Prevention Of Breast Cancer With Tamoxifen: Preliminary Findings From The Italian Randomised Trial Among Hysterectomised Women

Reviewers: John Han-Chih Chang, MD
Source: Lancet 1998; volume 352: pages 93 - 97

Introduction

It is projected that 178,700 new cases of breast cancer will be diagnosed this year. For noninvasive and early stage invasive disease, the cure rate is very high with thenumerous advances in surgery, radiation and chemotherapy. Hormonal therapy with tamoxifen (tam) has been utilized to palliate advanced disease and reduce theincidence of recurrence in the ipsilateral breast/chest wall or of a new second primary in the contralateral breast. No treatment is 100% effective, which is why attemptsare being made in advancing the field of cancer prevention. If tam can reduce the risk of second breast cancers in women with previous breast malignancies, perhaps itcan prevent first primaries in women at higher risk. Unfortunately, from the NSABP B14 trial, one of the major side effects of tam is the increased incidence ofendometrial carcinoma (from 0.2/1000 in the placebo group to 1.6/1000 in the tam group), albeit the risk is still well less than 1%. For this reason, the breast cancerchemoprevention trial detailed in this article enrolled post hysterectomy patients.

Materials and Methods

The Italian Tamoxifen Prevention Study's preliminary results were reported in this Lancet article. Patients enrolled were women who had total hysterectomies (TAH) forreasons other then neoplasm. Most of the patients were recruited from Italy. Eighteen percent had a first degree relative or an aunt with breast cancer. Excluded were patientswith significant history of severe vascular or cardiac compromise. Patients from 35 - 70 years of age were entered from October 1992 to June 1997. Initial intent was for 5years of enrollment, 5 years of intervention and 5 years of follow-up (twice yearly). Recruitment was ended early secondary to poor compliance to remaining on the tam orplacebo. Mammograms were done on an annual basis.

Thirteen thousand four hundred and nineteen women were recruited, but 8011 patients were not randomized secondary to refusal, ineligible or lost to follow-up. Five thousand four hundred and eight were randomized to either daily placebo or tam (20 mg per day x 5 years). This multicenter trial had 51 participating centers in Italy, 3 inSouth America and one in Greece.

Results

Median age was 51 years of age. Three thousand eight hundred and thirty-seven are currently on treatment (either placebo or tam) and 149 have completed their assigned treatment. Twenty-six percent had only a TAH, while 19% had unilateral salpingo-oophorectomies (USO) along with TAH and the rest had TAH and bilateralsalpingo-oophorectomies (BSO).

One thousand four hundred and twenty-two have dropped out (placebo 670 and tam 752). A little over one thousand dropped out on their own accord. The others didsecondary to some adverse event. The most common cause of voluntary withdrawal from the trial was side effects. Over half withdrew in the first year. There were 15 deathsin patients on the trial (9 in the placebo arm and 6 in the tam arm). These were not attributable to breast cancer.

Forty-one cases of breast cancer were recorded. There were 19 cases in the tam arm and 22 in the placebo treated group. Thirty-three occurred while treatment was ongoing(tam 14 and placebo 19). Among women who finished at least one year of treatment, 19 breast cancers were diagnosed in the placebo group and 11 in the tam group (p =0.16 not statistically significant). Of those that received hormone replacement therapy, 8 developed breast cancer in the placebo arm (390 patients) and only one did in thetam arm (362 patients)

The breast cancer characteristics were not different in the two arms. Specifically, no difference was seen in estrogen or progesterone receptor positivity, size, grade,peritumoral invasion, axillary involvement or in-situ disease. A significant increase in the vascular events seen with the addition of tam. Thirty-eight had thrombophlebitis,phlebothrombosis or embolus in the tam arm contrasted to only 18 in the placebo group (p = 0.0053). Likely underestimates, but hypertriglyceridemia was seen in 15 ofthe tam patient and 2 of the placebo group.

Discussion/Conclusions

This trial was an attempt at chemoprevention of breast cancer with tam. They targeted the wrong population. Most (over 85%) of the patients were 59 years old or less.Over 65% were 54 or less. Only 18% had a significant family history risk factor. All patients had had a hysterectomy so that endometrial cancer would not be an issue, butthese patients are at a lower risk for breast cancer than the general population (that is if they had their ovaries out also). Only 26% of the participants had both ovaries.Some (approximately 10 - 15%) were on hormone replacement therapy (simulating normal hormonal levels of the menstrual cycle). This subgroup demonstrated a benefitin utilizing tam. Perhaps it would also be interesting to compare TAH + BSO patients (lower risk) with the TAH only patients (normal risk). With approximately 26%percent not completing the assigned treatment, it also difficult to ascertain the power of this trial. The results are far too immature also to make any conclusions exceptperhaps for the side effects profile of increase vascular events and hypertriglyceridemia.

Two other trials of note in the area of chemoprevention for breast cancer are also important: Royal Marsden Hospital from the United Kingdom and NSABP P1 trial.The trial from UK took patients who had a strong family history of breast cancer and randomized them to tamoxifen or placebo. The frequency of breast cancer wasthe same in both groups. The NSABP trial randomized patients with high risk based on age, lobular carcinoma in-situ, and other risk factors. The benefit was reduction in theoccurrence of invasive breast cancer by 49% (43.4/1000 versus 22/1000) and in-situ carcinoma by 50%. The incidence of estrogen-receptor positive tumors were reduced by 69% with tam.

Thus, the benefits of tamoxifen as a chemo-protectant against breast cancer is still to be defined. The studied patients in the NSABP high risk group derived a benefit. TheItalian study seems to have targeted a group with an overall lesser risk than the general population. The fact that the results are too preliminary and the poor compliancewith assigned treatment make this a questionable study to evaluate the chemoprevention.