Japanese Journal of Clinical Oncology Advance Access published November 24, 2007

Transcription

1 Original Article Japanese Journal of Clinical Oncology Advance Access published November 24, 2007 Jpn J Clin Oncol doi: /jjco/hym109 Tumor Volume and Uterine Body Invasion Assessed by MRI for Prediction of Outcome in Cervical Carcinoma Treated with Concurrent Chemotherapy and Radiotherapy Hunjung Kim 1, Woochul Kim 2, Mijo Lee 2, Eunseop Song 3 and John J. K. Loh 2 1 Department of Radiation Oncology, Dongguk University Medical College, Gyongju, 2 Department of Radiation Oncology and 3 Department of Obstetrics and Gynecology, Inha University Medical College, Inchon, Korea Received April 3, 2007; accepted July 9, 2007 Objective: The aim of this study was to evaluate the prognostic significance of primary tumor volume and uterine body invasion assessed by pre-treatment MRI for uterine cervical cancer patient treated with concurrent chemotherapy and radiotherapy. Methods: A retrospective analysis of 106 patients with IB IIIB cervical carcinoma was performed. Potential prognostic factors were stage, clinical tumor diameter, histology, age, pelvic lymph node, vaginal extension, parametrial invasion, tumor volume and uterine body invasion status. Multivariate analyses were performed to identify the prognostic factor for overall survival (OS) and disease-free survival (DFS). Results: The 5-year OS, DFS rate were 59.7 and 56.6%. Using multivariate analyses, a large tumor volume (30 ml; P ¼ 0.012) and uterine body invasion (P ¼ 0.020) and positive pelvic lymph node (LN) enlargement (P ¼ 0.040) showed a significantly unfavorable influence on OS. Using these three factors, patients were divided into four subgroups: the OS rates of patients with risk 0 (volume,30 ml, no uterine body invasion, and negative LN), risk 1 (one of these three factors), risk 2 (two of these three factors) and risk 3 (volume 30 ml, uterine body invasion, and positive LN) were 96.3, 77.5, 53.0 and 14.8%, respectively (P, ). Conclusions: Tumor volume and uterine body invasion determined by MRI were significant prognostic factors for patients with cervical carcinoma. Pelvic lymph node enlargement diagnosed by CT also proved to be a significant prognostic factor in OS. Using these three parameters, we devised a practical and effective model to predict OS. The prognostic significance of primary tumor volume and uterine corpus invasion from MRI were evaluated from 106 patients with stage IB-IIIB uterine cervical cancer treated by RT. A large volume (30 ml) and corpus invasion showed a significantly unfavorable influence on OS. Key words: cervical cancer MRI tumor volume uterine body invasion INTRODUCTION Size and uterine body invasion of a primary tumor in surgical specimens are prognostically significant factors in determining relapse and survival of patients with invasive cervical carcinoma. In large sugicopathologic studies of patients with FIGO stage IB cervical cancer, increasing tumor size correlated with lymph node metastasis and shorter disease-free For reprints and all correspondence: Hunjung Kim, Department of Radiation Oncology, Dongguk University Gyongju Hospital, Sukjang-Dong, Gyongju-Si, Kyong-Buk , Korea. thinkonco@paran.com survival (1 5). In another study of early cervical cancer patients treated by radical hysterectomy and lymphadenectomy, survival was more closely related to tumor volume than to clinical or pathologic stage (6,7). On the other hand, Perez and colleagues showed that endometrial invasion by cervical cancer was a significant negative prognostic factor associated with a 10 20% decrease in overall survival among irradiated patients of the same stage (8). It was also associated with increased risk of distant metastatic disease (9). Surgical studies have also reported that uterine body invasion seen on the radical hysterectomy specimen is associated with a substantial risk of pelvic lymph node # 2007 Foundation for Promotion of Cancer Research

2 Page 2 of 9 MRI tumor volume and corpus invasion of cervical Ca metastases and a decreased 5-year survival (10). Endometrial extension in pathologic evaluation of the endometrium by biopsy or dilation and currettage correlated strongly with risk of FEG-PET detected lymph node metastases and poor prognosis (11). Although clinical stageisalsoanimportant factor, stage does not necessarily correlate with size and uterine body invasion of the primary tumor. However, an accurate assessment of tumor size and extension of the tumor in an inoperable patient who is candidate for radiotherapy may not always be optimal due to method limitations of pelvic examination and computed tomography (CT). Furthermore, clinical palpation is a subjective method and has significant interobserver variability (12 14). With the advent of magnetic resonance imaging (MRI), it has become possible to estimate the size and extension of primary tumors more accurately than was prior possible by clinical palpation in an inoperable patient (15 19). Tumor size and local invasion in the surgical specimen correlates well with the corresponding diameter measured in T 2 -weighted MRI (20). MRI provides a noninvasive method for tumor size and extension evaluation in cervical cancer (21). Thus, MRI can be used for an accurate estimation of tumor volume and infiltration to surrounding normal tissue in cervical cancer patients who are planned to be treated by concurrent chemotherapy and radiotherapy. In the present study, we have evaluated, in a retrospective data series from cervical carcinoma patients who underwent a curative course of concurrent chemotherapy and radiotherapy, the prognostic significance of uterine body invasion and tumor volume determined by MRI, in the presence of other more traditional prognostic factors, such as FIGO stage, vaginal extension, invasion of parametrium, histology, age, clinical tumor diameter and pelvic lymph node enlargement. PATIENTS AND METHODS This was a retrospective study of patients with stage IB IIIB cervical cancer who received an MRI scan before the initiation of radiation therapy. Tumor measurements were carried out in 106 consecutive patients undergoing serial MRI and treatment with concurrent chemotherapy and radiotherapy for cervical cancer. PATIENT POPULATION One-hundred and six patients with newly diagnosed cervical cancer, referred for definitive radiation therapy between 1999 and 2003 were registered. Ethical approval for the study was obtained from the institutional ethics committee. All patients met the following criteria: FIGO clinical stages IB IIIB, a previously untreated cervical cancer and pretreatment assessment by pelvic MRI. Cervical biopsies were obtained at the time of diagnosis. CLINICAL ASSESSMENT Patients were initially evaluated by a medical history, physical examination and routine hematologic and serum chemistry laboratory studies. All patients had a pelvic examination under anesthesia and underwent routine imaging with chest radiography, pelvis MRI and CT scan of the abdomen and pelvis. Cystoscopy and/or proctoscopy were performed when clinically indicated. Each patient s disease was staged according to the FIGO classification system. Lymph node status was assessed by CT scan and staging laparotomy was not performed. Lymph nodes greater than 10 mm in minimum diameter were interpreted as positive nodes. Evaluations during RT consisted of a physical examination, including a pelvic examination. Other studies such as complete blood count and serum chemistries were performed weekly during radiation therapy in selected patients when indicated. TREATMENT POLICY Patients were treated by concurrent chemotherapy and radiation therapy with a curative intent. Patients received both external beam radiation therapy (EBRT) and intracavitary brachytherapy (ICRT). The treatment planning was based on X-ray simulation. The external beam radiation therapy (EBRT) was administered using an isocentric set-up and a four-field box technique on a 10 MV photon linear accelerator. Patients were treated in the prone position with a full bladder. The superior margin of the radiation field was the L5 S1 junction, the lower margin was the caudal pole of the obturator foramen, the lateral boundaries were 1.5 cm outside of pelvic sidewalls and S2 S3 posteriorly. A dose of 30.6 Gy was delivered in 5 days a week at 1.8 Gy per fraction. After this, a midline block was added up to 45 Gy using parallel opposed anterior and posterior ports. Parametrial or nodal boost, if required, consisted of 9 10 Gy in Gy fractions given in between the ICR fractions using EBRT to the involved side. Intracavitary radiotherapy was given using an Ir-192 HDR brachytherapy remote afterloading unit (Nucletron, Micor-selectron HDR, The Netherlands) following 30.6 Gy EBRT, once a week while receiving EBRT and twice a week after finishing EBRT to a total dose of 30 Gy prescribed at point A in six insertions of 5 Gy each. In addition, patients with stage IB IIIB also received weekly cisplatin (40 mg/m 2 ) during the course of their external beam radiotherapy. MR IMAGING ANALYSIS The MRI studies were performed on a 1.5 T scanner (Philips Gyroscan, Best, The Netherlands) using the system s body coil. All patients were scanned supine with abdominal compression to minimize respiratory motion artifacts. High-resolution (512 matrix) T 2 -weighted scans were performed through the pelvis in the axial, transverse and

3 Jpn J Clin Oncol Page 3 of 9 coronal planes. MRI was used solely to estimate the tumor volume and uterine body extension. T 2 -weighted image was used for determining extension into the uterine body of primary tumor. Primary tumor over the isthmus of the uterus in the axial plane was interpreted as uterine body invasion. The cervical tumor was identifiable as a high-signal intensity on T 2 -weighted images. The largest diameter of the tumor was determined on the transverse images, but where tumor extension was clearly large in the craniocaudal direction, it was assessed on the sagital plane. Tumor volume (V) was assessed, using three-dimensional volumetric measurements according to the modified Simpson rule. In all contiguous transverse images, the tumor mass was outlined on the computer monitor using dedicated postprocessing software developed at the Division of Image Processing of the Department of Radiology. The area of tumor (A) ineach slice was multiplied by the slice profile (5 mm slice thickness plus 1 mm intersection gap), and total tumor volume was automatically calculated by summation of the adjacent volume according to the formula: VðmlÞ ¼0:6 where i is the individual slice number and n is the total number of slices. FOLLOW-UP The patients received follow-up by the radiation oncologist every 1 2 months for the first 6 months, then every 3 months for the first 2 years, and then every 4 months thereafter. Follow-up evaluations included a history and physical examination, pelvic examination, Pap smear, complete blood count and serum chemistries. Chest X-rays were obtained yearly. CT scans of the abdomen and pelvis, MR images or bone scans were obtained if clinically indicated. STATISTICAL ANALYSIS The patterns of failure, disease-free survival (DFS), and overall survival (OS) were the major endpoints of this study. OS and DFS rates were measured from the date of completion of irradiation. Survival curves were estimated with the Kaplan Meier method and were compared with the Mantel Hansel log-rank test. DFS was defined for all deaths and recurrences as the event. Cox multiple logistic regressions were performed to evaluate independent factors contributing to tumor control, and significance was determined with the logistic likelihood ratio test. Xn i¼1 Ai! RESULTS One-hundred and six patients were investigated for the analysis. The mean follow-up time for patients was 38 months (range 4 84 months). CLINICAL CHARACTERISTICS Baseline patients and tumor-related characteristics such as the FIGO stage, the histologic type and tumor size are recorded in Table 1. The median age of the women investigated was 56 years and ranged from 30 to 85 years. Ninety-three patients (87.7%) had squamous cell carcinoma, and 47 patients (44.3%) exhibited pelvic lymph node involvement as seen on abdomen pelvic CT. The tumors of 68 (64.2%) patients showed uterine body invasion by MRI. The median tumor volume was 29 ml, and ranged from 5.0 to 120 ml. Vaginal extension by pelvic examination was detected in 97 of 108 all patients. Parametrial invasion by MRI was detected in 85 of 108 all patients. Parametiral invasion was suggested by MRI in three of the 21 women in whom parametrial invasion was not detected by pelvic examination. No parametrial invasion was detected by MRI Table 1. Patient characteristics Variable Level Number of patients All patients Age (years) Median (range) 56 (30 85) IB IIA FIGO stage IIB IIIA IIIB MRI invasion of No parametrium Yes Squamous cell Ca Histology adenoca adenosquamous cell Ca Clinical diameter (cm) Median (range) 4 ( ) Uterine body invasion No Yes Tumor volume (ml) Median (range) 29 (5 120) Pelvic lymph node No involve Yes Treatment Radiotherapy alone Radiotherapy þ chemotherapy FIGO, Federation of Gynecologists and Obstetricians; MRI, magnetic resonance imaging. %

4 Page 4 of 9 MRI tumor volume and corpus invasion of cervical Ca in three of the 85 patients in whom parametrial invasion detected by pelvic examination. Ninety-eight patients (92.5%) were treated with radical concurrent chemoradiotherapy. Eight patients in stage IB IIA who did not want chemotherapy, who had a small tumor diameter, or who could not receive chemotherapy due to medical problem, were treated with radical radiation therapy alone. ASSOCIATIONS BETWEEN FACTORS The relationships between FIGO stage, tumor volume and uterine body invasion are shown in Table 2. Uterine body invasion was exhibited in 33.3, 63.3 and 83.3% of patients of FIGO stage I, II and III, respectively, and were strongly correlated with FIGO stage (P ¼ 0.007). Uterine body invasion was exhibited in 44.3 and 86.7% of patients with a small tumor volume (,30 ml; n ¼ 27) and a large tumor volume (30 ml; n ¼ 39), respectively; and were strongly correlated with tumor volume as measured by MRI (P, 0.001). Tumor volume was also strongly correlated with FIGO stage (P, 0.001). The clinical diameter of tumor was strongly correlated with