Sage Journals: Discover world-class research

Abstract

Background

Helicobacter pylori infection causes long-term chronic active gastritis, a risk factor for the intestinal and diffuse forms of gastric cancer. Most gastric cancers develop in a stepwise progression from chronic active gastritis to precursor lesions of gastric cancer. The early detection of gastric cancer improves survival. Studies with recent evidence have proposed circulating-microRNAs as biomarkers of cancer.

Objective

The purpose of this study was to explore the circulating-microRNA profile from H. pylori infection to gastric adenocarcinoma.

Methods

One hundred and twenty-three patients were enrolled and assigned to the discovery or the validation sets. In the discovery phase, circulating-microRNAs were measured by dye-based quantitative polymerase chain reaction and a selection of circulating-microRNAs was validated by probe-based quantitative polymerase chain reaction. A quality control protocol was used.

Results

One hundred and sixty-seven circulating-microRNAs were detected. Precursor lesions of gastric cancer and gastric cancer patients showed the downregulation of eight and five circulating-microRNAs, respectively. We further validated the deregulation of miR-196a-5p in precursor lesions of gastric cancer and the deregulation of miR-134-5p, miR-144-3p and miR-451a in gastric cancer. However, circulating-microRNAs exhibited moderate diagnostic performance due to the overlap of circulating-microRNA expression between non-cancer and cancer patients. miR-144-3p/miR-451a expression levels were correlated. Interestingly, these microRNAs are in 17q11.2, a site of rearrangements associated with gastric cancer.

Conclusion

Circulating-microRNAs are deregulated in precancerous and gastric cancer patients but efforts are needed to improve their diagnostic accuracy.

Keywords

Correa’s pathway circulating-microRNAs precursor lesions of gastric cancer gastric cancer sensitivity/specificity

Key summary

Established knowledge

The invasive nature of endoscopy limits its application as screening tool for the early detection of GC in areas of intermediate prevalence.

c-miRNAs have attracted attention in minimally invasive testing but there are inconsistencies in the literature regarding c-miRNA measurements.

c-miRNAs had not been evaluated in Correa's cascade to GC.

Using a strict QC protocol, the present study evaluated c-miRNAs in different stages of Correa's cascade to elucidate their diagnostic potential for the early detection of GC.

Significant findings

miR-196a-5p was deregulated in plasma of patients with precursor lesions of GC.

miR-134-5p, miR-144-3p and miR-451a were deregulated in PLGC.

miR-144-3p/miR-451a plasma levels were highly correlated. This pair is in 17q11.2, a site associated with GC.

c-miRNAs lacked the specificity/sensitivity to translate to the clinic.

Introduction

Gastric cancer (GC) is one of the most common cancers. Advances in diagnostic and treatment have increased survival for early-stage GC, but prognosis for those with advanced stages remains poor. Early stages are often asymptomatic and, currently, screening endoscopy in high-risk populations is the only way to diagnose GC. Its invasive nature limits its clinical application as screening tool in areas of low-to-intermediate GC prevalence. Therefore, the development of biomarkers has become a hot research topic.^1,2

Helicobacter pylori is the main aetiologic agent of GC.³ The infection causes a sequence of events known as Correa's cascade,³ a stepwise progression from non-active gastritis (NAG), chronic-active gastritis (CAG), precursor lesions of GC (PLGC) and GC. Circulating biomarkers would be helpful for monitoring PLGC and allow the early diagnosis of GC. In this sense, several biomarkers have been proposed but with moderate sensitivity/specificity.⁴

MicroRNAs (miRNAs) regulate gene expression and its expression is altered in cancers,⁵ acting as tumour suppressors or as oncogenes.⁵ Circulating-miRNAs (c-miRNAs) can be found in the blood and they have attracted attention as biomarkers.⁶ The discovery of c-miRNAs was first documented in a study demonstrating the existence of placental miRNAs in maternal plasma.⁷ Subsequent evidence has shown that serum/plasma contain large amounts of miRNAs derived from other origins, including cancer tissues,⁸ reflecting their potential as biomarkers.

Several studies have screened out hundreds of c-miRNAs in GC.⁹ However, there have been inconsistencies and discrepancies regarding the accuracy/reproducibility of c-miRNA measurements; in fact, none of them yet have a clinical use as a biomarker. Furthermore, few studies had evaluated tissue miRNA deregulation according to Correa's cascade and, to our knowledge, none of them has profiled c-miRNAs in plasma.

