Switching to anastrozole versus continued tamoxifen treatment of early breast cancer. Updated results of the Italian tamoxifen anastrozole (ITA) trial.

BACKGROUND: Tamoxifen, for many years the 'gold standard' in the adjuvant setting for the management of endocrine sensitive early breast cancer, is associated with an increased risk of endometrial cancer and other life-threatening events. Moreover, many women relapse during or after tamoxifen therapy due to the development of resistance. This provided the rationale for a switching trial with anastrozole, the updated results of which are reported here. PATIENTS AND METHODS: This trial investigated the efficacy of switching to anastrozole for women already receiving tamoxifen. After 2-3 years of tamoxifen treatment, postmenopausal, node-positive, ER-positive patients were randomized to receive either anastrozole 1 mg/day or to continue tamoxifen, 20 mg/day, giving a total duration of 5-years treatment. The primary end point was disease-free survival and secondary endpoints were event-free survival, overall survival and safety. RESULTS: A total of 448 patients were enrolled. At a median follow-up time of 64 months (range 12-93), 63 events had been reported in the tamoxifen group compared with 39 in the anastrozole group [HR 0.57 (95% CI 0.38-0.85) P = 0.005]. Relapse-free and overall survival were also longer in the anastrozole group [HR 0.56 (95% CI 0.35-0.89) P = 0.01 and 0.56 (95% CI 0.28-1.15) P = 0.1]. However, the latter difference was not statistically significant. Overall more patients in the anastrozole group experienced at least one adverse event (209 versus 151: P = 0.000). However, numbers of patients experiencing serious adverse events were comparable (37 versus 40, respectively: P = 0.7). CONCLUSIONS: Switching to anastrozole after the first 2-3 years of treatment was confirmed to improve event-free and relapse-free survival of postmenopausal, node-positive, ER-positive early breast cancer patients already receiving adjuvant tamoxifen.

Cyclophosphamide, methotrexate, and fluorouracil versus tamoxifen plus ovarian suppression as adjuvant treatment of estrogen receptor-positive pre-/perimenopausal breast cancer patients: results of the Italian Breast Cancer Adjuvant Study Group 02 randomized trial. boccardo@hp380.ist.unige.it.

PURPOSE: To compare the efficacy of chemotherapy versus that of tamoxifen plus ovarian suppression in pre-/perimenopausal estrogen receptor-positive patients with early breast cancer. PATIENTS AND METHODS: Patients were randomly assigned to receive either six cycles of a standard regimen of cyclophosphamide 100 mg/m(2) orally days 1 to 14, methotrexate 40 mg/m(2) intravenously (IV) days 1 and 8, and fluorouracil 600 mg/m(2) IV days 1 and 8 (CMF), with all drugs restarted on day 29, or 5 years of tamoxifen, 30 mg/d, plus ovarian suppression with surgical oophorectomy, ovarian irradiation, or monthly goserelin 3.6-mg injections. Disease-free survival was the main study end point. Overall survival and toxicity were additional end points. RESULTS: Between 1989 and 1997, 120 patients were assigned to CMF and 124 to tamoxifen and ovarian suppression (oophorectomy, n = 6; ovarian irradiation, n = 31; and goserelin injections, n = 87). At the time of analysis (median follow-up time, 76 months; range, 9 to 121 months), 82 patients had relapsed and 39 had died. No difference between groups had emerged with respect to either disease-free or overall survival. Treatments were comparable even in respect to age, tumor size, and nodal status, although a nonsignificant trend favored patients with poorly differentiated tumors treated with CMF. Leukopenia, nausea, vomiting, stomatitis, and alopecia were significantly more common in patients treated with CMF. There were few patients who developed benign gynecologic changes in either group, and numbers were comparable. CONCLUSION: The combination of tamoxifen with ovarian suppression seems to be safe and to yield comparable results relative to standard CMF.

The management of elderly patients with T1-T2 breast cancer treated with or without radiotherapy.

AIM: The aim of the current study is to identify a subgroup of patients with breast cancer who have a low risk of local recurrence after conservative surgery in order to avoid radiotherapy treatment. METHODS: A group of 472 patients underwent conservative surgery without radiotherapy, and it was compared to a second group of 755 patients with similar characteristics, but who had received radiotherapy treatment (RT) after conservative surgery. RESULTS: Breast relapse's univariate analysis demonstrated statistical significance for the following factors: radiotherapy treatment, clinical stage, pathological stage, positive axillary nodes and tumour grading. Different results were obtained studying breast relapse. In the no-RT group breast relapse was 10.6% while in the irradiated group it was 3.4%. The breast relapse incidence decreases as the age of the patients increases especially over 75 years of age. CONCLUSIONS: In conclusion, there is clinical evidence of avoiding adjuvant radiotherapy for patients over 75 years with T1-T2 cancer treated with quadrantectomy with a clear excision margin.

