High expression of CPNE5 and CPNE9 predicts positive prognosis in multiple myeloma

Abstract

BACKGROUND:

CPNEs are significant biomarkers which can affect the progression and prognosis of various tumor diseases. However, the prognosis role of CPNEs in multiple myeloma (MM) is still unclear.

OBJECTIVES:

To investigate the prognosis role of CPNEs in MM.

METHODS:

Seven hundred and thirty-five samples from two independent data sets were involved to analyze the clinical and molecular characteristics, and prognosis role of the expression of CPNE1-9 in MM.

RESULTS:

MM patients with higher expressions of CPNE5 and CPNE9 had longer event-free survival (EFS) and overall survival (OS) compared with CPNE5 ${}^{\mathrm{\text{low}}}$ and CPNE9 ${}^{\mathrm{\text{low}}}$ expression groups (EFS: $P=$ 0.0054, 0.0065; OS: $P=$ 0.015, 0.016, respectively). Multivariate regression analysis showed that CPNE5 was an independent favorable predictor for EFS and OS (EFS: $P=$ 0.005; OS: $P=$ 0.006), and CPNE9 was an independent positive indicator for EFS ( $P=$ 0.002). Moreover, the survival probability and the cumulative event of EFS and OS in CPNE5 ${}^{\mathrm{\text{high}}}$ CPNE9 ${}^{\mathrm{\text{high}}}$ group were significantly longer than other groups.

CONCLUSIONS:

High expressions of CPNE5 and CPNE9 might be used as positive indicators for MM, and their combination was a better predictor for the survival of MM patients.

1. Introduction

Multiple myeloma (MM) is a common malignant tumor disease of blood system, which is characterized by abnormal proliferation of clonal plasma cells. It is estimated that there were 176,404 new MM patients in 2020, accounting for 0.9% of all cancer patients, and 117,077 deaths of MM patients, accounting for 1.2% of all cancer deaths [1]. Although the prognosis of MM patients has improved with the improvement of MM diagnosis and treatment, only 45% of patients have an overall survival (OS) of 6 years [2]. It is currently believed that genetic changes, bone marrow microenvironmental effects and epigenetic modifications play an important role in the occurrence and development of MM [3, 4, 5]. Although genetic abnormalities have made great progress in the assessment of MM progression and prognosis, there is still an urgent need to find more effective prognostic biomarkers, which are of great clinical significance for optimizing the personalized treatment plan and improving the clinical efficacy of MM patients. It is also a current research hotspot.

CPNEs are highly conserved intracellular calcium-dependent lipid-binding proteins originally identified by Creutz et al. [6] from the extracts of paramecium tetraurelia, which play a fundamental role in eukaryotic cell biology. The N-terminal of CPNEs contains two C2 domains and a domain similar to the A domain, in which the C2 domain plays a role in membrane transport and the A-like domain mediates the interaction between integrins and extracellular ligands. The HUGO Gene Nomenclature Committee has identified nine homologues of CPNEs in humans, namely CPNE1-9. Jiang et al. [7] found that the down-regulation of CPNE1 can inhibit the proliferation and metastasis potential of osteosarcoma cells and reduce the expression of a variety of tumor-related genes. In non-small cell lung cancer (NSCLC), inhibition of CPNE1 expression can effectively reduce the proliferation and movement of cancer cells [8]. In addition, CPNE3 is also highly expressed in a variety of cancer cells and can regulate ErbB2 signaling pathway-mediated movement and migration of cancer cells [9, 10, 11]. Genetic variants in CPNE5 have been found to be associated with alcohol dependence and obesity [12]. As a protein specifically expressed in the brain, CPNE6 has a high degree of homology with human CPNE1, and is closely related to synaptic plasticity and memory formation [13, 14]. A recent study found that Ca ${}^{2+}$ -dependent CPNE6 can interact with synaptobrevin2, thereby selectively inhibiting spontaneous neurotransmission [15]. CPNE7 is mainly involved in the differentiation and formation of odontoblasts [16, 17]. The invasive acute myeloid leukemia with t(12;21) translocation may be closely related to the fusion of CPNE8 and AML1/Runx1 [18].

