ORIGINAL STUDY
The Role of Magnetic Resonance Imaging in Pretreatment Evaluation of Early-Stage Cervical Cancer Wei Zhang, MD,* Jie Zhang, MD,Þ Jiaxin Yang, MD,* Huadan Xue, MD,Þ Dongyan Cao, MD,* Huifang Huang, MD,* Ming Wu, MD,* Quancai Cui, MD,þ Jie Chen, MD,þ Jinghe Lang, MD,* and Keng Shen, MD* Objective: The aim of this study is to evaluate the accuracy of magnetic resonance imaging (MRI) in the preoperative assessments of primary tumor size, parametrial invasion, and pelvic lymph node metastasis in patients with early-stage cervical cancer. Materials and Methods: A cohort of 125 patients with International Federation of Gynecology and Obstetrics stage IA2 to IIA cervical cancer who had preoperative MRI and underwent radical hysterectomy were enrolled and analyzed. The accuracy of preoperative MRI scan and pelvic examination in the measurement of tumor size was assessed based on postoperative measurement and pathologic findings. The accuracy of detection of lymph node status and parametrial invasion was also assessed by comparing the MRI and pathologic findings. Results: The mean diameter of the tumor size measured by postoperative measurement, MRI, and pelvic examination was 2.97 T 1.39 cm, 2.78 T 1.24 cm, and 1.97 T 1.70 cm, respectively. There were significant differences in the mean diameter of the tumor size between pelvic examinations and MRI scan or postoperative measurement (P G 0.0001). Based on postoperative measurement findings, accuracy of tumor size measurement between pelvic examination and MRI was determined by the degree of agreement with a difference of less than 0.5 or 1.0 cm. Pelvic examination and MRI had an accuracy of 24.75% and 39.60%, respectively, with a difference of less than 0.5 cm, and had an accuracy of 43.56% and 61.39%, respectively, with a difference of less than 1.0 cm. Correlation with postoperative measurement in tumor size was higher for MRI (r[s] = 0.481) than that for pelvic examination (r[s] = 0.362). The sensitivity, specificity, and accuracy of MRI in detecting lymph node metastasis were 27.78%, 85.98%, 77.60%, respectively. The negative predictive value of MRI in detecting parametrial invasion is 100%. Conclusions: Magnetic resonance imaging is an accurate noninvasive modality for preoperative evaluation of tumor size and also gives important information to parametrial invasion and lymph node status in patients with early-stage cervical cancer. Key Words: Magnetic resonance imaging, Cervical cancer, Pelvic examination Received March 5, 2014, and in revised form April 1, 2014. Accepted for publication April 9, 2014. (Int J Gynecol Cancer 2014;24: 1292Y1298)
Departments of *Obstetrics and Gynecology, †Radiology, and ‡Pathology, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China. Address correspondence and reprint requests to Keng Shen, MD, Department of Obstetrics and Gynecology, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences and Copyright * 2014 by IGCS and ESGO ISSN: 1048-891X DOI: 10.1097/IGC.0000000000000169
1292
Peking Union Medical College, No.1. Shuaifuyuan, Dongcheng District, Beijing 100730, China. E-mail: [email protected]. Supported by grants from the National Science and Technology Infrastructure Program’s ‘‘The National Key Technologies R&D Program of China’’ (grant 2008BAI57B01), the National High Technology Research Development Program of China (863 program, grant 2012AA02A507), and the National Natural Science Foundation of China (grant 81172482 and grant 81372780). The authors declare no conflicts of interest.
International Journal of Gynecological Cancer
& Volume 24, Number 7, September 2014
Copyright © 2014 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.
International Journal of Gynecological Cancer
& Volume 24, Number 7, September 2014 1
cancer is the third most common cancer in women. C ervical Although screening programs have led to remarkable de-
crease of incidence and mortality, there is still high mortality in developing countries.1 The initial choice of treatment largely depends on the cancer stage at the time of diagnosis, and therefore, accurate staging is essential for appropriate treatment.2 The International Federation of Gynecology and Obstetrics (FIGO) recommends a clinical stage for cervical cancer based on clinical examination. However, clinical staging is less accurate than surgical staging.3,4 Pelvic examination is hard to accurately evaluate tumor size, parametrial invasion, and pelvic side wall invasion, and also cannot evaluate lymph node status. Although the revised system did not include imaging, FIGO now encourages the use of diagnostic imaging techniques to assess the size of the primary tumor, parametrial invasion, pelvic side wall invasion, adjacent organ invasion, and lymph node metastases.3 Magnetic resonance imaging (MRI), with good soft tissue contrast and multiplanar imaging capability, is now widely accepted as an optimal method for evaluation of gynecologic malignancies including cervical cancer.5 Magnetic resonance imaging seems to be better than computed tomography for locoregional disease assessment, especially for primary tumor and adjacent soft tissue extension.6 Some studies have demonstrated that MRI examination proves to be better than clinical examination in staging of cervical cancer,7,8 but other studies shown opposite conclusions.9,10 More and more studies are focusing on management of low-risk early-stage cervical cancer in which tumor size is an important criterion.11,12 Therefore, accurate preoperative evaluation of tumor size is very important. Although many studies have compared the staging of MRI to clinical examination, few of them evaluated their accuracy of tumor size measurement. The aim of this study was to assess the accuracy of MRI in preoperative measurement of primary tumor size and in detecting parametrial invasion and pelvic lymph node metastasis of patients with early-stage cervical cancer. The role of MRI in pretreatment evaluation of early-stage cervical cancer will be discussed.
MATERIALS AND METHODS Patients A cohort of 125 patients with FIGO stage IA2 to IIA cervical cancer who underwent radical hysterectomy and pelvic and/or para-aortic lymphadenectomy at the Department of Obstetrics and Gynecology at the Peking Union Medical College Hospital between September 2009 and December 2013 were enrolled and analyzed. All patients have pathologically confirmed diagnosis of cervical cancer and were staged according to FIGO criteria.7 The MRI examinations of the pelvis and lower abdomen within 2 weeks before surgery were carried out for every patient. Clinical examination and FIGO staging were performed by 2 experienced gynecologic oncologists. Radical hysterectomy and pelvic and/or para-aortic lymphadenectomy were performed by the same experienced gynecologic oncologists. Histopathologic diagnosis was made and reviewed by 2 experienced pathologists. Two independent radiologists with extensive backgrounds in gynecological MRI,
MRI in Evaluation of Cervical Cancer
who were blinded to the patient’s outcomes, reviewed all MRI slides and made a diagnosis according to the MRI findings. There was no difference between the 2 pathologists’ comments. Three patients had differences in the diagnosis of lymph node metastasis with the 2 MRI reviewers, and we adopted the views of the third expert in gynecological MRI. The clinical data, including clinical and pathological variables, preoperative MRI information, and surgical findings, were evaluated. The study protocol was approved by the ethics committee at the Peking Union Medical College Hospital, Beijing, China.
Preoperative Assessment All patients were admitted to the hospital for preoperative workup 3 days before the operation. Pelvic examination was carried out by 2 experienced gynecologic oncologists. The vagina and cervix were visualized by specula and inspection. The visible cervical tumor was measured by a scale, and the maximal tumor size was recorded in the medical files. Both of the parametrium were palpated by rectal-vaginal examination, and parametrial invasion was suspected when the parametrium was found with nodular thickness. If the pelvic examination was unsatisfactory, examination under anesthesia was considered. In the current study, 10 patients in all underwent examination under anesthesia.
