IJC International Journal of Cancer

Transcription

1 IJC International Journal of Cancer Persistent human papillomavirus DNA is associated with local recurrence after radiotherapy of uterine cervical cancer Yong Jung Song, Joo-Young Kim, Su-Kyoung Lee, Hyun-Sun Lim, Myong Cheol Lim, Sang-Soo Seo, Sokbom Kang, Dong Ock Lee and Sang-Yoon Park Research Institute and Hospital, National Cancer Center, Goyang, Gyeonggi, Korea Human papillomavirus (HPV) DNA is considered as a hallmark of cervical cancer. We investigated whether persistent HPV DNA at the cervix is associated with local recurrence after radiotherapy in patients with locally advanced cervical cancer. A total of 156 patients with HPV-positive cervical cancer (International Federation of Gynecology and Obstetrics stage IB IVB) treated with radiotherapy between July 2003 and December 2006 were analyzed. HPV DNA was measured prior to radiotherapy and after completion of radiotherapy. The results of HPV DNA test at postradiotherapy 1, 3, 6 and 12 months were analyzed individually for association with local recurrence-free survival (LRFS). In addition, the result of any last follow-up HPV test within 24 months postradiotherapy was defined as the overall status of HPV at 24 months and was also analyzed for association with LRFS. HPV DNA was cleared in 127 patients (81.4%) and persistent in 29 patients (18.6%) by 24 months. In 18 patients with local recurrences, 14 patients (78%) showed positive HPV tests at 1 3 months. Among the various time points analyzed, a positive HPV test at 3 months was the most accurate predictor of local recurrence. Multivariate analysis indicated that overall status of HPV at 24 months, low HPV viral load and histologic grade as being significantly related to poor LRFS. In HPV-positive cervical carcinoma treated primarily with radiotherapy, persistent HPV DNA within 24 months after treatment indicates a high risk of local recurrence. Diagnostic accuracy of HPV test was highest at 3 months. Key words: HPV, persistence, uterine cervical cancer, radiotherapy, local recurrence This work was presented as a poster at the 41th Society of Gynecologic Oncologists Annual Meeting for Women s Cancer on March 2010, San Francisco, CA, USA. Grant sponsor: National Cancer Center, Goyang, Korea; Grant number: DOI: /ijc History: Received 13 Jul 2010; Accepted 30 Sep 2010; Online 25 Oct 2010 Yong Jung Song s present address is: Department of Obstetrics and Gynecology, Pusan National University Yangsan Hospital, Yangsan, Korea Correspondence to: Joo-Young Kim, Research Institute and Hospital, National Cancer Center, 323 Ilsan-ro, Ilsandong-gu, Goyang, Gyeonggi , Korea, Tel.: þ / , Fax: þ , jooyoungcasa@ncc.re.kr Cervical cancer is a significant health problem worldwide, ranking second highest in cancer incidence and fourth highest in site-specific cause of cancer death in women. 1,2 While concurrent chemoradiotherapy is the main treatment modality for locally advanced cervical cancer, treatment failure in the central pelvis occurs in approximately 20 25% of patients. 3,4 Cytological tests are commonly inaccurate in detecting locally persistent or recurrent disease because the effect of radiation on the cells may result in ambiguous cell morphology in the early postradiotherapy period. 5 Because positive high-risk human papillomavirus (HPV) DNA is considered as an important tool in the diagnosis of both preinvasive and invasive cervical cancer, the usefulness of the HPV test has often also been considered as a method of post-treatment surveillance. Previous studies have examined the effectiveness of the HPV DNA test as a clinically useful marker for detecting residual disease or recurrence after conservative surgical procedures for cervical intraepithelial neoplasia 6 8 and also after radiotherapy In terms of radiotherapy outcome, several past studies showed that HPV persistence is associated with high rates of local recurrence and poor overall survival in patients with cervical cancer. 10,11 However, the status of HPV DNA was only examined at a single time point after radiotherapy in those studies, and hence, the pattern of HPV clearance after radiotherapy and the significance of positive HPV tests at individual time points was not subject to evaluation. In our study, HPV DNA was examined at multiple time points during the follow-up period after completion of radiotherapy. Absence or presence of HPV DNA at different time points was individually evaluated for association with local recurrence after radiotherapy. Initial HPV viral load was included along with persistent HPV DNA as one of the prognostic variables because a previous study revealed that lower HPV viral load than median value to be associated with a higher number of local as well as overall recurrences. 13 Material and Methods Study population, HPV DNA test and radiotherapy Our study was performed under the approval of our Institutional Review Board, and informed consent was obtained Int. J. Cancer: 129, (2011) VC 2010 UICC