The CPNEs have been deeply researched in a variety of diseases, especially for the pathogenesis, treatment and prognosis of malignant tumors. However, the prognostic value of CPNEs in MM is still unclear. This study explored the asscociation between the expression levels of CPNEs and survival of MM patients.

2. Subjects and methods

2.1 Patients

In this study, two independent microarray data sets (GSE39754 and GSE24080) were selected to analyze prognosis role of the expression of CPNEs in MM. The GSE39754 data set used Affymetrix Human Exon 1.0 ST Array to analyze the gene expression profiles of 6 healthy donors and 170 newly diagnosed MM patients [19, 20, 21]. The GSE24080 data set collected high-purity bone marrow plasma cell gene expression profiles of 559 newly diagnosed MM patients, and used Affymetrix human genome U133 Plus 2.0 array to analyze the expression level of CPNEs in MM patients [22, 23, 24]. All information related to the patients’ clinical and molecular data, as well as microarray data sets comes from Gene Expression Omnibus (GEO). All experimental design, quality control and data standardization are in compliance with the standard Affymetrix protocol. The study was conducted in accordance with the “Declaration of Helsinki”, and all patients had obtained written informed consent.

Figure 1.

Survival analysis of CPNE2, CPNE5 and CPNE9 in GSE24080. (A) There wasn’t any significant difference in EFS between CPNE2 ${}^{\mathrm{\text{High}}}$ and CPNE2 ${}^{\mathrm{\text{Low}}}$ patients ( $P=$ 0.91). (B) There wasn’t any significant difference in EFS between CPNE2 ${}^{\mathrm{\text{High}}}$ and CPNE2 ${}^{\text{𝐿𝑜𝑤}}$ patients ( $P=$ 0.28). (C) CPNE5 ${}^{\mathrm{\text{High}}}$ patients had a significantly longer EFS ( $P=$ 0.0054) than CPNE5 ${}^{\mathrm{\text{Low}}}$ patients. (D) CPNE5 ${}^{\mathrm{\text{High}}}$ patients had a significantly longer OS ( $P=$ 0.015) than CPNE5 ${}^{\mathrm{\text{Low}}}$ patients. (E) CPNE9 ${}^{\mathrm{\text{High}}}$ patients had a significantly longer EFS ( $P=$ 0.0065) than CPNE5 ${}^{\mathrm{\text{Low}}}$ patients. (F) CPNE9 ${}^{\mathrm{\text{High}}}$ patients had a significantly longer OS ( $P=$ 0.016) than CPNE5 ${}^{\mathrm{\text{Low}}}$ patients.

Figure 2.

Survival analysis of combined expression of CPNE5 and CPNE9. (A) CPNE5 ${}^{\mathrm{\text{High}}}$ CPNE9 ${}^{\mathrm{\text{High}}}$ expression group had longer EFS of the survival probability than other groups ( $P=$ 0.0011). (B) CPNE5 ${}^{\mathrm{\text{High}}}$ CPNE9 ${}^{\mathrm{\text{High}}}$ expression group had longer OS of the survival probability than other groups ( $P=$ 0.0016). (C) CPNE5 ${}^{\mathrm{\text{High}}}$ CPNE9 ${}^{\mathrm{\text{High}}}$ expression group had lower EFS of the cumulative events than other groups ( $P=$ 0.0011). (D) CPNE5 ${}^{\mathrm{\text{High}}}$ CPNE9 ${}^{\mathrm{\text{High}}}$ expression group had lower OS of the cumulative events than other groups ( $P=$ 0.0016).

