曾志仁醫師， 林政道醫師，周宏學醫師，黃寬仁醫師，張廷彰醫師，賴瓊慧醫師, 宋永魁醫師， 洪志宏醫師， 薛綏醫師， 包家駒教授。

INTRODUCTION

        The detection of circulating tumor cells using the reverse transcription assay followed by polymerase chain reaction (RT-PCR) has recently been applied for a variety of tumors. In gynecologic malignancies, we have recently reported that malignant cells in peripheral blood can be determined by means of the human papillomavirus (HPV) E6 transforming gene mRNA in cervical cancer patients with clinical evidence of distant metastasis; that cervical cancer cells could, indeed, be released into the circulatory system and that such cells could be identified by the RT-PCR method [17]. However, whether the detection of circulating cancer cells may predict the development of metastasis in the future, for those patients without distant metastasis at diagnosis, remains to be determined. The clinical significance and application of this technique will be much broader if circulating cervical cancer cells can be detected in patients with locally advanced cervical cancer, stage IIb to stage IVa, who would have a higher risk of distant metastasis. If the presence of tumor cells in the circulatory system is a prognostic indicator of disease recurrence or indicates a poor prognosis, the detection of circulating tumor cells may result in an earlier and more appropriate selection of cervical cancer patients for further adjuvant therapy, such as chemotherapy.

MATERIALS AND METHODS

        The presence of circulating tumor cells was detected through the identification of transcriptional products of the HPV type 16/18 E6 transforming gene in the peripheral blood, analyzed by a RT-PCR method. From January 1994 to November 1995, 35 locally advanced cervical cancer patients with positive HPV-16/18 DNA were included. All eligible patients received external beam radiation therapy followed by intracavitary brachytherapy. Fifty healthy volunteers donated their peripheral blood to serve as negative controls. Human cervical cancer cell line CaSki, HeLa, SiHa, MT-3, C4-I and C33-A cells were also obtained from the American Type Culture Collection and were used as HPV positive controls. Those patients with documented disease beyond the pelvis were excluded, as were patients with positive paraaortic lymph nodes. The patients could have no history of previous malignancy, chemotherapy, or radiotherapy. The disease was staged according to the International Federation of Gynecology and Obstetrics (FIGO).

Determination of the presence of HPV DNA by PCR: The procedures for the amplification and subsequent detection of HPV types 16 and 18 DNA in clinical specimens were those we described earlier in Cancer, 1997; 80: 91-97. Determination of the presence of circulating cervical cancer cells: we collected 20 ml of citrated venous blood prior to either biopsy or surgery in order to avoid possible introducing of cervical cells into blood stream because of the biopsy or surgical procedures. The procedures for the DNA/RNA extraction and reverse transcription of HPV type 16/18 E6 gene mRNA were described earlier in Journal of Clinical Oncology, 1997; 15:1008-1012.

RESULTS

        The sensitivity of our DNA amplification and detection methods for the HPV RT-PCR assay was that between 10 to 50 cervical cancer cells in 10 ml of blood could be detected. Of the healthy control group, none of the peripheral blood specimens from HPV-DNA positive volunteers contained HPV E6 mRNA. Peripheral blood specimens from 18 of the 35 (51.4%) cervical HPV DNA positive patients were found to contain HPV-specific mRNA in contrast to none for the rest 15 HPV-16/18 DNA negative cervical cancer patients. The frequency of the positive RT-PCR test in peripheral blood at each stage was 6 of 14 (42.9%) in Stage IIb, 10 of 18 (55.6%) in Stage III, and 2 of 3 (66.7%) in Stage IV.

         After a median follow-up of 22 months, the recurrence rate was significantly higher in patients who were RT-PCR positive for HPV E6 mRNA in circulating blood than patients who were RT-PCR negative (10/18, 55.6% versus 2/17, 11.8%, P=0.017, Table1). In addition, there was a statistically significant association with RT-PCR positivity and distant metastasis (P=0.02). Kaplan-Meier analysis showed that actuarial survival after 2 years was significantly worse in the RT-PCR positive patients as compared with the RT-PCR negative patients (mean survival time: 21 versus 33 months, P = 0.04, Fig 1)

CONCLUSION

         The study suggests that the detection of HPV E6 mRNA in circulating blood by a RT-PCR assay can be used to predict the presence of micrometastasis, to help make a prognosis for the patients, and thus might result in an earlier and more appropriate selection of patients for additional therapy.