Measurement of somatostatin receptor subtype 2 mRNA in breast cancer and corresponding normal tissue.

Somatostatin analogs are effective in inhibiting growth of human breast cancer cell lines. These antiproliferative effects are mediated by specific receptors located on cell membranes. The somatostatin receptor subtype 2 (sst2) is the principal mediator of somatostatin effects in normal and cancer cells, and its presence has already been demonstrated in breast cancer. The purpose of our study was to evaluate the clinical relevance of the expression of sst2 by quantifying its mRNA in a large group of infiltrating breast cancers and their corresponding normal tissues. The expression of sst2 mRNA was measured with quantitative real time RT-PCR in 169 breast cancers and in their corresponding unaffected tissues. We evaluated the association of sst2 expression with the commonest clinical-pathologic features of breast cancer. The correlation with a marker of cell proliferation (Ki-67) and with receptor concentration was also evaluated. In cancer tissues, we found that the absolute concentrations of sst2 mRNA were significantly higher in estrogen receptor (ER)-positive samples (P=0.002) as well as in lymph-node-negative cancers (P=0.04) (Student's t-test or one-way ANOVA). In addition, sst2 mRNA was significantly higher in breast cancers than in corresponding unaffected tissues (P=0.0002). However, when the clinical-pathologic parameters were considered, this gradient maintained its statistical significance only in tumors expressing positive prognostic markers, such as the presence of ER (P=0.0005) and progesterone receptors (PgR) (P=0005), and the lack of lymph-node involvement (P=0.0003). The same difference was also significant in postmenopausal women (P=0.001) and in T1 patients (P=0.001). In addition, sst2 mRNA expression was significantly higher (P=0.008) in low-proliferating breast cancers. Finally, we found that the quantitative expression of sst2 mRNA was directly related to the PgR concentration in breast cancer tissues (P<0.001). Our data seem to indicate that an upregulation of sst2 gene expression is a common feature of breast cancers which, on the basis of conventional predictive parameters, are expected to have a better prognosis. Featuring a possible role of somatostatin analogs in combined endocrine therapies for breast cancer, our results seem to confirm that the sst2 status of the tumor should be previously investigated.

Non-palpable lesions of the breast detected by mammography--review of 1182 consecutive histologically confirmed cases.

We report on 1182 consecutive histologically confirmed non-palpable breast lesions detected by mammography (infiltrating carcinoma 427, in situ carcinoma 121, benign 634). The proportion of cancer cases varied according to age (< 50 years = 33%; 50-59 years = 46%; > 59 years = 63%), mammographic pattern (regular opacities = 8%, parenchymal distortions = 20%, isolated calcifications = 42%, irregular opacities = 62%, stellate opacities = 73%), and calendar period (1970-1985 = 29%, 1986-1989 = 56%; 1990-1992 = 69%). A sharp decrease of the benign/malignant biopsy ratio was evident after routine fine-needle aspiration cytology (sonography-guided or stereotaxic) was introduced in 1986. The independent significant association of cancer frequency to age, calendar period and mammographic pattern was confirmed by multivariate analysis. A significant trend over time in favour of conservative surgery was also observed for cancer cases (1970-1979 = 6%, 1980-1985 = 41%, 1986-1992 = 83%). Among invasive cancers, node involvement was observed in 11.5% of cases, being associated with tumour size (pT1a = 0%, pT1b = 7%, pT1c = 13%, pT2a = 33%). Five-, ten- and fifteen-year overall survivals of invasive cancers were 98.1, 95.7 and 87.3%, respectively.

Monitoring surgical treatment of screen-detected breast lesions in Italy.

