The role of pretreatment squamous cell carcinoma antigen in predicting nodal metastasis in early stage cervical cancer

Transcription

1 Acta Obstet Gynecol Scand 2000; 79: Copyright C Acta Obstet Gynecol Scand 2000 Printed in Denmark All rights reserved Acta Obstetricia et Gynecologica Scandinavica ISSN ORIGINAL ARTICLE The role of pretreatment squamous cell carcinoma antigen in predicting nodal metastasis in early stage cervical cancer HAO LIN 1, CHAN-CHAO CHANGCHIEN 1, ENG-YEN HUANG 2, CHIH-WEN TSENG 1, HOCK-LIEW ENG 3 AND CHAO-CHENG HUANG 3 From the Departments of 1 Obstetrics and Gynecology, 2 Radiation Oncology, and 3 Pathology, Chang Gung Memorial Hospital, Kaohsiung, Taiwan Acta Obstet Gynecol Scand 2000; 79: C Acta Obstet Gynecol Scand 2000 Purpose. To evaluate whether the presence of pelvic lymph node metastasis can be predicted by pretreatment squamous cell carcinoma antigen (SCC-Ag) levels in early stage squamous cervical carcinoma. Materials and methods. Between 1994 and 1998, 284 patients with stage Ib and IIa cervical squamous cell carcinoma undergoing radical hysterectomy had preoperative SCC-Ag determination. The correlation between clinicopathological findings on SCC-Ag levels were examined. The Mann-Whitney U test was used to statistically analyze differences between node positive and negative patients. Multiple regression analysis and a multiple logistic model were employed to examine the effect of clinicopathological findings on SCC-Ag levels. Results. Of the 284 patients, 56 patients were found to have nodal metastasis. Median serum levels and 90% ranges of SCC-Ag were 0.74 mg/l ( ) in the 228 nodal negative patients and 4.33 mg/l ( ) in the 56 nodal positive patients (p 0.001). Lymph node metastasis and tumor size were found to have a significant impact on SCC-Ag levels. Around 86% of the patients with SCC-Ag levels below 8 mg/l showed no nodal metastasis, while about 65% of the patients with serum levels above 8 mg/l exhibited nodal metastasis. Multivariate analyses confirmed that only lymph node metastasis had a significant impact on the SCC-Ag levels exceeding 8 mg/l. Conclusion. For predicting nodal metastasis preoperatively, SCC-Ag levels greater than 8 mg/ l can be considered a high-risk zone for nodal metastasis. Key words: cervical cancer; lymph node metastasis; squamous cell carcinoma antigen Submitted 16 July, 1999 Accepted 30 September, 1999 Current therapeutic modalities for the cervical carcinoma classified as FIGO stage Ib and IIa consist mainly of radical hysterectomy with pelvic lymphadenectomy and radiotherapy. Surgery and radiation are considered equally effective, therefore, the choice of the treatment modalities is dependent on the physical condition of the patient and often also the preference of the patient or the Abbreviations: CT: computed tomography; MRI: magnetic resonance imaging; LAG: lymphangiograms. physician. Patients treated surgically are reevaluated postoperatively by a number of pathologic prognostic factors. Of all histopathologic features, lymph node metastasis is considered to be a poor prognostic factor (1 4). These patients usually have a poor clinical outcome despite aggressive adjuvant therapy (5, 6). Therefore, nowadays, the identification of nodal metastasis during operation will usually lead to modified therapy, which involves aborting the surgery and changing to radiotherapy or concurrent chemotherapy and radiation with uterus in place (7, 8). Unfortunately, a sig-