the FIGO stage (P, 0.001), tumor volume (P, 0.001), and uterine body invasion (P ¼ 0.004). Table 3 summarizes pelvic node positivity rates by FIGO stage, tumor volume and uterine body involvement. FIGO stage (P ¼ 0.001), tumor volume (P, 0.001), and uterine body invasion (P ¼ 0.005) were statistically significantly related to pelvic nodal involvement. Neither age (P ¼ 0.665) nor histology (P ¼ 0.594) was significantly related to uterine body invasion. PATTERNS OF FAILURE At the last follow-up, 31 patients had developed locoregional recurrence and 19 patients developed distant metastasis; six patients experienced both locoregional and distant metastasis. Table 2. Association between Uterien body invasion, FIGO stage and tumor volume Factor Uterine body invasion Uterine invasion (%) P-value No Yes FIGO stage P ¼ I II III Tumor volume (ml) P, 0.001, Table 3. Pelvic node positive rates by FIGO stage, tumor volume and uterine body invasion Factor Levels Uterine body invasion (%) No Yes FIGO stage I II III Tumor volume,30 ml ml OVERALL SURVIVAL For all 106 patients, the 5-year OS rate was 59.7% (95% confidence interval [CI] %). Table 4 gives estimated (Kaplan Meier) 5-year survival rates for various patient subgroups, and Figs 1 3 show Kaplan Meier plots of OS by uterine body invasion, tumor volume and FIGO stage. The 5-year survival rates according to FIGO stages I, II, and III were 88.8, 63.2 and 28.2%, respectively. By univariate analysis, increasing FIGO stage (P ¼ 0.002), larger clinical tumor diameter (P, 0.001), positive Table 4. Five-year OS for patient subgroup Factor Level N 5-Year OS %SE All patients Age, I FIGO stage II III Vaginal extension No Yes PE invasion of No parametrium Yes Histology SCC others Clinical tumor,5 cm diameter 5 cm Uterine body invasion No Tumor volume MRI invasion of parametrium Pelvic lymph node Yes ,30 ml ml No Yes Negative Positive PE, pelvic examination; SCC, squamous cell carcinoma.

5 Jpn J Clin Oncol Page 5 of 9 Figure 1. Overall survival curves of the groups divided by FIGO stage. FIGO, Federation of Gynecologists and Obstetricians. pelvic lymph node (P, ), parametrial invasion by pelvic examination (P ¼ 0.044), parametrial invasion by MRI (P ¼ 0.039), uterine body invasion (P, 0.001) and larger tumor volume (P, ) showed a statistically significantly relation to worse survival (Table 5). There was no evidence for a relationship between vaginal extension, histology or age and OS. In a multivariate analysis of OS with respect to these factors considered, uterine body invasion (P ¼ 0.020), larger tumor volume (P ¼ 0.012), and a positive pelvic lymph node (P ¼ 0.040) were shown to be independently related to OS in the presence of the other factors (Table 6). The hazard ratio for uterine body invasion was (95% CI ), for a larger tumor volume it was (95% CI ), and for a positive pelvic lymph node it was (95% CI ). Figure 2. Overall survival curves of the groups divided by uterine body invasion. Figure 3. Overall survival curves of the groups divided by tumor volume. Using the results of multivariate analysis, we devised a predictive model for OS. Patients were scored as being at risk with tumor with a volume 30 ml, positive lymph node enlargement or uterine body invasion. Patients were classified into four groups with zero, one, two and three risk factors. The 5-year OS rates of patients with risk 0 (volume,30 ml, no uterine body invasion and negative lymph node, LN), risk 1 (one of these three factors), risk 2 (two of these three factors) and risk 3 (volume 30 ml, uterine body invasion and positive LN) were 96.3, 77.5, 53.0 and 14.8%, respectively. The differences in OS among these four groups of the proposed model were statistically significant (P, ; Fig. 4). DISEASE-FREE SURVIVAL The 5-year DFS rate for all patients was 56.6% (95% CI %). Results for DFS closely paralleled those for OS. FIGO stage (P ¼ 0.001), clinical tumor diameter (P ¼ 0.001), parametrial invasion by pelvic examination (P ¼ 0.023), parametrial invasion by MRI (P ¼ 0.017), uterine body invasion (P, 0.001), pelvic lymph node (P ¼ 0.001) and tumor volume (P, 0.001) were again statistically significant factors by univariate analysis as with OS. The 5-year DFS rates according to FIGO stage I, II and III were 88.9, 58.5 and 32.4%, respectively. The 5-year DFS rate of patients with uterine body invasion was 39.3% compared with 86.4% in patients without uterine body invasion. Patients with larger tumor volume had a significantly worse DFS (5 years, 24.0%) compared with those with smaller volume (5 years, 79.6%). By multivariate analysis, only tumor volume (P ¼ 0.013) and uterine body invasion (P ¼ 0.014) were statistically significant factors (Table 6). FIGO stage, clinical tumor diameter and pelvic lymph node were not related to DFS.

