Early prediction of pathologic response to neoadjuvant treatment of breast cancer: use of a cell-loss metric based on serum thymidine kinase 1 and tumour volume

Abstract

Background: After neoadjuvant chemotherapy of breast cancer pathologic complete response (pCR) indicates a favorable prognosis. Among non-selected patients, pCR is, however, achieved in only 10-30%. Early evaluation of tumour response to treatment would facilitate individualized therapy, with ineffective chemotherapy interrupted or changed. The methodology for this purpose is still limited. Tumour imaging and analysis of macromolecules, released from disrupted tumour cells, are principal alternatives.

Objective: To investigate whether a metric of cell-loss, defined as the ratio between serum concentration of thymidine kinase1 (sTK1, ng x ml- 1) and tumour volume, can be used for early prediction of pathologic response.

Methods: One hunred four women with localized breast cancer received neoadjuvant epirubicin/docetaxel in 6 cycles, supplemented with bevacizumab in cycles 3-6. The cell-loss metric was established at baseline (n = 104), 48 h after cycle 2 (n = 104) and prior to cycle 2 (n = 57). The performance of the metric was evaluated by association with pathologic tumour response at surgery 4 months later.

Results: Treatment caused a rise in sTK1, a reduction in tumour volume and a marked increase in the cell-loss metric. Patients were subdivided into quartiles according to the baseline cell-loss metric. For these groups, baseline means were 0.0016, 0.0042, 0.0062, 0.0178 units. After subtraction of baselines, means for the quartiles 48 h after treatment 2 were 0.002, 0.011, 0.030 and 0.357 units. pCR was achieved in 24/104, their distribution in the quartiles (11, 11, 23 and 46%) differed significantly (p = 0.01). In 80 patients with remaining tumour, tumour size was inversely related to the metric (p = 0.002). In 57 patients studied before treatment 2, positive and negative predictive values of the metric were 77.8 and 83.3%, compared to 40.5 and 88.7% 48 h after treatment 2.

Conclusion: A cell-loss metric, based on serum levels of TK1, released from disrupted tumour cells, and tumour volume, reveal tumour response early during neoadjuvant treatment. The metric reflect tumour properties that differ greatly between patients and determine the sensitivity to cytotoxic treatment. The findings point to the significance of cell loss for tumour growth rate. The metric should be considered in personalized oncology and in the evaluation of new therapeutic modalities.

Trial registration: PROMIX (Clinical Trials.govNCT000957125).

Trial registration: ClinicalTrials.gov NCT00957125.

Keywords: Biomarker; Breast cancer; Cell-loss; Circulating thymidine kinase 1; Treatment response.

Conflict of interest statement

BT is a shareholder in AroCell Ab. The manuscript is written completely independent of the company.

Figures

Fig. 1

Receiver operating characteristic for distinguishing pCR from remaining tumour in 104 women, based on the cell-loss metric 48 h after the 2nd treatment cycle. At a cut-off value of 0.026 for the cell-loss metric, 1-specificity and sensitivity were 0.31 and 0.71, respectively. ROC Area = 0.714, p = 0.02

Fig. 2

Cell-loss metric 48 h after the 2nd treatment cycle in relation to pathologic tumour volume at surgery after six treatment cycles (p = 0.002)

Fig. 3

Percentage of pathological complete response in the breast after six cycles of chemotherapy among 104 women, grouped into quartiles according to the cell-loss metric obtained 48 h after the 2nd treatment cycle