MRI Examination Magnetic resonance imaging was performed with a 3.0-T MR scanner (HD750; General Electric Healthcare, Milwaukee, WI) with an 8-channel body array coil. The scan range was from the level of the umbilicus to the pubic symphysis. Routine MRI protocol used for the assessment of cervical cancer was performed as follows: the axial spin-echo T1-weighted imaging and the transverse, sagittal, and coronal fast spin-echo T2weighted imaging. The parameter details for routine MRI protocols were as follows: T1-weighted image (repetition time [TR]/echo time [TE], 400Y600/10Y20 ms; field of view (FOV), 30 30 cm; matrix, 256 256; slice thickness/gap, 4/1 mm) and T2-weighted image (TR/TE, 2500Y4000/60Y100 ms; FOV, 30 30 cm; matrix, 256 256; slice thickness/gap, 4/1 mm). Transverse diffusion-weighted images (DWIs) were also obtained. Imaging parameter details for DWI were as follows: b = 0.800 s/mm2; TR/TE, 4500/62.4 ms; FOV, 38 38 cm; matrix, 128 128; slice thickness/gap, 4/1 mm. T2-weighted images were used to determine the primary tumor size, location, and its extension to the parametria. The sagittal T2-weighted image was used to measure the tumor’s maximum craniocaudal and anteroposterior diameters, and the transverse T2-weighted image was used to measure the maximum latero-lateral diameter to the long axis of the cervix (Fig. 1). The highest b value diffusion-weighted data set and the T2-weighted data set were used to facilitate the detection of pelvic lymph nodes.8 Lymph nodes greater than 10 mm in short diameter were considered abnormal.9 The discontinuity of low signal in the outermost ring of the cervical stroma on MRI is considered as parametrial invasion.
Postoperative Measurement and Pathological Examination All specimens of the radical hysterectomy were carefully examined at the operation theater by operators after the
* 2014 IGCS and ESGO
Copyright © 2014 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.
1293
Zhang et al
International Journal of Gynecological Cancer
& Volume 24, Number 7, September 2014
FIGURE 1. A, T2-weighted sagittal image revealing a mass with a size of 4.01 cm (largest tumor diameter). B, T2-weighted axial imaging revealing invasion of the parametrium (arrow). C and D, T2-weighted axial image and DWI revealing a lymph node in the right external iliac area (arrow).The lymph node detected on MRI is a metastasis. completion of radical hysterectomy immediately. Opening the uterus and exposing the cervix, each visible cervical tumor was measured by a scale, and maximal tumor size and location of lesion were recorded (Fig. 2). All specimens of the lymphadenectomy and parametrical tissues were also palpated and examined carefully, and enlarged lymph nodes were selected for frozen section. All the specimens of radical hysterectomy were fixated in 10% buffered formalin after visible inspection for pathologic examination. Each visible tumor was described at macroscopic examination, including maximal tumor size, location of lesion, distance from margins, and gross depth of invasion. These findings were compared with those obtained from microscopic examination. Each specimen was sectioned transversally (the thickness of each transversal was approximately 3 mm). Other parameters assessed microscopically were depth of stromal invasion, parametrial involvement, and lymph node metastasis.
methods and to identify possible outliers. The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated by standard procedures.10 BlandAltman plot was performed using MedCalc version 11.4.2.0.
Statistics The maximal tumor diameter measured by pelvic examination and MRI was compared based on the maximal tumor diameter measured by pathological examination. McNemar test was used to compare the accuracy of pelvic examination and MRI in the measurement of tumor diameter. Agreement analysis and spearman correlation coefficient were used to determine the agreement of tumor size between different methods to pathological findings. The Bland-Altman plot was used to estimate the differences between the tumor diameters of different
1294
FIGURE 2. Postoperative measurement of cervical tumor diameter (4.20 cm). * 2014 IGCS and ESGO
Copyright © 2014 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.
International Journal of Gynecological Cancer
& Volume 24, Number 7, September 2014
TABLE 1. Characteristics of patients
TABLE 3. Accuracy of pelvic examination and MRI in the measurement of tumor size less than or equal to 2 cm with different error range
Characteristics 42.51 T 9.20 25Y72
Mean age, y Range, y FIGO stage IA2 IB1 IB2 IIA1 IIA2 Histology Squamous cell carcinoma Adenocarcinoma Others Mean tumor diameter, cm Range, cm Tumor location Exophytic Endocervical Imaging modality Routine MRI DW-MRI
13 97 6 6 3
(10.4%) (77.6%) (4.8%) (4.8%) (2.4%)
91 (72.8%) 25 (20%) 9 (7.2%) 2.97 T 1.39 0.7Y7 86 (85.1%) 15 (14.9%) 46 79
The other calculations were performed using SPSS software version 13.0. P G 0.05 was considered statistically significant.
RESULT A total 125 women with early-stage cervical cancer (IA2-IIA) were included in this study. Age, FIGO stage, histological type, tumor growth pattern, and mean maximal tumor diameter on postoperative inspection and final histological report are shown in Table 1. Except for 13 cases with stage IA2 and 11 cases with stage IB1 without macroscopic lesions, a total of 101 cases had visible cervical tumor including 86 cases with stage IB1, 6 with stage IB2, 6 with stage IIA1, and 3 with stage IIA2. The tumor size measured by pelvic examination ranged from 0 to 6.0 cm, and tumor size measured by MRI ranged from 0 to 6.55 cm.
TABLE 2. Accuracy of pelvic examination and MRI for the measurement of tumor diameter based on postoperative measurement Accuracy
Pelvic examination MRI *P G 0.001.
MRI in Evaluation of Cervical Cancer
T0.5 cm
T1.0 cm
24.75% 39.60%*
43.56% 61.39%*
Different Error Range
Pelvic examination MRI
T0.3 cm
T0.5 cm
T1.0 cm
11.49% 24.14%*
25.00% 40.63%
56.25% 65.63%
*P G 0.001.
The mean maximal tumor diameter was not significantly different between MRI (2.78 T 1.24 cm) and postoperative measurement (2.97 T 1.39 cm; P = 0.15). The difference between pelvic examinations (1.97 T 1.70 cm) and postoperative measurement (2.97 T 1.39 cm) was significant (P G 0.0001). Table 2 describes the accuracy of pelvic examination and MRI in the measurement of tumor diameter. The accuracy was estimated by the degree of agreement with a difference of 0.5 or 1.0 cm based on postoperative measurement and pathologic findings. The accuracy of MRI was significantly higher than that of pelvic examination with a difference of both 0.5 cm (39.60% vs 24.75%, P G 0.001) and 1.0 cm (61.39% vs 43.56%, P G 0.001). The accuracy of pelvic examination and MRI in the measurement of tumor size of less than or equal to 2 cm is showed in Table 3. The accuracy was also estimated by the degree of agreement with a difference of 0.5 or 1.0 cm based on postoperative measurement and pathologic findings. Although the accuracy of MRI were higher than that of pelvic examination, the difference was not significant with a difference of 0.5 cm (40.63% vs 25.0%, P = 0.052) and 1.0 cm (65.63% vs 56.25%, P = 0.25). To evaluate which measuring method is more accurate, a narrow difference of 0.3 cm was performed. The accuracy of MRI was significantly higher than that of pelvic examination with a difference of 0.3 cm in the group with tumor size less than or equal to 2 cm (24.14% vs 11.49%, P G 0.001). The accuracy of different methods in measuring tumor size in different growth pattern is shown in Table 4. The accuracy of MRI was significantly higher than that of pelvic examination with a difference of both 0.5 cm (exophytic: 40.70% vs 29.07%, P = 0.004; endocervical: 33.33% vs 0.00%, P = 0.002) and 1.0 cm (exophytic: 61.63% vs 48.84%, P = 0.001; endocervical: 60% vs 0.00%, P = 0.004) in each growth pattern. The degree of agreement of MRI and pelvic examination with postoperative measurement and pathological findings is shown in Table 5. The J value represents the degree of agreement with postoperative measurement and pathologic findings. The agreement of MRI with postoperative measurement is much higher than that of the pelvic examination (J value 0.88 vs 0.07). Correlation between postoperative measurement was higher for MRI than that for pelvic examination (r = 0.481 vs r = 0.362; Fig. 3). Bland-Altman plots (Fig. 4) also show that the difference of mean tumor diameters between MRI and postoperative measurement (0.5 cm) was smaller than that of pelvic examination with postoperative measurement and pathological findings (1.0 cm).
