The investigators design a phase 2, open labeled, randomized trial of Tamoxifen (20 mg/day) and Letrozole (2.5 mg) in treatment of squamous carcinoma of the cervix. Forty four patients with recurrent or persistent disease will be recruited, randomized, treated and followed three-monthly for 12 months. The primary end point is the treatment response rates. Secondary end points include survivals, ECOG performance status, quality of life and efficacy of biomarkers in predicting the responses. Candidate biomarkers including ER, PR, GPER and HPV genotype in paraffin cancer tissues as well as methylated genes in the blood will be studied in relation to the therapeutic outcomes.

Although human papillomavirus (HPV) is a necessary cause of cervical cancer, HPV infection itself is inefficient and insufficient to cause cancer. New evidences have suggested endogenous and exogenous sex hormones confer risk of developing cervical cancer. In unscreened populations, incidence of cervical cancer starts after menarche and constantly increases before menopause, after then the incidence is flattening and declines [21]. Indeed, after menopause, newly incident CIN3 is seldom detected[22]. Large-scale epidemiological studies also showed high number of full-term pregnancy[23] (rather than abortion) and long-term use of hormonal contraceptives[24] to be independent risk factors of cervical cancer. These evidences pointed to female sex hormones to be another culprit of cervical cancer.

The role of estrogen and ERα on HPV-induced cervical carcinogenesis is best demonstrated by the pK14-HPV E6/E7transgenic mice which, without estrogen exposure, develop benign skin tumors only. However, when these mice are treated with exogenous estradiol at physiological level, they develop cervical cancers in nearly 100% efficiency[25-28]. These cervical neoplasia recapitulate characteristics of human cervical cancer in all aspects: originated from the squamous-columnar junction, with early lesions of atypical squamous metaplasia, CIN and to invasive squamous cell carcinoma [29]. Most importantly, removal of exogenous estrogen or castration of these mice led to diminish of progression and partial regression of pre-existing neoplasia[30].

The investigators design an open, randomized, multi-center trial of tamoxifen and letrozole in treatment of recurrent or persistent squamous cell carcinoma of the cervix. Patients with recurrent or persistent SCC of cervix who are not amenable for further cytotoxic treatment will be randomized by block and by participating center to one of the two arms. The block size will be two . Medication will be given orally in daily dose of tamoxifen (Nolvadex) 20 mg, letrozole (Femara) 2.5 mg until disease progression or until the end of the study.

Primary end point of the study is the response rate (complete response and partial response rates) for tamoxifen and letrozole arms. Secondary end points include progression-free survival (PFS) and overall survival (OS) compared to the historical results, ECOG Performance Status, quality of life and outcome predictors (biomarkers and clinical characteristics) of responsiveness and survival. The experienced survival data of the participating center will be compared.

Inclusion Criteria:

1. With a histology proven primary squamous cell carcinoma of the cervix prior to the treatment failure
2. Must sign and date informed consent.
3. With age between 30 and 85
4. With tissue blocks of the recurrent cancer lesion or primary cancer lesion available for the study.
5. With a treatment-free interval of at least 4 weeks.
6. With currently (within 1 month) measurable (by CT) tumor of at least 2 cm in one diameter (at least twice the scan slice thickness), AND elevated SCC level over 2 folds of the institutional upper limit of normal (ULN),
7. With a ECOG performance status score of 0 to 2,
8. With adequate hematologic function (ANC≧500/uL and platelets≧50,000/uL),
9. With adequate renal function (serum creatinine≦2.0 mg/dL; if higher, then creatinine clearance≧40 mL/min was required),
10. With adequate hepatic function (ALT/AST ≦3.0 folds of ULN

Exclusion Criteria:

1. With histology type other than SCC
2. Had liver, brain metastasis or malignant ascites
3. Those having multiple metastasis (more than one metastasis lesion)
4. Whose cancer had been treated for more than three therapeutic courses [including 1 primary therapy (Operation+ CCRT is considered 1 primary therapy) and 2 secondary therapies] courses.
5. Who have received any investigational drugs within 30 days prior to enrollment
6. Who were pregnant or lactating
7. Who are taking selective serotonin receptor inhibitors (SSRI) (eg. Prozac, Celexa, Lexapro, Lubox, Paxi, Zoloft, etc.)
8. With pulmonary embolism or other veneous embolism
9. With uncontrolled medical conditions such as cardiac disease, cirrhosis of liver, active on chronic hepatitis, diabetes mellitus, autoimmune disease.
10. With current or prior therapy (less than 3 months ) of selective estrogen receptor modulators (SERMs) (tamoxifen, raloxifen, fulvestrant, etc.), or aromatase inhibitors (eg. Letrozole, Anastrozole, Exemestane, Vorozole, Formestane, Fadrozole, etc.)
11. Currently taking Warfarin or Rivaroxaben .
12. With history of malignant disease, except those had been disease-free for at least 5 years.
13. Patient who had allergy history to Tamoxifen or Letrozole