2 SCC-Ag and nodal status in cervical cancer 141 nificant number of complications may occur when staging laparotomy is followed by radiation (9, 10). To avoid these problems, we attempt to evaluate whether a serum marker squamous cell carcinoma antigen (SCC-Ag) can predict the presence of nodal metastasis preoperatively. Materials and methods Between 1994 and 1998, 284 patients with FIGO stage Ib and IIa cervical squamous cell carcinoma undergoing radical hysterectomy and pelvic lymphadenectomy were treated at Kaohsiung Chang Gung Memorial Hospital. None of the patients had received any treatment prior to surgery nor had any disease known to cause elevated serum SCC-Ag level. Blood for the analysis of SCC-Ag was taken from all patients prior to surgery. Serum levels of the marker were determined by a radioimmunoassay using the SCC-RI- ABEAD kit (Abbott Laboratories, USA). The effects of clinicopathological findings including age, stage, tumor diameter, depth of stromal invasion, parametrial invasion, lymph-vascular space invasion, and nodal status on serum SCC-Ag levels were examined. The Mann-Whitney U test was used to statistically analyze differences between nodal positive and negative patients. The tumor diameter was measured in formalin-fixed surgical specimens. The depth of stromal invasion was divided into four categories ( 1/3, 1/3 2/3, ±2/3, full thickness). All the pathologic slides were reviewed by two of the authors (HL Eng and CC Huang). To identify independent factors that might increase serum SCC-Ag level, a multiple regression analysis was used initially to examine the effects of clinicopathological findings on serum SCC-Ag levels. Since the 95th percentile of SCC- Ag levels in patients without nodal metastasis was 7.8 mg/l, a multiple logistic model was employed to examine the effects of clinicopathological findings on serum SCC-Ag levels exceeding 8 mg/l. All statistical analyses were performed with the SPSS for Windows version 7.5 program (SPSS Asia Pacific Pte.Ltd., Singapore). p values lesser than 0.05 were considered statistically significant. Results Of the 284 patients, 56 were found to have nodal metastasis. Median serum levels and 90% ranges (5 95%) of SCC-Ag were 0.74 mg/l ( ) in the 228 node negative patients and 4.33 mg/l ( ) in the 56 node positive patients. There was a significant difference in SCC-Ag levels between the two groups (p 0.001). The median age of the 228 node-negative and 56 node-positive patients was 52 and 45, respectively. The mean number of the metastatic lymph nodes was 1.69, ranging from 1 to 32. Table I shows the distribution of patients according to the nodal status and pathological findings. A significantly higher proportion of patients in the node positive group had advanced stage, large tumor size, deep cervical invasion, and parametrium invasion. Table II shows the results of the multiple regression analysis. Nodal metastasis and tumor diameter were significantly correlated with serum SCC-Ag levels. The multiple logistic model showed that only nodal metastasis had a significant effect on serum SCC-Ag level (Table III). When the cutoff value of SCC-Ag for lymph node metastasis was set at 8 mg/l, the sensitivity was 35.7% and the specificity was 95.2%. Table I. Distribution of patients according to the nodal status and pathological findings Lymph node status Negative Positive nω228 nω56 p-value FIGO stage Ib 200 (88%) 42 (75%) IIa 28 (12%) 14 (25%) Tumor diameter 4 cm 205 (90%) 40 (71%) ±4 cm 23 (10%) 16 (29%) Depth of cervix invasion 1/3 78 (34%) 1 (2%) /3 2/3 25 (11%) 3 (5%) ±2/3 40 (18%) 7 (13%) Full 85 (37%) 45 (80%) Parametrium invasion No 207 (90%) 33 (60%) Yes 21 (10%) 23 (40%) Lymph-vascular space invasion No 213 (93%) 49 (86%) Yes 15 (7%) 7 (14%) Table II. The multiple regression analysis of clinicopathological findings in relation to serum SCC-Ag levels Parameter Standard Variable estimates error t-value p-value Age ª Nodal metastasis FIGO stage Tumor diameter Depth of cervix invasion Parametrium invasion LVSI* ª * LVSI: Lymph-vascular space invasion.

3 142 Hao Lin et al. Table III. The multiple logistic modeling of clinicopathological findings in relation to serum SCC-Ag levels exceeding 8 mg/l Parameter Standard Variable estimates error Wald c 2 p-value Age Nodal metastasis FIGO stage Tumor diameter Depth of cervix invasion Parametrium invasion LVSI* * LVSI: Lymph-vascular space invasion. Twenty of the 31 (64.5%) patients with SCC-Ag level above 8 mg/l had nodal metastasis while 217 of 253 (85.8%) patients with a level below 8 mg/l were node negative. Therefore, the positive predictive value and the negative predictive value were 64.5% and 85.8%, respectively. The sensitivity, specificity and positive predictive value curves for different cutoff SCC-Ag levels are shown in Fig. 1. Discussion As described earlier, the survival rate of node positive patients is clearly inferior to that of node negative ones. If therapeutic procedures might be altered when nodes are involved, it is imperative to identify these patients prior to treatment so that therapy may be altered without surgical exploration. The efficacy of computed tomography (CT) as a predictive adjunct is uncertain. Computed tomography of pelvic nodes is still limited in that microscopic invasion and CT imaging cannot distinguish between reactive enlargement and metastatic lesions (11, 12). Matsukuma et al. (13) concluded that CT scan could be a useful method in the evaluation of paraaortic but not pelvic lymph node metastases. This might be due to the complex pelvic architecture containing normal ovaries or tumors, hydrosalpinx, and intraabdominal adhesions. Other radiographic methods such as lymphangiograms (LAG) and magnetic resonance imaging (MRI) have also been investigated. Scheidler et al. (14) undertook a meta-analysis to compare the utility of LAG, CT and MRI for the diagnosis of lymph node metastasis in patients with cervical cancer. They found that CT, MRI and LAG performed similarly in the detection of lymph node metastasis, however, MRI tended to perform better than LAG or CT. The reason may have been its better contrast resolution and the ability to differentiate nodes from vascular structures and ovaries. They concluded that CT or MRI Fig. 1. The sensitivity, specificity and positive predictive value curves for different cutoff SCC-Ag levels. was less invasive than LAG and could provide assessment of local tumor extent although LAG has a better sensitivity. Hence, CT or MRI should be regarded as the preferred adjunct to the clinical evaluation of lymph node status. Kato and Torigoe (15) first described a radioimmunoassay for the detection of tumor antigens (SCC-Ag) in the sera of women suffering from squamous cell carcinoma of uterine cervix. Since then, many investigators have reported on the potential clinical value of this new tumor marker. It is now well established that serum levels of SCC- Ag are basically correlated with tumor volume or size and extent of the disease (16, 17). Recently, investigators have started to evaluate whether this simple, non-invasive and cost-effective laboratory examination can predict lymph node metastasis. The results vary with respect to different cutoff values. With a cutoff at 2.5 mg/l, Patsner et al. (18) found that the sensitivity was 87% while the specificity was 86%. Fourteen of the 21 (67%) patients with SCC-Ag levels greater than 2.5 mg/l exhibited nodal metastasis. Bolger et al. (19) found that patients with greatly elevated SCC-Ag levels were highly predictive of nodal disease. If the cutoff point was set at 8 mg/l, the positive predictive value was 77.8% but the sensitivity was only 22.6%. Takeshima et al. (20) found that with a cutoff value of 4 mg/l, the positive predictive value and sensitivity were 65% and 59%, respectively. With a large sample size, we find that pretreatment SCC- Ag levels greater than 8 mg/l implies a high-risk for nodal metastasis but the sensitivity is only 36%. If

4 SCC-Ag and nodal status in cervical cancer 143 Table IV. Studies comparing sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of computed tomography (CT), magnetic resonance imaging (MRI), lymphangiography (LAG), and serum SCC-Ag (SCC) level for nodal metastasis in cervical cancer preoperative evaluation Study Method Patients Sensitivity (%) Specificity (%) PPV (%) NPV (%) Matsukuma et al. (13) a CT Subak et al. (21) CT Kim et al. (22) CT e 0.79 f Kim et al. (23) MRI e 0.39 f Subak et al. (21) MRI Hawnaur et al. (24) MRI e 0.28 f Heller et al. (25) LAG Swart et al. (26) LAG e 0.67 f Patsner et al. (18) b SCC Takeshima et al. (20) c SCC Bolger et al. (19) d SCC Present study d SCC a Pelvic lymph nodes only. b Cuffoff value: 2.5 mg/l. c Cuffoff value: 4 mg/l. d Cuffoff value: 8 mg/l. e Data are presented as positive likelihood ratio. f Data are presented as negative likelihood ratio. the cutoff value is set at 2 mg/l, the sensitivity will rise to 75%. However, both the specificity and positive predictive value will drop to 78% and 45.7%, respectively (Fig. 1). Furthermore, at this cutoff point, a logistic regression analysis will also show tumor diameter to be a significant independent variable in addition to nodal status. Similar results will be obtained if cutoff value is set at 3, 4, 5, 6 or 7 mg/l. For purpose of predicting nodal metastasis by using SCC-Ag alone preoperatively, a cutoff value of 8 mg/l will provide an optimal result. Studies comparing sensitivity, specificity, positive predictive value and negative predictive value of CT, MRI, LAG, and serum SCC-Ag level for nodal metastasis are shown in Table IV. With the advancement of modern technology and the understanding of the pathophysiology of cervical cancer, Stenman et al. (27) attempted to detect SCC-Ag gene mrna in peripheral blood by reverse transcriptase-polymerase chain reaction (RT-PCR). They concluded that the presence of SCC-Ag mrna in peripheral blood could predict disease recurrence. Tseng et al. (28) similarly detected human papillomavirus (HPV) (an oncogenic virus) E6 gene mrna in the peripheral blood by RT-PCR, they found that the presence of HPV E6 gene mrna could provide an early marker for metastasis. Both of the results may render interest to the clinicians and further studies are needed to clarify the correlation between SCC-Ag or HPV E6 gene mrna in peripheral blood and nodal metastasis. We concluded that preoperative serum SCC-Ag levels are not sensitive enough to screen for nodal disease in patients with early stage squamous cell carcinoma of uterine cervix. However, an elevated level might provide extra information, in additional to that from CT, MRI or LAG, to the surgeon and to alert them to the possibility of nodal metastasis. References 1. Shingleton HM, Gore H, Soong SJ, Orr JW Jr, Hatch KD, Austin JM et al. Tumor recurrence and survival in stage Ib cancer of the cervix. Am J Clin Oncol 1983; 6(3): Noguchi H, Shiozawa I, Sakai Y, Yamazsaki T, Fukuta T. Pelvic lymph node metastasis of uterine cervical cancer. Gynecol Oncol 1987; 27: Jobson VW, Girtanner RE, Averette HE. Therapy and survival of early invasive carcinoma of the cervix uteri with metastases to the pelvic nodes. Obstet Gynecol 1980; 151: Kinney WK, Alvarez RD, Reid GC, Schray MF, Soong SJ, Morley GW et al. Value of adjuvant whole-pelvic irradiation after Wertheim hysterectomy for early-stage squamous cell carcinoma of the cervix with pelvic nodal metastasis: a matched control study. Gynecol Oncol 1989; 34: Russell AH, Tong DY, Figge DC, Tamimi HK, Greer BE, Elder SJ. Adjuvant postoperative pelvic radiation for carcinoma of the uterine cervix: pattern of cancer recurrence in patients undergoing elective radiation following radical hysterectomy and pelvic lymphadenectomy. Int J Radiat Oncol Biol Phys 1984; 10: Hogan WM, Littman P, Griner L, Miller CL, Mikuta JJ. Results of radiation therapy given after radical hysterectomy. Cancer 1982; 49: Potter ME, Alvarez RD, Shingleton HM, Soong SJ, Hatch KD. Early invasive cervical cancer with lymph node involvement: to complete or not to complete radical hysterectomy? Gynecol Oncol 1990; 37: Bremer GL, van der Putten HW, Dunselman GA, de Haan J. Early stage cervical cancer: aborted versus completed radical hysterectomy. Eur J Obstet Gynecol Reprod Biol 1992; 47:

5 144 Hao Lin et al. 9. Weiser EB, Bundy BN, Hoskins WJ, Heller PB, Whittington RR, DiSaia PJ et al. Extraperitoneal versus transperitoneal selective paraaortic lymphadenectomy in the pretreatment surgical staging of advanced cervical carcinoma: a Gynecologic Oncology Group study. Gynecol Oncol 1989; 33: Stehman FB, Bundy BN, DiSaia PJ, Keys HM, Larson JE, Fowler WC. Carcinoma of the cervix treated with radiation therapy: A multivariate analysis of prognostic variables in the Gynecology Oncology Group. Cancer 1991; 67: Grumbine FC, Rosenshein NB, Zerhouni EA, Siegelman SS. Abdominopelvic computed tomography in the preoperative evaluation of early cervical cancer. Gynecol Oncol 1981; 12: Chen SS, Kumari S, Lee L. Contribution of abdominal computed tomography (CT) in the management of gynecologic cancer: Correlated study of CT image and gross surgical pathology. Gynecol Oncol 1980; 10: Matsukuma K, Tsukamoto N, Matsuyama T, Ono M, Nakano H. Preoperative CT study of lymph nodes in cervical cancer Its correlation with histological findings. Gynecol Oncol 1989; 33: Scheidler J, Hricak H, Yu KK, Subak L, Segal MR. Radiological evaluation of lymph node metastases in patients with cervical cancer. A meta-analysis. JAMA 1997; 278(13): Kato H, Torigoe T. Radioimmunoassay for tumor antigen of human cervical squamous cell carcinoma. Cancer 1977; 40: Kato H, Morioka H, Tsutsui H, Aramaki S, Torigoe T. Value of tumor antigen (TA-4) of squamous cell carcinoma antigen in predicting the extent of cervical cancer. Cancer 1982; 50: Duk JM, de Bruun HWA, Groenier KH, Hollema H, Hoor HAT, Krans M et al. Cancer of the uterine cervix: Sensitivity and specificity of serum squamous cell carcinoma antigen determinations. Gynecol Oncol 1990; 39: Patsner B, Orr Jr JW, Allmen T. Does preoperative serum squamous cell carcinoma antigen level predict occult extracervical disease in patients with stage Ib invasive squamous cell carcinoma of the cervix? Obstet Gynecol 1989; 75: Bolger BS, Dabbas M, Lopes A, Monaghan JM. Prognostic value of preoperative squamous cell carcinoma antigen level in patients surgically treated for cervical carcinoma. Gynecol Oncol 1997; 65: Takeshima N, Hirai Y, Katase K, Yano K, Yamauchi K, Hasumi K. The value of squamous cell carcinoma antigen as a predictor of nodal metastasis in cervical cancer. Gynecol Oncol 1998; 68: Subak LL, Hricak H, Powell CB, Azizi L, Stern JL. Cervical carcinoma: computed tomography and magnetic resonance imaging for preoperative staging. Obstet Gynecol 1995; 86: Kim SH, Choi BI, Lee HP, Kang SB, Choi YM, Han MC et al. Uterine cervical carcinoma: comparison of CT and MR findings. Radiology 1990; 175: Kim SH, Kim SC, Choi BI, Han MC. Uterine cervical carcinoma: evaluation of pelvic lymph node metastasis with MR imaging. Radiology 1994; 190: Hawnaur JM, Johnson RJ, Buckley CH, Tindall V, Isherwood I. Staging, volume estimation and assessment of nodal status in carcinoma of the cervix: comparison of magnetic resonance imaging with surgical findings. Clin Radiol 1994; 49: Heller PB, Maletano JH, Bundy BN, Barnhill DR, Okagaki T. Clinical-pathologic study of stage IIB, III and IVA carcinoma of the cervix: extended diagnostic evaluation for paraaortic node metastasis a Gynecologic Oncology Group study. Gynecol Oncol 1990; 38: Swart E, Bouma J, Schuur K. The clinical value of lymphography in cervical cancer, FIGO-stage Ib-IIa. Eur J Gynaecol Oncol 1989; 10: Stenman J, Lintula S, Hotakainen K, Vartiainen J, Lehvaslaiho H, Stenman UH. Detection of squamous-cell carcinoma antigen-expressing tumor cells in blood by reverse transcriptase-polymerase chain reaction in cancer of the uterine cervix. Int J Cancer 1997; 74: Tseng CJ, Pao CC, Lin JD, Soong YK, Hong JH, Hsueh S. Detection of human papillomavirus types 16 and 18 mrna in peripheral blood of advanced cervical cancer patients and its association with prognosis. J Clin Oncol 1999; 17: Address for correspondence: Dr. Chan-Chao ChangChien Department of Obstetrics and Gynecology Chang Gung Memorial Hospital 123, Ta Pei Road, Niao Sung Hsiang Kaohsiung Hsien, Taiwan