The object of this study was to assess quality of care and adherence to treatment guidelines of screen-detected lesions in Italy using a new audit system. Data on screen-detected cases surgically treated in 1997 were collected using a system (QT 2.3) developed within the Italian Group for Planning and Evaluating Mammographic Screening Programmes (GISMa) and the European Breast Cancer Screening Network. Results of 18 performance parameters were considered compared with the reference standards. In 1997, 515 lesions (335 invasive, 60 in situ and 120 benign) in 496 patients were collected from 14 departments in the Central and Northern area of Italy. The 18 indicators were analysed and grouped according to six quality objectives. Some results were good and others were excellent, such as intraoperative identification, breast conservation surgery, adequate axillary procedures and completeness of pathology reports, but most of them failed: waiting times, preoperative diagnosis, employment of frozen section on small lesions and avoiding axillary procedures in ductal carcinoma-in-situ. This work is a first attempt in Italy to evaluate and uniform the criteria adopted for quality control of breast cancer treatment, using a standardised system. Some results are good or excellent, the overall level of compliance with quality indicators is not satisfactory and corrective actions should be undertaken for a number of issues. A continuous monitoring should be performed and appropriate action taken in order to verify the effectiveness of the corrective actions and to provide screen-detected patients with the best quality of care.

Intraductal breast cancer: review of 183 consecutive cases.

The aim was to review a series of 183 intraductal breast cancer to define the diagnostic features and therapeutic outcome. Patients' average age was 54 years. Diagnostic procedures employed were clinical examination, mammography in 173 cases and fine-needle aspiration cytology in 98 cases. The sensitivity of clinical examination was 0.61, of mammography 0.74, of fine needle aspiration cytology 0.70. The sensitivity of clinical examination and mammography associated was 0.93. The surgical options adopted were: conservative surgery 80 cases, mastectomy 103 cases. Conservative surgery was followed by breast irradiation in 34 cases. Axillary dissection was performed in 122. 97 cases have been reviewed histologically. 60% of ductal carcinoma in situ (DCIS) were multifocal and 22% multicentric. Local recurrence, all infiltrating, occurred in the same breast after conservative surgery in 8 cases, 3 of which had received postoperative radiotherapy, and in 3 patients after mastectomy. Contralateral breast cancer was recorded in 13 cases, being synchronous in 4 (infiltrating in 3, DCIS in 1) and metachronous in 8 (all infiltrating). 3 patients died of breast cancer. The present series confirms the risk of ipsilateral cancer recurrence after conservative surgery but there are no significant differences relating to mammographic pattern, size, histological type, margin involvement and radiotherapy.

Prognostic significance of estrogen receptor determination in primary breast cancer.

The authors report on 767 consecutive primary Stage I-II breast cancer cases followed-up from 3 to 8 years. The estrogen receptor (ER) content was determined in all cases and did not influence the treatment choice. A correlation was attempted between ER and menstrual or pathological nodal status (N) or the 5-year disease-free survival (DFS). ER was correlated with menopausal status ER+ cases being more frequent in postmenopausal patients, whereas no correlation was observed between ER and nodal status. In absence of nodal involvement (N-) the prognosis was not influenced by the ER status. A significantly better DFS was evident for ER+ respect to ER- patients in the N+ series but such a correlation is questionable as the adjuvant treatment (hormone or chemotherapy) given to such patients may have influenced the DFS according to the ER status. According to the present study, ER determination should not be used as a discriminant in the performance of adjuvant postoperative treatment based on a prognostic judgment.

Cancer reoccurrence in the conserved breast - diagnostic features in a consecutive series of 102 cases.

Diagnostic features of 102 consecutive cases of cancer reoccurrence in the conserved breast are reported. The sensitivity of palpation, mammography, cytology or ultrasonography was 75.5%, 58.7%, 77.4% and 77.8%, respectively. Mammography failures were not due to natural breast density and could probably be explained by the masking effect of surgical scar and distortion. Most failures at cytology were due to inadequate sampling. Palpation should always be associated with mammography in the follow-up of the conserved breast. Aspiration cytology should.be performed incase of any abnormality seen at palpation or mammography.

Type-2 somatostatin receptor mRNA levels in breast and colon cancer determined by a quantitative RT-PCR assay based on dual label fluorogenic probe and the TaqMan technology.

