Preoperative clinical characteristics for differentiating malignant transformation from torsion of mature cystic teratoma

Abstract

This retrospective study was conducted in 27 patients with malignant transformation of mature cystic teratoma（MT-MCT）and 125 ovarian teratoma patients with torsion who underwent surgery in the First Affiliated Hospital of Wenzhou Medical University from 2008 to 2019. The incidence of MT-MCT in this study was 0.79%. The 3-year overall survival (OS) rate was 69.6 ± 9.6%. The 3-year progression-free survival (PFS) rate was 58.3 ± 9.6%. Kaplan–Meier survival analysis indicated that patients with squamous cell carcinoma (SCC) had significantly shorter OS compared with non-SCC patients. Older age (OR 1.076, 95% CI 1.041–1.111), higher platelet (PLT) level (OR 1.012, 95% CI 1.005–1.020) and lower neutrophil-to-lymphocyte ratio (NLR) level (OR 0.794, 95% CI 0.647–0.915) were independent predictors of MT-MCT. The area under the curve (AUC) for the combined use of age, PLT count and NLR was 0.921 (95% confidence interval 0.877–0.964; p < 0.001), with a sensitivity of 92.6% and a specificity of 80.8%.

Keywords:

• Diagnosis
• mature cystic teratoma
• malignant transformation
• torsion

Introduction

Mature cystic teratomas (MCTs) of the ovary, also known as dermoid cysts, are the most common ovarian germ cell tumours, accounting for 11–20% of ovarian neoplasms and 60% of benign ovarian tumours (Curling et al. 1979). They are benign tumours that may occur in women of any age, with the highest incidence during reproductive years (Goudeli et al. 2017). MCTs are composed of tissues derived from one or more of the three embryonic layers (ectoderm, mesoderm and endoderm) (Hackethal et al. 2008). Most patients with MCT are asymptomatic, but some experience torsion, rupture or malignant transformation. Torsion is the most frequent complication of MCT, occurring in approximately 3.2–16% of patients (Comerci et al. 1994, Artunc Ulkumen et al. 2013), with the highest incidence in women of reproductive age (White and Stella 2005), of which the most common symptom is lower abdominal pain. Emergency laparoscopic adnexal surgery is the gold standard treatment for torsion patients. However, only 46% of torsion patients have been reported in preoperative imaging examination (Koleli 2015). Moreover, abdominal pain is not specific to torsion, but could also occur in other clinical conditions, such as malignant transformation. As a rare and serious complication, malignant transformation occurs in only 0.17–3% of MCT cases (Comerci et al. 1994, Hackethal et al. 2008). Malignant transformation of mature cystic teratomas (MT-MCT) may occur in any of the three embryonic layers. The major histological type of malignant transformation is squamous cell carcinoma (SCC), accounting for approximately 80% of cases (Kikkawa et al. 1997), followed by carcinoid (Saunders and Hertzman 1960, Yan et al. 2015), adenocarcinoma (Chang et al. 2010), sarcoma (박미선 2007), thyroid carcinoma (O’Neill et al. 2012), melanoma (Kudva et al. 2015) and Oligodendroglioma (Opris et al. 2009). As it knows, malignant transformation of MCT has a poor prognosis. The estimated 5-year overall survival (OS) rates of these malignancies have been reported to be 31.2% (Qin et al. 2021). However, the mechanism and treatment of malignant transformation have not yet been demonstrated. It is reported that complete cytoreduction surgery including bilateral salpingo-oophorectomy, total hysterectomy and lymphadenectomy is associated with a better prognosis of patients with MT-MCT (Hackethal et al. 2008). Ovarian cancer scarcely presents clinical symptoms, even though tumour cells spread extensively on the peritoneum (Borner et al. 2018). The presence of fatty or organoid components (teeth, hair and bones) is a sensitive sign of teratoma. However, the sensitivity of the ultrasound does not make it possible to distinguish malignant degeneration from mature teratoma (Conte et al. 2020). Because of the non-specific ultrasound image and haematological tests, it is challenging to preoperatively diagnose MT-MCT. Patients with MT-MCT and torsion have similar symptoms, they are usually asymptomatic or experienced abdominal pain, distention or pelvic solid mass, it is challenging to distinguish MT-MCT and MCT with torsion preoperatively, particularly in these emergency patients with acute abdominal pain. It is a difficult problem whether emergency operation is needed. As reported by a systematic review in 2019 (Li et al. 2019), median preoperative CA125 was 64.4 U/mL, carbohydrate antigen (CA19-9) was 144.0 U/mL, which were both high in SCC transformation in MCT. There have also been reported that serum CA125, CA19-9 were elevated in 59.1%, 64.7%, respectively, of patients with SCC (Chen et al. 2008). In addition, Higher CA19-9 and CA125 were reported to be associated with MT-MCT compared with MCT (Nanki et al. 2017). Our previous study found that the serum levels of cancer antigen 125 (CA125) and CA19-9 were both significantly higher in the torsion patients with MCT (Wang et al. 2017). Therefore, similar elevated tumour markers made the diagnosis more difficult. Thus, incomplete diagnosis and suboptimal surgical tumour resection usually leads to poor prognosis of MT-MCT. Due to the lack of any distinguishable specific symptoms, ultrasound features and tumour markers, identifying patients with high-risk for MT-MCT is urgent for providing practical and clinical information for clinicians before surgery. In this article, we described the clinicopathologic characteristics and survival status of MT-MCT. And we aimed to investigate the preoperative indexes as diagnostic markers to distinguish MT-MCT from MCT with torsion.

