MS4-2: Preoperative Endocrine Therapy: New Approaches

Modern neoadjuvant endocrine therapy trials comparing aromatase inhibitors (AIs) with tamoxifen have shown that neoadjuvant AIs can downstage cancers and avoid mastectomy in around 50% of post menopausal women with large ER positive tumours, and are superior to tamoxifen in this respect.

The future potential of preoperative endocrine therapy is in the use of molecular biomarkers after starting treatment to predict outcome. This approach was pioneered in the IMPACT trial which showed that the reduction in tumour proliferation as measured by the biomarker KI67 comparing neoadjuvant anastrozole v tamoxifen v. the combination. was significantly greater for anastrozole than for the other 2 arms, predicting the same long-term outcome benefit in the equivalent adjuvant ATAC trial. The change in Ki67 with treatment could be used to predict potential benefit with other novel endocrine therapy approaches. Recently a similar neoadjuvant trial comparing anastrozole v letrozole v exemestane has shown no significant difference in Ki67 suppression between arms, predicting no long-term difference in clinical outcome with these agents.

Around 15% of patients with good initial tumour Ki67 suppression after 2 weeks of preoperative endocrine therapy show recovery of Ki67 by 12 weeks, providing early biomarker evidence of treatment resistance. Clinical outcome in these patients appears correspondingly worse. This provides an opportunity for biomarker led studies of agents targeting both innate and acquired resistance to endocrine therapy, defined by 2 week and 12 week Ki67 levels.. One such preoperative endocrine therapy trial with Ki67 as primary endpoint has already been carried out: patients were started on 2 weeks of preoperative anastrozole and then randomised to the addition or not of the tyrosine kinase inhibitor, gefitinib (IRESSA). This novel pre-operative endocrine therapy approach also allows the study of changes in other biomarkers and in gene expression as a means of elucidating underlying mechanisms of resistance

Finally the clinical potential of Ki67 after 2 weeks preoperative AI treatment is further demonstrated by the observation that higher 2 week values of Ki67 after treatment in the IMPACT trial predicted for a significantly worse recurrence-free survival. This raises the intriguing possibility that patients with ER positive breast cancer, including small cancers, could be treated with 2 weeks preoperative endocrine therapy before surgery and Ki67 levels at surgery be used to predict long-term outcome for the individual patient and in particular to identify those patients with a good prognosis who would not require chemotherapy. This hypothesis is being tested in the UK POETIC (Peri-Operative Endocrine Therapy for Individualised Care) trial which has so far accrued around 2,000 patients out of an overall target of 4,000. A similar approach in which Ki67, pathologic stage and ER at the end of a course of neoadjuvant endocrine therapy are combined to produce a Preoperative Endocrine Prognostic Index (PEPI) is being used to identify patients with such a low risk of relapse that adjuvant chemotherapy may not be necessary.

Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr MS4-2.

Genetic association study of GABRA2 single nucleotide polymorphisms and electroencephalography in alcohol dependence

The gamma aminobutyric acid (GABA) system has been implicated in the susceptibility to develop alcohol dependence and in determining electroencephalogram (EEG) beta activity. The role of the GABA receptor alpha-2 gene (GABRA2) in human alcohol dependence was determined in a genetic and electrophysiological study. The study population comprised 586 white UK individuals with alcohol dependence but a very low prevalence of co-morbid drug dependence, and 603 ancestrally matched healthy controls. Genotyping for seven GABRA2 single nucleotide polymorphisms (SNPs), identified from the literature as positively associated with alcohol dependence, was performed with success rates of 90% or greater. EEGs were available in 32 selected patients who had been abstinent from alcohol for a minimum of 24 months and in 138 ancestrally matched healthy controls. None of the SNPs showed allelic or haplotypic association with alcohol dependence. All markers were in Hardy Weinberg equilibrium (HWE) in the controls. HWE for marker rs279841 in the alcohol dependent sample was p=0.0199 and combined p=0.0166. Linkage disequilibrium patterns appear to be very similar to that observed in the HapMap CEU data. A significantly higher prevalence of excess EEG fast activity was found in the patients (31 vs. 14%, p=0.018). A significant relationship was found between the presence of excess EEG fast activity and GABRA2 SNPs rs548583, rs279871 and rs279841. This allelic association study provides no evidence for an association between GABRA2 polymorphisms and alcohol dependence. However, a significant relationship was identified between GABRA2 and excess EEG fast activity. This dissociation of effect may reflect the fact that the EEG is a more direct marker of phenotypic GABRA2 expression than the more heterogeneous alcohol dependence phenotype.

