cancer controls. F, females. M, males. (B) Mean % methylation (m/m $=$ pg of BCAT1 or IKZF1/pg of ACTB) of test positive cases ( $\pm$ standard error of the mean (SEM); log10 scale) of methylated BCAT1 and IKZF1 for non-cancer controls ( $N=$ 49), breast cancer ( $N=$ 3), prostate cancer ( $N=$ 7) and CRC ( $N=$ 175).

Figure 2.

Venn diagrams showing concordance for BCAT1 and IKZF1 positive cases for non-cancer controls (A), breast and prostate cancer (B), and CRC cases (C).

3.2 Methylated BCAT1 and IKZF1 DNA in blood

There were 9.4% (3/32) and 6.9% (7/101) breast and prostate cancer cases, respectively, who tested positive for methylated BCAT1 and/or IKZF1 DNA in blood (Table 1). The test positivity rates were no different to the test positivity rates observed for non-cancer controls at 8.0% for females (odds ratio (OR) $=$ 0.84, 95% confidence interval (CI) $=$ 0.2–3.1) and 7.6% for males (OR $=$ 0.86, 95%CI $=$ 0.7–2.1; Fig. 1A). A total of 60.3% (175/290) CRC cases were positive for BCAT1 and/or IKZF1 methylation in blood, which was significantly higher than non-cancer controls (males and females) with a positivity rate of 8.1% (49/606) (OR 18.2 95%CI 11.1–29.0; Fig. 1A). The blood test positivity rate was also significantly higher in CRC patients compared to the positivity rates measured in breast ( $p<$ 0.0001) and prostate ( $p<$ 0.0001) cancer patients (Fig. 1A).

Average levels of % methylated BCAT1 and IKZF1 in the blood from patients testing positive were over 100-fold higher in CRC patients (BCAT1: 5.4% $\pm$ 1.3%, IKZF1: 5.6% $\pm$ 1.2%) compared to breast (BCAT1: 0.02% $\pm$ 0.02%, IKZF1: 0.1% $\pm$ 0.1%) and prostate (BCAT1: 0.1% $\pm$ 0.1%, IKZF1: 1.0 $\times$ 10 ${}^{-3}$ % $\pm$ 1.0 $\times$ 10 ${}^{-3}$ %) cancer patients and non-cancer controls (BCAT1: 0.1 $\pm$ 2.0 $\times$ 10 ${}^{-2}$ %, IKZF1: 2.0 $\times$ 10 ${}^{-2}$ % $\pm$ 1.0 $\times$ 10 ${}^{-2}$ %) ( $p<$ 0.0001; Fig. 1B).

Table 2
Circulating methylated BCAT1/IKZF1 test result according to cancer stage

	Breast ${}^{1}$	Prostate ${}^{1}$	CRC
	Positive/all cases (%)	Positive/all cases (%)	Positive/all cases (%)	OR (95% CI) ${}^{3}$	$P$ -value ${}^{3}$
T stage (%) ${}^{2}$	32	101	278
T1	2/18 (11.1)	3/21 (14.3)	1/22 (4.5)	1.0
T2	0/10 (0)	2/58 (3.4)	18/49 (36.7)	13.0 (1.6 – 104.9)	0.016
T3	1/2 (50.0)	2/22 (9.1)	107/160 (66.9)	45.0 (5.8 – 343.0)	$<$ 0.0001
T4	0/2 (0)	0/0 (0)	38/47 (80.9)	93.0 (11.0 – 784.1)	$<$ 0.0001
N stage (%) ${}^{2}$	32	101	277
N0	3/26 (11.5)	7/99 (7.1)	72/151 (47.7)	1.0
N1	0/2 (0)	0/2 (0)	48/77 (62.3)	1.8 (1.0 – 3.2)	0.037
N2	0/0 (0)	N/A	43/49 (87.8)	8.0 (3.2 – 19.9)	$<$ 0.0001
N3	0/1 (0)	N/A	N/A
M stage (%) ${}^{2}$	32	101	290	1.0
M0	3/31 (9.7)	7/101 (6.9)	136/247 (55.1)
M1	0/1 (0)	0/0 (0)	39/43 (91.0)	7.9 (2.7 – 22.9)	$<$ 0.0001
Overall stage (%) ${}^{2}$	32	101	290
I	2/21 (9.5)	0/12 (0)	14/59 (23.7)	1.0
II	1/8 (12.5)	5/67 (7.5)	53/86 (61.6)	5.2 (2.5 – 10.9)	$<$ 0.0001
III	0/2 (0)	2/20 (10.0)	69/102 (67.6)	6.7 (3.2 – 13.9)	$<$ 0.0001
IV	0/1 (0)	0/2 (0)	39/43 (90.7)	31.3 (9.4 – 103.5)	$<$ 0.0001
Staged by surgery	29/32 (90.6)	38/101 (37.6)	265/290 (91.4%)	–	–

