Tamoxifen resistance in early breast cancer: statistical modelling of tissue markers to improve risk prediction.

BACKGROUND: For over two decades, the Nottingham Prognostic Index (NPI) has been used in the United Kingdom to calculate risk scores and inform management about breast cancer patients. It is derived using just three clinical variables - nodal involvement, tumour size and grade. New scientific methods now make cost-effective measurement of many biological characteristics of tumour tissue from breast cancer biopsy samples possible. However, the number of potential explanatory variables to be considered presents a statistical challenge. The aim of this study was to investigate whether in ER+ tamoxifen-treated breast cancer patients, biological variables can add value to NPI predictors, to provide improved prognostic stratification in terms of overall recurrence-free survival (RFS) and also in terms of remaining recurrence free while on tamoxifen treatment (RFoT). A particular goal was to enable the discrimination of patients with a very low risk of recurrence. METHODS: Tissue samples of 401 cases were analysed by microarray technology, providing biomarker data for 72 variables in total, from AKT, BAD, HER, MTOR, PgR, MAPK and RAS families. Only biomarkers screened as potentially informative (i.e., exhibiting univariate association with recurrence) were offered to the multivariate model. The multiple imputation method was used to deal with missing values, and bootstrap sampling was used to assess internal validity and refine the model. RESULTS: Neither the RFS nor RFoT models derived included Grade, but both had better predictive and discrimination ability than NPI. A slight difference was observed between models in terms of biomarkers included, and, in particular, the RFoT model alone included HER2. The estimated 7-year RFS rates in the lowest-risk groups by RFS and RFoT models were 95 and 97%, respectively, whereas the corresponding rate for the lowest-risk group of NPI was 89%. CONCLUSION: The findings demonstrate considerable potential for improved prognostic modelling by incorporation of biological variables into risk prediction. In particular, the ability to identify a low-risk group with minimal risk of recurrence is likely to have clinical appeal. With larger data sets and longer follow-up, this modelling approach has the potential to enhance an understanding of the interplay of biological characteristics, treatment and cancer recurrence.

Poor survival outcomes in HER2-positive breast cancer patients with low-grade, node-negative tumours.

We present a retrospective analysis on a cohort of low-grade, node-negative patients showing that human epidermal growth factor receptor 2 (HER2) status significantly affects the survival in this otherwise very good prognostic group. Our results provide support for the use of adjuvant trastuzumab in patients who are typically classified as having very good prognosis, not routinely offered standard chemotherapy, and who as such do not fit current UK prescribing guidelines for trastuzumab.

Is the biology of breast cancer changing? A study of hormone receptor status 1984-1986 and 1996-1997.

Using archived tumours, those from 1984-1986 and 1996-1997 underwent immunohistochemistry for hormone receptors and grade analysis. A significant shift towards more ER-positive and low-grade disease was found; this appears to reflect screening practices, but could still influence survival.

Automated telephone follow-up after breast cancer: an acceptability and feasibility pilot study.

Traditional clinical follow-up after breast cancer is inefficient at detecting relapse and is poorly suited to detecting and ameliorating psychological problems. There is interest in developing more effective and efficient methods of follow-up. We report a prospective cohort study of the acceptability and feasibility of remote, automated telephone follow-up after breast cancer. Women with a history of breast cancer were approached at their annual follow-up visit. For participants, the follow-up questionnaire was administered on paper at baseline. In place of a clinic visit following year, the women completed the same questionnaire using an automated telephone system. All patients were given mammograms. A semi-structured interview was then conducted to assess the acceptability. The potential impact on clinic usage was assessed. In all, 110 of 121 women (91%) agreed to participate. Seventy-five patients (71%) completed follow-up using the new automated system 1 year later. Seventy-one of the 75 patients found the system easy to use. Forty-nine of the 75 (65.33%) liked the system and were happy to use it as their sole method of follow-up. A further 12% were happy to use it as part of their follow-up. In only 10.66% of participants were concerns raised which led to clinic attendance. Automated questionnaire-based telephone follow-up is acceptable to women and has the potential to reduce attendance at clinic. Further studies to validate this method further are planned.

Phosphorylated c-Src in the nucleus is associated with improved patient outcome in ER-positive breast cancer.