Non-CPAP therapies in obstructive sleep apnoea

In view of the high prevalence and the relevant impairment of patients with obstructive sleep apnoea syndrome (OSAS) lots of methods are offered which promise definitive cures for or relevant improvement of OSAS. This report summarises the efficacy of alternative treatment options in OSAS. An interdisciplinary European Respiratory Society task force evaluated the scientific literature according to the standards of evidence-based medicine. Evidence supports the use of mandibular advancement devices in mild to moderate OSAS. Maxillomandibular osteotomy seems to be as efficient as continuous positive airway pressure (CPAP) in patients who refuse conservative treatment. Distraction osteogenesis is usefully applied in congenital micrognathia or midface hypoplasia. There is a trend towards improvment after weight reduction. Positional therapy is clearly inferior to CPAP and long-term compliance is poor. Drugs, nasal dilators and apnoea triggered muscle stimulation cannot be recommended as effective treatments of OSAS at the moment. Nasal surgery, radiofrequency tonsil reduction, tongue base surgery, uvulopalatal flap, laser midline glossectomy, tongue suspension and genioglossus advancement cannot be recommended as single interventions. Uvulopalatopharyngoplasty, pillar implants and hyoid suspension should only be considered in selected patients and potential benefits should be weighed against the risk of long-term side-effects. Multilevel surgery is only a salvage procedure for OSA patients.

The views of pharmacists supplying emergency hormonal contraception through a patient group direction

Focal points

The impact of the deregulation of emergency hormonal contraception on a patient group direction scheme in the north-west of England

Focal points

Day case laparoscopic nephrectomy: initial experience

RATIONALE

Laparoscopic nephrectomy tends to become the new gold standard surgical technique in a selected population (non-functioning kidney, localised renal cell carcinoma). Day surgery is a popular pathway of care and, procedures of ever-increasing complexity are being considered.

OBJECTIVE

The aim of the study was to report the postoperative complications of day case laparoscopic nephrectomy, according to the Clavien system, and, to assess the feasibility of the procedure performed as a day case.

MATERIAL AND RESULTS

This study included all the patients considered for day case transperitoneal laparoscopic nephrectomy between May 2008 and November 2009. Sixteen consecutive patients were enrolled in this retrospective study. There were ten procedures on the left hand-side and six on the right hand-side. Age ranges from 22 to 77 years old. Male to female ratio was 9:7. The preoperative diagnosis was non-functioning kidney in 9 cases and kidney tumour in the other 7 cases. All but two patients have been discharged in the same day (87.5%). The readmission rate was of 12.5%. One wheel-chair bonded patient was readmitted four days after the procedure, because of adynamic ileus, and another one three days later because of wound infection. There were two grade I and one grade IV complications (Clavien system). The patient readmitted with grade IV complication, wheel-chair bonded because of cerebral palsy, was not a typical day surgery patient.

DISCUSSION

The vast majority of complications were minor and resulted in no residual disability. In our small series, the day case laparoscopic nephrectomy was feasible and safe.

Abstract P6-11-11: Trastuzumab beyond Progression in HER2-Positive Advanced Breast Cancer: The Royal Marsden Experience

Background: Routine clinical practice and a growing body of evidence support the continuation of trastuzumab (T) in patients with HER-2 positive (+ve) advanced breast cancer progressing on previous T-based therapy. Despite this, recent UK clinical guidance advises against continuing T on evidence of disease progression (PD) in the absence of brain metastases. This retrospective study therefore evaluated the outcome of patients (pts) with HER-2+ve locally advanced (LA) or metastatic breast cancer (MBC) who continued T beyond PD, treated in our Unit. Patients and methods: HER-2+ve pts receiving T for LA or MBC were identified from our prospectively maintained database and pharmacy records. Those receiving T beyond PD after adjuvant or one line of T for advanced disease were assessed for response and outcome. From thetimepoint of T continuation beyond PD we calculated the overall disease control rate (response or stable disease), time to progression (TTP), and overall survival (OS).