6 Page 6 of 9 MRI tumor volume and corpus invasion of cervical Ca Table 5. Univariate analysis of the clinical outcome according to predictive variables (Cox proportinal hazard) Factor OS DFS P HR 95% CI P HR 95% CI Uterine body invasion, , Tumor volume (,30, 30 ml), , MRI invasion of parametrium FIGO stage (per unit stage) Vaginal extension PE invasion of parametrium Pelvic lymph node, , Clinical tumor diameter (,5 cm, 5 cm), SCC histology Age (,60, 60 years) HR, hazard ratio; CI, confidence interval; OS, overall survival; DFS, disease-free survival. DISCUSSION Computed tomography has been routinely used as an imaging modality in cancer of the cervix (22,23), but its role in providing anatomical details of the uterus and cervix is limited because of the lack of tissue contrast and multiplanar imaging capability (16,24). Magnetic resonance imaging has been reported to be superior to CT in evaluation of pelvic anatomy in gynecological malignancies and in the delineation of tumors of the uterine cervix (16,24,25). With its distinctive tissue contrast and multiplanar imaging capability, MR imaging provides excellent differentiation of tumor tissue from the normal tissue (15,19,25 27) in terms of excellent soft tissue contrast resolution, three-dimensional measurement and accurate judgment of invasion of the surrounding normal tissue (19,28,29). Surgical specimens have been proven to correlate well with the findings obtained by MRI in cervical cancer (15,18,30,31). Thus, MRI can be used for the accurate estimation of tumor volume and uterine body invasion in cervical cancer patients who are being planned to be treated by radiotherapy in this study. In this study, we examined the relationship of seven prognostic factors (FIGO stage, clinical tumor diameter, age, histology, pelvic lymph node, tumor volume and uterine body invasion) to radiation therapy in 106 newly diagnosed cervical carcinomas treated with a curative intent. Our results suggested that a larger tumor volume and uterine body invasion which were determined by using MRI had a statistically significant relationship to a worse OS and DFS, and pelvic lymph node invasion, which was determined by using CT, was related to a worse OS. Additionally, our study indicated that the FIGO stage, clinical tumor diameter and tumor histology had a poor relationship with prognosis by multivariate analysis. Tumor volume is believed to be an independent prognostic factor of outcome for patients with cervical carcinoma (5,27,32 36). Classically, tumor diameter has been measured by a physical pelvic examination. Although this is convenient and cost-effective, the accuracy of tumor size measurement is unsatisfactory (17). Clinical palpation is also a subjective method and has significant interobserver variability (12 14). However, several reports have noted that the Table 6. Multivariate analysis of factor according to OS and DFS Factor OS DFS P HR 95% CI P HR 95% CI Uterine body invasion Tumor volume (,30, 30 ml) MRI invasion of parametrium FIGO stage (per unit stage) PE invasion of parametrium Pelvic lymph node Clinical tumor diameter (,5, 5 cm)

7 Jpn J Clin Oncol Page 7 of 9 Figure 4. Overall survival curves of the groups divided by the predictive prognostic model. findings obtained from by MRI in cervical cancer proved to be well correlated with surgical confirmation (15,31). Our study showed that a lager tumor volume measured using MRI was associated with decreased OS and DFS in multivariate analyses. Thus, we assume that patients with a large tumor volume will develop recurrence more frequently than with a small tumor volume. The tumor volume calculated from MRI could thus providing promising information for predicting survival. In previous studies, uterine body invasion has been known to be associated with a poor prognosis (8) and to be associated with an increased incidence of lymph node positivity and poor survival (10,36). In our institution, we have seen a correlation between the presences of CT detected lymph node metastasis and uterine body invasion. Hope et al. has shown that pelvic lymph node metastasis was three times more frequent in patients with evidence of uterine body invasion as compared with those without uterine body invasion, and patients with endometrial invasion also had a significantly increased risk of paraaortic and supraclavicular metastasis and decreased OS and DFS (11). In our study, we assume that uterine body invasion might develop metastasis more frequently than no uterine body invasion and might also decrease OS and DFS. The uterine body invasion determined by MRI could thus provide promising information for predicting survival for cervical cancer patients treated by radiation therapy. Metastatic