Value of routine follow-up procedures for patients with stage I/II cervical cancer treated with combined surgery radiation therapy

Transcription

1 Original article Annals of Oncology 15: , 2004 DOI: /annonc/mdh050 Value of routine follow-up procedures for patients with stage I/II cervical cancer treated with combined surgery radiation therapy P. Morice 1 *, C. Deyrolle 1, A. Rey 2, D. Atallah 1, P. Pautier 2, S. Camatte 1, A. Thoury 1, C. Lhomme 3, C. Haie-Meder 4 & D. Castaigne 1 Departments of 1 Gynecologic Surgery, 2 Biostatistics, 3 Oncology and 4 Radiotherapy, Institut Gustave Roussy, Villejuif, France Received 16 June 2003; accepted 9 September 2003 Background: The aim of this study was to determine the value of routine follow-up for the detection of recurrence in patients treated for cervical cancer. Patients and methods: From 1986 to 1998, 583 women with stage I and II cervical carcinoma were treated with combined surgery radiation therapy. After treatment, follow-up was based on clinical examination, a systematic Pap smear and radiography (chest X-ray and abdomino-pelvic ultrasonography). Results: Forty-five patients had recurrence observed with a delay 6 months following the end of treatment. Thirty-eight patients had symptoms and seven were asymptomatic at the time of their recurrence. Among asymptomatic patients only two recurrences were diagnosed following routine examinations. Survival is similar in asymptomatic and symptomatic recurrent patients. Conclusions: In conclusion, follow-up of patients treated for cervical cancer based on routine Pap smears and systematic radiography does not permit earlier detection of recurrence and does not increase survival. Key words: cervical cancer, chest X-ray, follow-up, pap smears, recurrence Introduction Routine follow-up of cancer patients after the end of their treatment has two major objectives: (i) to diagnose complications related to treatment; and (ii) to detect recurrence earlier in order to, ideally, improve the survival of patients with recurrent disease. The follow-up procedure is based on the history and physical examinations and on the realization of supplementary examinations (radiology, markers or cytology). With respect to gynecological tumors, numerous studies have focused on endometrial cancer, but have failed to demonstrate the usefulness of laboratory examinations (radiology and pap smears) in improving the survival of patients treated for endometrial carcinoma [1 9]. Literature on the value of follow-up procedures in cervical cancer is rarer and more discordant than in endometrial cancer. Most of these publications involve a relatively small number of treated patients (<350), with a low number of recurrences (between 19 and 40) [10 15]; only two series included a large (>500 patients) number of patients treated for cervical cancer [16, 17]. Moreover, several of these papers focus on the study of prognostic factors of relapse rather than the means of recurrence detection [10 13]. The largest series recently published suggests that survival of asymptomatic patients detected during a surveillance program is increased and thus confirms the benefit of follow-up examinations [17]. *Correspondence to: Dr P. Morice, Institut Gustave Roussy, 39 rue Camille Desmoulins, Villejuif, France. Tel: ; Fax: ; morice@igr.fr The aim of this retrospective study, based on a large series of 583 consecutive patients, was to evaluate the usefulness of routine follow-up procedures for the detection of recurrent disease following treatment of stage I/II cervical carcinoma. Patients and methods From January 1986 to December 1998, 583 patients with stage I or II cervical carcinoma (squamous cell, adenocarcinoma or adenosquamous) were treated at the Institut Gustave-Roussy (Villejuif, France) by a combination of surgery and radiation therapy. All patients underwent a radical hysterectomy with pelvic lymphadenectomy by laparotomy. Para-aortic lymphadenectomy was performed systematically during the first 10 years of this study and thereafter for patients with bulky tumors or with small tumors associated with metastatic pelvic nodes determined using frozen section analysis [18]. The modalities of the radio-surgical combinations depended on the age of the patients, tumor stage, tumor size and the