Elevated c-Src protein expression has been shown in breast cancer and in vitro evidence suggests a role in endocrine resistance. To investigate whether c-Src is involved in endocrine resistance, we examined the expression of both total and activated c-Src in human breast cancer specimens from a cohort of oestrogen receptor (ER)-positive tamoxifen-treated breast cancer patients. Tissue microarray technology was employed to analyse 262 tumour specimens taken before tamoxifen treatment. Immunohistochemistry using total c-Src and activated c-Src antibodies was performed. Kaplan-Meier survival curves were constructed and log-rank test were performed. High level of nuclear activated Src was significantly associated with improved overall survival (P=0.047) and lower recurrence rates on tamoxifen (P=0.02). Improved patient outcome was only seen with activated Src in the nucleus. Nuclear activated Src expression was significantly associated with node-negative disease and a lower NPI (P<0.05). On subgroup analysis, only ER-positive/progesterone receptor (PgR)-positive tumours were associated with improved survival (P=0.004). This shows that c-Src activity is increased in breast cancer and that activated Src within the nucleus of ER-positive tumours predicts an improved outcome. In ER/PgR-positive disease, activated Src kinase does not appear to be involved in de novo endocrine resistance. Further study is required in ER-negative breast cancer as this may represent a cohort in which it is associated with poor outcome.

Genetic alterations of CCND1 and EMSY in breast cancers.

AIMS: CCND1 and EMSY, on 11q13, are frequently amplified in breast cancer. CCND1 is implicated in cell cycle progression and EMSY is a BRCA2-associated repressor protein. The aim was to investigate gene copy numbers of CCND1 and EMSY and to determine if CCND1 amplification is associated with reduced survival of tamoxifen-treated breast cancer patients. METHODS AND RESULTS: Fluorescence in situ hybridization (FISH) was performed on 111 consecutive and 354 oestrogen receptor (ER)+ tamoxifen-treated breast cancers. In the consecutive set, CCND1 and EMSY were amplified in 14.8% and 7.2%, respectively, and deleted in 8.7% and 13.5%, respectively. In the ER+ set, CCND1 and EMSY were amplified in 20.6% and 9.6%, respectively, and deleted in 1.7% and 4.2%, respectively. CCND1 and EMSY gene amplifications were associated with decreased overall survival (OS) (P = 0.03 and P = 0.04, respectively) of patients in the ER+ set. CONCLUSION: As hypothesized, CCND1 amplifications are associated with poor OS in ER+ patients. EMSY amplification is also associated with poor OS. However, as >70% of EMSY amplifications were CCND1 amplified, EMSY may not have any additional effect on survival of ER+ breast cancer.

Patients' expectations for follow-up in breast cancer--a preliminary, questionnaire-based study.

The current National Institute for Clinical Excellence (NICE) guidelines recommends that only 2-3 years of follow-up after breast cancer be provided. Clinicians are unwilling to implement these guidelines. However, little has been done to establish patients expectations before they embark on their regular follow-up programme. We have sought the opinions of patients in a questionnaire-based prospective cohort study. Expectations for length and frequency of follow-up were established in women prior to attending their first scheduled follow-up review 1 year after treatment. In addition, patients were asked their opinions on what clinics were designed to achieve. An attempt was made to establish whether patients would be happy with less follow-up when informed of the inefficiency of routine clinic visits. Most women expect some follow-up, but expectations for length and frequency vary dramatically. Most believe follow-up is for the detection of relapse, but very few see psychological support or side effect detection as being central to clinicians' aims. One third of women would be happy to not come back to clinic at all when told how infrequently routine clinical examination detects metastatic disease.

An allelotype of squamous carcinoma of the head and neck using microsatellite markers.

The detection in tumors of genomic regions with a high frequency of loss of heterozygosity has led to the localization and subsequent cloning of a number of tumour suppressor genes. To identify such regions involved in the development of squamous carcinoma of the head and neck we have analyzed 28 paired normal and tumor DNA samples. Using the polymerase chain reaction to amplify 50 simple sequence repeats or microsatellite markers we have studied all 22 q limbs and 17 of the p limbs in 21 patients. In informative cases we observed a high incidence of loss of heterozygosity at five specific chromosomal regions: 3p (44%); 5q (43%); 9q (35%); 11q (45%); and 17p (31%). In addition, further analysis of tumors showing loss of heterozygosity at 5q suggests that a gene at or near the APC locus is involved in squamous carcinoma of the head and neck.

