Objective
Myomatous erythrocytosis syndrome is a rare complication of uterine leiomyoma caused by erythropoietin (EPO) that is produced by tumor cells. We assessed the EPO expression in leiomyomas and investigated the effects of EPO on the tumor growth.
Study Design
Tissue samples were collected from 114 patients with uterine leiomyomas who underwent myomectomy or hysterectomy in Yokohama City University Hospital. From 17 patients, the corresponding normal myometrium was also collected. All samples were analyzed for EPO messenger RNA (mRNA) expression by real-time reverse transcription-polymerase chain reaction. EPO protein expression was determined by an enzyme-linked immunosorbent assay. The relationships between EPO expression and clinicopathological features were retrospectively analyzed using the patients’ charts. Blood vessel density and maturity were assessed using hematoxylin-eosin staining and CD34 immunohistochemistry.
Results
EPO mRNA expression was detected in 108 of 114, or 95%, of the leiomyomas. The mean EPO mRNA expression in the leiomyoma was higher than the corresponding normal myometrium (3836 ± 4122 vs 1455 ± 2141; P = .025 by Wilcoxon rank test). The EPO mRNA expression in the leiomyomas varied extensively among samples, ranging from undetectable levels to 18-fold above the mean EPO mRNA of normal myometrium. EPO protein production was observed concomitant with mRNA expression. A positive correlation of leiomyoma size and EPO mRNA expression was shown by Spearman rank correlation coefficient (ρ = 0.294; P = .001), suggesting the involvement of EPO in leiomyoma growth. The blood vessel maturity was also significantly increased in EPO-producing leiomyomas (high vessel maturity in high vs low EPO group: 67% vs 20%; P = .013 by Fisher exact test).
Conclusion
This report demonstrates that EPO is produced in most of conventional leiomyomas and supports a model in which EPO accelerates tumor growth, possibly by inducing vessel maturity. Our study suggests one possible mechanism by which some uterine leiomyomas reach a large size, and the understanding of EPO expression patterns in these tumors may be useful for management of the patients with leiomyomas.
Uterine leiomyoma is a benign tumor, frequently affecting women of reproductive age. The estimated cumulative incidence of uterine leiomyoma by the age of 50 years is greater than 70%. The progression of uterine leiomyomas can follow a variety of courses. Some tumors stay small for years, whereas others may enlarge tremendously and rapidly. Estrogen and progesterone are known to support the growth of uterine leiomyoma ; however, the mechanism by which some leiomyomas grow to a great size remains unclear.
Erythropoietin (EPO) is a glycoprotein hormone essential to the regulation of erythrocyte production. The possible association between EPO and leiomyoma was first observed in the case of myomatous erythrocytosis syndrome (MES) by Thomson and Marson in 1953. This rare clinical condition was defined as erythrocytosis with uterine leiomyoma and diagnosed by restoration of normal hematological values following the removal of the uterine leiomyoma.
The recent studies of MES have reported the expression of EPO protein or messenger RNA (mRNA) in the uterine leiomyomas of MES patients, which suggests that EPO produced by leiomyoma cells may play a role in the development of this syndrome. According to the largest clinicopathological review of MES cases by LevGur and Levie, the anatomical sizes of the leiomyoma were often very large, and half the MES cases were postmenopausal. Therefore, we hypothesize that EPO may act as a factor that stimulates the enlargement of leiomyomas in both premenopausal and postmenopausal women.
In this study, we assessed the expression of EPO mRNA and protein in a set of 114 uterine leiomyomas. We determined the correlation between EPO expression and tumor size to explore a potential role of EPO in the growth of uterine leiomyomas. Moreover, the patients’ clinicopathological features, including intratumor blood vessels, were assessed in the context of tumor EPO production to determine a possible mechanism by which EPO functions to induce growth in leiomyomas.
Materials and Methods
Patients and tissue samples
One hundred fourteen uterine leiomyoma tissue samples were collected from women who underwent hysterectomy or myomectomy between 2005 and 2012 at Yokohama City University Hospital. The reasons for the surgery included leiomyoma, uterine prolapse, ovarian cancer, and cervical cancer. In the cases of samples complicated with malignant disease, the absence of myometrial invasion was confirmed microscopically. Among the leiomyoma nodules, the largest leiomyoma nodule of the patient was chosen, and the center part of the nodule with minimum degeneration was sampled for analysis.
Corresponding normal myometrium tissue samples were also collected from 17 patients selected randomly. The patients’ clinicopathological features were assessed retrospectively using medical records. Menopause was defined as the absence of menstrual period for over 12 months.
To determine the largest leiomyoma nodule, diameters were measured on magnetic resonance images taken within 1 year prior to surgery. The leiomyoma was measured by ultrasonography at least 1 month prior to the surgery and reexamined by magnetic resonance imaging when a change in tumor size was detected. The samples were divided and (1) snap frozen in liquid nitrogen and stored at –80°C or (2) fixed in formalin and embedded in paraffin, for subsequent analysis.
This study was approved by the Ethical Committee of Yokohama City University Graduate School of Medicine. Written informed consent was obtained from each patient for the research-based use of their tissue samples and data.
Real-time reverse transcription–polymerase chain reaction
Total RNA was extracted from frozen tissue samples using Sepasol-RNA I Super G (Nacalai Tesque, Kyoto, Japan) and the Illustra RNAspin mini-RNA isolation kit (GE Healthcare, Buckinghamshire, UK) following the manufacturers’ protocols. RNA concentrations were determined spectrophotometrically at 260 nm using a NanoDrop 2000 (Thermo Scientific, Waltham, MA).
Reverse transcription (RT) was carried out at 25°C for 10 minutes, 42°C for 60 minutes, and 85°C for 5 minutes. The RT mixture contained up to 2.5 μg total RNA, 4 μL SuperScript VILO master mix (Life Technologies, Carlsbad, CA), and RNase-free water to a total volume of 20 μL. RT products were diluted to a final concentration of 10 ng/μL, and then stored at –30°C.
TaqMan Fast advanced master mix (Life Technologies) and commercial primers and probes from TaqMan gene expression assays were used for amplification of EPO (assay identification Hs00171267_m1; Life Technologies) and β-actin (assay identification 4326315E; Life Technologies).
The polymerase chain reaction (PCR) consisted of single steps at 50°C for 2 minutes and 95°C for 20 seconds, followed by 40 cycles of 95°C for 1 second and 60°C for 20 seconds. PCRs were performed on the 7900HT Fast real-time PCR system (Life Technologies). A standard curve was generated using a cloned 180 bp portion of the human EPO mRNA sequence inserted into pEX-A by FASMAC (FASMAC, Kanagawa, Japan). An absolute quantification analysis was performed using the software provided with the instrument using this standard curve.
Enzyme-linked immunosorbent assay
One hundred milligram frozen samples of each leiomyoma and myometrium tissue specimen were homogenized in 500 μL of lysis buffer containing T-PER tissue protein extraction reagent and protease inhibitor cocktail (Thermo Scientific). The supernatant was collected by centrifugation at 10,000 × g for 5 minutes, and protein concentrations were determined using Coomassie Plus (Bradford) assay kit (Thermo Scientific). The EPO concentrations were analyzed by a human erythropoietin platinum enzyme-linked immunosorbent assay (ELISA) (short incubation) (Bender MedSystems, Vienna, Austria) according to the manufacturer’s instructions.
Immunohistochemistry
Three micrometer thick sections were deparaffinized in xylene and rehydrated with graded ethanol series. Antigen was retrieved by treating with 0.1% trypsin in phosphate-buffered saline (PBS) at 37°C for 30 minutes and then self-cooled to room temperature. Endogenous peroxidase activity was blocked with 0.3% H 2 O 2 in PBS. Nonspecific protein binding was attenuated by incubation for 15 minutes with 10% rabbit serum. After washing with PBS, slides were incubated at 4°C for overnight with mouse anti-CD34 monoclonal antibody (diluted 1:50, IM0787; Beckman Coulter, Brea, CA).
The slides were washed in PBS, treated with secondary biotinylated antibody at room temperature for 10 minutes, washed in PBS, and then treated with streptoavidin peroxidase at room temperature for 5 minutes (Histofine SAB-PO[M]; Nichirei Bioscience Inc, Tokyo, Japan). Immunoreactivity was visualized by 3-3′ diaminobenzidine solution as the chromogen. Hematoxylin was used for counterstaining. For a negative control, primary antibodies were replaced with PBS. CD34 was used as an endothelial marker.
Measuring vessel density and assessment of the blood vessel maturity
To investigate the effects of EPO on tumor blood vessels, among the 114 samples, the intratumor vessel density and maturity were compared between the lowest EPO samples (n = 15) and the highest EPO samples (n = 15).
To obtain vessel density of the tumor, the immunostained tumor sections were observed at ×100 magnification to identify the areas with the most CD34 staining, indicating the highest number of vessels. Within this location, all vessels were counted within a 0.74 mm 2 area (×200 field) in 3 random fields and averaged.
Vessel maturation was assessed by the observation of blood vessels in the center part of the leiomyoma without severe degeneration. Blood vessels were observed using CD34 immunohistochemistry to identify the endothelial cells and hematoxylin and eosin (HE) staining to visualize the existence of pericytes and the thickness of tunica media (smooth muscle and elastic layer).
The maturity of the blood vessels in the tumor was evaluated according to a scoring system as follows ( Figure 1 ): score 1, microvessels are scattered; score 2, microvessels are moderately distributed; score 3, microvessels are notable; score 4, microvessels with mature vein and arteries with a luminal space, 20–50 μm in minimum diameter and smooth muscle layers; and score 5, notable dilated veins and arteries with a luminal space greater than 50 μm in minimum diameter with smooth muscle layers, with or without microvessels. The score was determined by agreement of 3 blinded researchers.
Statistical analyses
Statistical analyses were performed using IBM SPSS Statistics 22.0 (SPSS Inc, Chicago, IL). The association of EPO mRNA expression with clinicopathological features and results of ELISA for EPO was assessed using Spearman rank correlation coefficient. Wilcoxon signed-rank test was performed to analyze the EPO mRNA expression of the paired samples of uterine leiomyoma and myometrium. Univariate analyses were performed using the Mann-Whitney U test or Fisher exact test. Multivariate analysis was performed using multiple regression analysis. Values of P < .05 were considered statistically significant.
Results
The clinicopathological backgrounds of the patients enrolled in this study are summarized in Table 1 . Myomectomy or hysterectomy for uterine leiomyoma is usually performed on premenopausal women; therefore, most of the patients analyzed in this study were premenopausal (100 of 114, 88%). No patients were pregnant within the 6 months prior to surgery. Two to 6 cycles of gonadotropin-releasing hormone (GnRH) agonist was administered preoperatively in 19 patients (range, 0–2 months; mean, 1.0 ± 0.3 months). Four patients had received combination therapy of estrogen and progestin, and 1 patient had GnRH agonist therapy followed by progestin (dienogest) as indicated for endometriosis. Among the cases examined in this study, there were no cases with a histopathological diagnosis of lipoleiomyoma, or cellular, or bizzare leiomyoma.
| Parameter | Mean (range) | n (%) |
|---|---|---|
| Age, y | 44 (25–71) | |
| BMI, kg/m 2 | 23.5 (16.2–33.6) | |
| Parity | ||
| 0 | 48 (42) | |
| ≥1 | 66 (58) | |
| Menstrual status | ||
| Premenopausal | 100 (88) | |
| Postmenopausal | 14 (12) | |
| Diameter of largest leiomyoma nodule, cm | 10 (3–30) | |
| Hypermenorrhea in premenopausal patients | ||
| Yes | 56 (49) | |
| No | 44 (39) | |
| Hormone therapy prior to surgery | 24 (21) | |
| GnRHa therapy only | 19 (17) | |
| ≥3 cycles | 17 (15) | |
| <3 cycles | 2 (2) | |
| Combined or other hormone therapy | 5 (4) | |
| Preoperative hemoglobin, g/dL | ||
| <11 | 19 (17) | |
| ≤11 × <16 | 94 (82) | |
| ≥16 | 1 (1) |
There was 1 case clinically diagnosed as MES. The patient was 37 years old with a leiomyoma 30 cm in diameter and the preoperative hemoglobin (Hb) of 19.4 g/dL, and the serum EPO level was elevated to 32.6 IU/L (normal range, 3–16 IU/L). The patient underwent abdominal myomectomy, and 1 month after the operation, the Hb level decreased to 12.7 g/dL and the serum EPO level decreased to 15.0 IU/L. The histopathological examination of the resected tumor by HE staining revealed a growth of leiomyoma cells without conspicuous atypia or mitosis and the presence of remarkably dilated and well-muscularized blood vessels ( Figure 1 , E).
EPO mRNA and protein expression in leiomyomas
EPO mRNA was detectable by real-time reverse transcription–polymerase chain reaction (RT-PCR) in 108 of 114, or 95%, leiomyomas and all of the myometrium. Comparing paired samples of leiomyoma and normal myometrium, the mean EPO mRNA expression in leiomyoma tissues was significantly higher than that of the corresponding normal myometrium (3836 ± 4122 vs 1455 ± 2141; P = .025 by Wilcoxon rank test) ( Figure 2 ). EPO mRNA expression in the leiomyoma varied extensively by the sample, ranging from undetectable level to 18-fold of the mean EPO mRNA in the normal myometrium ( Figure 3 ).
To determine the production of EPO protein in leiomyoma samples, we selected 33 samples from the 114 samples for an ELISA analysis. We chose samples at the extremes of full EPO mRNA expression range: 15 samples with the highest and the lowest EPO mRNA expression leiomyoma samples as well as 3 samples from normal myometrium. There was a significant difference in EPO protein levels between the highest and lowest EPO mRNA group (18.7 ± 16.5 vs 1.0 ± 1.0 mIU/mg protein, P < .001 by Mann-Whitney U test) ( Figure 4 , A).
EPO protein expression correlated strongly with EPO mRNA expression (ρ = 0.828; P = .001, Spearman rank correlation coefficient), suggesting that EPO protein was produced concomitant with EPO mRNA expression ( Figure 4 , B). Therefore, we elected to use EPO mRNA real-time RT-PCR data for additional computational analysis.
Relationships between EPO expression and clinicopathological features
We examined the correlation between tumor size and EPO expression. As shown in Figure 5 , a positive correlation between leiomyoma size and EPO mRNA expression was observed (ρ = 0.294; P = .001 by Spearman rank correlation coefficient) ( Figure 5 ). These data suggest a possibility that EPO expression and large tumor size are linked.