period of treatment. The management of stage I/II cervical cancer during the period of this study has been previously described [19]. Briefly, for patients with stage IB or II tumors of 4 cm in diameter, the treatment consisted of pre-operative utero-vaginal brachytherapy (60 Gy) approximately 6 weeks prior to the operation. Up until 1995, patients with bulky tumors ( 5 cm) received pre-operative external pelvic irradiation (20 Gy) followed by vaginal brachytherapy (40 Gy); since 1996, the treatment has been external radiation therapy (45 Gy) with concomitant chemotherapy (cisplatin 40 mg/m 2 /weekly) for patients with stage >IB2 and/or tumor size >4 cm. Young patients (<40 years) with a small tumor (<2 cm) underwent initial surgery with ovarian conservation and transposition, followed, in most cases, by brachytherapy (60 Gy). Post-operative pelvic irradiation (40 45 Gy) 2004 European Society for Medical Oncology

2 219 Table 1. Characteristics of seven asymptomatic patients with recurrence Case No. Age Tumor characteristics Characteristics of disease recurrence Stage Nodal status Examination Location Treatment Delay (months) a Outcome (months) 1 58 IB1 N Pap smears Vagina S+RT 42 Deceased (26) 2 33 IB2 N Clin. exam. Vulva S+RT+CT 23 Deceased (15) 3 45 IB2 N Comp. tomogr. Pelvic RT+CT 9 Deceased (<24) 4 74 IIA N Clin. exam. Vagina plus vagina CT 26 Deceased (10) 5 54 IB2 N Comp. tomogr. Para-aortic RT+CT 13 Deceased (16) 6 35 IB1 N Chest X-ray Chest CT 7 Deceased (14) 7 39 IB2 N Chest X-ray Chest S+CT 9 Deceased (15) a Delay between the end of treatment and the diagnosis of recurrent disease. S, surgery; RT, radiation therapy; clin. exam., clinical examination; comp. tomogr., computed tomography; CT, chemotherapy. was delivered to patients with histologically proven pelvic node involvement (in absence of pre-operative external radiation therapy). Pelvic irradiation (40 45 Gy) and cisplatin-containing chemotherapy (40 mg/m 2 /weekly) were given to patients with metastatic common iliac and/or para-aortic nodes. Postoperative para-aortic irradiation was not performed after complete para-aortic lymphadenectomy. Women with no residual tumor after the completion of treatment were followed up according to the protocol of our institution. A clinical examination, consisting of a gynecological examination and a Pap smear of the vaginal cuff, was performed every 3 months during the first year, every 4 months during the second year, every 6 months during the third year, and every year thereafter. Chest X-ray and abdomino-pelvic ultrasonography were performed annually until 1995 and thereafter for patients with unfavorable prognostic factors (nodal involvement). Squamous cell carcinoma (SCC) antigen determination and magnetic resonance imaging (MRI) examinations were not routinely used during the period of this study. Recurrence was defined as the reappearance of cervical cancer 6 months after the end of treatment. Patients with recurrence occurring within 6 months of the completion of their treatment were not included in the present series. Each recurrence that developed in the patients considered as cured was reviewed. The clinical examination, radiographic studies and the histological and cytological specimens were analyzed. The duration of symptoms before detection of recurrence, the site and treatment of recurrences and survival were also studied. Prognostic factors for the occurrence of recurrence were not studied in the present series because they have been precisely defined previously [19]. Statistical analysis Survival curves were calculated using the Kaplan Meier method. Medians were compared using the Wilcoxon Rank test. Results The number of patients with residual disease, evident (after initial radiation therapy) on the histological specimen taken following radical hysterectomy, was 112 of 253 (44%) patients who underwent pre-operative brachytherapy and 86 of 131 (67%) patients who underwent external radiation therapy (± brachytherapy). Thirty patients had recurrent disease <6 months after the end of treatment and 28 patients had a follow-up <6 months after the end of treatment and therefore were excluded. Forty-five patients had a recurrence 6 months after the end of treatment. The mean age of these patients at the time of treatment of their