Relationship of endothelial cell proliferation to tumor vascularity in human breast cancer.

Current studies of tumor angiogenesis rely on the concept that endothelium proliferates 30-40 times faster in tumors than in normal tissues. This evidence is based on histological autoradiographic data largely from animal studies. To assess endothelial cell proliferation in human cancer we used the more sensitive and specific technique of immunohistochemistry. We measured the frequency and distribution of endothelial cell proliferation and examined their relationship to tumor cell proliferation. For the first time, we also correlated endothelial and tumor cell proliferation with tumor vascularity. Twenty breast carcinomas from patients exposed to bromodeoxyuridine 3-8 h prior to surgery were double immunostained using antibodies to CD31 (as a marker of endothelium) and bromodeoxyuridine (as a marker of proliferation). The labeling index (LI) for both tumor and endothelial cells was determined and tumor vascularity was assessed by counting the number of CD31 positive vessels. Endothelial cell proliferation was predominantly at the tumor periphery while tumor cell proliferation occurred throughout the lesion. The mean LIs for endothelium and tumor were 2.2% (range, 0.8-5.3) and 7.3% (range, 1.3-17.1), respectively. There was no correlation between tumor and endothelial cell LI (P = 0.414) or between the tumor LI or endothelial cell LI and tumor vascularity (P = 0.08 and P = 0.39, respectively). These findings suggest that previous studies in animal tumors have significantly overestimated endothelial cell proliferation and that its importance in tumor angiogenesis may be related more to continual remodeling and migration of vessels than to proliferation alone.

Quantitative estimation of epidermal growth factor receptor and c-erbB-2 in human breast cancer.

Epidermal growth factor receptor (EGFR) expression by human breast cancer has been shown to predict poor patient outcome, as has amplification of the c-erbB-2 proto-oncogene. We have developed a quantitative immunohistochemical method for measuring protein levels of both receptors and have applied this to a series of 123 breast primaries. We find EGFR expression is substantially lower than normal in nearly all breast cancers (97%). Quantification of p185erbB-2 indicates overexpression in 91% of the tumors. Two separate tumor populations are apparent with levels of c-erbB-2 expression ranging from 0.33 to 19 and 45 to 480 times normal, respectively. Within the lower population, p185erbB-2 expression is inversely related to EGFR expression (rank correlation, P < 0.0005). Using fluorescent in situ hybridization we show that tumors in the latter population have c-erbB-2 amplification and that amplification is restricted to this group. Our findings indicate that significant overexpression of p185erbB-2 occurs in the absence of amplification; these lower levels of expression may have functional significance. Fifty-three patients underwent in vivo bromodeoxyuridine labeling, allowing flow cytometric analysis of tumor cell cycle kinetics. EGFR expression correlates directly to the labeling index (P = 0.011) and indirectly to potential doubling time (P = 0.010), but not to the duration of the S-phase (P = 0.502). Conversely, p185erbB-2 expression does not relate to indices of proliferation. Our results have important implications for the use of both receptor types as therapeutic targets.

Long-term prognostic significance of thymidine labelling index in primary breast cancer.

Tumour growth rates, as measured by incorporation of tritiated thymidine, have been reported as being of prognostic importance in breast cancer. We have measured the thymidine labelling index (TLI) of 185 early breast cancers, followed-up for a minimum of 8 years. Above median TLI was associated with higher tumour grade, but not with other prognostic factors. TLI was not predictive of survival in either univariate or multivariate analysis. The inter- and intra-observer reproducibilities of TLI measurements were poor, which may be a factor limiting its usefulness as a prognostic indicator in breast cancer.

Localization of EGF receptors in frozen tissue sections by antibody and biotinylated EGF-based techniques.