2.2 Statistical analysis

The event-free survival (EFS) and OS were used as the study endpoints. EFS refers to the time from the diagnosis of the disease to the first event such as relapse and death. OS refers to the time from the diagnosis of the disease to the final death of the patient. In GSE39754 ( $n=$ 176), Wilcoxon rank sum tests were used to analyze the expression differentiation of CPNE1-9 between healthy donors ( $n=$ 6) and MM patients ( $n=$ 170). In GSE24080 ( $n=$ 559), numerical data were compared using Wilcoxon rank sum test, and categorical data were compared using Fisher’s exact test, Kaplan-Meier method and log-rank test were used for survival analysis, univariate and multivariate Cox proportional hazards models of EFS and OS were established with a confidence interval of 95%. All statistical analyses were performed using r3.5.0.

3. Results

3.1 The expression of CPNE1-9

In this study, GSE39754 was used to explore the expression levels of CPNE1-9 in healthy donors and MM patients. And we found that the expression levels of CPNE1 and CPNE5 in the MM group were significantly higher than that in the healthy group (CPNE1: $P=$ 0.006; CPNE5: $P=$ 0.002), while the expression of CPNE4 in the MM group was significantly lower than that in the healthy group (CPNE4: $P=$ 0.016, Fig. S1).

Table 1

Comparison of EFS and OS between the high and low expression levels of CPNE family

	EFS		OS
	${\chi }^2$	$P$ -Value	${\chi }^2$	$P$ -Value
CPNE1 (High vs. Low)	0.093	0.761	0.237	0.627
CPNE2 (High vs. Low)	0.00	1.000	5.379	0.020
CPNE3 (High vs. Low)	0.188	0.665	1.353	0.245
CPNE4 (High vs. Low)	1.839	0.175	0.092	0.762
CPNE5 (High vs. Low)	5.447	0.020	5.379	0.020
CPNE6 (High vs. Low)	0.00	1.000	2.516	0.113
CPNE7 (High vs. Low)	0.454	0.501	0.00	1.000
CPNE8 (High vs. Low)	1.372	0.241	2.340	0.126
CPNE9 (High vs. Low)	8.337	0.004	7.213	0.007

EFS event-free survival, OS overall survival.

Table 2

Patients’ characteristics of 559 MM patients in GSE24080

	CPNE5			CPNE9
	Low ( $n=$ 279)	High ( $n=$ 280)	$P$ -value	Low ( $n=$ 279)	High ( $n=$ 280)	$P$ -value
AGE (mean (range))	57.10 (24.83–75.94)	57.26 (29.7–76.50)	0.480	57.21 (24.83–76.50)	57.15 (35.23–75.00)	0.680
Gender, no, (%)			0.247			0.904
Female	118 (42.29)	104 (37.14)		112 (40.14)	110 (39.29)
Male	161 (57.71)	176 (62.86)		167 (59.86)	170 (60.71)
Race, no, (%)			0.675			1.000
White	246 (88.17)	251 (89.64)		248 (88.89)	249 (88.93)
Other	33 (11.83)	29 (10.36)		31 (11.11)	31 (11.07)
ISS, no, (%)			0.155			0.559
I	148 (52.86)	170 (60.93)		165 (58.93)	153 (54.84)
II	68 (24.29)	54 (19.35)		60 (21.43)	62 (22.22)
III	64 (22.86)	55 (19.71)		55 (19.64)	64 (22.94)
Isotype, no, (%)			0.223			0.002
FLC	35 (12.5)	49 (17.56)		37 (13.21)	47 (16.85)
IgA	62 (22.14)	71 (25.45)		84 (30)	49 (17.56)
IgD	2 (0.71)	1 (0.36)		0 (0)	3 (1.08)
IgG	170 (60.71)	143 (51.25)		142 (50.71)	171 (61.29)
B2M (mean (sd))	4.701 (5.379)	4.776 (5.35)	0.870	5.204 (6.42)	4.275 (3.99)	0.041
CRP (mean (sd))	12.781 (25.172)	10.491 (20.40)	0.238	13.687 (27.46)	9.58 (17.04)	0.034
CREAT (mean (sd))	1.332 (1.296)	1.315 (1.25)	0.874	1.408 (1.44)	1.238 (1.076)	0.114
LDH (mean (sd))	181.509 (74.44)	162.479 (54.70)	0.001	177.849 (68.15)	166.125 (63.23)	0.035
ALB (mean (sd))	4.109 (0.533)	3.989 (0.62)	0.014	4.028 (0.64)	4.07 (0.52)	0.398
HGB (mean (sd))	11.325 (1.807)	11.181 (1.82)	0.349	11.165 (1.74)	11.341 (1.88)	0.252
MRI (mean (sd))	11.5 (14.186)	10.583 (13.91)	0.441	11.547 (14.49)	10.51 (13.59)	0.383
High CCND1, no.(%)	140 (50.18)	140 (50)	0.510	146 (52.33)	134 (47.86)	0.182
High LIG4, no. (%)	123 (44.09)	157 (56.07)	0.195	137 (49.1)	143 (51.07)	0.863
High TP53, no. (%)	132 (47.31)	148 (52.86)	0.564	139 (49.82)	141 (50.36)	0.938
High CDK4, no. (%)	150 (53.76)	130 (46.43)	0.023	138 (49.46)	142 (50.71)	0.421
High FGFR3, no. (%)	153 (54.84)	127 (45.36)	0.031	126 (45.16)	154 (55)	0.700
High CDK5, no. (%)	132 (47.31)	148 (52.86)	0.472	140 (50.18)	140 (50)	0.976
High HK2, no. (%)	157 (56.27)	123 (43.93)	0.059	142 (50.9)	138 (49.29)	0.018