* 2014 IGCS and ESGO
Copyright © 2014 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.
1295
International Journal of Gynecological Cancer
Zhang et al
TABLE 4. Accuracy of different growth pattern of pelvic examination and MRI for the measurement of tumor size based on postoperative measurement Exophytic Accuracy with a difference of T0.3 cm Pelvic examination 15.12% MRI 26.74%* Accuracy with a difference of T0.5 cm Pelvic examination 29.07% MRI 40.70%† Accuracy with a difference of T1.0 cm Pelvic examination 48.84% MRI 61.63%†
Endocervical 0.00% 13.33%* 0.00% 33.33%† 0.00% 60%†
*P G 0.001. †P G 0.05.
Lymph node statuses of all the 125 patients were assessed by MRI before surgery including 79 cases by DWI in addition to routine MRI (DW-MRI) and 46 patients by routine MRI. Suspicions of lymph node metastases were detected in 20 patients, 13 cases were detected by DW-MRI, and 7 cases were suspected by routine MRI. Lymphadenectomy were performed on all of the 125 patients. Totally, 18 cases were confirmed to have lymph node metastasis by the following pathology, and only 5 of them were also suspected lymph node metastases on MRI examination. The remaining 13 cases (72.22%) were false negative on MRI. In 20 patients with MRI-suspected lymph node metastasis, only 5 of them were confirmed to have lymph node metastasis by pathology, and the other 15 cases (75%) were false positive. In 13 cases that suspected lymph node metastasis with DW-MRI, 4 cases were pathologically confirmed lymph node metastases. The routine MRI showed lymph node metastasis in 7 patients, only 1 of them was pathologically confirmed. The sensitivity, specificity, accuracy, PPV, and NPV of MRI in detecting lymph node metastasis was 27.78%, 85.78%, 77.60%, 25.0%, and 87.62%, respectively. The NPV of MRI in detecting parametrial invasion in our study is 100% but had 6.4% false positive.
DISCUSSION The FIGO stage for cervical cancer is actually a clinical stage based on pelvic examination.3 Tumor size is closely associated with FIGO stage and is also an important prognostic factor. Exact evaluation of tumor size played a very important role in making clinical decision and predicting the prognosis for patients with cervical cancer. International Federation of Gynecology and Obstetrics also recommended that the diagnostic imaging techniques can be used to assess the size of primary tumor, such as MRI/computed tomography scanning.3 Some studies have performed MRI assessment of tumor size, and most of them compared the accuracy of the tumor size measurement of MRI to other imaging techniques or pathological findings,11Y13 but there are still some inevitable discrepancies between clinical staging and pathologic findings.14,15 To
1296
& Volume 24, Number 7, September 2014
assess the value of MRI in preoperative evaluation of cervical cancer, we compared the accuracy of MRI and clinical examination in preoperative measurement of primary tumor size based on the postoperative measurement and pathologic findings. It has been demonstrated that the accuracy of pelvic examination in the measurement of primary tumor size of cervical cancer is approximately around 50% with a difference of 0.5 cm16 and 50% to 72% with a difference of less than 1.0 cm.16,17 The accuracy of MRI in the measurement of primary tumor size ranged from 39.8%16 to 70%18 with a difference of 0.5 cm and 55.3% with a difference of 1.0 cm.16 In comparison with other studies, we also used the difference of 0.5 or 1.0 cm as an error range to evaluate the measurement accuracy. The accuracy of MRI in preoperative measurement of the primary tumor size is 39.60% (with error range T0.5 cm) and 61.39% (with error range T1.0 cm), whereas the accuracy of pelvic examination is 24.75% (T0.5 cm) and 43.56% (T1.0 cm), respectively. There was a significant difference in measurement accuracy between MRI and pelvic examination, which was similar to previous studies.16Y20 Our study demonstrated again that the accuracy of MRI is higher than that of pelvic examination. With the valuable-based medicine introduced and the individualization treatment developed, more and more studies are focusing on management of low-risk early-stage cervical cancer.19Y21 Definition of low-risk patients with cervical cancer includes the following criteria: squamous carcinoma, adenocarcinoma or adenosquamous carcinoma, tumor size less than or equal to 2 cm, stromal invasion less than 10 mm, and no lymphovascular space invasion.20Y22 Radical trachelectomy is a common fertility-sparing surgery, which most guidelines recommend 2 cm as the cutoff value. Therefore, accurate assessment of the primary tumor size before surgery is a crucial significance in identifying the low-risk patients with early-stage cervical cancer. However, there were few studies focused on the preoperative assessment of tumor size less than or equal to 2 cm. Therefore, to find a more accurate measuring method for the preoperative assessment of tumor size less than or equal to 2 cm, we compared the accuracy of MRI and pelvic examination in measuring tumor diameter less than or equal to 2 cm. However, in measurement of tumor size less than or equal to 2 cm, the error range of 0.5 to 1.0 cm seems too large to be TABLE 5. Degree of agreement of MRI and pelvic examination with postoperative measurement and pathological findings Tumor Size, cm e1 1Y2 2Y3 3Y4 Q4 Measure of agreement J value
Pelvic Postoperative MRI Examination Examination 7 19 32 31 12 0.88
41 16 22 14 8 0.07
5 27 26 30 13
* 2014 IGCS and ESGO
Copyright © 2014 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.
International Journal of Gynecological Cancer
& Volume 24, Number 7, September 2014
MRI in Evaluation of Cervical Cancer
FIGURE 3. The correlation between the measurements of MRI and pelvic examination based on postoperative measurement and pathologic findings. A, Scatter gram between the measurements estimated by MRI and postoperative measurement and pathologic findings in patients with cervical cancer. B, Scatter gram between the measurements estimated by pelvic examination and postoperative measurement and pathologic findings in patients with cervical cancer. clinically unacceptable. Thus, we compared the accuracy of MRI and pelvic examination in the group with tumor size less than or equal to 2 cm with a narrow difference of 0.3 cm. The accuracy of MRI was significantly higher than that of pelvic examination with a difference of 0.3 cm in the group with tumor size less than or equal to 2 cm (24.14% vs 11.49%, P G 0.001). These results suggest that MRI is more precise than pelvic examination in the measurement of tumor size when tumor size is less than or equal to 2 cm. This suggests that MRI is helpful in planning for fertility-sparing surgery such radical trachelectomy. Few studies had evaluated the accuracy of MRI and pelvic examination in measuring tumor size with different growth pattern. The present study divided all patients with measurable tumor into exophytic tumor group and endocervical tumor group based on pathological findings. The accuracy of MRI in measuring tumor size was significantly higher than that of pelvic examination both in exophytic and endocervical tumor. Those findings suggest again that MRI is a better choice for the evaluation of cervical tumor size, especially for endocervical tumors.