The present study evaluated c-miRNAs at different stages of progression to GC, to help to elucidate their diagnostic potential for early detection of GC in a European country with an intermediate H. pylori prevalence (50%) and an estimated GC incidence of 7.8 per 100,000 habitants.

Materials and methods

Patients and samples

Blood was collected from patients referred to endoscopy for evaluation of dyspepsia and from patients with GC undergoing preoperative endoscopic ultrasound (EUS). The study was approved by the Ethics Committee of the Corporació Sanitària Parc Taulí (Institut Universitari Parc Taulí, Sabadell, Spain) (code: 2005511; approval date: 2006/1/11) in accordance with the Declaration of Helsinki. All subjects provided signed informed consent.

This study enrolled 123 patients classified into: Non-active gastritis and H. pylori negative (NAG; n = 31), H. pylori-infected patients with CAG (n = 35), patients with PLGC (n = 31) and GC (n = 26). GC patients were at early TNM stages I/II (in 15 cases) and in advanced (III+IV) (in nine cases); in two patients TNM could not be determined.

Patients were randomly assigned to discovery and validation sets, except for GC patients, in which patients with early stages were selected for the discovery set (Figure 1). The discovery and validation cohorts were composed of 40 and 83 independent patients, respectively. Four cases were excluded from the final analysis after quality control (QC) checking. Additional data on diagnostic tests, sample collection and RNA isolation are in the Supplementary Material.

Figure 1.

Study diagram showing discovery and validation cohorts and the quality control steps used to determine circulating-microRNA (c-miRNA) levels.

MicroRNA discovery profiling

MiRNA discovery profiling was performed using Exiqon panel I, V4.M (Exiqon, Denmark). Each plate interrogates 372 human miRNAs plus control assays. Details are in the Supplementary Material.

MicroRNA validation

The miRNA levels of 11 selected miRNAs were analysed by quantitative real-time polymerase chain reaction (qRT-PCR) using TaqMan Advanced miRNA Assays (ThermoFisher), in accordance with the manufacturer's instructions. Details about the protocol used can be found in the Supplementary Material.

Data analysis and statistics

Discovery phase

Raw qRT-PCR data from the 38 discovery 384-plates were processed using QuantStudio (ThermoFisher) and raw data imported to ExpressionSuite (ThermoFisher). For all 384 miRNAs and controls, the software calculated the thresholds across all samples. A file containing the data was exported to work in an R statistical environment. As no consensus on the use of internal normalization control is defined for plasma miRNA quantitative polymerase chain reaction (qPCR) studies, we used geNorm¹⁰ to determine the reference miRNA in our cohort. miR-185-5p resulted in the most stable endogenous control identified by geNorm. MiRNAs that were not amplified at least in 35 of the 38 analysed samples were not considered. Significance was set at p < 0.05 with fold-changes >1.4. The Mann–Whitney U test was used without false discovery rate (FDR) correction.

Validation phase

Individual and mean polymerase chain reaction (PCR) efficiencies for amplicon, Cq and N₀ values were obtained using LinRegPCR software¹¹ from raw data exported in RDLM language from QuantStudio. Details about LinRegPCR and subsequent analysis are in the Supplementary Material.

Results

Patient demographic and clinical data are in Tables 1 and 2.

Table 1.

Demographic and clinical data of dyspeptic patients.

	Discovery cohort			Validation cohort
	NAG	CAG	PLGC	NAG	CAG	PLGC
n	10	8	10	21	24	21
Age (years)	48 ± 10.7	48 ± 8.2	53 ± 14.3	45 ± 11.4	50 ± 12.0	54 ± 14.0
Sex (male/female)	3/7	3/5	5/5	7/13	12/12	8/13
Histology
Gastritis (n (%))	9 (90.0%)	8 (100%)	9 (90.0%)	18 (90.0%)	24 (100%)	21 (100%)
CAG (n (%))	0 (0%)	8 (100%)	6 (60.0%)	20 (0%)	23 (95.8%)	11 (52.4%)
Intestinal metaplasia and/or atrophy (n (%))	0 (0%)	0 (0%)	32 (100%)	0 (0%)	0 (0%)	21 (100%)
Endoscopic diagnosis
Normal	2	2	3	8	5	6
Hiatal hernia or oesophagitis	4	2	3	8	11	6
Antroduodenitis	2	0	0	3	3	2
Duodenal ulcus	0	2	1	0	3	2
Gastric ulcus	0	0	1	0	0	0
Other	2	2	1	2	2	5

CAG: chronic active gastritis; NAG: non-active gastritis; PLGC: precursor lesions of gastric cancer.

Table 2.

Demographic and clinical data of gastric cancer patients.

	Discovery cohort	Validation cohort
n	10	16
Age (years)	70 ± 13.5	66 ± 11.9
Sex (male/female)	5/5	9/7
Tumour localization
Antrum	5	6
Corpus	3	4
Cardias	2	3
Not known	-	3
Tumour type
Intestinal	5	2
Diffuse	3	8
Intestinal + diffuse	1	-
Undifferentiated adenocarcinoma	-	2
Medullar	1	-
Not known	-	4
Histologic grade
Grade 1	2	1
Grade 2	4	1
Grade 3	2	3
Grade 4	1	6
Not known	1	5
Perineural invasion (yes/no/not known)	(4/6/-)	(7/3/6)
TNM stage
I	7	1
II	3	4
III	-	2
IV	-	7
Not known	-	2

Discovery phase

The QC protocol used indicated that the groups of patients had equivalent technical variability (Supplementary Material Results, Supplementary Material Figure 1). Differences in c-miRNA expression in the studied cohort are therefore attributable to biological variability.

In the discovery phase, real-time PCR was performed to identify differential expression patterns of miRNAs in plasma between the four patient groups (NAG, CAG, PLGC and GC). Of all 372 miRNAs included in the panel, 167 (45%) were detected in all 38 samples and 209 (56%) in more than 35. Twenty-seven miRNAs were not detected in any plasma sample (Supplementary Material Figure 2(a)). A total of 142 miRNAs (42%) were detected at mean Cq values greater than 35, which were mainly detected only in a few samples and were therefore not considered in the final analysis. In summary, 203 miRNAs (64%) were correctly quantified in the 25-35 Cq range, among which 18 (5%) had a Cq below 25 (Supplementary Material Figure 2(b)).

Regarding differential c-miRNA expression, the first objective was to determine the expression levels of miRNAs at the four stages of Correa's cascade. To do so, we used NAG patients as the control group and compared the normalised miRNA levels against CAG, PLGC and GC. The results are summarised in Table 3.

Table 3.

Fold changes of candidate circulating-microRNAs (c-miRNAs) as identified by the quantitative polymerase chain reaction (qPCR) panel in the discovery phase.

Group comparison	c-miRNA ID	Fold change	p Value^a
NAG vs CAG	miR-128-3p	–1.58	0.0152
	miR-135a-5p	–2.54	0.0499
	miR-182-5p	–1.61	0.0242
	miR-196b-5p	–1.69	0.0143
NAG vs GC	miR-144-3p	–1.91	0.0024
	miR-205-5p	–2.48	0.0190
	miR-451a	–1.75	0.0005
	miR-500a-5p	–1.84	0.0195
	miR-501-5p	–1.74	0.0070
NAG vs PLGC	miR-128-3p	–1.46	0.0515
	miR-141-3p	–1.94	0.0071
	miR-143-3p	–2.00	0.0101
	miR-16-1-3p	–1.61	0.0054
	miR-182-5p	–1.44	0.0314
	miR-186-5p	–1.49	0.0259
	miR-196b-5p	–1.80	0.0237
	miR-598-3p	–1.46	0.0436
CAG vs PLGC	-	-	-
CAG vs GC	miR-205-5p	–3.18	0.0205
PLGC vs GC	miR-130b-3p	1.77	0.0288
	miR-134-5p	1.93	0.0360
	miR-140-5p	1.92	0.0115
	miR-204-5p	2.22	0.0288
	miR-21-5p	1.45	0.0433
	miR-376b-3p	2.70	0.0435
	miR-409-3p	2.79	0.0433
	miR-451a	–1.60	0.0039
	miR-877-5p	2.41	0.0133
Non-cancer (NAG+CAG+PLGC) vs GC	miR-140-5p	1.44	0.0282
	miR-144-3p	–1.54	0.0212
	miR-205-5p	–2.78	0.0118
	miR-215-5p	–1.45	0.0436
	miR-219a-5p	1.64	0.0282
	miR-451a	–1.63	0.0007
	miR-501-5p	–1.60	0.0224

CAG: chronic active gastritis; GC: gastric cancer; NAG: non-active gastritis; PLGC: precursor lesions of gastric cancer.