AGE: age at registration (years); B2M: beta-2 microglobulin (mg/l); CRP: C-reactive protein (mg/l); CREAT: creatinine (mg/dl); LDH: lactate dehydrogenase (U/l); ALB: albumin (35 g/l); HGB: hemoglobin (g/dl); MRI: number of magnetic resonance imaging (MGI)-defined focal lesions (skull, spine, pelvis); no.: number of patients.

Table 3

Univariate and multivariate cox regression analyses of EFS and OS

	Univariate cox regression				Multivariate cox regression
	EFS		OS		EFS		OS
Characteristics	HR (95%CI)	$P$	HR (95%CI)	$P$	HR (95%CI)	$P$	HR (95%CI)	$P$
CPNE1 (High vs. Low)	1.00 (0.74–1.35)	0.991	0.98 (0.73–1.32)	0.901	1.09 (0.80–1.50)	0.582	1.05 (0.77–1.43)	0.753
CPNE2 (High vs. Low)	0.98 (0.73–1.33)	0.911	0.74 (0.55–1.00)	0.049	0.91 (0.66–1.25)	0.554	0.63 (0.46–0.87)	0.005
CPNE3 (High vs. Low)	1.03 (0.76–1.40)	0.845	1.13 (0.84–1.53)	0.415	0.96 (0.70–1.33)	0.811	1.11 (0.80–1.52)	0.536
CPNE4 (High vs. Low)	1.20 (0.88–1.62)	0.249	0.89 (0.66–1.20)	0.455	1.20 (0.87–1.66)	0.264	0.97 (0.71–1.33)	0.845
CPNE5 (High vs. Low)	0.65 (0.48–0.88)	0.006	0.71 (0.52–0.96)	0.025	0.63 (0.46–0.87)	0.005	0.63 (0.46–0.87)	0.006
CPNE6 (High vs. Low)	0.98 (0.72–1.32)	0.889	0.77 (0.57–1.04)	0.091	1.13 (0.81–1.58)	0.478	0.95 (0.68–1.33)	0.773
CPNE7 (High vs. Low)	0.97 (0.71–1.31)	0.829	1.04 (0.77–1.40)	0.817	0.99 (0.72–1.35)	0.927	1.10 (0.81–1.50)	0.549
CPNE8 (High vs. Low)	1.19 (0.88–1.62)	0.253	1.24 (0.92–1.67)	0.166	1.07 (0.77–1.49)	0.680	1.09 (0.79–1.51)	0.602
CPNE9 (High vs. Low)	0.65 (0.48–0.89)	0.007	0.72 (0.53–0.97)	0.031	0.61 (0.44–0.84)	0.002	0.74 (0.54–1.02)	0.065
AGE ( $⩾$ 60 vs. $<$ 60)	0.97 (0.71–1.32)	0.839	1.40 (1.04–1.89)	0.028	0.85 (0.61–1.17)	0.313	1.30 (0.95–1.77)	0.096
Gender	1.05 (0.77–1.43)	0.750	0.97 (0.72–1.32)	0.850	1.28 (0.92–1.78)	0.137	1.21 (0.88–1.66)	0.244
ALB	0.72 (0.53–0.97)	0.033	0.48 (0.35–0.64)	$<$ 0.001	0.74 (0.54–1.02)	0.067	0.48 (0.35–0.67)	$<$ 0.001
B2M	1.72 (1.27–2.33)	$<$ 0.001	2.21 (1.64–3.00)	$<$ 0.001	1.40 (0.97–2.00)	0.069	1.81 (1.27–2.57)	0.001
HGB	0.54 (0.39–0.74)	$<$ 0.001	0.62 (0.45–0.84)	0.002	0.70 (0.49–1.01)	0.059	1.09 (0.76–1.56)	0.654