The Bland-Altman plot is a statistic approach used in clinical measurement comparison of a new measurement technique with an established one to see whether they agree sufficiently for the new one to replace the old one.23 We used the Bland-Altman plot to estimate the differences between the tumor diameters of different methods and to identify possible outliers. The result also suggested that MRI agreed with postoperative measurement was better than clinical examination with a smaller difference (0.2 vs 1.1 cm). Magnetic resonance imaging may also be used to evaluate the status of lymph node and parametrial invasion, which is very difficult to be detected in clinical examination, so these 2 parameters were taken into consideration in this study. The sensitivity, specificity, accuracy, PPV, and NPV of MRI in detecting lymph node metastasis have been reported to be 24% to 89%,6,24 78% to 99%,6,24 67% to 86%,6,25 47.1% to 78%,24,26 and 57% to 94.4%,6,26 respectively. Results of this study are consistent with previous studies except for PPV, which may be influenced by the low morbidity (18/125) of lymph node metastasis. Some studies recommended the use
FIGURE 4. The Bland-Altman plots. A, Bland-Altman plots between the measurements estimated by MRI and pathologic findings. B, Bland-Altman plots between the measurements estimated by pelvic examination and pathologic findings. * 2014 IGCS and ESGO
Copyright © 2014 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.
1297
International Journal of Gynecological Cancer
Zhang et al
of functional MRI in the evaluation of cervical cancer in detecting suspicious metastatic lymph node.8,27 Therefore, we divided 125 cases into routine MRI group and MRI/DWI group to compare the accuracy of the 2 modalities in detecting lymph node metastasis. Similar with previous studies, MRI/DWI was proven to be better than routine MRI in lymph node detection (82.28% vs 69.56%) in our study. Because the sample size is too small, whether additional DWI can increase the detection of positive lymph node still needs to be confirmed by further study. The NPV of MRI in detecting parametrial invasion in our study is 100%, which was similar to another report,28 but had 6.4% false positive. This illustrates that the use of MRI in assessing parametrial invasion in cervical cancer is reasonable, but some limitations should be considered.
CONCLUSIONS Magnetic resonance imaging is an accurate noninvasive modality for preoperative evaluation of tumor size and also gives important information to parametrial invasion and lymph node status in patients with early-stage cervical cancer. Magnetic resonance imaging is a choice for preoperative evaluation of cervical cancer.
REFERENCES 1. Ferlay J, Shin HR, Bray F, et al. Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008. Int J Cancer. 2010;127:2893Y2917. 2. Amendola MA, Hricak H, Mitchell DG, et al. Utilization of diagnostic studies in the pretreatment evaluation of invasive cervical cancer in the United States: results of intergroup protocol ACRIN 6651/GOG 183. J Clin Oncol. 2005;23:7454Y7459. 3. Pecorelli S, Zigliani L, Odicino F. Revised FIGO staging for carcinoma of the cervix. Int J Gynaecol Obstet. 2009;105:107Y108. 4. Lagasse LD, Creasman WT, Shingleton HM, et al. Results and complications of operative staging in cervical cancer: experience of the Gynecologic Oncology Group. Gynecol Oncol. 1980;9:90Y98. 5. Hricak H, Yu KK. Radiology in invasive cervical cancer. AJR Am J Roentgenol. 1996;167:1101Y1108. 6. Petsuksiri J, Jaishuen A, Pattaranutaporn P, et al. Advanced imaging applications for locally advanced cervical cancer. Asian Pac J Cancer Prev. 2012;13:1713Y1718. 7. Pecorelli S. Revised FIGO staging for carcinoma of the vulva, cervix, and endometrium. Int J Gynaecol Obstet. 2009;105:103Y104. 8. Klerkx WM, Veldhuis WB, Spijkerboer AM, et al. The value of 3.0Tesla diffusion-weighted MRI for pelvic nodal staging in patients with early stage cervical cancer. Eur J Cancer. 2012;48:3414Y3421. 9. Loubeyre P, Navarria I, Undurraga M, et al. Is imaging relevant for treatment choice in early stage cervical uterine cancer? Surg Oncol. 2012;21:e1Ye6. 10. Griner PF, Mayewski RJ, Mushlin AI, et al. Selection and interpretation of diagnostic tests and procedures. Principles and applications. Ann Intern Med. 1981;94:557Y592. 11. Epstein E, Testa A, Gaurilcikas A, et al. Early-stage cervical cancer: tumor delineation by magnetic resonance imaging and
1298
12.
13.
14.
15. 16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
& Volume 24, Number 7, September 2014
ultrasoundVa European multicenter trial. Gynecol Oncol. 2013;128:449Y453. Shweel MA, Abdel-Gawad EA, Abdelghany HS, et al. Uterine cervical malignancy: diagnostic accuracy of MRI with histopathologic correlation. J Clin Imaging Sci. 2012;2:42. Kim SH, Lee HJ, Kim YW. Correlation between tumor size and surveillance of lymph node metastasis for IB and IIA cervical cancer by magnetic resonance images. Eur J Radiol. 2012;81:1945Y1950. Van Nagell JR Jr, Roddick JW Jr, Lowin DM. The staging of cervical cancer: inevitable discrepancies between clinical staging and pathologic findings. Am J Obstet Gynecol. 1971;110:973Y978. Averette HE, Ford JH Jr, Dudan RC, et al. Staging of cervical cancer. Clin Obstet Gynecol. 1975;18:215Y232. Lee YK, Han SS, Kim JW, et al. Value of pelvic examination and imaging modality for the evaluation of tumor size in cervical cancer. J Gynecol Oncol. 2008;19:108Y112. Alvarez RD, Potter ME, Soong SJ, et al. Rationale for using pathologic tumor dimensions and nodal status to subclassify surgically treated stage IB cervical cancer patients. Gynecol Oncol. 1991;43:108Y112. Hricak H, Lacey CG, Sandles LG, et al. Invasive cervical carcinoma: comparison of MR imaging and surgical findings. Radiology. 1988;166:623Y631. Palaia I, Musella A, Bellati F, et al. Simple extrafascial trachelectomy and pelvic bilateral lymphadenectomy in early stage cervical cancer. Gynecol Oncol. 2012;126:78Y81. Ramirez PT, Pareja R, Rendon GJ, et al. Management of low-risk early-stage cervical cancer: should conization, simple trachelectomy, or simple hysterectomy replace radical surgery as the new standard of care? Gynecol Oncol. 2014;132:254Y259. Kodama J, Fukushima C, Kusumoto T, et al. Stage IB1 cervical cancer patients with an MRI-measured tumor size G or = 2 cm might be candidates for less-radical surgery. Eur J Gynaecol Oncol. 2013;34:39Y41. Ramirez PT, Pareja R, Rendon GJ, et al. Management of low-risk early-stage cervical cancer: should conization, simple trachelectomy, or simple hysterectomy replace radical surgery as the new standard of care? Gynecol Oncol. 2014;132:254Y259. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986;1:307Y310. Kim SH, Han MC. Invasion of the urinary bladder by uterine cervical carcinoma: evaluation with MR imaging. AJR Am J Roentgenol. 1997;168:393Y397. Yang WT, Lam WW, Yu MY, et al. Comparison of dynamic helical CT and dynamic MR imaging in the evaluation of pelvic lymph nodes in cervical carcinoma. AJR Am J Roentgenol. 2000;175:759Y766. Kim SK, Choi HJ, Park SY, et al. Additional value of MR/PET fusion compared with PET/CT in the detection of lymph node metastases in cervical cancer patients. Eur J Cancer. 2009;45:2103Y2109. Kuang F, Ren J, Zhong Q, et al. The value of apparent diffusion coefficient in the assessment of cervical cancer. Eur Radiol. 2013;23:1050Y1058. Subak LL, Hricak H, Powell CB, et al. Cervical carcinoma: computed tomography and magnetic resonance imaging for preoperative staging. Obstet Gynecol. 1995;86:43Y50.
* 2014 IGCS and ESGO
Copyright © 2014 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.
