Journal of Cancer Research and Therapeutics Medknow Publications on behalf of the Association of Radiation Oncologists of India (AROI) ISSN: 0973-1482 EISSN: 1998-4138 Vol. 1, Num. 2, 2005, pp. 103-107

Journal of Cancer Research and Therapeutics, Vol. 1, No. 2, April-June, 2005, pp. 103-107

Original Article

Magnetic resonance imaging in carcinoma cervix- Does it have a prognostic relevance

Sethi TejinderKataria, Bhalla NK, Jena AN, Rawat S, Oberoi R

Department of Radiation Oncology, Rajiv Gandhi Cancer Institute and Research Centre, New Delhi
Correspondence Address:Department of Radiation Oncology, Rajiv Gandhi Cancer Institute and Research Centre, New Delhi, t_kataria@rediffmail.com

Code Number: cr05021

Abstract

Keywords: M. R. I., Radiotherapy, Cervical carcinoma

INTRODUCTION

Carcinoma cervix is the most common cancer of females in the Indian subcontinent [1]. FIGO (International Federation of Gynecology and Obstetrics) staging system is the recommended staging procedure for these cancers. FIGO system allows clinical pelvic examination under anaesthesia and the stage is defined by the extent of disease spread beyond the cervix to the surrounding tissues including parametria, pelvic side walls, vagina, bladder or rectum. Chest x-ray is done to assess the lung metastasis and an intravenous pyelogram to assess hydronephrosis and thereby retroperitoneal lymph node disease. Pelvic lymphnode status and the tumour size is not evaluated to designate stages II - IV. The accuracy of such a clinical method depends largely upon the experience of the examining physician and leads to an intra- and interobserver variation in staging.

CT and MRI are valuable imaging techniques in the pretreatment evaluation of cervical cancer allowing direct visualization of the tumour. MRI, however, can better delineate cervical tumour size, location and extension into adjacent structures and lymph node metastasis with its excellent soft-tissue contrast resolution and multiplanar scanning capabilities.[2] MRI provides accurate staging information in 70-90% of the cases.[3]

Results of therapy for patients with carcinoma of uterine cervix are most often reported by clinical stage of the tumour. The pretreatment evaluation of prognostic factors helps in decision-making for treatment. Amongst the various prognostic factors, tumour size has emerged as the most important factor to which other morphological risk factors like tumour invasion depth, loco-regional extent and lymph node involvement are related.[3], [4], [5], [6] In patients treated by radical surgery or radiotherapy alone, size of the primary tumour is of prognostic significance irrespective of the clinical stage of the tumour.[7], [8] Recently, some studies have shown that tumour diameter/volume and pelvic lymph node status assessed by MRI significantly correlated with the outcome for patients with uterine cervical cancer treated with irradiation.[5], [6]

We conducted a study to assess the prognostic significance of MRI mapped initial tumour volumes on local control in patients of carcinoma cervix treated by surgery alone, surgery with radiation, concurrent chemo-radiotherapy or radiotherapy alone.

METHODS AND MATERIALS

Patients
Records of 70 consecutive patients with carcinoma of uterine cervix, treated at the department of Radiation Oncology between 1999 and 2001, were taken up for the study. All patients had undergone pre treatment evaluation including a complete history and physical examination, complete blood counts, serum chemistries, chest radiograph and cystoscopy. Cervical biopsies were obtained for histological classification in all patients. The patient′s disease was classified according to the FIGO staging. An abdomino-pelvic MRI was done in all patients to evaluate the tumour size and locoregional spread of disease.

MRI Protocol
Serial MRI scans were performed in all patients before and after completion of treatment on a 1.0 Tesla MRI scanner (Magnetom Impact Expert, Siemens AG. Germany) and images were taken using axial T1W SE (weighted spin echo) and STIR (short tau inversion recovery), sagittal T2W TSE (weighted turbo spin echo) and high resolution (512x512) T2W TSE in axial plane with 5mm slice thickness and 10% inter slice gap. The tumour diameter measurements were obtained by two experienced radiologists. The largest diameter of the tumour was determined on the transverse images, however, if the tumour extension was larger in the craniocaudal direction it was assessed on sagittal plane. Tumour volume (V) was calculated in a consensus reading by the two radiologists, using 3-dimensional volumetric measurements according to the modified Simpson rule. In all T2W contiguous transverse images and sagittal images tumour mass was outlined on the computer monitor using in-house developed software. The area of the tumour (A) in each slice was multiplied by the slice profile (5mm slice thickness and 10% interslice gap), and total tumour volume was automatically calculated by summation of the adjacent slice volumes according to -

n
V(ml) = 0.6 x ( Σ A_i)
i =1
where i is the individual slice number and n is the total number of slices. Inter observer difference was within ±7% between two independent observers.