2 Song et al. 897 Table 1. Patient characteristics HPV cleared (n 5 127) HPV persistent (n 5 29) Variable No. % No. % p value Age (yrs) 0.303* < FIGO stage IB IB IIA, IIB IIIA, IIIB, IVA IVB Histology SCC AD/ADS Grade 0.592* I, II III Node 0.307* Negative Positive Size 0.088* <4 cm cm Smoke Never Ever HPV viral load (RLU/CO) 0.284* >Median (387; ) Median Local recurrence 0.001* Yes No *Pearson s v 2 test. Fisher s exact test. Abbreviations: SCC: squamous cell carcinoma; AD: adenocarcinoma; ADS: adenosquamous carcinoma; RLU/CO: relative light unit/cutoff. from all patients. A total of 169 consecutive patients with International Federation of Gynecology and Obstetrics (FIGO) stage IB1 IVB cervical cancer were treated at the National Cancer Center Korea between July 2003 and December After excluding 13 patients who never took the HPV DNA test after completion of radiotherapy, 156 consecutive patients were included in this analysis (Table 1). All patients were clinically staged using physical examination, bimanual pelvic examination, posterior anterior chest radiograph, cystoscopy and rectosigmoidoscopy. Nodal status was determined by magnetic resonance imaging of the pelvis 6 positron emission tomography except for 36 patients who also underwent laparoscopic lymph node staging as a part of our previous clinical trial. 14,15 HPV status was evaluated from the cervical smear collected prior to radiotherapy using the Hybrid Capture II system (HC2; Digene Corporation, Gaithersburg, MD). HC2 test uses an RNA probe mixture of 13 high-risk HPV types and detects HPV types of 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59 and 68. The patients were followed up for a median period of 41 months (range, 6 66 months). Radiotherapy consisted of Gy whole pelvic teletherapy and five to seven fractions of high-dose rate

3 898 Persistence of HPV infection after radiotherapy brachytherapy, as previously described. 13 All patients were treated concomitant weekly by cisplatin 40 mg/m 2 during teletherapy, except the nine patients with FIGO stage IVB who received 5-fluorouracil þ cisplatin chemotherapy and 20 patients who received radiotherapy alone because of their poor physical condition. Local failure included the presence of residual disease at the end of radiotherapy that was not resolved by 1 month and confirmed by biopsy, and also included patients who showed complete response but relapse at the cervix, initially involved vaginal mucosa and medial parametria. HPV test was performed using HC2, with relative light unit (RLU)/cutoff (CO) ratio 1.0 considered positive. Follow-up visits were made at 1 month and then every 3 months for the first 2 years, every 4 months for the 3rd year and at 6-month intervals thereafter. HPV test was attempted to be performed on every visit, and the results at 1, 3, 6 and 12 months after radiotherapy were evaluated in our study for an association with local recurrence. Statistical analyses Demographic and clinical characteristics are presented as counts and percentages for categorical variables; median and range are supplied for continuous variables. Distributional differences between groups were assessed using the Pearson v 2 test or Fisher s exact test. The Kaplan Meier method and the log-rank test were used to compare local recurrence-free survival (LRFS) between subjects with negative versus positive HPV DNA tests after radiotherapy. Cases diagnosed as local recurrence at the time of detection of persistent HPV DNA were censored. Two patients who died of unknown causes were censored at the date of death. The multivariate Cox proportional hazards model was used to adjust for potential confounding factors such as age, stage group, histologic type, histologic grade, nodal involvement, tumor size, smoking status and initial HPV viral load. The final multivariate Cox model was chosen based on the combination of a stepwise procedure and the hierarchical selection method, as well as consideration of the clinical or biological importance of the variables in the model. Persistent HPV DNA at each time point (1, 3, 6 and 12 months) was examined for association with the incidence of local recurrences using sensitivity, specificity, accuracy, positive predictive value, negative predictive value and C statistics. Patients who were diagnosed with recurrence but who took the next HPV test were excluded from analysis of diagnostic accuracy at the next time point. The result of the last HPV test, performed within 24 months, was defined as the overall status of HPV at 24 months (HPV [24 mo]) and was also analyzed to take all the results of positive HPV test at any given time points within the first 2 years into consideration. For example, a patient whose HPV test was positive but turned negative within 24 months was termed overall status-negative, and a patient with a negative result who turned positive within this period was termed overall status-positive at 24 months. For the patients who were diagnosed with locally recurrent diseases at any given time point, the results of HPV test at that time point were recorded as their overall HPV status. Patients who has been HPV negative for the first 24 months but showed a positive HPV test 24 months after radiotherapy were not included in this analysis. HPV [24 mo] was examined for association with LRFS by uni- and multivariate analysis. All statistical analyses were performed using SPSS version 12.0 (Chicago, IL). All reported p values are two sided. Results Frequency of persistent HPV DNA after completion of radiotherapy Overall HPV DNA status was cleared in 123 patients (78.8%) and persistent in 33 patients (21.2%) after radiotherapy for a median period of 41 months (range, 6 66 months). HPV [24 mo] was cleared in 127 patients (81.4%) and persistent in 29 patients (18.6%). There was no significant difference with regard to clinicopathologic characteristics between the two groups (Table 1). One hundred twenty-one patients attended a follow-up examination at 1 month after radiotherapy, of which 34 patients had a positive HPV test (34/122 patients, 27.9%). The corresponding figures were 28/ 125 (22.4%), 21/116 (18.1%) and 23/123 (18.7%) at 3, 6 and 12 months, respectively. The reasons of different number of patients who underwent HPV test at each time point were lack of outpatient visit and/or HPV test not being performed in some patients, and also the occasions where the diagnosis of recurrent diseases was given in the previous follow-up visit and further HPV tests were not performed. Association of persistent HPV DNA with recurrence after radiotherapy The patients were followed up for a median period of 41 months (range, 6 66 months). The median follow-up for patients without recurrence was 43 months (range, 8 66 months). During the observed study period until the time of analysis, 18 patients were found to develop local recurrence. Fourteen patients relapsed in the first 6 months after radiotherapy and 4 patients showed relapse at 6, 9, 12 and 24 months. Seventy-eight percent (14/18) of the patients with local recurrence showed at least one positive HPV test within the first 3 months follow-up. However, there were two patients who were HPV DNA negative at 1 month but showed grossly persistent disease and received salvage operations. Both of them showed very low pretreatment HPV viral load (16.3 and 43.3 RLU/CO, respectively). Apart from these four patients, all local recurrences were preceded by positive HPV test at any given time before diagnosis of local recurrences was given. The results of cytologic examinations in the 14 recurrent patients whose HPV tests were positive were high-grade squamous intraepithelial lesion (HSIL) in nine patients and negative for malignancy in five patients. Typically, repeated cytologic examinations were performed before the first cytologic diagnosis of HSIL in these patients, and the most frequent diagnosis was radiation-associated cellular

4 Song et al. 899 Table 2. Univariate and multivariate analysis of local recurrence-free survival with clinicopathologic prognostic factors Patient number (n) Univariate HR 95% CI p value Multivariate HR 95% CI p value Age FIGO stage I, II III, IVA IVB Histology SCC AD/ADS Grade I, II III Node Negative Positive Size (cm) < Smoke Never Ever HPV viral load < HPV [24 mo] Clear Persist Abbreviations: SCC: squamous cell carcinoma; AD: adenocarcinoma; ADA: adenosquamous carcinoma; HR: hazard ratio; CI: confidence interval; HPV [24 mo]: result of the last HPV test within 24 months after completion of radiotherapy. changes. All these patients were finally diagnosed as invasive cancer by punch biopsies. The median elapse time between the positive HPV test and cytologic diagnosis of HSIL was 48.5 days (range, days). Local recurrence was significantly associated with HPV [24 mo] (Table 2) with a hazard ratio (HR) of 4.39 (95% CI, , p ¼ 0.003) and 4.22 (95% CI, , p ¼ 0.008) in univariate and multivariate analysis, respectively. In multivariate analysis, HPV [24 mo] and lower initial HPV viral load were significantly related to LRFS. Lower initial HPV viral load remained significant in multivariate analysis for overall survival, but the significance of HPV [24 mo] was lost (data not presented). Histologic differentiation was an additional factor associated with poor LRFS. Tumor size was significant poor prognostic factor for local recurrence in univariate analysis, but its significance was reduced in multivariate analysis. The log-rank test comparing survival distribution between clearance and persistence of HPV DNA showed a significant difference for LRFS (p ¼ 0.001; Fig. 1). The sensitivity, specificity, accuracy, positive predictive value, negative predictive value and C statistics of persistent HPV DNA for local recurrence at 1, 3, 6 and 12 months are shown in Table 3. Discussion The results of our study indicate that HPV test is a useful surveillance tool for diagnosis of recurrent cervical cancer primarily treated with radiotherapy. Multivariate analysis revealed that overall status of HPV DNA at 24 months after radiotherapy was associated with a high incidence of local relapse after radiotherapy. Among the results at the various time points examined, the diagnostic yield of the HPV test for detection of postradiotherapy recurrence was highest at 3 months; however, follow-up HPV tests beyond this time point further detected local recurrences. There were seven patients who showed reappearance of HPV DNA beyond 24 months; however, none of the positive tests were associated

5 900 Persistence of HPV infection after radiotherapy Figure 1. Local recurrence-free survival rate by HPV persistence. The HPV persistence group represents patients who showed positive HPV DNA at any previous HPV test performed within 24 months after completion of radiotherapy. Table 3. Diagnostic accuracy of persistent HPV state at each time point HPV test Time (mo) P/C/NA No. of patients SN (%) SP (%) Accuracy (%) PPV (%) NPV (%) C statistics (p value) 1 34/88/ % 78% 76% 32% 93% 0.71 (0.005) 3 28/97/ % 82% 82% 25% 98% 0.80 (0.003) 6 21/95/ % 84% 82% 19% 96% 0.71 (0.07) 12 23/100/ % 82% 81% 9% 98% 0.66 (0.27) Abbreviations: P: persistent HPV DNA test; C: clearance HPV DNA test; NA: not available; SN: sensitivity; SP: specificity; PPV: positive predictive. with local recurrence. The new appearance of HPV DNA may have been caused by reinfection of HPV by sexual activity or by reactivation and proliferation of HPV from dormant status in immunologically less competent populations, such as in elderly patients or those with systemic disease. We did not examine the genotypes of HPV during the follow-up period. Examining the HPV genotype might have raised the diagnostic accuracy of follow-up HPV test in detection of recurrent disease in our study. Several studies showed a possible correlation between post-treatment HPV status and prognosis after radiotherapy in patients with cervical cancer ,16 Nagai et al. 10 reported the adverse prognostic effect of persistence of HPV infection after radiotherapy in 97 patients with FIGO stage I IVA disease. HPV test performed at 1 month after radiotherapy was analyzed for an association with radiation response (complete or partial), local recurrence and overall survival. The local recurrence rate was 7.1% for patients with HPV clearance and

6 Song et al % for those with HPV persistence in 92 patients who achieved complete response at the end of radiotherapy. Their reported persistent rate of HPV was 56.7% (55/97), which is much higher than our result of 27.9% (34/122) at 1 month. A potential reason for the high clearance rate of HPV infection in our study compared to their result is that most of the present subjects were treated with concurrent chemoradiotherapy when only 35% of the patients included in their study received concurrent chemotherapy. Datta et al. 9 also evaluated the value of HPV test performed at 1 month postradiotherapy and found more than 99.5% decrease in HPV DNA titer value from preradiotherapy to postradiotherapy was significantly associated with local recurrence-free and overall survival in 21 radiotherapy patients. In our study, HPV test at 3 months resulted in the highest C statistics, sensitivity, specificity and accuracy. HPV DNA may be persistently detected after radiotherapy in tumors that were bulky at initial presentation because HPV DNA fragments may be present in degraded tumor cells even after all of the cancer cells have been killed. 17 In this case, detection of a small amount of HPV DNA may be falsely interpreted as inadequate ablation of the tumor cells. In our study population, 14 of 28 patients (50.0%) with persistent HPV DNA at 3 months after radiotherapy showed further clearance of HPV DNA on the following examinations. Twelve out of 14 showed HPV value below 10 RLU/CO, suggesting the presence of small amount of HPV DNA in degraded tumor cells after radiotherapy. The accuracy of the HPV test at 1 month in predicting local recurrence shows lower sensitivity, specificity, accuracy and positive predictive value compared to those at 3 months. Taken together, our results indicate that the positive HPV tests in early post-radiotherapy period should be taken with caution and should be meticulously followed up if it is not associated with other clinical evidences suggesting the presence of locally persistent or recurrent disease. According to a recent study, radiation atypia is detected in 28% of pap smears taken during the first 4 months after radiotherapy, with the rate decreasing thereafter. 18 Recent data demonstrated that the sensitivity, specificity and accuracy of cervical smears is 0.16, 0.96 and 0.46, respectively. 5 Our results of cytological tests before the diagnosis of local recurrences also indicate low discriminatory ability of cytological test for invasive recurrence. In light of the high false-positive rate of cytological test, HPV DNA test appears to be a suitable surveillance method for detecting local recurrence of cervical cancer treated primarily by radiotherapy. Local control is an important factor influencing the improved cure rates in patients with cervical cancer. 19,20 In our study, persistent HPV DNA was highly predictive of local recurrence in uni- and multivariate analysis. It was also significantly associated with overall survival in univariate analysis, however, the significance was lost in multivariate analysis as opposed to in the other two studies mentioned already in the discussion. 9,10 The one reason behind this is probably due to the relatively short follow-up time and/or high salvage rate of the locally recurrent diseases. Seven out of the 18 patients with local recurrence in our study were still alive at the time of this analysis because salvage treatment was given at early follow-up time (six by surgery and one by chemotherapy). The incidence of local relapse is small (18/ 156, 11.5%) compared to the other studies [25.8% (25/97) local persistent or relapse in Nagai s study and 28.6% (6/21) patients achieved complete response at 1 month of radiotherapy] and hence, the effect of local relapse on overall survival could have been obscured. Persistent disease at the end of radiotherapy and locally relapsed diseases would have been better to be separately analyzed; however, the two categories had to be analyzed all together because of the low patient number in each category of recurrence. Eight patients with stage IVB patients were included in our study. Although the prognosis of stage IVB cervical cancer is considered to be grave, the patients were included in our study because the patients were treated with curative aim for their central pelvic disease. All metastatic lesions were also irradiated with curative intent. As a result, six of the eight patients were locally controlled. One of the findings that is worth noting is that although tumor size was a significant prognostic factor in univariate analysis, the prognostic significance was reduced in multivariate analysis. Including the patients with advanced stage might have been one of the reasons of this finding because stage and tumor size are generally related to each other. In summary, a positive HPV test after completion of radiotherapy is strongly associated with local recurrence; however, patients may need to be followed up with HPV test for at least the first 2 years after completion of radiotherapy. Our study suggests that HPV tests may need to be performed as a part of routine follow-up examination in HPVpositive cervical cancer patients after radiotherapy. Given that half of the patients with persistent HPV will clear with further follow-up as observed in our study, performing HPV test on every follow-up visit might not be justified solely on the basis of cost effectiveness. However, considering the generally poor salvage rate of the central recurrences of locally advanced cervical cancer and grave symptoms of the patients with uncontrolled recurrent pelvic disease, the value of detection of central recurrence at its early stage could be tremendous because the lesions can be effectively salvaged with more conservative types of surgery as our result shows. If the tests cannot be performed on every visit for socioeconomic reasons, the reasonable time point for the test would be 3 months postradiotherapy from our result. Acknowledgement This work was conducted under Research Agreement E33024 (Optimizing Treatment of Cervix Cancer Using Radiotherapy and Analysis of Virally- Associated Cellular Resistance) of the International Atomic Energy Agency, Vienna, Austria.

7 902 Persistence of HPV infection after radiotherapy References 1. Parkin DM, Bray F, Ferlay J, Pisani P. Estimating the world cancer burden: Globocan Int J Cancer 2001;94: Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, CA Cancer J Clin 2005;55: Morris M, Eifel PJ, Lu J, Grigsby PW, Levenback C, Stevens RE, Rotman M, Gershenson DM, Mutch DG. Pelvic radiation with concurrent chemotherapy compared with pelvic and para-aortic radiation for high-risk cervical cancer. N Engl J Med 1999;340: Rose PG, Bundy BN, Watkins EB, Thigpen JT, Deppe G, Maiman MA, Clarke-Pearson DL, Insalaco S. Concurrent cisplatin-based radiotherapy and chemotherapy for locally advanced cervical cancer. N Engl J Med 1999;340: Zannoni GF, Vellone VG. Accuracy of Papanicolaou smears in cervical cancer patients treated with radiochemotherapy followed by radical surgery. Am J Clin Pathol 2008;130: Nagai Y, Maehama T, Asato T, Kanazawa K. Persistence of human papillomavirus infection after therapeutic conization for CIN 3: is it an alarm for disease recurrence? Gynecol Oncol 2000;79: Park JY, Bae J, Lim MC, Lim SY, Lee DO, Kang S, Park SY, Nam BH, Seo SS. Role of high risk-human papilloma virus test in the follow-up of patients who underwent conization of the cervix for cervical intraepithelial neoplasia. J Gynecol Oncol 2009;20: Park JY, Kim DY, Kim JH, Kim YM, Kim YT, Nam JH. Human papillomavirus test after conization in predicting residual disease in subsequent hysterectomy specimens. Obstet Gynecol 2009;114: Datta NR, Kumar P, Singh S, Gupta D, Srivastava A, Dhole TN. Does pretreatment human papillomavirus (HPV) titers predict radiation response and survival outcomes in cancer cervix? a pilot study. Gynecol Oncol 2006;103: Nagai Y, Toma T, Moromizato H, Maehama T, Asato T, Kariya K, Kanazawa K. Persistence of human papillomavirus infection as a predictor for recurrence in carcinoma of the cervix after radiotherapy. Am J Obstet Gynecol 2004;191: Singh RK, Maulik S, Mitra S, Mondal RK, Basu PS, Roychowdhury S, Panda CK. Human papillomavirus prevalence in postradiotherapy uterine cervical carcinoma patients: correlation with recurrence of the disease. Int J Gynecol Cancer 2006;16: Harima Y, Sawada S, Nagata K, Sougawa M, Ohnishi T. Human papilloma virus (HPV) DNA associated with prognosis of cervical cancer after radiotherapy. Int J Radiat Oncol Biol Phys 2002;52: Kim JY, Park S, Nam BH, Roh JW, Lee CH, Kim YH, Shin HJ, Lee SK, Kong SY, Seong MW, Han TJ, Lee MY, et al. Low initial human papilloma viral load implicates worse prognosis in patients with uterine cervical cancer treated with radiotherapy. J Clin Oncol 2009;27: Chung HH, Lee S, Sim JS, Kim JY, Seo SS, Park SY, Roh JW. Pretreatment laparoscopic surgical staging in locally advanced cervical cancer: preliminary results in Korea. Gynecol Oncol 2005;97: Lim MC, Bae J, Park JY, Lim S, Kang S, Seo SS, Kim JY, Rho JW, Park SY. Experiences of pretreatment laparoscopic surgical staging in patients with locally advanced cervical cancer: results of a prospective study. J Gynecol Oncol 2008;19: Ngelangel C, Munoz N, Bosch FX, Limson GM, Festin MR, Deacon J, Jacobs MV, Santamaria M, Meijer CJ, Walboomers JM. Causes of cervical cancer in the Philippines: a case control study. J Natl Cancer Inst 1998;90: Czegledy J, Iosif C, Hansson BG, Evander M, Gergely L, Wadell G. Can a test for E6/ E7 transcripts of human papillomavirus type 16 serve as a diagnostic tool for the detection of micrometastasis in cervical cancer? Int J Cancer 1995;64: Rintala MA, Rantanen VT, Salmi TA, Klemi PJ, Grenman SE. PAP smear after radiation therapy for cervical carcinoma. Anticancer Res 1997;17: Eifel PJ. The importance of locoregional control of cervical cancer why is it still controversial? Cancer J Sci Am 1996;2: Stock RG, Chen AS, Karasek K. Patterns of spread in node-positive cervical cancer: the relationship between local control and distant metastases. Cancer J Sci Am 1996;2: Trimble EL. Cervical cancer state-of-theclinical-science meeting on pretreatment evaluation and prognostic factors, September 27 28, 2007: proceedings and recommendations. Gynecol Oncol 2009;114: Narayan K, Fisher RJ, Bernshaw D, Shakher R, Hicks RJ. Patterns of failure and prognostic factor analyses in locally advanced cervical cancer patients staged by positron emission tomography and treated with curative intent. Int J Gynecol Cancer 2009;19:912 8.