The Role of Magnetic Resonance Imaging in Pretreatment Evaluation of Early-Stage Cervical Cancer Wei Zhang, MD,* Jie Zhang, MD,Þ Jiaxin Yang, MD,* Huadan Xue, MD,Þ Dongyan Cao, MD,* Huifang Huang, MD,* Ming Wu, MD,* Quancai Cui, MD,þ Jie Chen, MD,þ Jinghe Lang, MD,* and Keng Shen, MD* Objective: The aim of this study is to evaluate the accuracy of magnetic resonance imaging (MRI) in the preoperative assessments of primary tumor size, parametrial invasion, and pelvic lymph node metastasis in patients with early-stage cervical cancer. Materials and Methods: A cohort of 125 patients with International Federation of Gynecology and Obstetrics stage IA2 to IIA cervical cancer who had preoperative MRI and underwent radical hysterectomy were enrolled and analyzed. The accuracy of preoperative MRI scan and pelvic examination in the measurement of tumor size was assessed based on postoperative measurement and pathologic findings. The accuracy of detection of lymph node status and parametrial invasion was also assessed by comparing the MRI and pathologic findings. Results: The mean diameter of the tumor size measured by postoperative measurement, MRI, and pelvic examination was 2.97 T 1.39 cm, 2.78 T 1.24 cm, and 1.97 T 1.70 cm, respectively. There were significant differences in the mean diameter of the tumor size between pelvic examinations and MRI scan or postoperative measurement (P G 0.0001). Based on postoperative measurement findings, accuracy of tumor size measurement between pelvic examination and MRI was determined by the degree of agreement with a difference of less than 0.5 or 1.0 cm. Pelvic examination and MRI had an accuracy of 24.75% and 39.60%, respectively, with a difference of less than 0.5 cm, and had an accuracy of 43.56% and 61.39%, respectively, with a difference of less than 1.0 cm. Correlation with postoperative measurement in tumor size was higher for MRI (r[s] = 0.481) than that for pelvic examination (r[s] = 0.362). The sensitivity, specificity, and accuracy of MRI in detecting lymph node metastasis were 27.78%, 85.98%, 77.60%, respectively. The negative predictive value of MRI in detecting parametrial invasion is 100%. Conclusions: Magnetic resonance imaging is an accurate noninvasive modality for preoperative evaluation of tumor size and also gives important information to parametrial invasion and lymph node status in patients with early-stage cervical cancer. Key Words: Magnetic resonance imaging, Cervical cancer, Pelvic examination Received March 5, 2014, and in revised form April 1, 2014. Accepted for publication April 9, 2014. (Int J Gynecol Cancer 2014;24: 1292Y1298)
Departments of *Obstetrics and Gynecology, †Radiology, and ‡Pathology, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China. Address correspondence and reprint requests to Keng Shen, MD, Department of Obstetrics and Gynecology, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences and Copyright * 2014 by IGCS and ESGO ISSN: 1048-891X DOI: 10.1097/IGC.0000000000000169
1292
Peking Union Medical College, No.1. Shuaifuyuan, Dongcheng District, Beijing 100730, China. E-mail: [email protected]. Supported by grants from the National Science and Technology Infrastructure Program’s ‘‘The National Key Technologies R&D Program of China’’ (grant 2008BAI57B01), the National High Technology Research Development Program of China (863 program, grant 2012AA02A507), and the National Natural Science Foundation of China (grant 81172482 and grant 81372780). The authors declare no conflicts of interest.
International Journal of Gynecological Cancer
& Volume 24, Number 7, September 2014
Copyright © 2014 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.
International Journal of Gynecological Cancer
& Volume 24, Number 7, September 2014 1
cancer is the third most common cancer in women. C ervical Although screening programs have led to remarkable de-
crease of incidence and mortality, there is still high mortality in developing countries.1 The initial choice of treatment largely depends on the cancer stage at the time of diagnosis, and therefore, accurate staging is essential for appropriate treatment.2 The International Federation of Gynecology and Obstetrics (FIGO) recommends a clinical stage for cervical cancer based on clinical examination. However, clinical staging is less accurate than surgical staging.3,4 Pelvic examination is hard to accurately evaluate tumor size, parametrial invasion, and pelvic side wall invasion, and also cannot evaluate lymph node status. Although the revised system did not include imaging, FIGO now encourages the use of diagnostic imaging techniques to assess the size of the primary tumor, parametrial invasion, pelvic side wall invasion, adjacent organ invasion, and lymph node metastases.3 Magnetic resonance imaging (MRI), with good soft tissue contrast and multiplanar imaging capability, is now widely accepted as an optimal method for evaluation of gynecologic malignancies including cervical cancer.5 Magnetic resonance imaging seems to be better than computed tomography for locoregional disease assessment, especially for primary tumor and adjacent soft tissue extension.6 Some studies have demonstrated that MRI examination proves to be better than clinical examination in staging of cervical cancer,7,8 but other studies shown opposite conclusions.9,10 More and more studies are focusing on management of low-risk early-stage cervical cancer in which tumor size is an important criterion.11,12 Therefore, accurate preoperative evaluation of tumor size is very important. Although many studies have compared the staging of MRI to clinical examination, few of them evaluated their accuracy of tumor size measurement. The aim of this study was to assess the accuracy of MRI in preoperative measurement of primary tumor size and in detecting parametrial invasion and pelvic lymph node metastasis of patients with early-stage cervical cancer. The role of MRI in pretreatment evaluation of early-stage cervical cancer will be discussed.
MATERIALS AND METHODS Patients A cohort of 125 patients with FIGO stage IA2 to IIA cervical cancer who underwent radical hysterectomy and pelvic and/or para-aortic lymphadenectomy at the Department of Obstetrics and Gynecology at the Peking Union Medical College Hospital between September 2009 and December 2013 were enrolled and analyzed. All patients have pathologically confirmed diagnosis of cervical cancer and were staged according to FIGO criteria.7 The MRI examinations of the pelvis and lower abdomen within 2 weeks before surgery were carried out for every patient. Clinical examination and FIGO staging were performed by 2 experienced gynecologic oncologists. Radical hysterectomy and pelvic and/or para-aortic lymphadenectomy were performed by the same experienced gynecologic oncologists. Histopathologic diagnosis was made and reviewed by 2 experienced pathologists. Two independent radiologists with extensive backgrounds in gynecological MRI,
MRI in Evaluation of Cervical Cancer
who were blinded to the patient’s outcomes, reviewed all MRI slides and made a diagnosis according to the MRI findings. There was no difference between the 2 pathologists’ comments. Three patients had differences in the diagnosis of lymph node metastasis with the 2 MRI reviewers, and we adopted the views of the third expert in gynecological MRI. The clinical data, including clinical and pathological variables, preoperative MRI information, and surgical findings, were evaluated. The study protocol was approved by the ethics committee at the Peking Union Medical College Hospital, Beijing, China.
Preoperative Assessment All patients were admitted to the hospital for preoperative workup 3 days before the operation. Pelvic examination was carried out by 2 experienced gynecologic oncologists. The vagina and cervix were visualized by specula and inspection. The visible cervical tumor was measured by a scale, and the maximal tumor size was recorded in the medical files. Both of the parametrium were palpated by rectal-vaginal examination, and parametrial invasion was suspected when the parametrium was found with nodular thickness. If the pelvic examination was unsatisfactory, examination under anesthesia was considered. In the current study, 10 patients in all underwent examination under anesthesia.