Mann–Whitney U test.

Compared to NAG, CAG patients showed four down-regulated miRNAs (miR-128-3p, miR-135a-5p, miR-182-5p and miR-196b-5p). Compared to the same control group, PLGC patients showed the down-regulation of 12 miRNAs (miR-16-1-3p, miR-128-3p, miR-141-3p, miR-143-3p, miR-182-5p, miR-186-5p and miR-196b-5p). Finally, five miRNAs (miR-144-3p, miR-205-5p, miR-451a, miR-500a-5p and miR-501-5p) were down-regulated in plasma samples of GC.

The second objective was to compare the miRNA profile between PLGC and GC patients. Expression levels of six miRNAs were up-regulated (miR-130b-3p, miR-140-5p, miR-204-5p, miR-21-5p, miR-376b-3p, miR-409-3p and miR-877-5p) and one miRNA down-regulated (miR-451a).

The final objective was to obtain the c-miRNA profile of GC patients when using all non-cancer patients (NAG+CAG+PLGC) as the control group. Five miRNAs were down-regulated (miR-144-3p, miR-205-5p, miR-215-5p, miR-219a-5p, miR-451a, miR-501-5p) and two up-regulated (miR-140-5p, miR-219a-5p).

Validation phase

The 11 c-miRNAs selected for validation were: miR-134-5p, miR-141-3p, miR-144-3p, miR-182-5p, miR-196b-5p, miR-204-5p, miR-205-5p, miR-215-5p, miR-451a, miR-501-5p and miR-877-5p. Cel-miR-39 was used as an internal QC and c-miRNAs were normalised to endogenous housekeeping miR-185-5p.

miR-141-3p and miR-204-5p showed PCR efficiencies below 1.5 and were excluded from further analysis. miR-185-5p, miR-144-3p and miR-451a were the miRNAs with the lowest Cq values and better LinRegPCR quality checks and LinRegPCR sample use. miR-877-5p, miR-134-5p, miR-451a, miR-205-5p, miR-196b-5p, miR-182-5p, and miR-215-5p had baseline errors that were corrected manually (Supplementary Material Tables 3 and 4).

To validate the c-miRNAs identified using the Exiqon panel we used a separate, independent cohort of 83 patients composed of NAG (n = 21), CAG (n = 25), PLGC (n = 21) and GC patients (n = 16) (Figure 1). Two samples did not comply with the QC protocol and were excluded from the analysis.

We found that the expression levels of miR-144-3p and miR-451a were significantly down-regulated in plasma of GC patients. This was independent of whether the control group consisted of NAG patients (miR-144-3p: p = 0.046, one-way analysis of variance (ANOVA), Figure 2(a); miR-451a: p = 0.033, one-way ANOVA, Figure 2(b)) or all non-cancer patients (miR-144-3p: p = 0.033, t-Test, Figure 2(d); miR-451a: p = 0.042, t-Test, Figure 2(e)). When we used the noncancer group as the control group, a significant increase of miR-134 was also validated (miR-134-3p: p = 0.046, t-Test, Figure 2(f)). miR-196a was also down-regulated in PLGC with respect to NAG (p = 0.045, one-way ANOVA, Figure 2(c)). Five c-miRNAs (miR-182-5p, miR-205-5p, miR-215-5p, miR-501-5p and miR-877-5p) did not reach statistical significance in the validation set.

Figure 2.

Circulating-microRNA (c-miRNA) expression validation of miR-144-3p ((a) and (d)), miR-451a ((b) and (e)), miR-196b-3p (c) and miR-134-5p (f) in key steps of Correa’s cascade ((a)–(c)) and in the pool of non-gastric cancer patients ((d)–(f)).

Next we evaluated the correlations in the expression of c-miRNAs. miR-144-3p and miR-451a were highly correlated.

Figure 3.

Correlation analysis between circulating microRNAs, miR-144-3p and miR-451a.

Finally, we determined the diagnostic accuracy of the validated c-miRNAs for diagnosing GC. miR-144-3p and miR-451a showed low sensitivity and moderate specificity (Table 4). The combination of the two miRNAs did not improve diagnostic accuracy. When the non-cancer group was set as the reference group, the sensitivity and specificity of miR-144-3p were low (62% and 69%, respectively). miR-451a showed moderate sensitivity (87%) with low specificity (44%), whereas sensitivity values of miR-134 were low (37%) but with acceptable specificity (91%) (Table 4). Receiver-operating characteristic (ROC) curves are shown in Supplementary Material Figure 3.