Treatment interventions
Patients were treated by either Werthiems′s hysterectomy alone (Group A, n = 15), Werthiems′s hysterectomy followed by adjuvant radiotherapy (Group B, n = 27), concomitant chemo-radiation (Group C, n = 14) using weekly administration of cisplatin at a dose of 50 mg or by radical radiotherapy alone (Group D, n = 14). Radiation therapy consisted of a combination of pelvic external beam radiation on linear accelerators followed by high dose-rate (HDR) intracavitary brachytherapy (Microselectron, Ir-192). The external beam radiation was given using 6MV photons to a mid pelvis dose of 45 Gy in 25 fractions with 1.8 Gy per fraction, 5 fractions per week. Intra cavitary brachytherapy (ICBT) was performed by using either Nucletron standard applicator to a dose of 9 Gy x 2 fractions prescribed to point A at weekly interval for intact uterine cervix or by a Nucletron vaginal cylinder to a dose of 4 Gy x 3 fractions 5mm away from the surface of cylinder at weekly interval for vault irradiation in the operated subset.

Follow up
Clinical examination and MRI scan were performed 3 months after completion of treatment. If the clinical examination and/or MRI scan demonstrated high signal intensity on T2W (Weighted) images and the index of suspicion to have a residual tumour was high the same, was confirmed by histopathologic examination. These patients were referred for further treatment with salvage surgery or chemotherapy. The patients who showed no evidence of residual tumour on MRI scan were kept on regular follow up every 3 months for 2 years and then every 6 months thereafter. Every followup visit included a specific medical history and clinical examination. Routine blood counts, serum chemistries, MRI scan of pelvis and x-ray chest were done once a year.

Statistics
Treatment failures included pelvic failures (including primary tumour and lymphnodes) and distant metastasis. Primary tumour recurrence included cervical, corpus, vagina, parametria and vaginal vault in operated cases. FIGO stage and MRI volumes were correlated with patient outcome. The end points of this study were pelvic control rate, disease free survival rate, distant metastasis free rate and overall survival rate. These estimates were calculated using the Kaplan-Meier method and compared using log rank test.

RESULTS

The median age at presentation was 51 years (range 21-80 years) and the distribution according to FIGO stages was Stage I, 26 (37%); stage II, 35 (50%); stage III, 7 (10%) and stage IV 4 (5%). Histologically 68 patients had squamous cell carcinoma, one adeno carcinoma and one adeno squamous carcinoma.

Based on the MR volumes patients were grouped into Group I (MR volume < 40cc), Group II (MR volume 40 - 100cc) and Group III (MR volume >100cc). Forty three (61.4%) patients were in group I, 20 (28.5%) in Group II and 7 (10%) in group III. In group I, 44% of patients had FIGO stage I disease, 47% had stage II disease and 9% had stage III disease. In group II, 35% patients had stage I, 45% had stage II disease and 10% had stage III and IV disease each. In group III, 57% patients had stage II disease, 14% had stage III and 29% had stage IV disease [Table - 1].

Disease free survival (DFS) and overall survivals (OS) were estimated according to MR volume, FIGO stage, and treatment protocol. According to MR volume the 3 year DFS was 86% in Group I, 87% in Group II and 33% in Group III respectively (p value = 0.0015) [Figure - 3]. 3 year OS was 100% in group I and II each and 62% in Group III respectively (p value = .0001) [Figure - 4]. According to FIGO staging the 3 year DFS was 96% in stage I, 78% in stage II, 63% in stage III and 33% in stage IV respectively (p value = .0094) and 3 year OS were 100% in stage I and II, 80% in stage III and 67% in stage IV respectively (p value = 0.0037). According to treatment protocol, the 3 year DFS for Group A was 92%, for Group B was 83%, for Group C was 92% and for Group D was 52% respectively (p value = .0685). 3 year OS for Group A and B were 100%, for Group C 92% and for Group D 86% respectively (p value = 0.3242). Thus, the disease free survival and overall survival did not achieve a level of statistical significance when evaluated as per protocol (p = .0685 for DFS and 0.3242 for OS), whereas a statistical significance was achieved when DFS and OS were evaluated according to MR volumes (p = 0.0015 for DFS & 0.0001 for OS) as compared to the conventional stage wise evaluation (p = 0.0094 for DFS & 0.0037 for OS).

Twelve patients (17.14%) developed recurrence during the follow up, of these, 14% were in Group I, 10% in Group II and 57% in Group III. 5/12 had local recurrence and 7/12 distant metastases. Three patients underwent salvage surgery for local recurrence and one of them patient received salvage chemotherapy to achieve a complete response. Two patients received pelvic radiation and achieved complete response. Seven patients of these twelve patients developed distant metastases and were offered salvage chemotherapy. Two patients died 6 months after completion of RT (lung metastasis), and five patients refused further treatment. "58 patients were alive and on regular follow up every 3 months up to 2 years for a median of 28 months (range 24-36 months). Two patients were lost to follow up."