disease in the pelvic nodes is a powerful prognostic factor (37,38). In our study, pelvic lymph node status correlated with OS by multivariate analyses and correlated strongly with tumor volume and uterine body involvement. Lymphatics within the endometrium or myometrium may be more accessible to tumor cells because of the uterine structure or close proximity of the uterine vasculature (11). There also may be direct lymphatic pathways for metastases to proceed from the uterine corpus to paraaortic regions, as has been seen in endometrial cancer (39). This would increase the risk of lymph node involvement. Therefore, we suggest that uterine body invasion and large tumor volume may be correlated with lymph node metastases, which correlates with a worse prognosis. The prognostic significance of clinical tumor diameter has been reported previously in cervical cancer patients treated by radical radiotherapy (40). It is known from previously reported studies that tumor size affects overall survival. Eifel et al. (5) studied the relationship between tumor diameter and survival in stage IB patients and demonstrated decreasing survival in concert with increasing tumor diameter. A similar finding was reported by Finan et al. (41) based on their own surgical series. A much better survival was reported for a tumor,4 cm than for those 4 cm. From our results, the presence of a larger clinical tumor diameter (5 cm) correlated with worse OS and DFS by univariate analysis, but there was no correlation with OS and DFS by multivariate analysis. One reason for this difference may be that it is influenced by the diameter measuring procedure. Clinical tumor diameter was measured by physical examination, so the result was unsatisfactory. Parametrial invasion and vaginal extension are wellknown prognostic factor inversely related to local and actuarial survival (42,43). In this study, parametrial invasion proved to be a significantly predictor of overall survival and disease-free survival by univeriate analyses, but not in multivariate analysis. The greater proportion of patients (85/106; 80.2%) in this study was referred for concurrent chemotherapy and radiotherapy due to parametrial invasion of primary tumor. So there might be no correlation between parametiral invasion and survival in multivariate analyses. FIGO stage has been reported to be a convincing predictor of outcome among patients treated with definitive radiotherapy. However, several factors such as tumor volume, lymph node status and uterine body invasion status, which are not assessed by this system, have also been reported to be meaningful factors (5,27,32,33). The FIGO system proved to be a significantly strong predictor of overall survival and diseasefree survival by univariate analyses, but not in multivariate analyses. One reason for this may be that it is influenced by the staging procedure. The majority of patients in this study (97/106; 91.5%) were stage IIA, IIB, IIIA and IIIB patients. The difference between these groups is defined by the presence or absence of parametrial invasion and/or pelvic fixation, which was determined by a physical examination. Thus the majority of these groups were diagnosed by subjective and nonreproducible methods. We believe that tumor volume and uterine body invasion evaluated by using MRI and pelvic lymph node enlargement evaluated by using CT are more objective and more consistent with survival when compared with the FIGO system and clinical tumor diameter.

8 Page 8 of 9 MRI tumor volume and corpus invasion of cervical Ca With these three prognostic variables that had shown a strong correlation with OS, we adapted a predictive model for OS. Four subgroups showed significant differences in OS (P, ). The 5-year OS rates of patients with risk 0, risk 1, risk 2 and risk 3 were 96.3, 77.5, 53.0 and 14.8%, respectively. With the aid of MRI and CT, noninvasive, accurate and useful information could be obtained before treatment, and we believe application of this model to clinical trials would be advantageous and meaningful. Finally, we also evaluated other clinical parameters such as patient age and histological differentiation. An age younger than 60 was not a significant adverse factor for OS and DFS in this study. This finding has been reported in another study (44), but others have reported different results that, for the youngest women, the risk of death was significantly increased (40,45,46). Reagan and Fu (47) and Wentz and Reagan (48) demonstrated a prognostic value of histological differentiation in patients treated with irradiation, while in other reports, such as Goellner (49), Gunderson et al. (50), and Crissman et al. (51), a correlation between histological parameters and