Table 4.

Sensitivity, specificity, positive and negative predictive values of circulating-microRNAs (c-miRNAs) for the diagnosis of gastric cancer using non-active gastritis (NAG) or non-cancer patients (NAG+chronic active gastritis (CAG)+precursor lesions of gastric cancer (PLGC)) as reference groups.

Reference group	c-miRNA	Cut-off	AUC	Sensitivity* (95% CI)	Specificity* (95% CI)	PPV	NPV	LR+	LR–
NAG	miR-144-3p	<71.1	0.7688 p < 0.05	62.5 (35–85)	90.0 (68–99)	83.3	75.0	6.3	0.42
	miR-451a	<47.9	0.7656 p < 0.05	62.5 (35–85)	80.0 (56–94)	71.4	72.7	3.1	0.47
	miR-144-3p +miR-451a	>0.53	0.7440 p < 0.05	43.7 (19–70)	85.7 (64–97)	70.0	66.7	3.1	0.66
Non-cancer patients	miR-144-3p	<71.1	0.6548 p < 0.05	62.5 (35–84)	69.2 (56–80)	33.3	88.2	2.0	0.5
	miR-451a	<60.4	0.6596 p < 0.05	87.5 (62–98)	44.6 (32–57)	28.0	93.5	1.6	0.3
	miR-134-5p	>–3.9	0.6644 p < 0.05	37.5 (15–65)	90.7 (81–97)	50.0	85.5	4.1	0.7

AUC: area under the curve; CI: confidence interval; LR+; positive likelihood ratio; LR–: negative likelihood ratio; NPV: negative predictive value; PPV: positive predictive value.

*Values are percentages and 95% CIs.

Discussion

Increasingly, reports indicate that circulating miRNAs have the potential to be biomarkers in patients with different diseases, including digestive cancers.¹² In the present study, we have assessed the use of c-miRNAs as potential non-invasive biomarkers for the diagnosis of patients with PLGC and GC. Two cohorts of patients with NAG, CAG, PLGC and GC were studied. Three hundred and seventy-two miRNAs were screened in the plasma of the discovery cohort by qPCR and validation was performed in an independent cohort using a different qPCR platform. Quantification was technically controlled by the addition of RNA spike-ins to key steps of c-miRNA quantification, and individual qPCR amplifications were checked for quality parameters. Using this strategy, we have found and validated that miR-196a-5p is down-regulated in PLGC, miR-134-5p is up-regulated in GC patients, and miR-144-3p and miR-451a are highly correlated and down-regulated in GC. However, the magnitude of change in expression levels was small, with overlapping expression patterns between groups, leading to low diagnostic accuracy.

Early GC diagnosis remains an important clinical problem. These patients have no obvious symptoms and dyspeptic symptoms correlate poorly with histological findings.¹³ Dyspeptic symptoms can only be associated with H. pylori gastritis if successful eradication is followed by remission of symptoms.¹³ CAG is a common pathological event that occurs after H. pylori infection; after years of sustained inflammation, atrophic gastritis and intestinal metaplasia may appear. A low proportion of infected individuals may finally develop GC. Gastric atrophy and intestinal metaplasia are considered as PLGC. There is no general agreement regarding whether patients with PLGC may benefit from endoscopy surveillance since the risk of progression from this step to GC is low. The risk increases when atrophy and/or metaplasia extend to the corpus. In this case, endoscopic surveillance should be offered to the patient, although it is not cost effective.¹⁴ The evaluation of such lesions is made by extensive biopsy sampling and OLGA (Operative Link for Gastritis Assessment) / OLGIM (Operative Link on Gastric Intestinal Metaplasia Assessment) classification.² Non-invasive methods based on pepsinogen levels had moderate diagnostic accuracy for detecting PLGC and low accuracy for GC.¹ Our goal was to fill the gap of minimally invasive markers for surveillance from PLGC to GC by measuring c-miRNAs.