We developed a sensitive EGF receptor detection method for frozen tissue sections using biotinylated EGF as the primary reagent. The method was directly compared with an immunohistochemical technique based on an anti-EGF receptor monoclonal antibody (MAb EGFR1) in normal human and rat tissues and in human tumors. The method was more sensitive than a previously published biotinylated EGF-based technique. In normal human tissues and in 37 of the 50 tumors, the binding pattern mirrored that of positive staining with EGFR1. Five further tumors showed weak immunoreactivity, but in these no binding of biotinylated EGF was detected. The remaining eight tumors were negative by both techniques. The discordant cases may reflect a lower level of sensitivity of the ligand-binding technique or, alternatively, abnormal receptors may have been expressed in these tissues. EGF receptors could be detected in rat liver with biotinylated EGF but not with the antibody, indicating the usefulness of the ligand-based technique in cross-species studies.

Quantitative radioimmunohistochemical measurements of p185(erbB-2) in frozen tissue sections.

The relationship between expression of the c-erbB-2 proto-oncogene and the biology of breast cancer has been investigated widely, most studies using immunohistochemistry in formalin-fixed, paraffin-embedded tissues. This technique is at best semiquantitative and there is a high degree of interstudy variability because of its subjective nature and poor methodological standardization. The relationship between the levels of expression and biology can be examined thoroughly only with an accurately quantitative technique. We have developed a radioimmunohistochemical assay to measure p185(erbB-2) in tissue biopsy specimens. The method involves incubating frozen sections with 125I-labeled monoclonal antibody, microautoradiograpy, and grain counting with image analysis. Sections of cell pellets with known c-erbB-2 levels are processed with each batch of samples as internal calibration standards. We have quantified c-erbB-2 expression in 60 breast carcinomas and compared the results with conventional immunohistochemistry. Radioimmunohistochemistry measured receptor levels throughout the range of expression in breast carcinomas, whereas conventional immunohistochemistry detected the protein only in the highest expressing tumors. The quantitative, objective data produced by radioimmunohistochemistry allow a more thorough evaluation of the relationship between c-erbB-2 expression and tumor biology. This technique may have applications in other fields where quantitative data is required and relevant monoclonal antibodies are available.

Glass yttrium-90 microspheres for patients with colorectal liver metastases.

Total calculated uniform liver doses of up to 150 Gy were achieved using glass yttrium-90 microspheres administered via the hepatic artery and targeted to tumour using angiotensin II in seven patients with colorectal liver metastases. No toxicity was observed. Hepatic metastatic progression was delayed in six patients. Median survival was 11 months (range 5-25 + months).

Influences on measurement of cellular proliferation by histology and flow cytometry in mammary carcinomas labeled in vivo with bromodeoxyuridine.

Field-selection procedures strongly influence histologically determined bromodeoxyuridine (BrdU) labeling indices (HLIs): For a series of 52 breast carcinomas, "high-labeling" fields had mean 2,000-cell HLIs that were, on average, more than double those for random fields, but the ratio was extremely variable. Although there may be a priori reasons for preferring particular field-sampling techniques, nonstandardized kinetic data from different laboratories will be uninterpretable and therefore of limited clinical application. Also, if heterogeneity of tumor cell proliferation is neglected, it is possible that incomplete or even misleading data may be gained. Total flow cytometric BrdU labeling indices (TLIs) were not, on average, higher or lower than HLIs for the same cases, but individual discrepancies could be pronounced. The best correlation of flow cytometric LIs with HLIs was obtained for aneuploid carcinomas with a restricted labeling index derived for aneuploid (presumptively neoplastic) cells only. Although flow cytometric analysis avoids the statistical problems associated with accurately determining a small percentage from relatively few cells, it lacks precise definition of the character of the cells counted, but even histologic methods are not entirely free of this difficulty.

Prognostic value of ploidy of primary tumour and nodal secondaries in colorectal cancers.

Tumour ploidy is of prognostic value in colorectal cancer, DNA aneuploid tumours having a worse outlook. Nearly all studies have concentrated on the DNA content of the primary tumour. We have examined the ploidy of the primary tumour and its lymph node metastases in 71 cases of Dukes' stage C disease, to see whether this provides greater prognostic information than the primary alone. Analysis was performed using formalin-fixed, paraffin-embedded tumour sections. Ploidy of primary and metastases was different in 20 cases (28%), aneuploid nodes being seen with diploid primaries and vice versa. Ploidy of both the primary (chi 2 = 4.86, P = 0.03) and secondary (chi 2 = 4.86, P = 0.03) tumours predicted survival in univariate analysis. Combining the ploidy of primary and nodes, three prognostic groups could be defined--diploid primaries with diploid metastases (hazard relative to both aneuploid, 0.36) had significantly better survival than cases where the ploidy of the primary and nodes were mixed (relative hazard 0.47-0.56), which did better than cases with aneuploid primary and nodes. This study demonstrates that ploidy variation between primary and secondary tumours is common, and better prognostic information may be gained by studying both.