patients survival was not seen. From our findings, there was no correlation between histology and outcome. CONCLUSION Tumor volume and uterine body involvement determined by pre-treatment MRI examinations were significant prognostic factors for patients with invasive cervical carcinoma. Lymphatic involvement diagnosed by CT was also a significant prognostic factor for OS. Using these three parameters from MRI and CT, we have devised a practical and effective model to predict OS. Acknowledgement This work was supported by the Dongguk University Research Fund. Conflict of interest statement None declared. References 1. Delgado G, Bundy B, Zaino R, Sevin BU, Creasman WT, Major F. Prospective surgical pathological study of disease-free interval in patients with stage IB squamous cell carcinoma of the cervix: a Gynecologic Oncology Group study. Gynecol Oncol 1990;38: Zaino RJ, Ward S, Delgado G, et al. Histopathologic predictor of the behavior of surgically treated stage IB squamous cell carcinoma of the cervix. A Gynecologic Oncology Group study. Cancer 1992;69: Gauthier P, Gore I, Shingleton HM, Soong SJ, Orr JW, Jr, Hatch KD. Identification of histopathologic risk groups in stage IB squamous cell carcinoma of the cervix. Obstet Gynecol 1985;66: Hopkins MP, Morley GW. Stage IB squamous cell cancer of the cervix: clinicopathologic features related to survival. Am J Obstet Gynecol 1991;164: Eifel PJ, Morris M, Wharton JT, Oswald MJ. The influence of tumor size and morphology on the outcome of patients with FIGO stage IB squamous cell carcinoma of the uterine cervix. Int J Radiat Onocol Biol Phys 1994;29: Burghardt E, Baltzer J, Tulusan AH, Haas J. Results of surgical treatment of 1028 cervical cancers studied with volume try. Cancer 1998;70: Kovalic JJ, Perez CA, Grigsby PW, Lockett M. The effect of volume of disease in patients with carcinoma of the uterine cervix. Int J Radiat Oncol Biol Phys 1991;21: Perez CA, Camel HM, Askin F, Breaux S. Endometrial extension of carcinoma of the uterine cervix: a prognostic factor that may modify staging. Cancer 1981;48: Fagundes H, Perez CA, Grisby PW, Lockett MA. Distant metastases after irradiation alone in carcinoma of the uterine cervix. Int J Radiat Oncol Biol Phys 1992;24: Noguchi H, Shiozawa I, Kitahara T, Yamazaki T, Fukuta T. Uterine body invasion of carcinoma of the uterine cervix as seen form surgical specimens. Gynecol Oncol 1988;30: Hope AJ, Saha P, Grigsby PW. FDG-FET in carcinoma of the uterine cervix with endometrial extension. Cancer 2006;106: Averette HE, Ford JH, Dudan RC, Girtanner RE, Hoskins WJ, Lutz MH, Staging of cervical cacner. Clin Obstet Gynecol 1975;18: Brunschwig A. The surgical treatment of cancer of the cervix: stage I and II. AJR 1968;102: Heyman J, Kottmeier HL, Segerdahl CO, An investigation of the reliability of stage grouping in cancer of the uterine cervix. Acta Obstet Gynecol Scand 1953;32: Burghardt E, Hofmann HM, Ebner F, Haas J, Tamussino K, Justich E. Magnetic resonance imaging in cervical cancer: a basis for objective classification. Gynecol Oncol 1989;33: Subak L, Hricak H, Powell C, Azizi L, Stern J. Cervical carcinoma: Computed tomography and magnetic resonance imaging for preoperative staging. Obstet Gynecol 1995:86; Mayr NA, Yuh WT, Zheng J, et al. Tumor size evaluated by pelvic examination compared with 3-D quantitative analysis in the prediction of outcome for cervical cancer. Int J Radiat Oncol Biol Phys 1997:39; Martin AJ, Poon CS, Thomas GM, et al. MR evaluation of cervical cancer in hysterectomy specimens: correlation of quantitative T 2 measurement and histology. J Magn Reson Imag 1994;4: Hricak H, Quivey J, Campos Z, Gildengorin V, Hindmarsh T, Bis K, Stern J, Phillips T. Carcinoma of the cervix: Predictive value of clinical and magnetic resonance (MR) imaging assessment of prognostic factors. Int J Radiat Oncol Biol Phys 1993:27; Narayan K, McKenzie A, Fisher R, Susil B, Jobing T, Bernshaw D. Estimation of tumor volume in cervical cancer by magnetic resonance imaging. Am J Clin Oncol 2003;26: Hofman HM, Ebner F, Haas J, et al. Magnetic resonance imaging in clinical cervical cancer: pretherapeutic tumor volumetry. Baillieres Clin Obstet Gynaecol 1988;2: Kilcheski TS, Arger PH, Mulhern CBJ, Coleman BG, Kressel HY, Mikuta JI. Role of computed tomography in the presurgical evaluation of carcinoma of the cervix. JComputAssist Tomogr 1981;5: Vas W, Wolverson M, Freel J, Salimi Z, Sundaram M. Computed tomography in the pretreatment assessment of carcinoma of the cervix. J Comput Tomogr 1985;9: Kim SH, Choi BI, Lee HP, Kang SB, Choi YM, Han MC, Kim CW. Uterine cervical carcinoma: Comparison of CT and MR findings. Radiology 1990;175: Greco A, Mason P, Leung AWL, Dische S, Mclndoe GAJ, Anderson MC. Staging of carcinoma of the uterine cervix: MRI-surgical correlation. Clin Radio 1989;40: Rubens D, Thornbury JR, Angel C, Stoler MH, Weiss SL, Lerner RM, Beecham J. Stage IB cervical carcinoma: Comparison of clinical, MR, and pathologic staging. AJR 1998;150: Sironi S, Belloni C, Taccagni G, DelMaschio A. Carcinoma of the cervix: Value of MR imaging in detecting parametrial involvement. AJR 1991;156: Toita T, Kakinohana Y, Shinzato S, et al. Tumor diameter/volume and pelvic node status assessed by magnetic resonance imaging (MRI) for