${}^{1}$ Cannot perform statistics due to small sample size. ${}^{2}$ American Joint Committee on Cancer staging (assessed by MRI or surgery). ${}^{3}$ Logistical regressions for CRC cases only, with reference set as T1 (T stage), N0 (N stage), M0 (M stage) or stage I. OR $=$ odds ratio. CI $=$ confidence interval.

Concordance of blood test positive cases across the different patient groups for BCAT1 and IKZF1 and a combination of the two genes are shown in Fig. 2. Non-cancer controls (Fig. 2A) and breast and prostate cancer patients (Fig. 2B) were more likely to test positive for BCAT1 than for IKZF1 alone or for BCAT1 and IKZF1 combined ( $p<$ 0.01). Conversely, CRC cases were more likely to test positive for both BCAT1 and IKZF1 over BCAT1 or IKZF1 alone ( $p<$ 0.001; Fig. 2C).

Table 3

Methylated BCAT1 and IKZF1 in matched normal and cancer patient tissues

		Breast ( $N=$ 26)		Prostate ( $N=$ 9)		Colorectal ( $N=$ 15)
		Positive (%)	$P$ -value	Positive (%)	$P$ -value	Positive (%)	$P$ -value
BCAT1	Normal	0 (0)	0.317	0 (0)	0.317	5 (33.3)	0.003
	Cancer	1 (3.8)		1 (11.1)		14 (93.3)
IKZF1	Normal	0 (0)	0.157	1 (11.1)	0.158	0 (0)	0.001
	Cancer	2 (7.7)		3 (33.3)		11 (73.3)
BCAT1 and/or IKZF1	Normal	0 (0)	0.083	1 (11.1)	0.083	5 (33.3)	0.004
	Cancer	3 (11.5)		4 (44.4)		14 (93.3)

Figure 3.

DNA methylation levels for BCAT1 (A) and IKZF1 (B) in matched normal and cancer tissues. Test positivity cut-off was set at 10% methylation. Cases were defined as not detectable where there was no signal from three replicates in 50 PCR amplification cycles.

3.3 Association between methylated BCAT1 and IKZF1 DNA in blood and clinicopathology findings

Blood test positivity for each cancer type and for each primary tumor, lymph node and metastasis (TNM) stage and AJCC (American Joint Committee on Cancer) stage is presented in Table 2. For CRC patients, there were strong associations between detection of methylated BCAT1/IKZF1 DNA in blood and advancing tumor stage (T1 4.3% vs T4 80.9%; OR $=$ 23.0, 95%CI $=$ 11.0–784.1), nodal stage (N0 47.7% vs N2 87.8%; OR $=$ 8.0, 95%CI $=$ 3.2–19.9), metastasis stage (M0 55.1% vs M1 90.7%; OR $=$ 7.9, 95%CI $=$ 2.7–22.9) and AJCC stage (Stage I 23.7% vs Stage IV 90.7%; OR $=$ 21.9, 95%CI $=$ 9.4–103.5; Table 2). For breast and prostate cancer, there were no obvious associations between stage and detection of BCAT1 and IKZF1, with the three breast cancer positive cases being stage I and II, and 5/7 of the prostate cancer positive cases being stage II (Table 2).

