Screening microRNAs as potential prognostic biomarkers for lung adenocarcinoma

Figures & data

Table 1. Basic clinical characteristics of initial sample group.

Basic information	Early recurrence and metastasis group	No recurrence and metastasis group	p value
	n = 40	n = 40
Age (year)			0.946^a
Mean ± SD	60.2 ± 11.0	60.0 ± 8.5
Gender			0.171^b
Male	19(47.5%)	13(32.5%)
Female	21(52.5%)	27(67.5%)
Smoking history			0.112^b
Yes	20(50.0%)	13(32.5%)
No	20(50.0%)	27(67.5%)
Surgical method			0.356^c
Lobes of lung	36(90.0%)	39(97.5%)
Wedge	4(10.0%)	1(2.5%)
Tumor location			0.700^d
Right upper	11(27.5%)	16(40.0%)
Right middle	3(7.5%)	3(7.5%)
Right bottom	13(32.5%)	8(20.0%)
Left upper	9(22.5%)	9(22.5%)
Left bottom	4(10.0%)	4(10.0%)
Degree of differentiation			0.267^e
Low + Moderately and low	10(25.0%)	6(15.0%)
Moderately + Moderately high + High	30(75.0%)	34(85.0%)
Pleural involvement			1.000^b
Yes	31(77.5%)	31(77.5%)
No	9(22.5%)	9(22.5%)
Postoperative chemotherapy			0.736^b
Yes	18(46.2%)	19(50.0%)
No	21(53.8%)	19(50.9%)
Pathologic subtype			0.044^d
Mural predominance	3(5.1%)	6(15.0%)
Papillary predominant type	4(10.0%)	10(25.0%)
Acinar predominant type	24(60.0%)	11(27.5%)
Solid predominant type	5(12.5%)	5(12.5%)
Micropapillary predominance	0(0.0%)	0(0.0%)
Other	4(10.0%)	8(20.0%)

Note: a: unpaired T-test; b: Pearson x2 test; c: continuity corrected Chi-square test; d: Fisher exact test; e: Wilcoxon rank sum test (Mann-Whitney U test), SD (standerd deviation).

Table 2. Basic clinical characteristics of validation sample group.

Basic information	Early recurrence and metastasis group	No recurrence and metastasis group	p value
	n = 41	n = 39
Age (year)			0.251^a
Mean ± SD	60.8 ± 10.3	58.1 ± 10.7
Gender			0.271^b
Male	20(51.3%)	16(39.0%)
Female	19(48.7%)	25(61.0%)
Smoking history			0.075^b
Yes	20(51.3%)	13(31.7%)
No	19(48.7%)	28(68.3%)
Surgical method			1^c
Lobes of lung	36(92.3%)	38(92.7%)
Wedge	3(7.7%)	3(7.3%)
Tumor location			0.241^d
Right upper	14(35.9%)	6(14.6%)
Right middle	2(5.1%)	2(4.9%)
Right bottom	7(17.9%)	13(31.7%)
Left upper	9(23.1%)	11(26.8%)
Left bottom	7(17.9%)	9(22.0%)
Degree of differentiation			0.258^e
Low + Moderately and low	8(20.5%)	13(31.7%)
Moderately + Moderately high + High	31(79.5%)	28(68.3%)
Pleural involvement			0.381^b
Yes	34(87.2%)	32(78.0%)
No	5(12.8%)	9(22.0%)
Postoperative chemotherapy			0.84^b
Yes	21(53.8%)	23(56.1%)
No	18(46.2%)	18(43.9%)
Pathologic subtype			0.436^d
Mural predominance	2(5.1%)	3(7.3%)
Papillary predominant type	6(15.4%)	11(26.8%)
Acinar predominant type	24(61.5%)	20(48.8%)
Solid predominant type	6(15.4%)	3(7.3%)
Micropapillary predominance	0(0.0%)	1(2.4%)
Other	1(2.6%)	3(7.3%)

Note: a: unpaired T-test; b: Pearson x2 test; c: continuity corrected Chi-square test; d: Fisher exact test; e: Wilcoxon rank sum test (Mann-Whitney U test), SD (standerd deviation).

Figure 1. miRNA microarray quality control images. A. Reflects correct grid selection. B: Reflects clean chip cleaning. C. Reflects chip signal distribution range. D. Reflects background noise of chip. E. Reflects fluorescence signal distribution. F. Quality Control Metrics-plot of miRNA microarray data. Indicators are as follows: Any color PrcntFeatPopnOL: percentage content of Feature escape value; gTotalSignal75pctiIe: 75% value of all gene signals; Add Error Estimate Green: additional error value; gNon Ctrl Med Prcnt CVBG Sub Sig: coefficient of variation value to remove background signal. G. Each point represents one case. Red: early recurrence metastasis group. Blue: no recurrence metastasis group.