MRI Examination Magnetic resonance imaging was performed with a 3.0-T MR scanner (HD750; General Electric Healthcare, Milwaukee, WI) with an 8-channel body array coil. The scan range was from the level of the umbilicus to the pubic symphysis. Routine MRI protocol used for the assessment of cervical cancer was performed as follows: the axial spin-echo T1-weighted imaging and the transverse, sagittal, and coronal fast spin-echo T2weighted imaging. The parameter details for routine MRI protocols were as follows: T1-weighted image (repetition time [TR]/echo time [TE], 400Y600/10Y20 ms; field of view (FOV), 30 30 cm; matrix, 256 256; slice thickness/gap, 4/1 mm) and T2-weighted image (TR/TE, 2500Y4000/60Y100 ms; FOV, 30 30 cm; matrix, 256 256; slice thickness/gap, 4/1 mm). Transverse diffusion-weighted images (DWIs) were also obtained. Imaging parameter details for DWI were as follows: b = 0.800 s/mm2; TR/TE, 4500/62.4 ms; FOV, 38 38 cm; matrix, 128 128; slice thickness/gap, 4/1 mm. T2-weighted images were used to determine the primary tumor size, location, and its extension to the parametria. The sagittal T2-weighted image was used to measure the tumor’s maximum craniocaudal and anteroposterior diameters, and the transverse T2-weighted image was used to measure the maximum latero-lateral diameter to the long axis of the cervix (Fig. 1). The highest b value diffusion-weighted data set and the T2-weighted data set were used to facilitate the detection of pelvic lymph nodes.8 Lymph nodes greater than 10 mm in short diameter were considered abnormal.9 The discontinuity of low signal in the outermost ring of the cervical stroma on MRI is considered as parametrial invasion.
Postoperative Measurement and Pathological Examination All specimens of the radical hysterectomy were carefully examined at the operation theater by operators after the
* 2014 IGCS and ESGO
Copyright © 2014 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.
1293
Zhang et al
International Journal of Gynecological Cancer
& Volume 24, Number 7, September 2014
FIGURE 1. A, T2-weighted sagittal image revealing a mass with a size of 4.01 cm (largest tumor diameter). B, T2-weighted axial imaging revealing invasion of the parametrium (arrow). C and D, T2-weighted axial image and DWI revealing a lymph node in the right external iliac area (arrow).The lymph node detected on MRI is a metastasis. completion of radical hysterectomy immediately. Opening the uterus and exposing the cervix, each visible cervical tumor was measured by a scale, and maximal tumor size and location of lesion were recorded (Fig. 2). All specimens of the lymphadenectomy and parametrical tissues were also palpated and examined carefully, and enlarged lymph nodes were selected for frozen section. All the specimens of radical hysterectomy were fixated in 10% buffered formalin after visible inspection for pathologic examination. Each visible tumor was described at macroscopic examination, including maximal tumor size, location of lesion, distance from margins, and gross depth of invasion. These findings were compared with those obtained from microscopic examination. Each specimen was sectioned transversally (the thickness of each transversal was approximately 3 mm). Other parameters assessed microscopically were depth of stromal invasion, parametrial involvement, and lymph node metastasis.
methods and to identify possible outliers. The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated by standard procedures.10 BlandAltman plot was performed using MedCalc version 11.4.2.0.
Statistics The maximal tumor diameter measured by pelvic examination and MRI was compared based on the maximal tumor diameter measured by pathological examination. McNemar test was used to compare the accuracy of pelvic examination and MRI in the measurement of tumor diameter. Agreement analysis and spearman correlation coefficient were used to determine the agreement of tumor size between different methods to pathological findings. The Bland-Altman plot was used to estimate the differences between the tumor diameters of different
1294
FIGURE 2. Postoperative measurement of cervical tumor diameter (4.20 cm). * 2014 IGCS and ESGO
Copyright © 2014 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.
International Journal of Gynecological Cancer
& Volume 24, Number 7, September 2014
TABLE 1. Characteristics of patients
TABLE 3. Accuracy of pelvic examination and MRI in the measurement of tumor size less than or equal to 2 cm with different error range
Characteristics 42.51 T 9.20 25Y72
Mean age, y Range, y FIGO stage IA2 IB1 IB2 IIA1 IIA2 Histology Squamous cell carcinoma Adenocarcinoma Others Mean tumor diameter, cm Range, cm Tumor location Exophytic Endocervical Imaging modality Routine MRI DW-MRI
13 97 6 6 3
(10.4%) (77.6%) (4.8%) (4.8%) (2.4%)
91 (72.8%) 25 (20%) 9 (7.2%) 2.97 T 1.39 0.7Y7 86 (85.1%) 15 (14.9%) 46 79
The other calculations were performed using SPSS software version 13.0. P G 0.05 was considered statistically significant.
RESULT A total 125 women with early-stage cervical cancer (IA2-IIA) were included in this study. Age, FIGO stage, histological type, tumor growth pattern, and mean maximal tumor diameter on postoperative inspection and final histological report are shown in Table 1. Except for 13 cases with stage IA2 and 11 cases with stage IB1 without macroscopic lesions, a total of 101 cases had visible cervical tumor including 86 cases with stage IB1, 6 with stage IB2, 6 with stage IIA1, and 3 with stage IIA2. The tumor size measured by pelvic examination ranged from 0 to 6.0 cm, and tumor size measured by MRI ranged from 0 to 6.55 cm.
TABLE 2. Accuracy of pelvic examination and MRI for the measurement of tumor diameter based on postoperative measurement Accuracy
Pelvic examination MRI *P G 0.001.
MRI in Evaluation of Cervical Cancer
T0.5 cm
T1.0 cm
24.75% 39.60%*
43.56% 61.39%*
Different Error Range
Pelvic examination MRI
T0.3 cm
T0.5 cm
T1.0 cm
11.49% 24.14%*
25.00% 40.63%
56.25% 65.63%
*P G 0.001.
The mean maximal tumor diameter was not significantly different between MRI (2.78 T 1.24 cm) and postoperative measurement (2.97 T 1.39 cm; P = 0.15). The difference between pelvic examinations (1.97 T 1.70 cm) and postoperative measurement (2.97 T 1.39 cm) was significant (P G 0.0001). Table 2 describes the accuracy of pelvic examination and MRI in the measurement of tumor diameter. The accuracy was estimated by the degree of agreement with a difference of 0.5 or 1.0 cm based on postoperative measurement and pathologic findings. The accuracy of MRI was significantly higher than that of pelvic examination with a difference of both 0.5 cm (39.60% vs 24.75%, P G 0.001) and 1.0 cm (61.39% vs 43.56%, P G 0.001). The accuracy of pelvic examination and MRI in the measurement of tumor size of less than or equal to 2 cm is showed in Table 3. The accuracy was also estimated by the degree of agreement with a difference of 0.5 or 1.0 cm based on postoperative measurement and pathologic findings. Although the accuracy of MRI were higher than that of pelvic examination, the difference was not significant with a difference of 0.5 cm (40.63% vs 25.0%, P = 0.052) and 1.0 cm (65.63% vs 56.25%, P = 0.25). To evaluate which measuring method is more accurate, a narrow difference of 0.3 cm was performed. The accuracy of MRI was significantly higher than that of pelvic examination with a difference of 0.3 cm in the group with tumor size less than or equal to 2 cm (24.14% vs 11.49%, P G 0.001). The accuracy of different methods in measuring tumor size in different growth pattern is shown in Table 4. The accuracy of MRI was significantly higher than that of pelvic examination with a difference of both 0.5 cm (exophytic: 40.70% vs 29.07%, P = 0.004; endocervical: 33.33% vs 0.00%, P = 0.002) and 1.0 cm (exophytic: 61.63% vs 48.84%, P = 0.001; endocervical: 60% vs 0.00%, P = 0.004) in each growth pattern. The degree of agreement of MRI and pelvic examination with postoperative measurement and pathological findings is shown in Table 5. The J value represents the degree of agreement with postoperative measurement and pathologic findings. The agreement of MRI with postoperative measurement is much higher than that of the pelvic examination (J value 0.88 vs 0.07). Correlation between postoperative measurement was higher for MRI than that for pelvic examination (r = 0.481 vs r = 0.362; Fig. 3). Bland-Altman plots (Fig. 4) also show that the difference of mean tumor diameters between MRI and postoperative measurement (0.5 cm) was smaller than that of pelvic examination with postoperative measurement and pathological findings (1.0 cm).