9 Jpn J Clin Oncol Page 9 of 9 uterine cervical cancer treated with irradiation. Int J Radiat Oncol Biol Phys 1999;43: Hricak H, Lacey CG, Sandles LG, et al. Invasive cervical carcinoma: comparison of MR imaging and surgical findings. Radiology 1988;166: Oellinger JJ, Blohmer JU, Michniewicz K, et al. Pre-operative staging of cervical cancer: comparison of magnetic resonance imaging (MRI) and computed tomography (CT) with histologic results. Zentralbl Gynackol 2000;122: Manfredi R, Maresca G, Smaniotto D, et al. Cervical cancer response to neoadjuvant therapy: MR imaging assessment. Radiology 1998;209: Perez CA, Grigsby PW, Nene SM, et al. Effect of tumor size on the prognosis of carcinoma of the uterine cervix treated with irradiation alone. Cancer 1992;69: Kodaira T, Fuwa N, Kamata M, et al. Clinical assessment by MRI for patients with stage II cervical carcinoma treated by radiation alone in multicenter analysis: are all patients with stage II disease suitable candidate for chemoradiotherapy? Int J Radiat Oncol Biol Phys 2002;52: Kodaira T, Fuwa N, Toita T, et al. Comparison of prognostic value of MRI and FIGO stage among patients with cervical carcinoma treated with radiotherapy. Int J Radiat Oncol Biol Phys 2003;56: Narayan K, Fisher R, Bernshaw D. Significance of tumor volume and corpus uteri invasion in cervical cancer patients treated by radiotherapy. Int J Gynecol Cancer 2006;16: Narayan K, McKenzie AF, Hicks RJ, et al. Relation between FIGO stage, primary tumor volume, and presnece of lymph node metastases in cervical cancer patients referred for radiotherapy. Int J Gynecol Cancer 2003;13: Kamura T, Shigematsu T, Kaku T, Shimamoto T, Saito T, Sakai K, et al. Histopathologic factor influencing pelvic lymph node metastases in two or more sites in patients with cervical carcinoma undergoing radical hysterectomy. Acta Obstet Gynecol Scand 1999;78: Noguchi H, Shiozawa K, Tsukamoto T, Tsukahara Y, Iwai S, Fukuta T. The postoperative classification for uterine cervical cancer and its clinical evaluation. Gynecol Oncol 1893:16: Creasman WT, Morrow CP, Bundy BN, Homesley HD, Graham JE, Heller PB. Surgical pathologic spread patterns of endometrial cancer. A Gynecologic Oncology Group Study. Cancer 1987;60(8 Suppl): Logsdon MD, Eifel PJ. FIGO IIIB squamous cell carcinoma of the cervix: an analysis of prognostic factors emphasizing the balance between external beam and intracavitary radiation therapy. Int J Radiat Oncol Biol Phys 1999;43: Finan MA, DeCesare S, Fiorica JV, et al. Radical hysterectomy for stage IB1 vs IB2 carcinoma of the cervix: dose the new staging system predict morbidity and survival? Gynecol Oncol 1996;62: Zreik TG, Chambers JT, Chambers SK. Parametrial involvement, regardless of nodal status: a poor prognostic factor for cervical cancer. Obstet Gynecol 1996;87: Delgado G, Bundy B, Fowler WC, et al. A prospective surgical pathological study of stage I squamous carcinoma of cervix: a Gynecologic Oncology Group study. Gynecol Oncol 1989; 35: Rutledge FN, Mitchell MF, Nunsell S, et al. Youth as a prognostic factor in carcinoma of the cervix: a matched analysis. Gynecol Oncol 1992;44: Fyles AW, Pintilie M, Kirkbride P, et al. Prognostic factors in patients with cervix cancer treated by radiation therapy: results of a multiple regression analysis. Radiother Oncol 1995;35: Delaloye JF, Pampallona S, Coucke PA, et al. Younger age as a bad prognostic factor in patients with carcinoma of the cervix. Eur J Obstet Gynecol Reprod Biol 1996;64: Reagan JW, Fu YS. Histologic types and prognosis of cancers of the uterine cervix. Int J Radiat Oncol Biol Phys 1979;5: Wentz WB, Reagan JW. Survival in cervical cancer with respect to cell type. Cancer 1959;12: Goellner JR. Carcinoma of the cervix: clinicopathologic correlation of 196 cases. Am J Clin Pathol 1976;66: Gunderson LL, Weems WS, Hebertson RM, et al. Correlation of histopathology with clinical results following radiation therapy for carcinoma of cervix. AJR Am J Roentgenol Radium Ther Nucl Med 1974;120: Crissman JD, Budhraja M, Aron BS, et al. Histopathologic prognostic factors in stage II and III squamous cell carcinoma of the uterine cervix. Int J Gynecol Pathol 1987;6: