The specimens in this study were obtained from 41 patients who were diagnosed in accordance with the National Comprehensive Cancer Network (NCCN) clinical practice guidelines for MM at the 1st Affiliated Hospital of Nanchang University from October 2015 to July 2017 (12). Among the 41 plasma specimens from MM patients, 6 were used for microarray analysis, and 35 served as the validation set. All patients who receiving chemotherapy and/or biotherapy were excluded, and patients with other types of malignant tumors were also eliminated. From healthy controls, we used 20 plasma samples in the validation set, who range in age from 17 to 63. 20 healthy individuals were recruited to the 1st Affiliated Hospital of Nanchang University between October 2015 to July 2017, including 8 males and 12 females. Venous blood was collected into tubes with EDTA (BD Biosciences, Franklin Lakes, NJ, USA). Plasma was transferred to a sterile tube and then immediately stored at −80°C after being frozen in liquid nitrogen. The study was approved by the Medical Research Ethics Committee of The First Affiliated Hospital of Nanchang University (Nanchang, China), and written informed consent was obtained from all study subjects. Additionally, FISH analysis was performed in 35 patients. Purified MM patient PCs (~10⁵ cells) were classified using the FISH technique. The following probe sets obtained from Vysis (Abbott Diagnostics, Berkshire, UK) were used for classification. D13S319 Spectrum-Orange/LSI-13q34-Spectrum-Green (del(13q) probe), IgH Spectrum-Green/FGFR3 Spectrum-Orange (t(4;14) fusion probe), IgH Spectrum-Green/CCND1 Spectrum-Orange (t(11;14) fusion probe), IgH Spectrum-Green/MAF Spectrum-Orange (t(14;16) fusion probe), and IgH dualcolor (Sperctrum-Green/Spectrum-Orange) break-apart probe. At least 100 nuclei were scored for each probe. 35 MM patients were cytogenetically classified using FISH, then we investigate the miR-125b-5p expression in the different genetic subtypes of MM patients. The 35 MM patient samples were from the original cohort of 41 patients with MM that were recruited to the study. The patients details are described in Table I.

Total RNA was extracted from samples from 6 MM patients and 6 healthy controls using phenol-chloroform (TRIzol; Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA). The quality of RNA was assessed by capillary electrophoresis (Agilent Technologies, Inc., Santa Clara, CA, USA). Libraries for small RNA sequencing were prepared using NEB kits (New England Biolabs, Inc., Ipswich, MA, USA). Libraries were quantified using Agilent Bioanalyzer 2100 system by DNA high sensitivity chips. Clean reads 21–22 nucleotides in length were screened for miRNAs and mapped onto a reference genome with Bowtie. The functions of new miRNAs were analyzed by miRDeep2 software. DE sequence was used to quantify differentially expressed miRNAs and to assess statistical significance.

TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc.) was used to extract total RNA. cDNA was reverse transcribed with the M-MLV Reverse Transcriptase kit (GeneCopoeia, Rockville, MD, USA) following the manufacturer's protocol. Approximately 1,200 µl plasma was employed for miRNA isolation. We employed quantitative reverse transcription PCR (qRT-PCR) with SYBR-Green (Takara, Osaka, Japan) to detect miR-125b-5p, miR-483-3p, miR-4326, miR-6894-3p, miR-4498, miR-490-3p, miR-7155-5p, and miR-937-3p expression levels, U6 was used as an internal control. The reaction was performed via 40 amplification cycles using the following protocol: 95°C for 3 min, 95°C for 45 sec, 55°C for 15 sec, and 72°C for 50 sec. There was no significant difference in the U6 Ct value between control and MM samples. The ΔCt=average Ct (miRNA of target)-average Ct (U6). The primers used in PCR were as follows: miR-125b-5p forward, 5′-TGCGCTCCCTGAGACCCTAAC-3′ and reverse, 5′-CCAGTGCAGGGTCCGAGGTATT-3′; miR-483-3p forward, 5′-TGCGCTCACTCCTCTCCTCCC-3′ and reverse, 5′-CTCAAGTGTCGTGGAGTCGGCAA-3′; miR-4326 forward, 5′-GCCCGCTGTTCCTCTGTCTCCC-3′ and reverse, 5′-GTGCAGGGTCCGAGGT-3′; miR-6894-3p forward, 5′-TGCGCTTGCCTGCCCTCTTCC-3′ and reverse, 5′-CTCAAGTGTCGTGGAGTCGGCAA-3′; miR-4498 forward, 5′-TGCGCTGGGCTGGTAGGGCAAG-3′ and reverse, 5′-CTCAAGTGTCGTGGAGTCGGCAA-3′; miR-490-3p forward, 5′-TGCGCCAACCTGGAGGATCCA-3′ and reverse, 5′-CTCAAGTGTCGTGGAGTCGGCAA-3′; miR-7155-5p forward, 5′-TGCGCTCTGGGGTCTTGGG-3′ and reverse, 5′-CTCAAGTGTCGTGGAGTCGGCAA-3′; miR-937-3p forward, 5′-TGCGCATCCGCGCTCTGACTCT-3′ and reverse, 5′-CTCAAGTGTCGTGGAGTCGGCAA-3′; and U6 forward, 5′-CTCGCTTCGGCAGCACA-3′ and reverse, 5′-AACGCTTCACGAATTTGCG-3′. Samples were analyzed in triplicate, and gene expression was quantified by normalizing target gene expression to that of the internal control using the 2-^ΔΔCt formula.

Data were analyzed using SPSS 19.0 software. All assays were repeated three times, and data were represented as the mean ± SD. Two-tailed t-tests were performed to compare the difference in plasma miRNA expression levels. Student's t-test was used to determine the significance of difference between two groups, and ROC curves were constructed to evaluate the prediction ability of the miRNAs for MM patients. EFS curves were constructed using the Kaplan-Meier method and compared using the log-rank test. P<0.05 was considered to indicate a statistically significant difference.

We carried out an miRNA microarray analysis to screen candidate plasma miRNAs in the treatment group (MM patients) and the control group. After data processing and analysis, we identified a set of 12 miRNAs that were differentially expressed between the treatment group (MM patients) and the control group, among which, 8 were significantly upregulated, and 4 were significantly downregulated (Fig. 1).

To further validate the differentially expressed miRNA identified by the miRNA microarray analysis, we used qRT-PCR to confirm the eight elevated miRNAs (miR-125b-5p, miR-483-3p, miR-4326, miR-6894-3p, miR-4498, miR-490-3p, miR-7155-5p, and miR-937-3p) in plasma samples from 35 MM patients and 20 healthy individuals. The results suggested that the plasma levels of five miRNAs, namely, miR-125b-5p, miR-4326, miR-4498, miR-490-3p and miR-7155-5p (Fig. 2), were significantly increased between the two groups. In addition, miR-125b-5p expression was associated with extramedullary infiltration and was markedly higher in stage III patients (1.76±0.03) than in stage I/II patients (1.62±0.04) (both P<0.05). However, miR-125b-5p expression was not correlated with patient age, sex or karyotype (all P>0.05) (Table II).

To compare the diagnostic value of miRNAs with that of typical tumor markers such as carcinoembryonic antigen (CEA) and cancer antigen 199 (CA199), we constructed ROC curves to validated plasma miRNAs in identifying MM patients. As shown in Fig. 3, miR-125b-5p and miR-490-3p were accurate in distinguishing MM patients from healthy controls. The area under the curve (AUC) for miR-125b-5p was 0.954, with 86% sensitivity and 96% specificity [95% confidence interval (CI) 0.901–1.006; P<0.001]; the AUC for miR-490-3p was 0.866, with 60% sensitivity and 85% specificity (95% CI 0.522–0.836; P=0.028).

We further assessed the correlation between miRNA expression and MM prognosis by Kaplan-Meier analysis. MM patients receive the chemotherapy of bortezomib, thalidomide plus dexamethasone (VTD) regimen, followed by thalidomide maintenance therapy. MM patients were divided into either high or low miR-125b-5p/miR-490-3p expression subgroups. Using this approach, we compared EFS between high- and low-expression subgroups. EFS durations were calculated from the time of diagnosis to the date of relapse or death. After a median follow-up 9.5 months (range, 1–26 months) from diagnosis, the median EFS duration in patients expressing high levels of miR-125b-5p was 8 months, and that in patients expressing low levels of miR-125b-5p was 13 months. Our results indicated that patients with higher miR-125b-5p expression had a shorter EFS than patients with lower miR-125b-5p expression (P=0.02, Fig. 4). However, there was no difference between high and low miR-490-3p expression groups, with median EFS of 12 and 13 months, respectively (P=0.23; Fig. 4).