Blood samples were taken at a median time of 12 days after biopsy for breast cancer patients (range 1–39 days) and 36 days after biopsy for prostate cancer patients (range 13–796 days), where 1 day is sufficient time for clearance of $>$ 95% of ctDNA from circulation post-surgery [31]. Additionally, there were no significant differences between the time from biopsy until blood test for those that were positive versus negative for the blood biomarkers for either breast or prostate cancer ( $p=$ 0.866 and $p=$ 0.472, respectively), which indicates that the biopsy procedure did not contribute to the small proportion of ccfDNA positive cases observed.

All BCAT1/IKZF1 positive breast cancer cases were diagnosed as invasive ductal carcinoma (IDC), two positive cases were estrogen receptor- $\alpha$ positive (ER $+$ ) and one ER-(Supplementary Table 1). One progesterone receptor positive (PR $+$ ) and two PR-breast cancer cases returned positive blood tests for BCAT1 and/or IKZF1. The only patient diagnosed with triple negative breast cancer (ER-/PR-/human epidermal growth receptor (HER2)-) tested positive for IKZF1. For prostate cancer patients (Supplementary Table 2), there was no association of BCAT1/IKZF1 test positivity with low versus high prostate specific antigen (PSA) levels (RR $=$ 0.7, 95%CI 0.15–3.4), or low/medium versus high Gleason score (RR $=$ 0.73, 95%CI $=$ 0.08–6.4) (Supplementary Table 2).

3.4 BCAT1 and IKZF1 methylation in tissues

Normal and matched cancer tissues were collected from 26 breast, 9 prostate and 15 CRC patients. Percent DNA methylation of BCAT1 and IKZF1 (Fig. 3A and B, respectively) were significantly higher in CRC tissue samples compared to their matched normal samples ( $p<$ 0.001). No differences in % methylation for either biomarker was found for breast and prostate cancer tissues compared to their matched normal tissue samples. For cancer tissues, mean % methylation in CRC tissues (BCAT1 64.4% $\pm$ 10.2% and IKZF1 68.1% $\pm$ 13.6%) were significantly higher than levels for breast cancer (BCAT1 0.8%% $\pm$ 0.7%% and IKZF1 7.0%% $\pm$ 5.0%) and prostate cancer (BCAT1 1.9%% $\pm$ 1.6%% and IKZF1 8.4%% $\pm$ 4.3%) tissues ( $p<$ 0.0001 for BCAT1 and $p<$ 0.0003 for IKZF1).

Percent DNA methylation was reported as a binary outcome using a 10% methylation cut-off value to assess hypermethylation of cancer versus matched normal tissues for breast, prostate and CRC (Fig. 3 and Table 3). Breast and prostate cancers were not significantly hypermethylated for BCAT1 and/or IKZF1 compared to matched normal tissue samples. However, CRC tissues were significantly hypermethylated for BCAT1 ( $p=$ 0.003), IKZF1 ( $p=$ 0.001) and for BCAT1 and/or IKZF1 ( $p<$ 0.004) compared to matched normal colonic mucosa in the cancer patients.

4. Discussion

This study explored whether detection of methylated BCAT1 and IKZF1 DNA in blood was specific for CRC by comparing levels of these biomarkers in plasma from patients with breast, prostate or CRC, as well as control cases without evidence of any cancer. The blood test positivity rates in breast and prostate cancer patients were not significantly different from those of controls, whereas blood test positivity was significantly higher in CRC patients. These findings were consistent with levels of methylated BCAT1 and IKZF1 in cancer tissue from the three organs; levels were considerably higher in CRC tissues $.$ Most CRC cancer tissues (93.3%) were hypermethylated for one or both biomarkers, in agreement with our previous findings [8, 27], with methylation levels around 100-fold higher compared to breast and prostate cancer cases. There was no evidence of significant hypermethylation in either gene for breast or prostate cancer tissues. Therefore, this study further supports the use of assaying circulating DNA for methylation in BCAT1 and IKZF1 DNA for detection of CRC.