Table 3. Differentially expressed miRNAs in the initial sample group (based on p value and Fold change).

Systematic name	p value	Fold change
has-miR-4453	0.028	1.87
has-miR-183-5p	0.014	1.78
has-miR-4685-5p	0.022	1.76
has-miR-3651	0.007	1.76
has-miR-1260b	0.013	1.72
has-miR-21-3p	0.013	1.71
has-miR-6511a-5p	0.038	1.66
has-miR-4514	0.032	1.61
has-miR-4659a-3p	0.001	1.57
has-miR-2467-3p	0.038	1.55
has-miR-3621	0.010	1.55
has-miR-1471	0.001	1.53
has-miR-3692-5p	0.047	1.53
has-miR-92a-3p	0.013	1.53

Figure 2. Relationship between expression levels of eight candidate differentially expressed miRNAs and progression-free survival (PFS) in the initial sample group. A. Relationship between miR-1260b expression and prognosis of patients, p = 0.001. B. Relationship between miR-21-3b expression and prognosis of patients, p = 0.02. C. Relationship between miR-4514 expression and prognosis of patients, p = 0.024. D. Relationship between miR-4659a-3p expression and prognosis of patients, p = 0.009. E. Relationship between miR-2467-3p expression and prognosis of patients, p < 0.001. F. Relationship between miR-3621 expression and prognosis of patients, p = 0.026. G. Relationship between miR-1471 expression and prognosis of patients, p = 0.013. H. Relationship between miR-92a-3p expression and prognosis of patients, p = 0.011. I. Heatmap of eight candidate miRNAs. The abscissa is the case and the ordinate is eight candidate miRNAs.

Table 4. Differentially expressed miRNAs in the initial sample group (based on p value, Fold change and patient survival analysis).

Systematic name	p value	Fold change
miR-1260b	0.013	1.72
miR-21-3p	0.013	1.71
miR-4514	0.032	1.61
miR-4659a-3p	0.001	1.57
miR-2467-3p	0.038	1.55
miR-3621	0.01	1.55
miR-1471	0.001	1.53
miR-92a-3p	0.013	1.53

Figure 3. Real-time PCR validation of candidate miRNAs in validation sample group. A. miR-2467-3p expression level in validation sample group, p > 0.05. B. miR-21-3p expression level in validation sample group, *: p = 0.02. C. miR-1260b expression level in validation sample group, p > 0.05. D. miR-92a-3p expression level in validation sample group, p > 0.05. E. miR-3621 expression level in validation sample group, p > 0.05.

Figure 4. The ROC curve for the early recurrence and metastasis group and the non-recurrence and metastasis group. A. The area under the ROC curve was 0.917, p < 0.001. B. Prognostic index in the initial sample group and progression-free survival in patients with lung adenocarcinoma, p < 0.001. C. Prognostic index in the validation sample group and progression-free survival in patients with lung adenocarcinoma, p < 0.001.

Table 5. Multivariate Cox regression analysis results for the initial sample group.

	B	SE	Wald	df	Sig.	Exp(B)
ZhsamiR1260b	.499	.280	3.163	1	.075	1.647
ZhsamiR24673p	−.853	.302	7.957	1	.005	.426
ZhsamiR92a3p	.641	.302	4.524	1	.033	1.899
Zgrade	−.496	.278	3.171	1	.075	.609

Note: B: standardized regression coefficient; SE: standard error; Wald: Wald test statistic; df: degrees of freedom; Sig.: p-value; Exp (B): relative risk.

Table 6. Chi-square test for paired design between PI prediction results of initial sample group and actual patient survival.

	Prognostic index (PI)			p value
	Poor prognosis (PI high)		Poor prognosis (PI low)
Actual Survival Results	Poor prognosis	19	4	0.549^a
	Good prognosis	7	21	<0.001^b

Note: ^aMcNemer test, p > 0.05, not statistically significant.

^bKappa agreement test, p < 0.05, statistically significant.

Table 7. Chi-square test for paired design between PI prediction results of validation sample group and actual patient survival.

	Prognostic index (PI)			p value
	Poor prognosis (PI high)		Poor prognosis (PI low)
Actual Survival Results	Poor prognosis	29	10	0.678^a
	Good prognosis	13	28	<0.001^b

Note: ^aMcNemer test, p > 0.05, not statistically significant.

^bKappa agreement test, p < 0.05, statistically significant.

Data availability statement

All data generated or analyzed during this study are included in this article.