Results: 114 pts with HER-2+ve LA or MBC treated with T beyond PD were identified. At the time of analysis 35 (31%) pts were still alive with a median follow up of 20 months (mo). The main site of disease was visceral in 84 (74%) pts, including 37 (32%) pts with CNS involvement. 30 (26%) pts had soft tissue or bone metastases only. Fifty nine (52%) pts had received adjuvant chemotherapy and 13 (11%) pts had received adjuvant T. Seventy six (66%) pts had 1 line of chemotherapy prior to continuation of T beyond PD and 21 (19%) had 2 or more lines. Fifty three (46%) pts had previously received taxanes + T for their LA or MBC. Post-progression, 66 (58%) pts received T combined with chemotherapy; 12 (11%) taxane-based, 32 (28%) capecitabine and 22 (19%) vinorelbine. Information regarding response was not available in 21(18%) pts. Of the ninety three (82%) pts with documented clinical (n=16) or radiological (n=77) response evaluation, 68 (60%) pts were considered as having stable disease (SD) or better and 25 (22%) as having PD. The median duration of T was 10 mo (95% CI: 8-11 mo), the median TTP was 24wks (95% CI: 21-28 wks) and the median OS was 19 mo (95% CI:12-24mo). In a sub-group analysis of the 81(71%) pts who received T as first-line Rx or relapsed within 12 wks of adjuvant T, overall disease control was achieved in 50 (61%) pts, the median TTP was 25wks (95% CI:18-33 wks) and the median OS was 22 mo(95% CI:17-27mo). In terms of safety, only 6 (5%) pts overall had to discontinue T secondary to decline in left ventricular ejection fraction.

Conclusion: Our results from an unselected group of patients are supported by positive results from other studies and provide additional evidence that continuation of trastuzumab beyond disease progression is of clinical benefit.

Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P6-11-11.

Metabolic effects of low dose angiotensin converting enzyme inhibitor in dietary obesity in the rat

BACKGROUND AND AIMS

Given the recent observation of a local renin-angiotensin system (RAS) in adipose tissue, and its association with obesity-related hypertension, the metabolic effects of treatment with a low dose angiotensin converting enzyme inhibitor (ACEI) were investigated in a rodent model of diet-induced obesity.

METHODS AND RESULTS

Male Sprague Dawley rats were exposed to either standard laboratory chow (12% calories as fat) or palatable high fat (30% calories as fat) diet for 12 weeks. A subset from both dietary groups was given low dose ACEI in drinking water (perindopril, 0.3 mg/kg/day) throughout the study. The high fat diet increased body weight, adiposity, circulating leptin and insulin and in the liver we observed fat accumulation and increased tissue ACE activity. Treatment with perindopril decreased food intake and circulating insulin in both diet groups, and hepatic ACE activity in high fat fed animals only. Decreased plasma leptin concentration with ACE inhibition was only evident in chow fed animals. These effects were independent of any blood pressure lowering effect of ACE inhibition.

CONCLUSION

Chronic low dose ACEI treatment reduced circulating insulin and leptin levels with some reduction in food intake in chow fed rats. Fewer beneficial effects were observed in obesity, and further work is required to investigate higher ACEI doses. Our data suggest a reduction in hepatic ACE activity may affect lipid accumulation and other inflammatory responses, as well as improving insulin resistance. Our findings may have implications for maximizing the clinical benefit of ACEI in patients without overt cardiovascular complications.

Analysis of Bevacizumab (Bev) Therapy, Bisphosphonate Use and Osteonecrosis of the Jaw (ONJ) in >1900 Patients Treated in Two Randomized, Controlled Trials