The literature suggests a role for IKZF1 and BCAT1 in breast and prostate cancers, and therefore it was important to determine whether the biomarkers sequences applied within the assay for CRC detection are also hypermethylated in these other common cancers. This could have implications for the exclusive use of the circulating methylated BCAT1/IKZF1 biomarkers for CRC detection. The transcription factor IKZF1 has been implicated as a tumor suppressor in breast and prostate cancer [32, 33, 34], but also as a prostate cancer oncogene [17, 35], depending on the type of cancer cell-line being tested. Over-expression of BCAT1, a protein that catalyzes transamination of branched chain amino acids, leading to cell growth, is associated with poorer survival in breast cancer patients and enhanced malignancy of breast cancer cell lines in vitro [13, 14, 15, 36]. In prostate cancer, higher expression of BCAT1 promotes cancer cell survival [16] and decreases sensitivity to chemotherapy treatments of prostate cancer cells in vitro [18]. Although there is a suggestive role of BCAT1 and IKZF1 in breast and prostate cancer progression, less is understood whether this is driven by aberrant methylation in these genes. Our findings suggest this is not the case as there was no significant difference in the levels of methylation in tumor versus normal tissues at the CpG sites interrogated by the PCR assay studied here for either BCAT1 or IKZF1. Thus, this assay is not likely to have any translational potential as an effective screening tool for breast or prostate cancer.

Detection of methylated BCAT1/IKZF1 in CRC blood samples was strongly associated with higher TNM and AJCC stage, agreeing with our previous findings showing that the test is more sensitive for detection of advanced disease [9, 37]. A lower sensitivity for early-stage cancer, particularly T1 cancers, is a well-described limitation of ctDNA tests, including for breast and prostate cancer [38]. This study was limited by the number of advanced cancers cases for both breast and prostate cancer patients. However, the trends observed in blood testing were the same for these cancers and non-cancer controls in that we found circulating BCAT1 methylation was more frequently detected than IKZF1 methylation in blood. This aligns with our previous findings showing that methylation of BCAT1 is the main source of false positives in patients without colonoscopically-evident CRC [39]. The low frequency of methylated BCAT1/IKZF1 in blood collected from patients without CRC may stem from lysed white blood cells [30].

The main limitation of the current study was the small number of breast and prostate cancer cases. Additionally, statistical comparisons for breast and prostate cancer cases were constrained as there were zero cases for some subgroups (e.g. nodal stage and tumor stage). Larger-cohort studies are warranted to confirm the results reported herein. A further limitation of this study was that it did not assess sensitivity of the biomarkers in other common adenocarcinomas, such as lung cancer. This warrants further investigation as lung cancer is also commonly diagnosed in both men and women [1]. A potential weakness of the study was the use of adjacent normal tissues as opposed to using tissue samples from patients without cancer. Therefore, we cannot exclude the possibility that some cancer cells were in the “normal” samples or that there was a field effect in in adjacent non-neoplastic tissues. Despite these limitations, there was a clear differentiation between biomarker hypermethylation levels in cancer vs non-neoplastic tissue from CRC patients, in contrast to the low levels of methylation in these genes observed in tissues from prostate and breast cancer patients.

This study concludes that detection of circulating DNA methylated in BCAT1 and IKZF1 is both sensitive and specific for detection of CRC, supporting its use as a valuable non-invasive tool for detection of CRC. The likelihood of a positive test arising as a result of unsuspected prostate or breast cancer in a CRC screening population is very low.

Footnotes

Acknowledgments

Archival biological samples and data used in this research were provided by the Flinders Tissue Bank.

Author contributions

Jean M. Winter, Lorraine Sheehan-Hennessy, Beibei Yao, Susanne K. Pedersen, Molla M. Wassie, Michael Eaton, Michael Chong, Graeme P. Young, Erin L. Symonds.

Conception: Susanne K. Pedersen, Graeme P. Young, Erin L. Symonds Interpretation or analysis of data: Jean M. Winter, Beibei Yao, Susanne K. Pedersen, Molla M. Wassie, Graeme P. Young, Erin L. Symonds.