We have validated four miRNAs. Both discovery and validation cohorts showed the down-regulation of miR-144-3p and miR-451a. This pair forms a cluster in 17q11.2 and are separated from each other by 100 bp. Interestingly, this region had been reported to be involved in recurrent chromosomal rearrangements associated with GC¹⁵ and it contains oncogenes.¹⁶

In agreement with our results in plasma, miR-451a is down-regulated in GC tissue.^17,18 A study shows that patients with down-regulated miRNA-451a in the tumour have poorer survival,^18,19 but increased circulating levels of miR-451a have also been described by the same authors.²⁰ miR-144-3p is a tumour suppressor gene aberrantly expressed in GC.²¹ miR-144-3p is significantly decreased in GC tissue and diminished miR-144 expression is associated with a poor prognosis.²² Its loss promotes proliferation, epithelial-to-mesenchymal transition and it has been associated with metastasis.²¹ MET proto-oncogene (MET) and Pre-B-Cell Leukemia Transcription Factor 3 (PBX3) are the target of miR-144-3p in GC cells.^21,23 Overexpression of MET is correlated with tumour invasion and prognosis in GC.²⁴ PBX3 was reported to play an oncogenic role in human cancers.²⁵ Therefore, it has been suggested that miR-144 may inhibit GC tumourigenesis by targeting MET and subsequently the PIK3/Akt pathway.^16,26

Here, we have found that circulating miR-144-3p/miR-451a expression is diminished in GC patients when compared to a control group. The control group is a representation of Correa’s cascade towards GC, with patients with NAG, CAG and PLGC. Interestingly, plasma miR-144-3p/miR-451a are highly correlated and correlation is maintained when the dataset is split into NAG, CAG, PLGC and GC, supporting the concept that they are coexpressed because they belong to the same cluster.

To our knowledge, this is the first report to show that circulating miR-134-5p is deregulated in GC. We have found that miR-134-5p is up-regulated in GC patients when compared to the non-GC group. However, miR-134 has been found to be down-regulated^27,28 in GC tissues and has been proposed to act as a tumour suppressor. The overexpression of miR-134 in gastric cell lines reduces proliferation by targeting Kirsten ras oncogene (KRAS)²⁸ and Golgi Phosphoprotein 3 (GOLPH3).²⁷ The inverse relationship between tissue and circulating levels has been found in certain miRNAs acting as tumour suppressors, though the origin of this discrepancy is unknown.²⁶

Although our findings support the deregulation of miR-144-3p, miR-451a and miR-134, their use in the clinical setting is limited by the moderate area under the curve (AUC). This is caused by the overlapping expression pattern between GC and non-GC.

The last c-miRNA validated is miR-196b-5p. miR-196b-5p is down-regulated in PLGC compared with NAG patients. Increased c-miR-196a/b has been previously associated to GC and its levels decrease after surgery; the combined miR-196a/b sensitivity and specificity was 71% and 98%.²⁹ In the present study, miR-196b was not significantly deregulated in GC patients. Multiple factors may be involved in this divergence. First, in our discovery set, the mean Cq for miR-196b-5p was rather high and detected in only 21% plasmas. Second, the normalization method was different. Tsai et al.²⁹ use miR-16, a miRNA associated with GC,^19,20,30,31 although others have also used it.^32–34

Our study has strengths. The first is that few studies have analysed c-miRNAs in European populations, the vast majority of studies being conducted in Asian populations where GC is more prevalent. This study also includes two independent cohorts, with patients from each stage of Correa’s cascade. Another strength concerns the methodology used. Sample processing was controlled by the use of spike-ins and the QC of data by determining PCR efficiency for each amplicon. The use of spike-ins technically controls the process of miRNA quantification, but it cannot correct other sources of variability such as the total concentration of the miRNA fraction that likely changes inter-individually. As a consequence, external controls are not useful for normalization and endogenous controls are required. Some authors use plasma volume³⁵ and others use referenced c-miRNAs.³⁶ As there is no consensus on how to normalise, we used the geNorm algorithm in our discovery cohort and applied the resulting stable c-miRNA in the validation set. This study also has some drawbacks. GC patients’ age may be a confounding variable. As the cascade from CAG to GC may take several years, some c-miRNAs may be affected by age. Using mirnacon,³⁶ a tool that tests miRNAs for their likelihood of being influenced by age, mir-144-3p levels (but not miR-185-5p/miR-451a) may be influenced by age.