* 2014 IGCS and ESGO
Copyright © 2014 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.
1295
International Journal of Gynecological Cancer
Zhang et al
TABLE 4. Accuracy of different growth pattern of pelvic examination and MRI for the measurement of tumor size based on postoperative measurement Exophytic Accuracy with a difference of T0.3 cm Pelvic examination 15.12% MRI 26.74%* Accuracy with a difference of T0.5 cm Pelvic examination 29.07% MRI 40.70%† Accuracy with a difference of T1.0 cm Pelvic examination 48.84% MRI 61.63%†
Endocervical 0.00% 13.33%* 0.00% 33.33%† 0.00% 60%†
*P G 0.001. †P G 0.05.
Lymph node statuses of all the 125 patients were assessed by MRI before surgery including 79 cases by DWI in addition to routine MRI (DW-MRI) and 46 patients by routine MRI. Suspicions of lymph node metastases were detected in 20 patients, 13 cases were detected by DW-MRI, and 7 cases were suspected by routine MRI. Lymphadenectomy were performed on all of the 125 patients. Totally, 18 cases were confirmed to have lymph node metastasis by the following pathology, and only 5 of them were also suspected lymph node metastases on MRI examination. The remaining 13 cases (72.22%) were false negative on MRI. In 20 patients with MRI-suspected lymph node metastasis, only 5 of them were confirmed to have lymph node metastasis by pathology, and the other 15 cases (75%) were false positive. In 13 cases that suspected lymph node metastasis with DW-MRI, 4 cases were pathologically confirmed lymph node metastases. The routine MRI showed lymph node metastasis in 7 patients, only 1 of them was pathologically confirmed. The sensitivity, specificity, accuracy, PPV, and NPV of MRI in detecting lymph node metastasis was 27.78%, 85.78%, 77.60%, 25.0%, and 87.62%, respectively. The NPV of MRI in detecting parametrial invasion in our study is 100% but had 6.4% false positive.
DISCUSSION The FIGO stage for cervical cancer is actually a clinical stage based on pelvic examination.3 Tumor size is closely associated with FIGO stage and is also an important prognostic factor. Exact evaluation of tumor size played a very important role in making clinical decision and predicting the prognosis for patients with cervical cancer. International Federation of Gynecology and Obstetrics also recommended that the diagnostic imaging techniques can be used to assess the size of primary tumor, such as MRI/computed tomography scanning.3 Some studies have performed MRI assessment of tumor size, and most of them compared the accuracy of the tumor size measurement of MRI to other imaging techniques or pathological findings,11Y13 but there are still some inevitable discrepancies between clinical staging and pathologic findings.14,15 To
1296
& Volume 24, Number 7, September 2014
assess the value of MRI in preoperative evaluation of cervical cancer, we compared the accuracy of MRI and clinical examination in preoperative measurement of primary tumor size based on the postoperative measurement and pathologic findings. It has been demonstrated that the accuracy of pelvic examination in the measurement of primary tumor size of cervical cancer is approximately around 50% with a difference of 0.5 cm16 and 50% to 72% with a difference of less than 1.0 cm.16,17 The accuracy of MRI in the measurement of primary tumor size ranged from 39.8%16 to 70%18 with a difference of 0.5 cm and 55.3% with a difference of 1.0 cm.16 In comparison with other studies, we also used the difference of 0.5 or 1.0 cm as an error range to evaluate the measurement accuracy. The accuracy of MRI in preoperative measurement of the primary tumor size is 39.60% (with error range T0.5 cm) and 61.39% (with error range T1.0 cm), whereas the accuracy of pelvic examination is 24.75% (T0.5 cm) and 43.56% (T1.0 cm), respectively. There was a significant difference in measurement accuracy between MRI and pelvic examination, which was similar to previous studies.16Y20 Our study demonstrated again that the accuracy of MRI is higher than that of pelvic examination. With the valuable-based medicine introduced and the individualization treatment developed, more and more studies are focusing on management of low-risk early-stage cervical cancer.19Y21 Definition of low-risk patients with cervical cancer includes the following criteria: squamous carcinoma, adenocarcinoma or adenosquamous carcinoma, tumor size less than or equal to 2 cm, stromal invasion less than 10 mm, and no lymphovascular space invasion.20Y22 Radical trachelectomy is a common fertility-sparing surgery, which most guidelines recommend 2 cm as the cutoff value. Therefore, accurate assessment of the primary tumor size before surgery is a crucial significance in identifying the low-risk patients with early-stage cervical cancer. However, there were few studies focused on the preoperative assessment of tumor size less than or equal to 2 cm. Therefore, to find a more accurate measuring method for the preoperative assessment of tumor size less than or equal to 2 cm, we compared the accuracy of MRI and pelvic examination in measuring tumor diameter less than or equal to 2 cm. However, in measurement of tumor size less than or equal to 2 cm, the error range of 0.5 to 1.0 cm seems too large to be TABLE 5. Degree of agreement of MRI and pelvic examination with postoperative measurement and pathological findings Tumor Size, cm e1 1Y2 2Y3 3Y4 Q4 Measure of agreement J value
Pelvic Postoperative MRI Examination Examination 7 19 32 31 12 0.88
41 16 22 14 8 0.07
5 27 26 30 13
* 2014 IGCS and ESGO
Copyright © 2014 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.
International Journal of Gynecological Cancer
& Volume 24, Number 7, September 2014
MRI in Evaluation of Cervical Cancer
FIGURE 3. The correlation between the measurements of MRI and pelvic examination based on postoperative measurement and pathologic findings. A, Scatter gram between the measurements estimated by MRI and postoperative measurement and pathologic findings in patients with cervical cancer. B, Scatter gram between the measurements estimated by pelvic examination and postoperative measurement and pathologic findings in patients with cervical cancer. clinically unacceptable. Thus, we compared the accuracy of MRI and pelvic examination in the group with tumor size less than or equal to 2 cm with a narrow difference of 0.3 cm. The accuracy of MRI was significantly higher than that of pelvic examination with a difference of 0.3 cm in the group with tumor size less than or equal to 2 cm (24.14% vs 11.49%, P G 0.001). These results suggest that MRI is more precise than pelvic examination in the measurement of tumor size when tumor size is less than or equal to 2 cm. This suggests that MRI is helpful in planning for fertility-sparing surgery such radical trachelectomy. Few studies had evaluated the accuracy of MRI and pelvic examination in measuring tumor size with different growth pattern. The present study divided all patients with measurable tumor into exophytic tumor group and endocervical tumor group based on pathological findings. The accuracy of MRI in measuring tumor size was significantly higher than that of pelvic examination both in exophytic and endocervical tumor. Those findings suggest again that MRI is a better choice for the evaluation of cervical tumor size, especially for endocervical tumors.