We reported previously that the expression of type 2 somatostatin receptor (sst2) was positively related to patient outcome in the childhood tumor neuroblastoma. To quantitate the expression of mRNA sst2 expression, we used a competitive RT-PCR assay. To improve the practicability of this measurement and its applicability to large groups of patients, we present here an original 'real-time' quantitative RT-PCR method, based on a dual-labeled fluorogenic probe and the TaqMan technology. By this method, we have measured sst2 mRNA expression in 24 breast cancer samples and 26 colon carcinomas as well as on the corresponding non-adjacent non-neoplastic tissue of the same patients. The proposed method has a dynamic range of 4 x 10(4) to 4 x 10(8) molecules of sst2 mRNA. The intra-assay precision of the test, evaluated as signal detection variability, was 2.4%. Accuracy, evaluated by the addition of standard RNA to unknown samples, provided a mean recovery of 98+/-2%. A significant correlation has been observed in a study performed in 24 neuroblastoma samples measured both with the proposed method and with a competitive RT-PCR assay (r=0.913, p<0.001). In our preliminary clinical study, no significant differences were observed in sst2 mRNA levels between normal and tumor specimens in both colorectal (normal tissue 5.1 x 10(7)+/-2.0 x 10(7) molecules/microg total RNA, cancer tissue 9.7 x 10(7)+/-4.2 x 10(7)) and breast tumors (normal tissue 5.5 x 10(8)+/-2.0 x 10(8), cancer tissue 4.4 x 10(8)+/-3,7 x 10(8)).However, in colorectal cancer, sst2 mRNA values of subjects with high circulating carcinoembryonic antigen (CEA) levels (>5 ng/ml) were statistically lower (2.3 x 10(7)+/-6.2 x 10(6) molecules/, microg total RNA; p<0.05) than those of subjects with low CEA concentration (1.4 x 10(8)+/-6.7 x 10(7)). Also, the sst2 mRNA ratio between normal and tumor tissue (N/T ratio) resulted significantly inversely related to CEA levels. In breast cancer, a significant difference was found between the mean N/T ratio of negative (below 10 fmol/mg protein) and positive estrogen receptor tumors (p<0.05). Analogous results were found selecting breast tumors on the basis of the progesterone receptor status (p<0.05). The proposed method is accurate, precise, sensitive and less labor-intensive than the competitive RT-PCR assay. For a correct evaluation of sst2 mRNA expression, it seems very important to measure the sst2 expression both in tumor and in the non-tumoral non-adjacent tumor specimens.

The role of a new monoclonal antibody assay in the detection of recurrent breast cancer.

The immunoradiometric assay "CA 15-3", recently developed to measure a breast tumor-associated antigen, gave a mean serum value of 13.8 U/ml (S.D. 6.2) for this antigen in 156 non-cancer controls (36 biopsies for a benign breast lesion and 120 healthy controls). Setting a cut-off value of 30 U/ml (specificity 99.3%), only 3 out of 58 primary breast cancer cases were positive. In metastatic breast cancer, 11 out of 33 cases with limited recurrence (33.3%) and 36 out of 56 cases with extensive recurrence (64.3%) gave abnormal values in this assay, above the cut-off point, with an overall sensitivity of 52.8%; the difference between the sensitivity values in the two groups of recurrent cases was statistically significant (P less than 0.01). According to the findings of the present study, CA 15-3 has no role in the detection of primary breast cancer, but its usefulness in disease monitoring can be hypothesized, as circulating levels of the antigen seem to be dependent on the tumor mass.

Cell kinetics and biochemical parameters in breast cancer.

The study analyzes biochemical and cell kinetic parameters to characterize solid tumor growth in humans. The concentrations of polyamines, CEA, the thymidine labeling index (T.L.I.) and the mitotic index (M.I.) were determined on fragments of neoplastic tissue from 18 patients with breast carcinoma. Urinary polyamines were evaluated in the same patients. Two groups of patients were distinguished according to the median value of the with high T.L.I., M.I. and tissue polyamines were significantly higher than in the group with low T.L.I., whereas tissue CEA was lower, though in a not statistically significant way. Urinary polyamines showed no variations between groups. These preliminary results showed that T.L.I. levels were higher in patients who relapsed during a 4-year follow-up than in patients achieving complete remission and remaining disease free. Results concerning polyamine concentration showed that the tissue polyamine level in breast carcinoma indicated proliferative activity, but this does not seem to be valuable for current prognostic purposes.

[Somatostatin receptors in non-endocrine tumours]. / Recettori per la somatostatina in tumori non endocrini.