In conclusion, c-miRNAs are altered with the progression from H. pylori infection to PLGC and GC, suggesting four c-miRNAs as potential biomarkers for PLGC and GC. However, higher sensitivity/specificity should be achieved for translation to clinical applications. Several points may help achieve this goal such as the analysis of miRNA isoforms, the study of extracellular vesicles containing miRNAs, or the use of innovative and sensitive methodologies with less steps.^37,38

Supplemental Material

Supplemental material for Expression profile of circulating microRNAs in the Correa pathway of progression to gastric cancer

Supplemental material for Expression profile of circulating microRNAs in the Correa pathway of progression to gastric cancer by Sergio Lario, Anna Brunet-Vega, María E Quílez, María J Ramírez-Lázaro, Juan J Lozano, Lorena García-Martínez, Carles Pericay, Mireia Miquel, Félix Junquera, Rafael Campo and Xavier Calvet in United European Gastroenterology Journal

Footnotes

Acknowledgments

The authors are indebted to N Piqué for her help with English in this article.

Declaration of conflicting interests

The authors declare that there is no conflict of interest.

Funding

This study was funded by Fundació la Marató de TV3 (1007/C/2013) and Fondo de Investigación en Salud (FIS), Instituto de Salud Carlos III, ISCiii (PI12/01802).

Ethics approval

The study was approved by the Ethics Committee (code: 2005511; approval date: 2006/1/11) in accordance with the Declaration of Helsinki.

Informed consent

All subjects provided signed informed consent.

References

Leja

You

Camargo

et al.

Implementation of gastric cancer screening – the global experience. Best Pract Res Clin Gastroenterol 2014; 28: 1093–1106.

Rugge

Capelle

Fassan

. Individual risk stratification of gastric cancer: Evolving concepts and their impact on clinical practice. Best Pract Res Clin Gastroenterol 2014; 28: 1043–1053.

Moss

. The clinical evidence linking helicobacter pylori to gastric cancer. Cell Mol Gastroenterol Hepatol 2016; 3: 183–191.

Sturgeon

Duffy

Hofmann

et al.

National Academy of Clinical Biochemistry laboratory medicine practice guidelines for use of tumor markers in liver, bladder, cervical, and gastric cancers. Clin Chem 2010; 56: e1–48.

Farazi

Hoell

Morozov

et al.

MicroRNAs in human cancer. Adv Exp Med Biol 2013; 774: 1–20.

Lin

Kong

et al.

Current state of circulating microRNAs as cancer biomarkers. Clin Chem 2015; 61: 1138–1155.

Chim

Shing

Hung

et al.

Detection and characterization of placental microRNAs in maternal plasma. Clin Chem 2008; 54: 482–90.

Mitchell

Parkin

Kroh

et al.

Circulating microRNAs as stable blood-based markers for cancer detection. Proc Natl Acad Sci U A 2008; 105: 10513–8.

Tsujiura

Ichikawa

Komatsu

et al.

Circulating microRNAs in plasma of patients with gastric cancers. Br J Cancer 2010; 102: 1174–1179.

10.

Perkins

Dawes

McMahon

et al.

ReadqPCR and NormqPCR: R packages for the reading, quality checking and normalisation of RT-qPCR quantification cycle (Cq) data. BMC Genomics 2012; 13: 296–296.

11.

Ramakers

Ruijter

Deprez

RHL

et al.

Assumption-free analysis of quantitative real-time polymerase chain reaction (PCR) data. Neurosci Lett 2003; 339: 62–66.

12.

Shigeyasu

Toden

Zumwalt

et al.

Emerging role of microRNAs as liquid biopsy biomarkers in gastrointestinal cancers. Clin Cancer Res Off J Am Assoc Cancer Res 2017; 23: 2391–2399.

13.

Sugano

Tack

Kuipers

et al.

Kyoto global consensus report on Helicobacter pylori gastritis. Gut 2015; 64: 1353–1367.

14.

Areia

Dinis-Ribeiro

Rocha Gonçalves

. Cost-utility analysis of endoscopic surveillance of patients with gastric premalignant conditions. Helicobacter 2014; 19: 425–436.

15.

Yamashita

Nishida

Okuda

et al.

Recurrent chromosomal rearrangements at bands 8q24 and 11q13 in gastric cancer as detected by multicolor spectral karyotyping. World J Gastroenterol 2005; 11: 5129–5135.

16.

Grinchuk

Motakis

Kuznetsov

. Complex sense-antisense architecture of TNFAIP1/POLDIP2 on 17q11.2 represents a novel transcriptional structural-functional gene module involved in breast cancer progression. BMC Genomics 2010; 11: S9–S9.

17.

Chen

Saad

Jia

et al.