Preparation of the manuscript: Jean M. Winter, Lorraine Sheehan-Hennessy, Susanne K. Pedersen, Molla M. Wassie, Graeme P. Young, Erin L. Symonds.

Revision for important intellectual content: Jean M. Winter, Lorraine Sheehan-Hennessy.

Susanne K. Pedersen, Molla M. Wassie, Michael Eaton, Michael Chong, Graeme P. Young, Erin L. Symonds.

Supervision: Graeme P. Young, Erin L. Symonds.

Supplementary data

Supplementary Table 1

Breast cancer patients: circulating methylated BCAT1/IKZF1 positivity for different clinicopathological findings

	All cases ${N}$	BCAT1/IKZF1 test positivity $N$ (%)
Sub-type
Non-invasive ductal carcinoma	9	0
Invasive ductal carcinoma	23	3 (13)
Histological Grade
Low (1–5)	7	1 (14.3)
High (6–9)	22	1 (4.5)
Hormone receptor status
ER $+$	31	2 (6.5)
ER $-$	1	1 (100)
PR $+$	25	1 (4)
PR $-$	7	2 (28.6)

ER $=$ Estrogen receptor. PR $=$ Progesterone receptor.

Supplementary Table 2

Prostate cancer patients: circulating methylated BCAT1/IKZF1 positivity for different clinicopathological findings

	$N$	$N$ (%)	P ${}^{1}$	RR	95% CI
	All cases	BCAT1/IKZF1 test positivity
PSA
Low (4–9.99 ng/mL)	56	5 (8.9)
High ( $\geqslant$ 10 ng/mL)	43	2 (4.7)	0.662	0.70	0.15 – 3.4
Gleason score ${}^{2}$
Low/medium (6–7)	78	6 (7.7)
High (8–10)	21	1 (4.8)	0.776	0.73	0.08 – 6.4

${}^{1}$ Age-adjusted Poisson regression. ${}^{2}$ Gleason score was stratified into low/medium grade that are well-moderately differentiated or high grade that are poorly differentiated (American Society of Clinical Oncology). PSA $=$ Prostate specific antigen. RR $=$ Relative Risk.

References

Sung

Ferlay

Siegel

R.L.

Laversanne

Soerjomataram

Jemal

and Bray

, Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries, CA Cancer J Clin 71 (2021), 209–249.

N.C.I. Surveillance Research Program, SEER*Explorer: An interactive website for SEER cancer statistics, in, 2015.

Noffsinger

A.E.

, Serrated polyps and colorectal cancer: new pathway to malignancy, Annu Rev Pathol 4 (2009), 343–364.

Heichman

K.A.

and Warren

J.D.

, DNA methylation biomarkers and their utility for solid cancer diagnostics, Clin Chem Lab Med 50 (2012), 1707–1721.

Symonds

E.L.

and Young

G.P.

, Blood tests for colorectal cancer screening in the standard risk population, Current Colorectal Cancer Reports 11 (2015), 397–407.

Constâncio

Nunes

S.P.

Henrique

and Jerónimo

, DNA methylation-based testing in liquid biopsies as detection and prognostic biomarkers for the four major cancer types, Cells 9 (2020).

Pedersen

S.K.

Baker

R.T.

McEvoy

Murray

D.H.

Thomas

Molloy

P.L.

Mitchell

Lockett

Young

G.P.

and LaPointe

L.C.

, A two-gene blood test for methylated DNA sensitive for colorectal cancer, PLoS One 10 (2015), e0125041.

Symonds

E.L.

Pedersen

S.K.

Baker

R.T.

Murray

D.H.

Gaur

Cole

S.R.

Gopalsamy

Mangira

LaPointe

L.C.

and Young

G.P.

, A blood test for methylated BCAT1 and IKZF1 vs. a fecal immunochemical test for detection of colorectal neoplasia, Clin Transl Gastroenterol 7 (2016), e137.

Symonds

E.L.

Pedersen

S.K.