LDH	2.58 (1.65–4.05)	$<$ 0.001	3.68 (2.53–5.37)	$<$ 0.001	2.27 (1.41–3.68)	0.001	2.90 (1.92–4.37)	$<$ 0.001
CCND1	0.65 (0.48–0.88)	0.006	0.74 (0.55–1.00)	0.053	0.78 (0.45–1.36)	0.386	1.14 (0.66–1.99)	0.636
CDK4	1.22 (0.90–1.66)	0.197	1.36 (1.01–1.84)	0.043	1.18 (0.84–1.65)	0.347	1.37 (0.98–1.93)	0.065
CDK5	0.88 (0.65–1.19)	0.396	1.09 (0.81–1.47)	0.572	0.76 (0.54–1.08)	0.127	0.84 (0.60–1.17)	0.304
FGFR3	0.90 (0.67–1.22)	0.508	0.80 (0.59–1.08)	0.150	0.82 (0.58–1.15)	0.249	0.87 (0.62–1.21)	0.396
LIG4	0.84 (0.62–1.14)	0.268	0.84 (0.63–1.14)	0.269	0.98 (0.71–1.37)	0.918	1.04 (0.75–1.45)	0.800
TP53	1.06 (0.78–1.44)	0.700	0.85 (0.63–1.15)	0.284	1.21 (0.87–1.68)	0.254	0.84 (0.61–1.17)	0.306

EFS: event-free survival; OS: overall survival; CR: complete remission; HR: hazard ratio; CI: confidence interval; ALB: Albumin, 35 g/l; B2M: Beta-2 microglobulin, mg/l; HGB: Haemoglobin, g/dl; LDH: Lactate dehydrogenase, U/l.

3.2 Survival analysis of CPNE2, CPNE5 and CPNE9

We analyzed the association between expression levels of CPNEs and survival of MM patients in GSE24080. According to the median expression level of each CPNE member, all patients were divided into high and low expression groups. EFS and OS were compared between the two groups in Table 1, the OS and EFS between high and low CPNE5/9 expression were significantly different (OS: $P=$ 0.020, 0.007; EFS: $P=$ 0.020, respectively). The OS of CPNE2 high expression group and low expression group was significantly different ( $P=$ 0.020). In the GSE24080 data set, the effects of CPNE2, CPNE5, and CPNE9 expression on the prognosis of MM patients were analyzed, and it was found that there was no significant difference between CPNE2 high expression group and low expression group in OS and EFS (OS: $P=$ 0.28; EFS: $P=$ 0.91). Patients with high expression of CPNE5 or CPNE9 had longer OS and EFS than those with CPNE5 ${}^{\mathrm{\text{low}}}$ or CPNE9 ${}^{\mathrm{\text{low}}}$ (OS: $P=$ 0.00554, 0.0065; EFS: $P=$ 0.015, 0.016, respectively, Fig. 1).

3.3 Clinical and molecular characteristics

The clinical and molecular characteristics were shown in Table 2. Compared with the CPNE5 ${}^{\mathrm{\text{low}}}$ , the lactate dehydrogenase (LDH) and albumin (ALB) in the CPNE5 ${}^{\mathrm{\text{high}}}$ group were significantly lower ( $P=$ 0.001, $P=$ 0.014, respectively). And the expression of CDK4 and FGFR3 in the CPNE5 high expression group were significantly lower than in the CPNE5 low expression group ( $P=$ 0.023, $P=$ 0.031, respectively). Beta-2 microglobulin (B2M), CRP (C-reactive protein), LDH, and HK2 expression were lower in the CPNE9 high expression group than in the CPNE9 low expression group ( $P=$ 0.041, $P=$ 0.034, $P=$ 0.035, $P=$ 0.018, respectively), and the CPNE9 high expression group may also be associated with isotype ( $P=$ 0.002).

3.4 Univariate and multivariate analysis of EFS and OS

To further assess the prognostic significance of CPNE expression, univariate and multivariate COX proportional hazard regression models were constructed (Table 3). Univariate analysis showed that ALB, B2M, HGB, LDH, CPNE5 and CPNE9 were closely related to EFS and OS (all $P<$ 0.05). AGE, CPNE2 and CDK4 were closely related to OS ( $P=$ 0.028, 0.049, 0.043, respectively), and CCND1 was closely related to EFS ( $P=$ 0.006). In addition, in multivariate analysis, we found that LDH and CPNE5 were independent prognostic factors for EFS and OS (OS: $P<$ 0.001, $P=$ 0.006; EFS: $P=$ 0.001, $P=$ 0.005, respectively). Meanwhile, ALB, B2M and CPNE2 were independent prognostic factors for OS ( $P<$ 0.001, $P=$ 0.001, $P=$ 0.005), and CPNE9 was an independent prognostic factor for EFS ( $P=$ 0.002).

3.5 Co-expression of CPNE1-9

Correlations between the expression levels of CPNE members were analyzed in our study. As shown in Fig. S2, the expressions of CPNE members were not significantly associated with each other (all $P<$ 0.001, $|R_{\text{𝑝𝑒𝑎𝑟𝑠𝑜𝑛}}|⩽$ 0.25).

3.6 Survival analysis of combined expression of CPNE5 and CPNE9

The effect of the combination of CPNE5 and CPNE9 on prognosis was further explored in MM, the results showed that patients with CPNE5 ${}^{\mathrm{\text{High}}}$ CPNE9 ${}^{\mathrm{\text{High}}}$ have significantly higher survival probability on EFS and OS compared with other groups ( $P=$ 0.0011, $P=$ 0.0016, Fig. 2A and B). Meanwhile, EFS and OS of the cumulative events of CPNE5 ${}^{\mathrm{\text{High}}}$ CPNE9 ${}^{\mathrm{\text{High}}}$ were also significantly lower than other groups ( $P=$ 0.0011, $P=$ 0.0016, Fig. 2C and D).

4. Discussion

In our study, a total of 729 patients were enrolled to explore the prognosis role of CPNE1-9 in MM. High expression of CPNE5 and CPNE9 were proved to be positive prognosis indicators. CPNE5 was an independent favorable predictor for EFS and OS, and CPNE9 was an independent positive indicator for EFS. Combination of high expressions of CPNE5 and CPNE9 might be a better predictor for the survival of MM patients.