initial tumor was 42 years (range 22 74). The International Federation of Gynecology and Obstetrics (FIGO) stage of the initial tumors were as follows: IB, n = 27; IIA, n = 3; and IIB, n = 15. Forty patients had squamous cell carcinoma, four patients an adenocarcinoma and one patient an adenosquamous lesion. Nodal status of these patients were negative nodes (n = 26), positive pelvic nodes (n = 13) and positive para-aortic nodes (n = 6). The location of the recurrence was as follows: local pelvic, 25 patients (with pelvic and/or paraortic nodal metastases); distant metastases (except para-aortic node), 13 patients; and local plus metastatic, seven patients. Thirty-two patients had local recurrence and 20 patients a distant metastasis. Patients with local pelvic recurrence had histological confirmation of the recurrent disease. The location of 32 local recurrences was as follows: pelvic alone, 21 patients (centropelvic, 10; lateropelvic, three and extensive pelvic infiltration, eight); nodal metastases (three pelvic and three para-aortic), six patients; and centropelvic plus pelvic nodes, five patients. In 10 patients with centropelvic extension, six had a vaginal tumor, two vaginal with extension to the vulva and perineum, one had a recurrence on the vulva, and one rectal involvement. Of 20 patients with metastases, 12 had isolated metastases (chest, six; supra-clavicular nodes, three; inguinal nodes, two; and in one patient a superior mesenteric node was invaded) and eight multiple metastases (chest, four; liver, four; bone, three; splenic, one; peritoneum, one; skin, one; mediastinal, one; distant nodes, two). Among the patients with recurrence seven patients were asymptomatic (Table 1; patients 1 7). In four patients, recurrent disease was discovered at the time of the routine follow-up consultation. Two asymptomatic patients (Table 1; patients 2 and 4) had pelvic recurrences diagnosed during clinical examination (with normal pap smears). Only two asymptomatic patients had recurrence diagnosed using cytological or radiological examinations: the first one had an abnormal pap smear (with a normal clinical examination of the vagina; Table 1, patient 1) and the other had metastases following a routine chest X-ray (Table 1; patient 7). In three asymptomatic patients (Table 1, patients 3, 5 and 6), recurrences (chest, one; upper pelvic cavity, one; and para-aortic nodes, one) were fortuitously diagnosed during a radiological examination

3 220 Table 2. Characteristics of symptomatic patients with isolated local recurrence Case No. Age Tumor characteristics Characteristics of recurrence disease Stage Nodal status Symptoms Location Treatment Delay (months) a Outcome (months) Local pelvic 8 47 IB2 N Pain, sciatic Centropelvic S+CT 7 Deceased (14) 9 50 IIB N Pain, bleeding Lateropelvic S 128 Deceased (28) IIB N+ Pain, sciatic Pelvic CT 24 Deceased (7) IB1 N+ Pain, sciatic Lateropelvic S 15 Deceased (14) IIB N Pain, lymphoedema Lateropelvic Unknown 24 Lost to FU IIB N Pain, sciatic Pelvic nodes RT 73 Deceased (37) IIB N+ Pain, lymphoedema Pelvic CT 35 Deceased (3) IIB N Pain Pelvic + nodes RT+CT 12 Deceased (23) IB2 N+ Pain, lymphoedema Pelvic Corticoids 24 Deceased (14) IIB N+ Pain, bowel obstruction Pelvic S 2 Deceased (1) IB2 N+ Pain Pelvic nodes Unknown 6 Deceased (17) IB1 N Pain Pelvic + nodes S+CT 108 Alive without disease IB2 N+ Pain, sciatic Pelvic nodes RT 6 Unknown IB1 N+ Pain, fever Pelvic + nodes CT 7 Deceased (6) IIB N+ Pain, lymphoedema Pelvic nodes S+RT 10 Deceased (3) IB1 N Bleeding Centropelvic Brachytherapy 15 Alive with recurrence IIB N Bleeding Pelvic RT+CT 11 Deceased (28) IIB N+ Bleeding Centropelvic S 17 Deceased (30) IIB N+ Bleeding, sciatic Centropelvic + nodes None 7 Deceased (3) IB1 N Pain Centropelvic CT 15 Deceased (12) a Delay between the end of treatment and the diagonsis of recurrent disease FU, follow-up; S, surgery; RT, radiation therapy; CT, chemotherapy. (chest X-ray, one; abdominal CT scan, two) performed for other reasons (before a surgical procedure for a pathology not related to the initial cancer). Thirty-eight were seen for various symptoms before their scheduled follow-up consultation. Characteristics of these patients are given in Table 2 (patients with isolated local recurrence) and Table 3 (patients with distant ± local recurrence). The most common symptom was pain (Tables 2 and 3). In eight other patients symptoms included vaginal bleeding (n = 5), cough (n = 3), two with self-detection of mass (supra-clavicular and inguinal nodes: Table 3, patients 30 and 44). Twenty-seven of 32 (84%) patients with local recurrences and 18 of 20 (90%) patients with distant metastasis were symptomatic. The median disease-free interval (DFI) (delay between the end of treatment and the discovery of recurrence) was 16 months (range 2 128). The DFI of asymptomatic and symptomatic patients were 13 (range 7 42) and 16 (range 2 128) months, respectively (NS). The DFI of patients with distant metastasis and local recurrences were 34 (range 7 67) and 15 months (range 2 128) (NS). Five patients (11%) (all with symptoms) had a recurrence diagnosed after a DFI >5 years (Tables 2 and 3, patients 9, 13, 19, 41 and 42). Location of the recurrent disease in these five patients was as follows: centropelvic, one; lateropelvic, one; pelvic and para-aortic nodes, one; and supra-clavicular nodes, two. Treatment of patients with asymptomatic recurrence is detailed in Table 1. Treatment of patients with symptomatic recurrences is detailed in Tables 2 and 3 and comprised surgery (14 patients), radiation therapy (seven patients), chemotherapy (22 patients) and radiation therapy (external and/or brachytherapy) (eight patients). In 12 patients more than one modality was combined. Overall survival for all patients at 2 and 3 years after diagnosis of the recurrence was 25% [95% confidence interval (CI) 14% to 41%] and 11% (95% CI 4% to 25%), respectively. Only three patients (with local recurrences) were alive after 3 years: one of them died at 68 months, the other currently has progressive disease at 56 months and one is alive without apparent disease 44 months after the treatment of nodal pelvic and para-aortic recurrence with chemotherapy. An exploratory laparotomy was performed in this last patient and confirmed remission of the recurrent disease. The median survival is similar in asymptomatic and symptomatic patients (14 months; Figure 1). All seven patients with asymptomatic recurrence died (within 26 months following the discovery of the recurrence). Only two symptomatic patients

4 221 Table 3. Characteristics of symptomatic patients with distant metastases (± local recurrence) Case No. Age Tumor characteristics Characteristics of recurrence disease Stage Nodal status Symptoms Location Treatment Delay Outcome (months) a (months) Local pelvic + distant IB1 N Bleeding Centropelvic + S+RT+B 20 Deceased (21) inguinal nodes IB2 N+ Pain Nodes + bone + liver CT 31 Unknown IB2 N Discovered by patient Centropelvic + S+RT+B 38 Deceased (68) inguinal nodes IIA N+ Anemy Pelvic + mesenteric CT 8 Deceased (7) nodes IB2 N+ Pain Pelvic + supra-clavicular CT 16 Deceased (5) nodes IIB N+ Pain, lymphoedema Pelvic + supra-clavicular CT 32 Unknown nodes + liver IB2 N+ Pain, bowel obstruction Pelvic + chest S+CT 6 Deceased (3) Distant IIB N Pain, cough, dyspnea Chest + mediastinal Unknown 11 Deceased (1) IB1 N Pain Peritoneum S+CT 36 Deceased (16) IIB N Pain Chest S 13 Deceased (22) IB2 N Pain, dyspnea Chest + bone CT 16 Deceased (6) IIA N Cough Chest S+CT 15 Deceased (11) IB1 N Pain Liver + peritoneum S+CT 50 Deceased (24) IB1 N+ Pain Supra-clavicular nodes CT 63 Deceased (10) IB1 N+ Pain, cough Supra-clavicular nodes CT 67 Deceased (22) IB1 N Cough, fever Chest CT 50 Deceased 20) IB2 N+ Discovered by patient Chest + liver + spleen CT 34 Deceased (13) + nodes IIB N Pain, cough, dyspnea Chest + bone RT+CT 49 Deceased (4) a Delay between the end of treatment and the diagnosis of recurrent disease. B, brachytherapy; S, surgery; RT, radiation therapy; CT, chemotherapy. (nos 19 and 23) are currently alive (one of them, patient 23, with evolutive disease). Discussion The results of the present series confirm the conclusions of others, that most patients with recurrent cervical carcinoma have symptoms, and following the onset of new symptoms, receive medical advice before their routine follow-up visit (interval visit). In the present series, the rate of symptomatic patients with recurrent disease was 85% (38/45). Such results are similar to those observed in patients with endometrial cancer [9]. In contrast to endometrial cancer, cervical cancer is a virus-induced disease related to human papilloma virus (HPV) infection. In the natural history of this tumor, this infection could persist and cause recurrence in the vagina (in patients treated with hysterectomy) or in the vulva. Thus, we undertook a surveillance program with routine pap smears that could theoretically increase the rate of early recurrence and improve survival of recurrent patients. In the series of Soisson et al., the sensitivity and specificity of vaginal cytology for the detection of recurrence were 13% and 100%, respectively [12]. Two of four patients with pelvic recurrence detected using vaginal cytology were salvaged using radiation therapy. The authors concluded that even if vaginal cytology is not sensitive this is the technique most likely to detect recurrent tumor while it is localized and potentially curable [12]. Nevertheless, in the series of Bodurka- Bevers et al., no patient was diagnosed on routine pap smears [17]. In our study, only one patient with a normal clinical examination had recurrence diagnosed using vaginal cytology, but this patient died 26 months after the discovery of recurrent disease. Furthermore, two patients with vaginal recurrence diagnosed during clinical examination had normal pap smears. This clinical issue was previously observed in endometrial cancer [5, 9]. According to Owen and Duncan, the recurrent tumor invades the entire thickness of the vagina before cells can be sampled by routine cytology, and at this stage, the recurrence becomes symptomatic [5]. This explains why cytology could be normal in patients with clinical evidence of recurrent disease on the vaginal vault. These several

5 222 the subgroup of patients with important fibrosis and a potential differential diagnosis of recurrent disease. In conclusion, clinical examination is the most effective method for follow-up of patients with cervical carcinoma treated by radiosurgical combination (with radical hysterectomy). In addition, this follow-up should be performed after 5 years following the end of treatment because late recurrences are observed (11% in present series) [24]. Routine vaginal cytology and radiological studies do not permit earlier detection of recurrences. In the future, the development of new techniques (PET imaging) may be helpful in the follow-up of patients treated for cervical carcinoma. References Figure 1. Overall survival of 45 cervical cancer patients with recurrent disease. points suggest that routine cytology should probably be abandoned in patients treated with radio-surgical combination using radical surgery (hysterectomy). But the interest of routine vaginal cytology should be evaluated in patients treated radically using surgery exclusively without radio-surgical combination. The most recent large series by Bodurka-Bevers et al., suggests that surveillance could improve survival by detecting recurrence earlier [17]. The median survival from recurrence of asymptomatic versus symptomatic patients was 42 versus 11 months (P <0.001) [17]. Samlal et al. reached similar conclusions [14]. Furthermore, in the series of Bodurka-Bevers et al., the median survival of asymptomatic patients with pulmonary recurrence (eight patients) was substantially improved (3 years versus 1 year in symptomatic patients) [17]. Both these authors, along with Soisson et al., recommend the inclusion of a routine chest X-ray in follow-up programs [12, 17]. Nevertheless, in our series the median survival of asymptomatic and symptomatic patients is identical (14 months; Figure 1). All four asymptomatic patients with recurrence diagnosed during a radiological examination died. Those results seem to suggest, as in endometrial cancer, that routine radiology should not be performed in asymptomatic patients. But perhaps, in the near future, PET imaging could be helpful in detecting early recurrent disease [20]. Recent studies suggest that serum SCC measurements could be useful in optimizing treatment and detecting recurrences in patients treated for cervical cancer [21]. In the series of Bolli et al., SCC levels were elevated in 81% of recurrences and the delay between elevation and the clinical diagnosis of recurrence was 7 months [22]. Post-treatment follow-up of this marker could therefore be a predictor of recurrence. In a recent paper, even if recurrence was detected earlier, monitoring of SCC levels did not increase survival [22, 23]. Thus, this marker does not seem to be a cost-effective procedure [24]; nevertheless, it could be helpful in 1. Podczaski E, Kaminski P, Gurski K et al. Detection and patterns of treatment failure in 300 consecutive cases of early endometrial cancer after primary surgery. Gynecol Oncol 1992; 47: Shumsky AG, Stuart GCE, Brasher PM et al. An evaluation of routine follow-up of patients treated for endometrial carcinoma. Gynecol Oncol 1994; 55: Berchuck A, Anspach C, Evans AC et al. Post surgical surveillance of patients with FIGO stage I/II endometrial adenocarcinoma. Gynecol Oncol 1995; 59: Reddoch JM, Burke TW, Morris M et al. Surveillance for recurrent endometrial carcinoma: development of a follow-up scheme. Gynecol Oncol 1995; 59: Owen P, Duncan ID. Is there any value in the long term follow up of women treated for endometrial cancer? Br J Obstet Gynaecol 1996; 103: Agboola OO, Grunfeld E, Coyle D, Perry G. Costs and benefits of routine follow-up after curative treatment for endometrial cancer. Can Med Assoc J 1997; 157: Allsop JR, Preston J, Cocker S. Is there any value in the long term follow up of women treated for endometrial cancer? Br J Obstet Gynaecol 1997; 104: Salvesen HB, Akslen LA, Iversen T, Iversen OE. Recurrence of endometrial carcinoma and the value of routine follow up. Br J Obstet Gynaecol 1997; 104: Morice P, Levy-Piedbois C, Ajaj S et al. Value and cost-effectiveness of routine follow-up of patients treated for endometrial cancer. Eur J Cancer 2001; 37: Larson ML, Copeland LJ, Stringer CA et al. Recurrent cervical carcinoma after radical hysterectomy. Gynecol Oncol 1988; 30: Krebs HB, Helmkamp F, Sevin BU et al. Recurrent cancer of the cervix following radical hysterectomy and pelvic node dissection. Obstet Gynecol 1982; 59: Soisson AP, Geszler G, Soper JT et al. A comparison of symptomatology, physical examination, and vaginal cytology in the detection of recurrent cervical carcinoma after radical hysterectomy. Obstet Gynecol 1990; 76: Gerdin E, Cnattingius S, Johnson P, Pettersson B. Prognostic factors and relapse patterns in early-stage cervical carcinoma after brachytherapy and radical hysterectomy. Gynecol Oncol 1994; 53: Samlal RAK, Van der Velden J, Van Eerden T et al. Recurrent cervical carcinoma after radical hysterectomy: an analysis of clinical aspects and prognosis. Int J Gynecol Cancer 1998; 8: Duyn A, Van Eijkeren MV, Kenter G et al. Recurrent cervical cancer: detection and prognosis. Acta Obstet Gynecol Scand 2002; 81:

6 Ansink A, De Barros Lopes A, Naik R, Monaghan JM. Recurrent stage IB cervical carcinoma: evaluation of the effectiveness of routine follow up surveillance. Br J Obstet Gynaecol 1996; 103: Bodurka-Bevers D, Morris M, Eifel PJ et al. Posttherapy surveillance of women with cervical cancer: an outcomes analysis. Gynecol Oncol 2000; 78: Michel G, Morice P, Castaigne D et al. Lymphatic spread of stage IB/II cervical carcinoma: anatomy and surgical implications. Obstet Gynecol 1998; 91: Morice P, Castaigne D, Pautier P et al. Interest of para-aortic lymphadenectomy in patients with stage IB and II cervical carcinoma. Gynecol Oncol 1999; 73: Grigsby PW, Siegel BA, Dehdashti F, Mutch DG. Posttherapy surveillance monitoring of cervical cancer by FDG-PET. Int J Radiat Oncol Biol Phys 2003; 55: Pras E, Willemse PH, Canrinus AA et al. Serum squamous cell carcinoma after radical hysterectomy: an analysis of clinical aspects and prognosis. Int J Radiat Oncol Biol Phys 2002; 52: Bolli JA, Doering DL, Bosscher JR et al. Squamous cell carcinoma antigen: clinical utility in squamous cell carcinoma of the uterine cervix. Gynecol Oncol 1994; 55: Esajas MD, Duk JM, De Bruijn HW et al. Clinical value of routine serum squamous cell carcinoma antigen in follow-up of patients with early-stage cervical cancer. J Clin Oncol 2001; 19: Chan Y, Ng TY, Ngan HYS, Wong LC. Monitoring of serous squamous cell carcinoma antigens levels in invasive cervical cancer: is it costeffective? Gynecol Oncol 2002; 84: Takehara K, Shigemasa K, Sawasaki T et al. Recurrence of invasive cervical carcinoma more than 5 years after initial therapy. Obstet Gynecol 2001; 98: