Abstract P4-12-23: Subcutaneous trastuzumab (Herceptin™) in patients with HER-2 positive early breast cancer: A UK time and motion study in comparison with intravenous formulation

Aim:

To quantify active healthcare professional (HCP) time and costs associated with subcutaneous (SC) and intravenous (IV) infusion administration of trastuzumab (Herceptin) in the treatment of patients with HER2-positive early breast cancer within the adjuvant PrefHer trial; secondly, to measure patient time in the care unit and patient chair time for both routes of administration.

Methods:

A UK multi-centre prospective, observational time and motion study was conducted alongside the PrefHer trial (ClinicalTrials.gov id: NCT01401166). Trained observers measured the duration of each SC and IV related task HCPs undertook and recorded patient time in the chemotherapy unit and treatment chair. The type and quantity of medical consumables used with each route of administration were also collected. 24 patient episodes were recorded (12 SC, 12 IV). Mean total administration time was calculated as the mean sum of task times, for both IV and SC formulations. The mean cost of each route of administration was calculated as the mean cost of HCP time plus the mean cost of consumables used. HCP time was costed using data from the Personal Social Services Research Unit. Consumables were costed using hospital pharmacy data and online sources.

Results:

Mean active HCP time for IV preparation and administration was 92.6 minutes compared with 24.6 minutes for SC administration. The mean cost for IV preparation and administration was £144.96 (£132.05 of HCP time and £12.92 of consumables) versus £33.15 (£31.99 of HCP time and £1.17 of consumables) for SC administration. Mean time spent by patients in the care unit and treatment chair was 94.5 minutes and 75 minutes respectively for IV, and 30.3 minutes and 19.8 minutes for SC.

SC administration of trastuzumab could translate to a HCP time saving of 68 minutes (34.5 minutes of preparation time and 33.5 minutes of administration time) (versus IV) with a total cost saving of £111.81 per patient episode. This equates to a potential saving of £2012.58 over a full course of adjuvant trastuzumab treatment (18 cycles).

Conclusion:

Substituting IV infusion with SC administration of trastuzumab may lead to a substantial reduction in active HCP time, consumable use and overall cost. The reduced patient chair and unit time could provide increased capacity within existing resources.

Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P4-12-23.

Abstract PD07-04: Predicting Outcome with Ki67 in Primary Breast Cancer in the Neoadjuvant Setting

Background: The proliferation antigen Ki67 has been shown to be a reliable predictive marker of treatment efficacy in the neoadjuvant treatment of breast cancer but its prognostic significance remains uncertain. This study assesses Ki67 before and after neoadjuvant chemotherapy (NAC) in relation to long-term outcome. Methods: 117 patients with primary breast cancer due to undergo NAC were studied (median age 48 years, range 25-78; T2-4, N0-3, M0) and information gathered from a prospectively maintained database. Immunohistochemically derived Ki67 expression was obtained from pretreatment core biopsy and surgical specimens. Their relationship to DFS and OS was analyzed along with known prognostic variables (age, ER/PR/HER2 status, clinical and pathologic T and N stage, grade), NAC regimen, response and adjuvant treatment. Survival curves were estimated using the Kaplan-Meier method and a log-rank test used to determine significance using a two-tailed p-value of 0.05. A multivariate Cox proportional hazards regression model performed in a stepwise fashion was used to determine the prognostic value of each signifcant variable. Results: 84 patients with matched biopsy and surgical samples were assessable for pre and post NAC Ki67 levels. The majority (90%) received anthracycline based NAC (median number of cycles 6, range 2-6). On univariate analysis, the only significant pretreatment predictive factor for shorter DFS was higher clinical nodal stage (cN) (P<0.001). Posttreatment variables that predicted for worse DFS were: Ki67 at surgery (HR 1.52, p=0.048), pathologic nodal stage (pN) (p=0.001) and grade (p=0.013). On multivariate analysis, pN was the most powerful predictor for DFS (chi-squared test 19.8, 3 df, P<0.001). Univariate analysis of pretherapy factors for OS revealed that Ki67 at biopsy (HR 2.06, p=0.039), cN (HR 4.44, P<0.001) and PR positivity (HR 0.41, p=0.012) were significant. Significant posttreatment variables for OS were Ki67 at surgery (HR 2.01, p=0.006), pN (p=0.001), and grade (p=0.009). PR status and pN remained important predictors of OS on multivariate analyses. Conclusions: The expression of Ki67 is a widely accepted marker of cellular proliferation in breast cancer. Ki67 levels after NAC were a better predictor for long-term outcome than pretherapy Ki67, although nodal status appears to be the most powerful determinant overall. High Ki67 levels post NAC may identify patients with poorer outcomes who are candidates for further systemic therapy.

Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr PD07-04.

Pre-treatment haemoglobin levels and the prediction of response to neoadjuvant chemotherapy in breast cancer

AIMS

A low pre-treatment haemoglobin level has been shown to negatively influence outcome in the treatment of tumours of the cervix, bladder and head and neck by radiotherapy. The purpose of this study was to assess the influence of baseline haemoglobin levels on the response to neoadjuvant chemotherapy for breast cancer.

MATERIALS AND METHODS

One hundred and thirty-nine women receiving neoadjuvant chemotherapy for operable breast tumours (T2-4, N0-1, M0) were accessed from our prospective database. Women were treated between March 1999 and June 2004. The median age was 47 years (range 25-70). Most women were treated with 5-fluorouracil, epirubicin and cyclophosphamide chemotherapy (122/139 patients). Baseline haemoglobin levels were compared for clinical responders (partial or complete) and non-responders (stable or progressive disease) using Student's t test and logistic regression. The analysis was adjusted for nodal status, tumour size, tumour grade and menopausal status.

RESULTS

The overall response rate was 84.9% (118/139), with a complete clinical response in 24.5% (34/139). Mean haemoglobin levels were 13.3 g/dl in responders and 13.4 g/dl in non-responders (range 7.9-15.8). The distributions of haemoglobin levels were not significantly different when comparing either responders with non-responders or 'good' responders with 'poor' responders (P = 0.70 and P = 0.32, respectively). If haemoglobin is treated as a binary variable using 12.0 g/dl as the threshold, there is a non-significant trend towards a reduction in the probability of achieving a good response if baseline haemoglobin is below 12.0 g/dl (odds ratio = 0.26, confidence interval = 0.06-1.21; P = 0.086). The rate of complete pathological response was 4.3% (6/139). The mean haemoglobin level in these patients was 14.2 g/dl (range = 12.8-15.7), but the small numbers precluded further analysis.

CONCLUSIONS

There is no evidence for an influence of pre-treatment haemoglobin levels on the clinical response to neoadjuvant chemotherapy in breast cancer. It is unlikely that correction of anaemia above that which is warranted clinically will improve outcomes.

Clinical response assessment after 2 cycles of neoadjuvant chemotherapy for primary breast cancer fails to predict for final clinical response, but does predict for pathological complete response

667

Background: Neoadjuvant chemotherapy (NC) is increasingly being used for large primary breast carcinomas. Clinical trials have established its role in improving breast-conserving surgery (BCS) rates and have shown that complete pathological response (pCR) is associated with improved survival. Early evaluation of response is important for selecting patients with possible worse outcomes, who may benefit from alternative treatments. Methods: 107 women received NC for large operable breast cancers (T2–4, N1–2, M0). Patients received 6 cycles of FEC chemotherapy (5-FU 500mg/m2, epirubicin 75mg/m2, cyclophosphamide 500mg/m2 every 21 days) prior to a planned operation. Clinical response was recorded at baseline, after 2 cycles of NC and on completion of 6 cycles. Baseline and completion ultrasound and/or mammography were performed and a pathological assessment of response was made in those patients who underwent surgery. Results: Median age was 50 (range= 29–78). Overall clinical response rate after 2 cycles of chemotherapy was 59.8% (64/107) and after 6 cycles was 84.1% (90/107). 56 patients (52.3%) underwent BCS, 37 (34.6%) mastectomy and 14 (13.1%) no operation. Overall pCR rate was 15.0% (16/107). Of the 43 patients who failed to respond clinically after 2 cycles, 27 (62.8%) went on to exhibit a clinical response on completion of chemotherapy. 3 (7.0%) patients went on to have a complete clinical response and 21 (48.8%) underwent BCS. However, none of these 43 patients demonstrated a pCR. Conclusions: Lack of clinical response after 2 cycles of neoadjuvant chemotherapy does not preclude clinical response after further treatment with the same schedule, and many women will have sufficient down-staging to enable breast-conserving surgery. However, a pathological complete response is unlikely if no clinical response is observed after 2 cycles.

No significant financial relationships to disclose.