Murray

D.H.

Jedi

Byrne

S.E.

Rabbitt

Baker

R.T.

Bastin

and Young

G.P.

, Circulating tumour DNA for monitoring colorectal cancer-a prospective cohort study to assess relationship to tissue methylation, cancer characteristics and surgical resection, Clin Epigenetics 10 (2018), 63.

10.

Risbridger

G.P.

Davis

I.D.

Birrell

S.N.

and Tilley

W.D.

, Breast and prostate cancer: More similar than different, Nat Rev Cancer 10 (2010), 205–212.

11.

Lindsay

C.R.

Shaw

E.C.

Blackhall

Blyth

K.G.

Brenton

J.D.

Chaturvedi

Clarke

Dick

Evans

T.R.J.

Hall

Hanby

A.M.

Harrison

D.J.

Johnston

S.R.D.

Mason

M.D.

Morton

Newton-Bishop

Nicholson

A.G.

Oien

K.A.

Popat

Rassl

Sharpe

Taniere

Walker

Wallace

W.A.

West

N.P.

Butler

Gonzalez de Castro

Griffiths

and Johnson

P.W.M.

, Somatic cancer genetics in the UK: Real-world data from phase I of the cancer research UK stratified medicine programme, ESMO Open 3 (2018), e000408.

12.

Wilfred

Korbie

and Trau

, Regulation of canonical oncogenic signaling pathways in cancer via DNA methylation, Cancers (Basel) 12 (2020).

13.

Oktyabri

Ishimura

Tange

Terashima

and Suzuki

, DOT1L histone methyltransferase regulates the expression of BCAT1 and is involved in sphere formation and cell migration of breast cancer cell lines, Biochimie 123 (2016), 20–31.

14.

Song

Zhao

Ren

Lin

Zhou

and Sun

, Prognostic significance of branched-chain amino acid transferase 1 and CD133 in triple-negative breast cancer, BMC Cancer 20 (2020), 584.

15.

Zhang

and Han

, Branched-chain amino acid transaminase 1 (BCAT1) promotes the growth of breast cancer cells through improving mTOR-mediated mitochondrial biogenesis and function, Biochem Biophys Res Commun 486 (2017), 224–231.

16.

Chattopadhyay

Wang

Qin

Gao

Holtz

Vessella

R.L.

Leach

R.W.

and Gelman

I.H.

, Src promotes castration-recurrent prostate cancer through androgen receptor-dependent canonical and non-canonical transcriptional signatures, Oncotarget 8 (2017), 10324–10347.

17.

Kirby

M.K.

Ramaker

R.C.

Roberts

B.S.

Lasseigne

B.N.

Gunther

D.S.

Burwell

T.C.

Davis

N.S.

Gulzar

Z.G.

Absher

D.M.

Cooper

S.J.

Brooks

J.D.

and Myers

R.M.

, Genome-wide DNA methylation measurements in prostate tissues uncovers novel prostate cancer diagnostic biomarkers and transcription factor binding patterns, BMC Cancer 17 (2017), 273.

18.

Zhu

Shao

and Peng

, MicroRNA-218 inhibits tumor growth and increases chemosensitivity to CDDP treatment by targeting BCAT1 in prostate cancer, Mol Carcinog 56 (2017), 1570–1577.

19.

Amin

M.B.

Greene

F.L.

Edge

S.B.

Compton

C.C.

Gershenwald

J.E.

Brookland

R.K.

Meyer

Gress

D.M.

Byrd

D.R.

and Winchester

D.P.

, The eighth edition AJCC cancer staging manual: Continuing to build a bridge from a population-based to a more “personalized” approach to cancer staging, CA Cancer J Clin 67 (2017), 93–99.

20.

Buyyounouski

M.K.

Choyke

P.L.

McKenney

J.K.

Sartor

Sandler

H.M.

Amin

M.B.

Kattan

M.W.

and Lin

D.W.

, Prostate cancer – major changes in the American Joint Committee on Cancer eighth edition cancer staging manual, CA Cancer J Clin 67 (2017), 245–253.