The study of the antiproliferative action of somatostatin (ss) is important not only to understand the regulation of neuroendocrine tumours that express receptors (sst), but also non-endocrine tumours which express these receptors. We previously demonstrated the presence of sst2 in a wide panel of cell lines from human neuroblastoma. Although hypotheses have been put forward that treatment with ss or its analogs may be beneficial in oncological patients, this does not appear to be the case in neuroblastoma; patients with high sst2 levels (who are therefore sensitive to ss treatment) have per se a relatively positive outcome. Therefore, adjuvant treatment with ss is not necessary. Viceversa, patients with a poor prognosis are essentially characterized by a low expression of sst2 (and therefore are insensitive to a therapy with ss). In these patients adjuvant treatment with ss might be indicated, but would have little chance of success. Although the majority of neuroendocrine tumours expresses sst2, pancreas and prostate cancer express sst1 but not sst2, and are therefore insensitive to octreotide treatment which binds preferentially to sst2. Tumours like colorectal carcinoma and breast cancer also express sst2 in their more favourable forms. However, the concentration of sst2 in colorectal cancer is similar, if not lower than that in the surrounding normal tissue. Therefore, the probability of successful adjuvant therapy with ss is relatively low. In breast cancer, it is possible that sensitivity to estrogens may have a positive influence on the expression of sst2. This might justify clinical trials with ss in breast cancer.

New trends in the surgical treatment of primary breast cancer in the Florence District: 1979-1985.

The authors report on a consecutive series of 1979 TIS-T1A-T2A-T3A breast cancer cases surgically treated from 1979 to 1985. The study investigates the impact of recent reports on conservative surgery and of a national treatment protocol for breast cancer, encouraging the use of modified radical mastectomy (MRM) and quadrantectomy, axillary dissection and breast irradiation (QUART) as alternatives to Halsted's radical mastectomy (RM). RM frequency dropped from 52% to 8% in TIS-T1A and from 82% to 20% in T2A-T3A cases in the study period, in favor of MRM and of QUART, which increased from 10% to 54% in TIS-T1A cases. Although surgical choices differ widely according to individual surgeons, a definite trend towards more conservative surgery is evident, which demonstrates how the national treatment protocol has affected current surgical practice.

Lymphoscintigraphy and gamma probe tracing in detecting breast cancer lymph node involvement: can they replace axillary lymph node dissection?

BACKGROUND: Axillary lymph node status is the most important pathological determinant of prognosis in early breast cancer. However, axillary lymph node dissection (ALND) performed for pathological assessment is not without costs and morbidity. Recently, radioisotope-guided sentinel node biopsy (SNB) has been proposed as a promising technique for staging breast cancer patients. AIM OF THE STUDY: In this study we report our experience (76 patients) in radioguided sentinel node (SN) biopsy in breast cancer. The study was divided into two phases: the first represents our learning curve, necessary to establish our guidelines for its use in clinical practice, while the second phase was aimed at assessing the feasibility of SN localization using preoperative lymphoscintigraphy and intraoperative gamma probe (GP) detection. METHODS: All patients underwent lymphoscintigraphy (LS) up to two hours after tracer delivery (99mTc-micro-nanocolloid, four i.d. injections of 200 microCi/200 miccroL around the primary lesion) 24 hours before surgery and GP tracing during surgery. Subsequently ALND was performed for pathological assessment. RESULTS: SNs were identified in 73/76 patients using LS and in 72/76 using GP. In one case the SN was detected by GP alone while in two cases GP was not able to locate the SN although it had been identified by means of LS. Thirty-three of these 73 patients had axillary node involvement. In 31/33 cases the SN was the only positive node. No positive nodes were found in the remaining 40 ALNDs where SNs were identified. Thus, according to our experience 40/73 ALNDs could have been avoided. SNB seems to be a very interesting technique but further experience in lymph node radioisotope tracing is needed.

Long-term sequelae of breast cancer surgery.

BACKGROUND: Quality of care is today a major issue in oncology, and much attention is given to research on the outcome of breast cancer care. Too little attention has been devoted in the scientific literature to the consequences of treatment in long-term survivors, and in particular to the possible side effects. The specific aim of this contribution is to present population-based data about the long-term impact of breast cancer care in women who had an incident cancer in 1985/1986. PATIENTS AND METHODS: The cases are 476 breast cancers incident in the City of Florence in 1985-86. Women still living 5 years later were invited to have an interview and a physical examination. Lymphedema, peripheral nerve lesions and damage to the shoulder were assessed. RESULTS: Of the 346 5-year survivors, 238 accepted our invitation: 35.2% of the women reported some early postoperative sequelae, 30.2% had a chronic lymphedema and 18.9% a shoulder deficit. Comparing breast-conserving surgery with radical mastectomy, the risk of chronic lymphedema (OR = 1.62; 95% C1: 0.91-2.88) and other lesions was higher for women who had a radical surgery. Women who had a breast-conserving surgery more often reported an early lymphedema (OR = 1.60; 95% Cl: 0.88-2.88). CONCLUSIONS: The proportion of women who complained of (or manifested at the physical examination) a minor or major disability of the arm in our study was high. The impact of these functional problems in terms of quality of life should also be assessed, but it is our impression that there is need for much greater attention to the issue of long-term survivor sequelae.