Materials and methods

Patients

This retrospective, single-institution study was conducted on MCT patients who underwent surgery at the Department of Gynaecology in the First Affiliated Hospital of Wenzhou Medical University from January 2008 to November 2019. This study was approved by the Ethical Committee of the First Affiliated Hospital of Wenzhou Medical University with the number KY2022-R069 and written informed consent was obtained. A total of 27 cases histopathologically confirmed as malignant transformation of ovarian MCT were admitted into the MT-MCT group. A total of 125 cases with torsion during surgery were finally enrolled in the torsion group (Figure 1). Clinical and pathological data for these patients were extracted from the medical records. Haematological parameters, such as WBC count, haemoglobin (Hb), albumin, platelet (PLT) count, neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), lymphocyte-to-monocyte ratio (LMR) and tumour markers, such as Alpha fetoprotein (AFP), carcinoembryonic antigen (CEA), CA125, CA19-9 were recorded. All patients underwent pelvic ultrasonography (US) or computed tomography (CT). Tumour size was determined by maximum diameter of ovarian tumour based on preoperative images. Blood samples were collected from the patients 3 d prior to surgery. Follow-up data of MT-MCT was obtained by telephone interview. The last follow-up visit of MT-MCT was 3 March 2022. The OS was defined as the time (months) from the date of surgery to the date of death. The PFS was measured from the date of primary surgery to the date of disease recurrence or disease progression.

Figure 1. Flowchart of patients in this study.

Statistical analysis

Statistical analysis was performed using IBM SPSS Statistics version 20 (IBM Corp., Armonk, NY) and p < 0.05 was considered statistically significant. Diagnostic value of preoperative laboratory markers was assessed by receiver-operating characteristic (ROC) curve analysis and Youden’s index. The ROC area under the curve (AUC), sensitivity, specificity, 95% confidence interval (95% CI) was presented to describe the diagnostic value of each marker. Continuous variables are expressed as mean with standard deviation and compared using an independent-samples t test or Mann–Whitney test to evaluate the significant differences between groups. Survival analyses were performed using the Kaplan–Meier method. Univariable and multivariable analyses were performed using logistic regression model.