CPNE5 mainly exists in cytoplasm of human cells and can be detected in the heart, stomach, spleen, lymph nodes and testis [25]. In esophageal squamous cell carcinoma, the expression of CPNE5 in cancer tissues is lower than that in adjacent tissues, and patients with low CPNE5 expression have shorter OS. Moreover, low expression of CPNE5 may promote the local growth of esophageal cancer and increase the resistance of chemotherapy drugs [26]. These findings suggest that CPNE5 is closely related to tumorigenesis and may affect the prognosis of cancer patients. In our study, MM patients with high expression of CPNE5 had longer OS and EFS. And the clinical and molecular characteristics analysis found that LDH, ALB and the expressions of CDK4 and FGFR3 in the CPNE5 ${}^{\mathrm{\text{high}}}$ group were lower than that in the CPNE5 ${}^{\mathrm{\text{low}}}$ group. All these factors are related to MM progression or prognosis. For instance, LDH is an independent poor prognostic factor for MM, which promotes tumor growth, increases tumor burden and chemotherapy drug resistance [27]. ALB is an important indicator for evaluating the prognosis of MM in the International Staging System (ISS), when ALB $⩾$ 3.5 g/dL and B2M $<$ 3.5 mg/L, the patients in stage I have the longest median survival time [28]. As a target gene in MM, inhibition of CDK4 can promote apoptosis of cells, thereby exerting an anti-cancer effect in MM cells [29]. FGFR3 is an oncogene and plays an essential role in the progression of MM [30]. Therefore, the positive prognosis effect of CPNE5 may be partly due to the association of CPNE5 with these prognostic factors. To this end, we further conducted a multivariate analysis and revealed that CPNE5 is an independent prognosis indicator for MM patients. Collectively, high expression of CPNE5 might be an independent positive indicator in MM.

A study by Liu et al. [31] found that CPNE9 is specifically expressed in pancreatic tumor tissues, which indicates that CPNE9 is related to cancer progression. Moreover, CPNE9 was the top five of the prognostic 14-gene signatures that constructed the prognostic model with a high predictive ability for glioblastoma [32]. This suggests that CPNE9 has important clinical significance for the prognosis assessment of glioblastoma patients. In our study, MM patients with high expression of CPNE9 had longer OS and EFS. At the same time, the CPNE9 ${}^{\mathrm{\text{high}}}$ group had lower expressions of B2M, CRP, LDH and HK2 compared with the CPNE9 ${}^{\mathrm{\text{low}}}$ group. In R-ISS, patients with B2M $⩾$ 5.5 mg/L, high-risk chromosomal abnormalities and/or high LDH levels are in stage III and have the worst prognosis [33]. Elevated CRP can promote osteoclast differentiation and accelerate bone destruction mechanism, which also has a negative impact on the prognosis of MM [34, 35]. Meanwhile, the level of HK2 is negatively correlated with EFS and OS, and may become a potential therapeutic target for MM [36]. These clinical and molecular characteristics negatively correlated with CPNE9 expression are all poor prognostic factors for MM, which are consistent with our findings and further confirm that the high expression of CPNE9 is related to the positive prognosis of MM.

Multivariate Cox regression analysis showed that LDH and CPNE5 were independent prognostic factors for EFS and OS. ALB, B2M and CPNE2 were independent prognostic factors for OS, and CPNE9 was an independent prognostic factor for EFS. These results further confirmed that high expressions of CPNE5 and CPNE9 were positive prognostic factors in MM patients. The co-expression analysis of CPNE1-9 in MM showed that there was no obvious expression correlation between CPNE family members, indicating that there was no mutual regulation mechanism between CPNE5 and CPNE9. In addition, the survival analysis of the combined expression of CPNE5 and CPNE9 found that the survival probabilities and cumulative events of CPNE5 ${}^{\mathrm{\text{High}}}$ CPNE9 ${}^{\mathrm{\text{High}}}$ group were significantly higher than other groups. Therefore, the combination of CPNE5/CPNE9 may become a better positive prognostic factor for MM patients.

5. Conclusion

In summary, we have shown that high expressions of CPNE5 and CPNE9 are positive prognostic factors in MM patients, especially for the combination of CPNE5 and CPNE9, and they may be potential targets for MM therapy. But the specific molecular mechanisms of CPNE5 and CPNE9 in the development of MM are still unclear. Further studies with larger sample size are needed to validate our findings, and the pathogenesis of CPNE5 and CPNE9 in MM also needs additional exploration.