[Role of axillary dissection in breast carcinoma]. / Ruolo della dissezione ascellare nel carcinoma mammario.

Between 1986 and 1995, 965 patients with T1 breast cancer were treated with wide excision (796 pts) or mastectomy (138 pts). Axillary dissection was performed in 934 patients. Nodal positivity was 8.7% for T1a, 15.6% for T1b and 37.7% for T1c. 68.5% of cases had one to three positive nodes, 18.3% four to ten and 13% more than ten nodes. No differences were observed in the frequency of lymph nodes involved according to the age of patients, menopausal status, pattern at mammography, histology and receptor status. Favourable histology (tubular, cribriform, papillary, colloid) shows a significant difference when compared with all the others histotypes. On the basis of data collected it is impossible to anticipate the axillary status and avoid axillary dissection.

[Lipoma of the colon: apropos of 2 cases]. / I lipomi del colon: a proposito di due casi.

The lipoma of the colon is a rare benign tumour. The Authors describe two cases of colonic lipoma and stress diagnostic and therapeutic problems related to this condition. They also emphasize the importance of double contrast enema and colonoscopy associated to bioptic examination. The indication to lipomas removal is based on clinical remarks and differential diagnosis with adenomatous polyps. The modality of removal, where feasible for local conditions, should be preferably endoscopic.

Nonpalpable lesions detected with mammography: review of 512 consecutive cases.

The authors reviewed 512 consecutive cases in which biopsies were performed on nonpalpable lesions of the breast detected with mammography from 1970 to 1985. The positive predictive value (PPV) of different radiologic patterns was determined. Biopsy is advised in the presence of patterns such as parenchymal distortion, opacities with undefined borders, strongly suspect microcalcifications (rodlike or branching morphologic features, high spatial density), and stellate opacities, which yielded higher PPVs (0.11, 0.35, 0.56, and 0.75, respectively). The average detection rate of preclinical cancer was 0.7 per 1,000; wide variations were recorded during different periods in the study (0.2 per 1,000 and 1.3 per 1,000 in the first 4 years and the last 4 years of the study, respectively), for the different age groups considered (0.6 per 1,000 and 1.7 per 1,000 for the 30-39-year and over-59-year age groups, respectively), and for different reporting radiologists (0.5-1.9 per 1,000). The ratio of benign to malignant biopsy findings was 2.4:1 for the whole series, with wide variations encountered relative to the variables mentioned above.

Intensive diagnostic follow-up after treatment of primary breast cancer. A randomized trial. National Research Council Project on Breast Cancer follow-up.

OBJECTIVE: To evaluate the effectiveness of early detection of intrathoracic and bone metastases in reducing mortality in breast cancer patients. DESIGN: Randomized clinical trial allocating breast cancer patients to two alternative follow-up protocols (intensive vs clinical) for at least 5 years. SETTING: Twelve breast clinics (referral centers) in different areas in Italy. PATIENTS: A total of 1243 consecutive patients (either premenopausal or postmenopausal) surgically treated for unilateral invasive breast carcinoma with no evidence of metastases. The two study groups were well balanced in terms of clinical and prognostic characteristics. INTERVENTION: Patients in both treatment groups had physical examination and mammography, while patients of the intensive follow-up group had, in addition, chest roentgenography and bone scan every 6 months. MAIN OUTCOME MEASURES: Vital status at 5 years was the main outcome; information was available for all except five patients (0.4%). Relapse-free survival was also analyzed. RESULTS: Overall, 393 recurrences (104 local and 289 distant) were observed during the study. Increased detection of isolated intrathoracic and bone metastases was evident in the intensive follow-up group compared with the clinical follow-up group (112 vs 71 cases), while no difference was observed for other sites and for local and/or regional recurrences. The 5-year relapse-free survival rate was significantly higher for the clinical follow-up group, with patients in the intensive follow-up group showing earlier detection of recurrences. No difference in 5-year overall mortality (18.6% vs 19.5%) was observed between the two follow-up groups. CONCLUSIONS: Periodic chest roentgenography and bone scan allow earlier detection of distant metastases, but anticipated diagnosis appears to be the only effect of intensive follow-up, and no impact on prognosis is evident after 5 years. Periodic intensive follow-up with chest roentgenography and bone scan should not be recommended as a routine policy.