Discussion

Clinical pelvic examination is used to estimate primary tumour size and the extent of involvement in cervical cancer prior to treatment. Tumor size has been well established in numerous studies as an important independent predictor of response to radiation, local control and survival in cervical cancer.[7], [8], [9]

Recent studies have shown that 3 dimensional quantitative imaging-based method of tumor size assessment using MRI is highly accurate in determining actual tumor size and extent of disease and may be superior to clinical palpation in predicting local tumor control.[3], [6], [10], [11], [12] With its multiplanar capability, distinctive tissue characteristics and exquisite anatomical detail, MRI has been reported to be superior to CT in the evaluation of pelvic anatomy and malignancies of the uterus. On MRI, cervical cancer appears as a high signal intensity mass on T2W (weighted) images, and can be distinguished from the normal low signal intensity cervical stroma.[13], [14] MRI differentiation between tumour and fibrosis can be achieved with high degree of accuracy. Furthermore, the strength of MRI is its ability to directly visualize the tumour therefore allowing measurement of its size. MRI is the most appropriate study for the evaluation of the post irradiation pelvis when searching for suspected residual tumours or recurrence of cervical carcinoma [11]. Typically fibrosis has low signal intensity on both T1W (weighted) and T2W (weighted) images. On the other hand tumour exhibits high signal intensity on T2W images. However, the inflammation and edema associated with acute radiation changes may also demonstrate high signal intensity on T2W (weighted) images, sometime leading to an incorrect diagnosis.[11]

Our study indicates that FIGO staging may not be a true representation of the correct volume of disease and therefore, does not provide prognostic insight into the metastatic potential of the disease. Thus, a patient clinically assessed to have FIGO stage I disease may actually have a large volume [Figure - 1] and may require more aggressive treatment to achieve complete response and a patient with FIGO stage III disease may have small volume of disease [Figure - 2] and may actually have a better prognosis.

Flueckiger et al. noted that all tumours regressed completely within 6 months of radiotherapy, showing a characteristic drop in relative signal intensity following radiotherapy. This decline was most precipitous during the first 3 months and was almost complete at 6 months.[10] This suggests that it is useful to obtain MRI 3 months after completion of treatment to evaluate the effect of radiation. Several prognostic factors including clinical evaluation (Patients age and Karnofsky Performance status), histological finding (tumour grade) and morphological features such as tumour size, location, depth of stromal invasion, local tumour extension and lymph node metastasis have been shown to affect the therapeutic outcome.[15], [16] In our study, the volume of disease as assessed on MRI was thought to be a better prognostic indicator than FIGO staging. MRI can better delineate cervical tumour size, location and extension into adjacent structures with its superior soft tissue definition and multiplanar scanning capacities.[17]

Hricak, et al mentioned that tumor size and response to radiation showed a strong correlation.[6] In their report, a significant relationship was found between tumor size, disease stage and invasion to adjacent tissue and organs. However, in this report, the analysis was focused on tumor response with short follow-up, so survival influence was not mentioned. Mayr, et al also reported size and volume to be strong prognostic factors for local control.[5], [18] The three-dimensional tumor volume was calculated by the summation of all tumor areas and multiplication by the section profile. Using this volumetric technique, they reported that the volume regression rate during treatment would predict local control more accurately than the initial and residual volume. However, in another report,[18] DFS was influenced by both initial tumor size and follow-up examination, but not by regression rate. Toita, et al failed to show a significant influence for volume calculated in the same manner with DFS in their previous report, although tumor size was a significant prognostic factor for DFS.[15] This was chiefly because of their small patient numbers with a nonstage-limiting analysis. In our study the tumour volume was calculated in each case from T2W sagittal images using an in-house software that adds up the volumes for each slice from user marked area on the visualized tumour and taking care of the area due to inter slice gap through interpolation.

Our overall patient numbers are not high enough at this time to allow assessment of tumor volume parameter with respect to other prognostic variables. Our data will need further evaluation with larger number of patients.

Conclusion

The clinical assessment and FIGO staging is not enough to assess the magnitude of disease. Pretreatment MRI determines the actual tumor size, local tumor extension, depth of stromal invasion and lymph node metastasis with its superior soft tissue definition and multiplanar scanning capacities and has been shown to affect the therapeutic outcome. Thus, a patient clinically evaluated to be in an early stage may actually have a worse prognosis due to large volume of disease as seen on MRI and may be optimized for an aggressive treatment plan. The correlation between MR volume, FIGO stage and survivals (disease free and overall survival) suggests that MRI gives a better indication of the volume of disease than FIGO stage.

Volumetric study using MRI is thought to be clinically efficacious to estimate clinical results, and we believe this criterion can be meaningfully adapted for its use in better understanding of prognosis and devise an optimal treatment strategy. We also believe that MRI should be included as a routine investigation for staging in the FIGO staging system.

References

Copyright 2005 - Journal of Cancer Research and Therapeutics

The following images related to this document are available:

Photo images