Result

A total of 3420 MCT patients who underwent surgery in the First Affiliated Hospital of Wenzhou Medical University between 2008 and 2019. Among these patients, malignant transformation accounted for 27 (0.79%), of which 15 were SCC, torsion was founded in 164 (4.80%) patients, among whom 39 subjects were excluded from the analysis for the combination of pregnancy, inflammatory disease, malignancy and incomplete data (Figure 1). Of the patients in MT-MCT group, only 5 (18.52%) cases were promoted malignant possibility by ultrasonic examination. Twelve cases (44.44%) presented with abdominal pain and 6 (22.22%) with abdominal distension, 1 case (3.70%) had manifestation of fever, 8 cases had abdominal mass revealed by US when admitted to the hospital. Laparoscopic surgery was performed in 10 (37.04%) patients in this group. Of 9 (33.33%) patients were pathologically diagnosed at an advanced stage (Stage III–IV). 9 (33.33%) deaths were found during follow-up. It is estimated that the 3-year survival rate was 69.6 ± 9.6% for OS and 58.3 ± 9.6% for PSF. The median OS time was 71 months and the recurrent rate was 57 months. Kaplan–Meier survival analysis indicated that SCC patients had significantly shorter OS compared with non-SCC patients (Figure 2). In the torsion group, 99 (79.20%) patients presented with abdominal pain, 12 (9.60%) with distention and 14 (11.20%) with asymptomatic abdominal mass. Of 121 (96.80%) patients underwent laparoscopic procedures. As shown in Table 1, the average age of patients with MT-MCT was 54.81 ± 14.42, compared with 31.66 ± 14.83 in the torsion group (p < 0.01), which indicated that older patients were more prone to have malignant transformation. Baseline characteristic also showed indication of higher CEA levels (27.24 ± 122.17 vs. 5.48 ± 22.40, p < 0.001), CA125 levels (154.82 ± 369.08 vs. 31.85 ± 31.54) and PLT count (316.19 ± 118.25 vs. 242.70 ± 64.87 p < 0.001) in the MT-MCT group when compared with torsion group. Additionally, the comparison revealed lower level of albumin and NLR in MT-MCT group (p < 0.05 for both). Patients in the MT-MCT group had a larger tumour size than those in the torsion group (122.59 ± 53.23 vs. 101.72 ± 34.55, p = 0.011). No significant difference was observed among groups for preoperative haematological markers including Hb, PLR, LMR and tumour markers, such as AFP and CA19-9.

Figure 2. Kaplan–Meier survival curves for overall survival (OS) in patients with SCC and non-SCC, which showed that SCC were significantly associated with shorter OS.

Table 1. Baseline characteristic of patients with MT-MCT and torsion.

Characteristics	MT-MCT group (n = 27)	Torsion group (n = 125)	p Value
Age (years)	54.81 ± 14.42	31.66 ± 14.83	<0.001
Size (mm)	122.59 ± 53.23	101.72 ± 34.55	0.011
AFP (U/mL)	27.24 ± 122.17	5.48 ± 22.40	0.073
CEA (U/mL)	15.80 ± 31.83	1.72 ± 2.27	<0.001
CA125 (U/mL)	154.82 ± 369.08	31.85 ± 31.54	<0.001
CA19-9 (U/mL)	176.22 ± 577.56	122.12 ± 481.62	0.155
PLT count (×10^9 /L)	316.19 ± 118.25	242.70 ± 64.87	<0.001
Haemoglobin (g/L)	119.37 ± 14.47	122.94 ± 19.82	0.385
Albumin (g/L)	39.40 ± 6.15	44.05 ± 4.27	<0.001
NLR	4.23 ± 2.97	6.93 ± 2.97	0.023
PLR	223.30 ± 110.76	203.43 ± 141.65	0.495
MLR	0.37 ± 1.85	0.95 ± 3.78	0.428

AFP: Alpha fetoprotein; CEA: carcinoembryonic antigen; PLT: platelet; NLR: neutrophil-to-lymphocyte ratio; PLR: platelet to lymphocyte ratio; MLR: lymphocyte to monocyte ratio

Variables with statistical significance in the clinical characteristic and univariate analysis were progressed to logistic regression model by backward step wise selection. Multivariable analyses suggested age (OR 1.0746, 95% CI: 1.041–1.117, p < 0.001), PLT (OR 1.012, 95% CI: 1.005–1.020, p = 0.001) and NLR (OR 0.794, 95% CI: 0.647–0.915, p = 0.027) were significantly associated with malignant transformation (Table 2).