Mitoxantrone and methotrexate chemotherapy with and without mitomycin C in the regional treatment of locally advanced breast cancer.

Fifty patients with locally advanced breast cancer received regional chemotherapy delivered angiographically via the internal mammary artery and varying vessels supplying the lateral aspect of the breast. Thirty three patients received mitomycin C, methotrexate and mitoxantrone, and 17 patients received methotrexate and mitoxantrone only. There was no significant difference in clinical response between the two groups. However, in patients who received mitomycin C, severe local skin toxicity occurred in nine patients resulting in delay of further therapy and considerable morbidity. Mitomycin C should not be administered regionally in patients with locally advanced breast cancer.

Mucinous breast carcinoma.

A case of mucinous breast carcinoma is discussed in which several unusual features in presentation and difficulties in management are addressed.

The effect of portal venous flow on the washout of a regionally injected marker substance 99mTc-methylene diphosphonate after hepatic arterial blockade with degradable starch microspheres.

Patients with hepatic metastases derived from colorectal carcinoma have a poor prognosis. Regional chemotherapy, either alone, or combined with agents such as degradable starch microspheres (DSM) that reduce or abolish intrahepatic arterial flow and potentiate the delivery of cytotoxics to hepatic metastases, have not significantly improved survival. We have investigated one positive mechanism, namely the effect of portal venous washout of cytotoxics, for the poor efficacy of drugs administered either alone or in combination with DSM via the hepatic artery in the rat. Using a radiolabelled marker, 99mTc-methylene diphosphonate (MDP), to represent a cytotoxic drug, the initial studies indicated that with the hepatic artery and portal vein clamped, a volume of 0.05 ml of the marker administered via the hepatic artery resulted in the most uniform intrahepatic distribution with minimal washout into the systemic circulation (21 +/- 3.7%). When the hepatic artery was clamped, the washout of MDP was reduced from 100% (with clamps on the portal vein and hepatic artery) to 84.2 +/- 7.7%. DSM administered concomitantly with MDP, resulted in a greater reduction of the portal venous washout of the marker (63 +/- 2.4%). Administration of DSM and MDP via the hepatic artery and with the portal vein clamped further reduced the washout of the marker to (21 +/- 2.26), results similar to those observed with inflow vessel clamps. Following restoration of portal venous flow, there was a rapid washout of 53.7 +/- 7.6% of the marker into the systemic circulation. The results of this study suggest that portal venous washout of regionally delivered cytotoxics, either alone or in combination with DSM, offer an explanation for the poor efficacy of regional chemotherapy in improving the prognosis of patients with hepatic metastases.

A randomised controlled trial of adjuvant hormono-chemotherapy in Stage II breast cancer.

From 1979 to 1984 women with histologically proven Stage II breast cancer were entered into a randomised controlled trial of adjuvant hormono-chemotherapy following mastectomy. Adjuvant therapy was started within 2 weeks of surgery and was given for 6 months as continuous 'Tamoxifen' and 6-monthly cycles of vincristine, adriamycin and cyclophosphamide (VAC). Of the 157 women randomised 82 were assigned to the control group and 75 to the adjuvant group. Comparing the two groups by life table analysis the rate of recurrence was found to be significantly greater in the control group (n = 46) than in the adjuvant group (n = 25) (Peto log rank test chi 2 = 7.46 P less than 0.01). Although there were more deaths amongst the control patients (n = 31) than in those randomised to receive adjuvant therapy (n = 18), this difference did not achieve statistical significance (chi 2 = 2.80 P less than 0.1). This analysis has shown that although adjuvant hormono-chemotherapy may significantly delay recurrence the trial has failed to demonstrate any significant improvement in the survival of women with Stage II breast cancer following mastectomy.