Background: Long-term bisphosphonate therapy is known to increase the risk of ONJ. A 16% incidence of ONJ was reported in a retrospective analysis of 116 patients receiving bisphosphonates with anti-angiogenic therapy (Bev or sunitinib) for bone metastases from breast, colon, or renal cell cancers.Methods: To assess the incidence of ONJ with Bev, we analyzed data from >3500 patients with locally recurrent or metastatic breast cancer (LR/MBC) treated in three large trials of Bev-containing therapy: AVADO (Bev in combination with docetaxel); RIBBON-1 (Bev in combination with taxane, anthracycline-based combination therapy, or capecitabine); and MO19391 (single-arm safety study of >2000 patients receiving Bev-containing therapy in the general oncology practice context). The incidence of ONJ was compared in patients treated with Bev versus placebo and in patients with or without bisphosphonate exposure.Results: Data from the blinded phase of two randomized, placebo-controlled trials demonstrated an ONJ incidence of 0.3%. ONJ was more common in patients who also received bisphosphonate therapy than in those who received no bisphosphonates (Table). This observation is supported by data from 2216 patients treated in the single-arm MO19391 study (2.4% with bisphosphonate versus 0% without). AVADO*RIBBON-1**TotalIncidence of ONJ, no. of pts (%)Bev (n=492)Pla (n=238)Bev (n=817)Pla (n=412)Bev (n=1309)Pla (n=650)Overall population receiving Bev (n=1309)3 (0.6%)O1 (0.1%)O4 (0.3%)OBisphosphonate (n=233)1 (1.2%)O1 (0.6%)O2 (0.9%)ONo bisphosphonate (n=1076)2 (0.5%)OOO2 (0.2%)O *Bev 15 and 7.5 mg/kg arms pooled**Taxane/anthracycline and capecitabine cohorts pooledConclusions: This is the largest analysis of ONJ in patients receiving Bev for LR/MBC. The 0.3% incidence of ONJ with Bev is considerably lower than previously reported by Christodoulou et al. with anti-angiogenic therapy. As in the general population, the risk of ONJ is increased in patients exposed to bisphosphonates. The 0.9–2.4% incidence seen here in a large population of patients receiving Bev and bisphosphonate therapy is substantially lower than the 16% observed in a small cohort of patients from a retrospective analysis and within the range reported in the literature for bisphosphonates alone (1–4%). Good oral hygiene, dental examination and avoidance of invasive dental procedures remain important in patients receiving bisphosphonates, irrespective of Bev treatment.

Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 208.

Immunohistochemical Phenotype after 20-Year Follow-Up of the Royal Marsden Tamoxifen Breast Cancer Prevention Trial (RMTBCPT)

Background: 20-yrs after starting the randomized, double-blind RMTBCPT (13-year median follow-up) 186 women developed invasive breast cancer, 82 on tamoxifen and 104 on placebo1. There was a significant reduction in the incidence of invasive ER+ but not ER- breast cancer that was significant after but not during the 8yr treatment period. The phenotype of ER+ breast cancer is highly variable. We have therefore assessed PgR, HER2, EGFR and Ki67 expression in as many as possible of these tumours.Methods: Tumour blocks were available on 154 participants, 65 on tamoxifen and 89 on placebo a similar distribution to the whole population. Staining was conducted using the following antibodies: ER, clone 6F11(Vector); PgR, clone 16 (Vector); HER2, HercepTest + K5207 (Dako), FISH PathVysion (Abbott) for IHC 2+ cases; EGFR, clone 31G7 (Invitrogen); Ki67 clone Mib1(Dako). ER and PgR were quantified as H-scores, HER2 and EGFR as + or –, and Ki67 as % cells staining.Results: There were 47 and 18 ER+ and ER- tumours in the tamoxifen arm vs 76 and 13 in the placebo arm, showing a 37% (95% CI 10- 57%, p=0.01) reduction of ER+ tumours which was essentially the same as that in the whole population1.There were 38 and 27 PgR+ and PgR- tumours in the tamoxifen arm vs 58 and 31 in the placebo arm, showing a 33% (95% CI 0-56%, p=0.05) reduction of PgR+ tumours. The distribution of ER/PgR tumours according to arm and time on trial at tumour diagnosis is shown in the table (there were no ER-PgR+ cases).The decrement in ER+ tumours was predominantly in ER+PgR+ cases after 8yrs but extended to ER+PgR- cases. It is important to note that the post 8yr tumour phenotype cannot be affected by continued exposure to tamoxifen. ER levels were significantly lower in the tamoxifen-treated group even among tumours presenting as ER+ (median H-score 123 vs 161, p=0.02). There were 9 and 6 HER2+ and 11 and 12 EGFR+ cases in the tamoxifen and placebo arms, respectively (p=NS for both). The mean (95%CI) levels of Ki67 were 8.4% (6.3-11.1) and 8.5% (6.8-10.6) in the 2 arms, respectively.Discussion: The decrement in ER+ tumours in the tamoxifen-treated women was restricted to the post-treatment period and was similar to that seen in the overall population1. Among the ER+ group there was a similar proportional reduction of PgR+ and PgR- tumours by tamoxifen.There was no evidence of enhanced HER2 or EGFR expression or increased proliferation in tumours developing in the tamoxifen arm but ER expression was reduced even among ER+ tumours.1Powles et al, JNCI, 2007;99:283-90. Time on trialPlaceboTamoxifenER+PgR+<8 years2622ER+PgR+>8 years3216ER+PgR-<8 years106ER+PgR->8 years83ER-PgR-<8 years711ER-PgR->8 years67All<8 years4339All>8 years4626 Supported by the Da Costa International Fund for Breast Cancer Prevention

Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 1046.

Tumor Ki67 Proliferation Index within 4 Weeks of Initiating Neoadjuvant Endocrine Therapy for Early Identification of Non-Responders

Background: The Preoperative Endocrine Prognostic Index (PEPI) scores the independent prognostic effects of tumor pathologic staging and expression levels of ER and the “proliferation” marker Ki67 in the surgical sample to predict long term outcomes after completion of neoadjuvant endocrine treatment (Ellis et al JNCI 100:1380, 2008). A limitation of the PEPI is that the prognostic information becomes available only after 4 months of treatment. We therefore evaluated the value of an early assessment of the Ki67 level in a tumor biopsy sample taken two to four weeks after initiating treatment in two neoadjuvant endocrine therapy trials for the purposes of the early identification of non- respondersMethods: A Ki67 cut point of greater than 10% for poor outcome in ER+ breast cancer was derived by comparing the PAM50 intrinsic subtype profile using a qRT-PCR assay with Ki67 data in a 700+ sample data set. A baseline level of 10% or less correlated most closely with a PAM50-based definition of LumA breast cancer and above 10% LumB breast cancer. We subsequently applied the 10% cut point to the baseline and early on-treatment Ki67 data in two trials, POL (Olson et al JACS 208:906, 2009) and IMPACT (Smith et al JCO: 23, 5108, 2005).Results: At baseline the dichotomized Ki67 definition was not significantly predictive for surgical Ki67 level, PEPI score or RFS in this modest size sample set. In contrast, in a result that emphasizes the enhaced prognostic properties of the on-treatment Ki67 approach, the one month POL sample Ki67 values (62 patients) predicted a higher level of Ki67 in the surgical samples at four months after treatment initiation (P=.01), a poorer PEPI score (P=0.01), a smaller number of patients in the PEPI risk point zero group (P=0.08) and worse relapse free survival (P=0.003). The IMPACT data (153 patients) confirmed that a two week Ki67 >10% predicted higher Ki67 in the surgical specimen (P=0.001), a poorer PEPI score (P=0.001), smaller numbers of patients in the PEPI 0 risk point group (P= 0.004) and worse relapse free survival (P=0.008).Ki67 and OutcomePOL 4W Ki67% PEPI 0RFS (events)10%>1/19 (5%)5/21 (23%)10%≤10/36 (28%)1/41 (2.4%)P ValueP=0.08 (Fisher)P=0.003 (log rank)IMPACT 2W Ki67% PEPI 0RFS (events)10%>0/32 (0%)9/35 (26%)10%≤21/101 (21%)13/118 (11%)P ValueP=0.004 (Fisher)P=0.008 (log rank) Conclusions: A tumor Ki67 assessment taken a short time (2 to 4 week window) after the initiation of neoadjuvant AI identifies patients with poor outcome ER+ disease. Amendment 6 of the neoadjuvant endocrine therapy protocol ACOSOG Z1031 will triage patients with an “on treatment” Ki67 value above 10% to chemotherapy in order to assess the pathological response rate to cytotoxic therapy in this important tumor subset.Supported by R01 CA095614, Avon PFP award 3P50 CA68438-07S2, U01 CA114722, ACOSOG U10 CA 76001, Breakthrough Cancer UK and AstraZenica (IMPACT trial).

Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 78.

Aromatase Inhibitors Reduce the Expression of a Hypoxia Metagene in Oestrogen Receptor Positive Breast Cancer in Postmenopausal Women

Aims: To (1) define the effect of oestrogen deprivation on genes related to hypoxia in oestrogen receptor positive (ER+) breast cancer and (2) identify any link between hypoxia and proliferation.Background: The majority of breast cancer patients are postmenopausal women with ER+ tumours and at some point receive an aromatase inhibitor (AI) as part of their treatment. Hypoxia and proliferation are important factors in the progression of ER+ tumours. Proliferation is profoundly reduced in most ER+ cancers after treatment with AIs[1], however little is known on the effects of AIs on hypoxia.Materials and methods: 81 pre- and 2-week post-treatment core-cut tumor biopsies were obtained from postmenopausal women with ER+ breast cancer who received single agent neoadjuvant anastrozole (AI)[2], and from 20 of these patients after 16 weeks of AI treatment. RNA was extracted and analysed on Illumina 48K microarrays. A hypoxia metagene (MG) was developed by identification of genes whose expression clustered with the expression of classical hypoxia-regulated genes[3]. Genes associated with proliferation were removed from the MG. A proliferation MG was derived by selecting the intersection of proliferation clusters from three public breast cancer datasets.Results: Spearman correlations revealed a strong relationship between the hypoxia and proliferation MGs prior to AI treatment (r =0.61, p<10-3), and persisted after 2 weeks (r =0.77, p<10-3) and after 16 weeks of treatment (r=0.72, p =0.002). Baseline expression of the hypoxia MG was (1) positively correlated with 2-week Ki67 (Spearman r =0.37, p =0.002), (2) showed a trend for a positive correlation with poor 2-week Ki67 change (Spearman r=0.22, p =0.06) and with poor reduction in the mean expression of four classical oestrogen dependant genes (TFF1/pS2, GREB1, PDZK1 and PGR) known as AVERG[4] (Spearman r =0.22, p=0.06). Expression of the hypoxia MG was significantly down-regulated after 2 weeks of oestrogen deprivation using AI treatment (p<10-3). The 2-week change in hypoxia showed a positive correlation with the 2-week change in proliferation (Spearman r=0.58, p<10-3), and with the 2-week change in Ki67 (Spearman r=0.35, p=0.005).Conclusions: The expression of a hypoxia MG decreases after oestrogen deprivation. The hypoxia MG is strongly associated with proliferation prior to and after AI treatment. The data are consistent with hypoxia being a secondary effect of proliferation in ER+ breast cancer and could contribute in understanding the need to combine anti-proliferative drugs with anti-angiogenic agents for these patients. There may be a weak effect of hypoxia on de-novo resistance to AIs.1. Dowsett M, et al., Clin Cancer Res 2006; 12(3): p.1024-302. Smith, I.E, et al., J Clin Oncol 2007; 25(25): p.3816-223. Winter, S.C, et al., Cancer Res 2007; 67(7): p.3441-94. Dunbier, A.K, et al., Cancer Res 2009; 69(Suppl.), 78sSupported by The Mary-Jean Mitchell Green Foundation and Breakthrough Breast Cancer.

Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 408.

Alcohol and driving: is the 0.05% blood alcohol concentration limit justified?

This paper reviews the findings of experimental and laboratory research to determine whether there is sufficient evidence to support a uniform 0.05% blood alcohol concentration (BAC) as the legal limit for driving. Statistically significant and meaningful decrements in driving-related performance were identified at 0.05% BAC or below in many studies. Younger and inexperienced drinkers and drivers appear to be at greatest risk for alcohol-related traffic crashes. It is concluded that on scientific grounds, there is support for setting the legal limit at 0.05%. The setting of a uniform 0.05% BAC statutory limit should be but one effort within a comprehensive approach including other legal, social, behavioural and environmental strategies to deal with the drink driving problem.

Bone fractures among postmenopausal patients with endocrine-responsive early breast cancer treated with 5 years of letrozole or tamoxifen in the BIG 1-98 trial

BACKGROUND

To compare the incidence and timing of bone fractures in postmenopausal women treated with 5 years of adjuvant tamoxifen or letrozole for endocrine-responsive early breast cancer in the Breast International Group (BIG) 1-98 trial.

METHODS

We evaluated 4895 patients allocated to 5 years of letrozole or tamoxifen in the BIG 1-98 trial who received at least some study medication (median follow-up 60.3 months). Bone fracture information (grade, cause, site) was collected every 6 months during trial treatment.

RESULTS

The incidence of bone fractures was higher among patients treated with letrozole [228 of 2448 women (9.3%)] versus tamoxifen [160 of 2447 women (6.5%)]. The wrist was the most common site of fracture in both treatment groups. Statistically significant risk factors for bone fractures during treatment included age, smoking history, osteoporosis at baseline, previous bone fracture, and previous hormone replacement therapy.

CONCLUSIONS

Consistent with other trials comparing aromatase inhibitors to tamoxifen, letrozole was associated with an increase in bone fractures. Benefits of superior disease control associated with letrozole and lower incidence of fracture with tamoxifen should be considered with the risk profile for individual patients.