The Bland-Altman plot is a statistic approach used in clinical measurement comparison of a new measurement technique with an established one to see whether they agree sufficiently for the new one to replace the old one.23 We used the Bland-Altman plot to estimate the differences between the tumor diameters of different methods and to identify possible outliers. The result also suggested that MRI agreed with postoperative measurement was better than clinical examination with a smaller difference (0.2 vs 1.1 cm). Magnetic resonance imaging may also be used to evaluate the status of lymph node and parametrial invasion, which is very difficult to be detected in clinical examination, so these 2 parameters were taken into consideration in this study. The sensitivity, specificity, accuracy, PPV, and NPV of MRI in detecting lymph node metastasis have been reported to be 24% to 89%,6,24 78% to 99%,6,24 67% to 86%,6,25 47.1% to 78%,24,26 and 57% to 94.4%,6,26 respectively. Results of this study are consistent with previous studies except for PPV, which may be influenced by the low morbidity (18/125) of lymph node metastasis. Some studies recommended the use
FIGURE 4. The Bland-Altman plots. A, Bland-Altman plots between the measurements estimated by MRI and pathologic findings. B, Bland-Altman plots between the measurements estimated by pelvic examination and pathologic findings. * 2014 IGCS and ESGO
Copyright © 2014 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.
1297
International Journal of Gynecological Cancer
Zhang et al
of functional MRI in the evaluation of cervical cancer in detecting suspicious metastatic lymph node.8,27 Therefore, we divided 125 cases into routine MRI group and MRI/DWI group to compare the accuracy of the 2 modalities in detecting lymph node metastasis. Similar with previous studies, MRI/DWI was proven to be better than routine MRI in lymph node detection (82.28% vs 69.56%) in our study. Because the sample size is too small, whether additional DWI can increase the detection of positive lymph node still needs to be confirmed by further study. The NPV of MRI in detecting parametrial invasion in our study is 100%, which was similar to another report,28 but had 6.4% false positive. This illustrates that the use of MRI in assessing parametrial invasion in cervical cancer is reasonable, but some limitations should be considered.
CONCLUSIONS Magnetic resonance imaging is an accurate noninvasive modality for preoperative evaluation of tumor size and also gives important information to parametrial invasion and lymph node status in patients with early-stage cervical cancer. Magnetic resonance imaging is a choice for preoperative evaluation of cervical cancer.
REFERENCES 1. Ferlay J, Shin HR, Bray F, et al. Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008. Int J Cancer. 2010;127:2893Y2917. 2. Amendola MA, Hricak H, Mitchell DG, et al. Utilization of diagnostic studies in the pretreatment evaluation of invasive cervical cancer in the United States: results of intergroup protocol ACRIN 6651/GOG 183. J Clin Oncol. 2005;23:7454Y7459. 3. Pecorelli S, Zigliani L, Odicino F. Revised FIGO staging for carcinoma of the cervix. Int J Gynaecol Obstet. 2009;105:107Y108. 4. Lagasse LD, Creasman WT, Shingleton HM, et al. Results and complications of operative staging in cervical cancer: experience of the Gynecologic Oncology Group. Gynecol Oncol. 1980;9:90Y98. 5. Hricak H, Yu KK. Radiology in invasive cervical cancer. AJR Am J Roentgenol. 1996;167:1101Y1108. 6. Petsuksiri J, Jaishuen A, Pattaranutaporn P, et al. Advanced imaging applications for locally advanced cervical cancer. Asian Pac J Cancer Prev. 2012;13:1713Y1718. 7. Pecorelli S. Revised FIGO staging for carcinoma of the vulva, cervix, and endometrium. Int J Gynaecol Obstet. 2009;105:103Y104. 8. Klerkx WM, Veldhuis WB, Spijkerboer AM, et al. The value of 3.0Tesla diffusion-weighted MRI for pelvic nodal staging in patients with early stage cervical cancer. Eur J Cancer. 2012;48:3414Y3421. 9. Loubeyre P, Navarria I, Undurraga M, et al. Is imaging relevant for treatment choice in early stage cervical uterine cancer? Surg Oncol. 2012;21:e1Ye6. 10. Griner PF, Mayewski RJ, Mushlin AI, et al. Selection and interpretation of diagnostic tests and procedures. Principles and applications. Ann Intern Med. 1981;94:557Y592. 11. Epstein E, Testa A, Gaurilcikas A, et al. Early-stage cervical cancer: tumor delineation by magnetic resonance imaging and
1298
12.
13.
14.
15. 16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
& Volume 24, Number 7, September 2014
ultrasoundVa European multicenter trial. Gynecol Oncol. 2013;128:449Y453. Shweel MA, Abdel-Gawad EA, Abdelghany HS, et al. Uterine cervical malignancy: diagnostic accuracy of MRI with histopathologic correlation. J Clin Imaging Sci. 2012;2:42. Kim SH, Lee HJ, Kim YW. Correlation between tumor size and surveillance of lymph node metastasis for IB and IIA cervical cancer by magnetic resonance images. Eur J Radiol. 2012;81:1945Y1950. Van Nagell JR Jr, Roddick JW Jr, Lowin DM. The staging of cervical cancer: inevitable discrepancies between clinical staging and pathologic findings. Am J Obstet Gynecol. 1971;110:973Y978. Averette HE, Ford JH Jr, Dudan RC, et al. Staging of cervical cancer. Clin Obstet Gynecol. 1975;18:215Y232. Lee YK, Han SS, Kim JW, et al. Value of pelvic examination and imaging modality for the evaluation of tumor size in cervical cancer. J Gynecol Oncol. 2008;19:108Y112. Alvarez RD, Potter ME, Soong SJ, et al. Rationale for using pathologic tumor dimensions and nodal status to subclassify surgically treated stage IB cervical cancer patients. Gynecol Oncol. 1991;43:108Y112. Hricak H, Lacey CG, Sandles LG, et al. Invasive cervical carcinoma: comparison of MR imaging and surgical findings. Radiology. 1988;166:623Y631. Palaia I, Musella A, Bellati F, et al. Simple extrafascial trachelectomy and pelvic bilateral lymphadenectomy in early stage cervical cancer. Gynecol Oncol. 2012;126:78Y81. Ramirez PT, Pareja R, Rendon GJ, et al. Management of low-risk early-stage cervical cancer: should conization, simple trachelectomy, or simple hysterectomy replace radical surgery as the new standard of care? Gynecol Oncol. 2014;132:254Y259. Kodama J, Fukushima C, Kusumoto T, et al. Stage IB1 cervical cancer patients with an MRI-measured tumor size G or = 2 cm might be candidates for less-radical surgery. Eur J Gynaecol Oncol. 2013;34:39Y41. Ramirez PT, Pareja R, Rendon GJ, et al. Management of low-risk early-stage cervical cancer: should conization, simple trachelectomy, or simple hysterectomy replace radical surgery as the new standard of care? Gynecol Oncol. 2014;132:254Y259. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986;1:307Y310. Kim SH, Han MC. Invasion of the urinary bladder by uterine cervical carcinoma: evaluation with MR imaging. AJR Am J Roentgenol. 1997;168:393Y397. Yang WT, Lam WW, Yu MY, et al. Comparison of dynamic helical CT and dynamic MR imaging in the evaluation of pelvic lymph nodes in cervical carcinoma. AJR Am J Roentgenol. 2000;175:759Y766. Kim SK, Choi HJ, Park SY, et al. Additional value of MR/PET fusion compared with PET/CT in the detection of lymph node metastases in cervical cancer patients. Eur J Cancer. 2009;45:2103Y2109. Kuang F, Ren J, Zhong Q, et al. The value of apparent diffusion coefficient in the assessment of cervical cancer. Eur Radiol. 2013;23:1050Y1058. Subak LL, Hricak H, Powell CB, et al. Cervical carcinoma: computed tomography and magnetic resonance imaging for preoperative staging. Obstet Gynecol. 1995;86:43Y50.
* 2014 IGCS and ESGO
Copyright © 2014 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.