21.

Giuliano

A.E.

Edge

S.B.

and Hortobagyi

G.N.

, Eighth edition of the AJCC cancer staging manual: Breast cancer, Annals of Surgical Oncology 25 (2018), 1783–1785.

22.

Parker

Castro

Fizazi

Heidenreich

Ost

Procopio

Tombal

and Gillessen

, Prostate cancer: ESMO clinical practice guidelines for diagnosis, treatment and follow-up? Annals of Oncology 31 (2020), 1119–1134.

23.

Glimelius

Tiret

Cervantes

and Arnold

, Rectal cancer: ESMO clinical practice guidelines for diagnosis, treatment and follow-up? Annals of Oncology 24 (2013), vi81–vi88.

24.

Kam

Yuminaga

Krelle

Gavin

Koschel

Aluwihare

Sutherland

Skinner

Brennan

Wong

L.M.

and Louie-Johnsun

, Evaluation of the accuracy of multiparametric MRI for predicting prostate cancer pathology and tumour staging in the real world: An multicentre study, BJU Int 124 (2019), 297–301.

25.

Al-Sukhni

Milot

Fruitman

Beyene

Victor

J.C.

Schmocker

Brown

McLeod

and Kennedy

, Diagnostic accuracy of MRI for assessment of T category, lymph node metastases, and circumferential resection margin involvement in patients with rectal cancer: A systematic review and meta-analysis, Annals of Surgical Oncology 19 (2012), 2212–2223.

26.

Symonds

E.L.

Pedersen

S.K.

Murray

Byrne

S.E.

Roy

Karapetis

Hollington

Rabbitt

Jones

F.S.

LaPointe

Segelov

and Young

G.P.

, Circulating epigenetic biomarkers for detection of recurrent colorectal cancer, Cancer 126 (2020), 1460–1469.

27.

Jedi

Young

G.P.

Pedersen

S.K.

and Symonds

E.L.

, Methylation and gene expression of BCAT1 and IKZF1 in colorectal cancer tissues, Clinical Medicine Insights: Oncology 12 (2018), 1179554918775064.

28.

Qiu

Huang

Zhou

and Zheng

, Aberrant methylation of TRIM58 in hepatocellular carcinoma and its potential clinical implication, Oncol Rep 36 (2016), 811–818.

29.

Xie

Tubbs

and Yang

, Detection of MGMT promoter methylation in glioblastoma using pyrosequencing, Int J Clin Exp Pathol 8 (2015), 636–642.

30.

Mitchell

S.M.

Ross

J.P.

Drew

H.R.

Brown

G.S.

Saunders

N.F.

Duesing

K.R.

Buckley

M.J.

Dunne

Beetson

Rand

K.N.

McEvoy

Thomas

M.L.

Baker

R.T.

Wattchow

D.A.

Young

G.P.

Lockett

T.J.

Pedersen

S.K.

Lapointe

L.C.

and Molloy

P.L.

, A panel of genes methylated with high frequency in colorectal cancer, BMC Cancer 14 (2014), 54.

31.

Diehl

Schmidt

Choti

M.A.

Romans

Goodman

Thornton

Agrawal

Sokoll

Szabo

S.A.

Kinzler

K.W.

Vogelstein

and Diaz

L.A.

, Jr, Circulating mutant DNA to assess tumor dynamics, Nature Medicine 14 (2008), 985–990.

32.

Tong

Song

and Deng

, Combined analysis and validation for DNA methylation and gene expression profiles associated with prostate cancer, Cancer Cell International 19 (2019), 50.

33.

Wang

Zhu

Guo

Song

Fan

and Qin

, Identification of tumor microenvironment-related prognostic biomarkers in luminal breast cancer, Front Genet 11 (2020), 555865.

34.

Zhou

Yan

Zhang

Zhao

and Han

, Ginseng polysaccharide inhibits MDA-MB-231 cell proliferation by activating the inflammatory response, Exp Ther Med 20 (2020), 229.

35.

Kim

S.A.

S.-H.

Park

S.J.