Gastric adenocarcinoma has a unique microRNA signature not present in esophageal adenocarcinoma. Cancer 2013; 119: 1985–1993.

18.

Zhao

Rong

et al.

MiR-451, a potential prognostic biomarker and tumor suppressor for gastric cancer. Int J Clin Exp Pathol 2015; 8: 9154–9160.

19.

Ren

Chen

Han

et al.

High expression of miR-16 and miR-451 predicating better prognosis in patients with gastric cancer. J Cancer Res Clin Oncol 2016; 142: 2489–2496.

20.

Zhu

Ren

Han

et al.

A five-microRNA panel in plasma was identified as potential biomarker for early detection of gastric cancer. Br J Cancer 2014; 110: 2291–2299.

21.

Zhang

Shen

et al.

MicroRNA-144-3p suppresses gastric cancer progression by inhibiting epithelial-to-mesenchymal transition through targeting PBX3. Biochem Biophys Res Commun 2017; 484: 241–247.

22.

Akiyoshi

Fukagawa

Ueo

et al.

Clinical significance of miR-144-ZFX axis in disseminated tumour cells in bone marrow in gastric cancer cases. Br J Cancer 2012; 107: 1345–1353.

23.

Liu

Xue

Zhang

et al.

MicroRNA-144 inhibits the metastasis of gastric cancer by targeting MET expression. J Exp Clin Cancer Res CR 2015; 34: 35–35.

24.

Wang

Shao

et al.

MET amplification is not rare and predicts unfavorable clinical outcomes in patients with recurrent/metastatic gastric cancer after chemotherapy. Cancer 2014; 120: 675–682.

25.

Wang

et al.

PBX3 promotes gastric cancer invasion and metastasis by inducing epithelial-mesenchymal transition. Oncol Lett 2016; 12: 3485–3491.

26.

Wang

Zhang

K-Y

Liu

S-M

et al.

Tumor-associated circulating microRNAs as biomarkers of cancer. Mol Basel Switz 2014; 19: 1912–1938.

27.

Liu

Sun

Zhao

. MicroRNA-134 suppresses cell proliferation in gastric cancer cells via targeting of GOLPH3. Oncol Rep 2017; 37: 2441–2448.

28.

Qiu

Z-A

G-P

. MicroRNA-134 functions as a tumor suppressor gene in gastric cancer. Am J Transl Res 2016; 8: 4320–4328.

29.

Tsai

M-M

Wang

C-S

Tsai

C-Y

et al.

Circulating microRNA-196a/b are novel biomarkers associated with metastatic gastric cancer. Eur J Cancer Oxf Engl 1990 2016; 64: 137–148.

30.

Zhang

Song

Zhang

et al.

Circulating miR-16-5p and miR-19b-3p as two novel potential biomarkers to indicate progression of gastric cancer. Theranostics 2015; 5: 733–745.

31.

Wang

et al.

Three dysregulated microRNAs in serum as novel biomarkers for gastric cancer screening. Med Oncol Northwood Lond Engl 2014; 31: 298–298.

32.

Shiotani

Murao

Kimura

et al.

Identification of serum miRNAs as novel non-invasive biomarkers for detection of high risk for early gastric cancer. Br J Cancer 2013; 109: 2323–2330.

33.

Peng

W-Z

Wang

et al.

Role of miR-191/425 cluster in tumorigenesis and diagnosis of gastric cancer. Int J Mol Sci 2014; 15: 4031–4048.

34.

Song

Bai

Han

et al.

Identification of suitable reference genes for qPCR analysis of serum microRNA in gastric cancer patients. Dig Dis Sci 2012; 57: 897–904.

35.

Zhang

Wang

Chen

et al.

Expression profile of microRNAs in serum: A fingerprint for esophageal squamous cell carcinoma. Clin Chem 2010; 56: 1871–1879.

36.

Meder

Backes

Haas

et al.

Influence of the confounding factors age and sex on microRNA profiles from peripheral blood. Clin Chem 2014; 60: 1200–1208.

37.

Keller

Meese

. Can circulating miRNAs live up to the promise of being minimal invasive biomarkers in clinical settings? Wiley Interdiscip Rev RNA 2016; 7: 148–156.

38.

Backes

Meese

Keller

. Specific miRNA disease biomarkers in blood, serum and plasma: Challenges and prospects. Mol Diagn Ther 2016; 20: 509–518.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.24 MB