Table 2. Univariate and multivariate analysis.

Characters	Univariate analysis		Multivariate analysis
	OR (95% CI)	p Value	OR (95% CI)	p Value
Age (years)	1.055 (1.017–1.096)	<0.001	1.076 (1.041–1.111)	<0.001
Size (mm)	1.013 (1.003–1.024)	0.015	–
CEA (U/mL)	1.224 (1.033–1.450)	0.001	–
CA125 (U/mL)	1.012 (1.004–1.019)	0.003	–
PLT (×10^9 /L)	1.013 (1.007–1.019)	<0.001	1.012 (1.005–1.020)	0.001
Albumin (g/L)	0.834 (0.762–0.913)	<0.001
NLR	0.845 (0.729–0.978)	0.024	0.794 (0.647–0.915)	0.027

CEA: carcinoembryonic antigen; PLT: platelet; NLR: NLR: neutrophil-to-lymphocyte ratio

The ROC analysis of significant variables to diagnose MT-MCT is shown in Figure 3. The AUC for the comprising measurement of age, serum PLT count and serum NLR was 0.921 (95% confidence interval 0.877–0.964; p < 0.001), with a diagnostic sensitivity of 92.6%, a specificity of 80.80%. The optimal cut-off value was 38.5 for age, 317.5 for PLT count and 3.31 for NLR, based on ROC curve and Youden’s index (Table 3).

Figure 3. Receiver-operating curve of preoperative markers to diagnose MT-MCT.

Table 3. The diagnostic value of preoperative markers for MT-MCT.

Marker	Cut-off value	Sensitivity %	Specificity %	AUC (95% CI)	p Value
Age (years)	38.50	85.2	74.4	0.869 (0.808–0.929)	<0.001
PLT (×10^9 /L)	317.50	44.4	89.6	0.711 (0.591–0.831)	0001
NLR	3.31	73.6	59.3	0.657 (0.554–0.761)	0.010
Combination^a	NA	92.6	80.8	0.921 (0.877–0.964)	<0.001

PLT: platelet; NLR: neutrophil-to-lymphocyte ratio; NA: not applicable

^aComprising measurement of age, serum platelet count and serum neutrophil-to-lymphocyte ratio.

Discussion

MCTs of the ovary are benign germ cell tumours, the most are asymptomatic. Although abdominal pain is more prominent with torsion of this tumour, it can also be found in malignant transformation. Moreover, serum level of CA125 and CA19-9 has no significant difference between MT-MCT and torsion group. It is difficult to distinguish MT-MCT and torsion of MCT due to the atypical symptom and laboratory markers. Moreover, Ultrasound is insufficient in demonstrating malignant degeneration. Quality ultrasound evaluations are largely dependent on experienced sonographers, which may be the primary limiting factor in the performance of ultrasound evaluation. To the best of our knowledge, this is the first retrospective study to demonstrate the preoperatively diagnostic value of preoperative markers to distinguish MT-MCT from MCT with torsion. As the most severe complication, malignant transformation exists in 0.17–3% of MCT (Hackethal et al. 2008, Qin et al. 2021). In this article, the incidence of this rare malignancy was 0.79%, consistent with previous reports. Among the 27 patients of MT-MCT, 15 cases (55.56%) were pathologically confirmed as SCC, which is much lower than that of other studies (Kikkawa et al. 1997, Jamor et al. 2020). In our study, the mean age of MT-MCT was 54.81 years, of which 59.26% were over 50 years old, while patients in torsion group were generally younger, majority in childbearing period. Age is a vital risk factor for malignancy, which was confirmed earlier by previous literatures. In a study made by Kikkawa et al. (1998), the mean age of 37 cases with malignant transformation of SCC was 55.2 years, Chen et al. (2008) also reported a mean age of 55.0 ± 14.4 years for all 220 cases of SCC. It was also reported that the mean age of patients with MT-MCT was 51.3 years (Qin et al. 2021). Thus, for postmenopausal women diagnosed with MCT, we should pay more attention to the possibility of malignant transformation.

It has been clearly stated that PLT production and activation promoted tumour growth (Hufnagel et al. 2020). Multiple studies have demonstrated a significant relationship between thrombocytosis and malignant tumour (Hufnagel et al. 2020, Nakao et al. 2020). PLTs can stimulate cells in tumour microenvironment through paracrine mechanism, so as to promote the production of tumour promoting products (Seizer et al. 2013). This study found that the serum PLT count was higher in the MT-MCT group than in the torsion group, indicating that PLT plays an important role in MT-MCT.

Of note, patients with MT-MCT tended to have lower level of NLR. The result appears to be contrary to previous studies, which found that preoperative NLR, as an important inflammatory indicator, was shown to be promising screening and prognostic factors of ovarian cancer (Wang et al. 2016, Feng et al. 2016, Prodromidou et al. 2017). The systemic inflammatory response from cancer cells promotes the infiltration of neutrophils (Balkwill and Mantovani 2001, Yin et al. 2019). In addition, the production of an effective angiogenesis cytokine, vascular endothelial growth factor, promoted by elevated neutrophils, further increases the growth of cancer (Nie et al. 2017). Nanki et al. (2017) showed pre-treatment NLR might be a potential preoperative diagnostic marker of MT-MCT. NLR were found to be significantly elevated among MCT patients with torsion, compared with those without torsion (5.28 ± 3.33 vs. 2.15 ± 0.77 p < 0.001), and the cut-off value was 3.56 (Wang et al. 2017). No previous studies have compared MT-MCT and adnexal torsion of MCT. Our study demonstrated that the mean level of NLR was 4.23 ± 2.97 in MT-MCT group, lower than 6.93 ± 2.97 in torsion group. NLR elevated in both MT-MCT and MCT with torsion compared with MCT. Various inflammatory mechanisms that could be involved in the process of ovarian ischemia–reperfusion injury owing to adnexal torsion (Guven et al. 2015). This study confirmed this theory and speculated that the inflammatory response is more severe in torsion patients.

The serum level of tumour markers was higher in the MT-MCT group than in the torsion group, founded by our study. The serum levels of CA 19-9 were higher in MT-MCT, but without significance. The serum levels of CEA and CA125 were significantly higher in univariate analysis but meaningless in multivariate analysis. A few articles have reported elevated CEA in MT-MCT patients (Park et al. 2008, Chen et al. 2008). CEA has been found elevated in 65% patients with SCC (Chen et al. 2008). SCC antigen also has been reported to be a useful tumour marker for patients with SCC arising in MCT (Mori et al. 2003, Rim et al. 2006, Santos et al. 2007). But SCC antigen was not evaluated in our study because lots of patients did not complete SCC before operation because of ovarian neoplasm.

The limitation of this study is the small cohort of MT-MCT due to the rarity and the retrospective design. Thus, additional large prospective studies are needed to further validate the diagnostic value of age, preoperative PLT count and NLR.

Conclusion

This study demonstrated that clinical characteristic (age, PLT and NLR) can be used as preoperative markers for the diagnosis of MT-MCT. Combination of the above indicators is more sensitive and has better diagnostic value. There should be awareness of the malignant possibility of MCT complicated with abdominal pain if patient is older, especially postmenopausal; the PLT counts more than 317.5 × 10^9/L, as well as NLR level less than 3.3.

Disclosure statement

No potential conflict of interest was reported by the authors.