Jeon

Y.U.

Kim

J.E.

Lee

H.K.

Park

C.H.

Lee

C.-O.

Park

S.G.

Kim

Park

B.C.

Cho

S.Y.

Kim

J.D.

Kim

J.-H.

and Hwang

J.Y.

, A novel cereblon modulator for targeted protein degradation, European Journal of Medicinal Chemistry 166 (2019), 65–74.

36.

Thewes

Simon

Hlevnjak

Schlotter

Schroeter

Schmidt

Anzeneder

Wang

Windisch

Kirchgäßner

Melling

Kneisel

Büttner

Deuschle

Sinn

H.P.

Schneeweiss

Heck

Kaulfuss

Hess-Stumpp

Okun

J.G.

Sauter

Lykkesfeldt

A.E.

Zapatka

Radlwimmer

Lichter

, and Tönjes

, The branched-chain amino acid transaminase 1 sustains growth of antiestrogen-resistant and ERα-negative breast cancer, Oncogene 36 (2017), 4124–4134.

37.

Cock

Anwar

Byrne

S.E.

Meng

Pedersen

Fraser

R.J.L.

Young

G.P.

and Symonds

E.L.

, Low sensitivity of fecal immunochemical tests and blood-based markers of DNA hypermethylation for detection of sessile serrated adenomas/polyps, Dig Dis Sci 64 (2019), 2555–2562.

38.

Bettegowda

Sausen

Leary

R.J.

Kinde

Wang

Agrawal

Bartlett

B.R.

Wang

Luber

Alani

R.M.

Antonarakis

E.S.

Azad

N.S.

Bardelli

Brem

Cameron

J.L.

Lee

C.C.

Fecher

L.A.

Gallia

G.L.

Gibbs

Giuntoli

R.L.

Goggins

Hogarty

M.D.

Holdhoff

Hong

S.M.

Jiao

Juhl

H.H.

Kim

J.J.

Siravegna

Laheru

D.A.

Lauricella

Lim

Lipson

E.J.

Marie

S.K.

Netto

G.J.

Oliner

K.S.

Olivi

Olsson

Riggins

G.J.

Sartore-Bianchi

Schmidt

Shih l

Oba-Shinjo

S.M.

Siena

Theodorescu

Tie

Harkins

T.T.

Veronese

Wang

T.L.

Weingart

J.D.

Wolfgang

C.L.

Wood

L.D.

Xing

Hruban

R.H.

Allen

P.J.

Schmidt

C.M.

Choti

M.A.

Velculescu

V.E.

Kinzler

K.W.

Vogelstein

Papadopoulos

and Diaz

L.A.

, Jr, Detection of circulating tumor DNA in early- and late-stage human malignancies, Sci Transl Med 6 (2014), 224ra224.

39.

Saluja

Young

G.P.

Kholmurodova

and Symonds

E.L.

, Variables associated with detection of methylated BCAT1 or IKZF1 in blood from patients without colonoscopically evident colorectal cancer, Cancer Epidemiol Biomarkers Prev 30 (2021), 774–781.

Detection of hypermethylated BCAT1 and IKZF1 DNA in blood and tissues of colorectal,breast and prostate cancer patients

Abstract

BACKGROUND:

OBJECTIVE:

METHODS:

RESULTS:

CONCLUSIONS:

Keywords

1. Introduction

2. Materials and methods

2.1 Experimental subjects

2.2 Clinical procedures

2.3 Blood DNA methylation testing

2.4 Tissue DNA methylation testing

Table 1 Demographics and blood BCAT1 and IKZF1 positivity rates by cancer status

3. Results

3.1 Experimental subjects

Table 2 Circulating methylated BCAT1/IKZF1 test result according to cancer stage

3.4 BCAT1 and IKZF1 methylation in tissues

4. Discussion

Footnotes

Acknowledgments

Author contributions

Supplementary data

References

Table 1
Demographics and blood BCAT1 and IKZF1 positivity rates by cancer status

Table 2
Circulating methylated BCAT1/IKZF1 test result according to cancer stage