Reduction of breast cancer relapses with perioperative non-steroidal anti-inflammatory drugs: new findings and a review

Abstract

To explain a bimodal pattern of hazard of relapse among early stage breast cancer patients identified in multiple databases, we proposed that late relapses result from steady stochastic progressions from single dormant malignant cells to avascular micrometastases and then on to growing deposits. However in order to explain early relapses, we had to postulate that something happens at about the time of surgery to provoke sudden exits from dormant phases to active growth and then to detection. Most relapses in breast cancer are in the early category. Recent data from Forget et al. suggest an unexpected mechanism. They retrospectively studied results from 327 consecutive breast cancer patients comparing various perioperative analgesics and anesthetics in one Belgian hospital and one surgeon. Patients were treated with mastectomy and conventional adjuvant therapy. Relapse hazard updated Sept 2011 are presented. A common Non-Steroidal Anti-Inflammatory Drug (NSAID) analgesic used in surgery produced far superior disease-free survival in the first 5 years after surgery. The expected prominent early relapse events in months 9-18 are reduced 5-fold. If this observation holds up to further scrutiny, it could mean that the simple use of this safe, inexpensive and effective anti-inflammatory agent at surgery might eliminate early relapses. Transient systemic inflammation accompanying surgery could facilitate angiogenesis of dormant micrometastases, proliferation of dormant single cells, and seeding of circulating cancer stem cells (perhaps in part released from bone marrow) resulting in early relapse and could have been effectively blocked by the perioperative anti-inflammatory agent.

Fig. (1)

Hazard of relapse for premenopausal patients treated at Istituto Nazionale Tumori in Milan, Italy. Hazard is the number of events that occur in a time interval divided by the number of patients who enter that time as event free. Patients were treated by mastectomy well before the routine use of adjuvant therapy. The time interval in all hazard figures used here is 3 months. Average and standard deviations are indicated as diamonds and bars. The curve was obtained by a kernel-like smoothing procedure.

Fig. (2)

Same as fig. 1 except that these are postmenopausal patients.

Fig. (3)

These are independent data for mastectomy treated patients in disease free survival format as modified from Fisher et al. Cancer 1984 [37]. Patients are grouped by axillary lymph nodes invasion, with axillary lymph node tumor-free (N-) patients in the uppermost curve and patients with 10 or more invaded lymph nodes (N+) in the lower curve. The magnitude of the early relapse component may be visualized. For N- patients, after surgery alone 80% of patients are long lasting disease-free with half the relapses early and half late. For the poor prognosis patients (N+ with 10 or more involved lymph nodes), surgery alone is quite ineffective. Most of relapses (90%) are early and support the label of “poor prognosis”. These data are visually and quantitatively very similar to the Milan data shown in fig. 4. Poor prognosis appears to have less impact on late relapses.

Fig. (4)

Milan data for patients treated with mastectomy are presented in the more conventional format as disease free survival. The percentage disease free starts at 100% and rapidly drops until approximately 4 years where there seems to be a short plateau. Relapses start to happen again at about 5 years and slowly continue thereafter tapering off gradually at about 15 years. The plateau at 4 years corresponds to the end of the early relapses seen in figs. 1 and 2. Modified from Bonadonna et al. NEJM 1995 [38].

Fig. (5)

Computer simulations of early relapse events. Hazard of relapse for early events centered at 10 months and at 30 months post-surgery as proposed by computer simulation are shown. Simulations included effects of mastectomy and were based on Milan data shown in figs. 1 and 2. The 10 month and 30 month events may be distinguished in fig. 1 and are less clear but present in fig. 2.

Fig. (6)

The result of adjuvant CMF chemotherapy. The two early peaks in the untreated population coalesce into a single peak in this treated population at about 20 months. Apparently CMF chemotherapy acts to produce most extensive reduction in relapse hazard in the 1^st and 3^rd year [39].

Fig. (7)

Forget et al. [40] data from Universite catholique de Louvain in Brussels, Belgium. Relapse hazard is shown for mastectomy patients given ketorolac or not. Data are smoothed as indicated for fig. 1.

Fig. (8)

Forget et al. data were updated September 2011 and shown in hazard form but not smoothed as in fig. 7. Patient data are presented in the table. Patients included in this figure were less than 80 years of age, tumor less than 9 cm diameter and disease free survival greater than 2 months. It can be seen that relapses in months 9 -18 accounted for the major difference between ketorolac and non-ketorolac patients.

Fig. (9)

Symbolic description of proposed explanations for Forget et al. data. Early relapses are assumed to be related, at least in part, to the inflammatory process due to primary tumor surgical removal, directly or indirectly eliciting peritumoral endothelial cell and single cell proliferation. A few possible mechanisms are explained. A) Angiogenic factors, like VEGF and bFGF, are directly released by degranulated platelets or even produced via IL-6; B) Bone marrow derived CXCR-4 positive cells, acting both on tumor foci and on the inflammatory process, are mobilized by SDF-1 directly released or even produced via COX-2. Perioperative ketorolac would restrict both endocrine and cellular pathways, thus impairing the metastatic process. CTC refers to circulating tumor cells.

Fig. (10)

Hazard data from a Milan database for 121 TNBC patients with 10-11 years follow up. There are 50 relapse events within 5 years of surgery. The general similarity of these hazard data to Forget et al. data for no-ketorolac patients seen in Fig. 7 leads to the suggestion that TNBC may be the ideal study group with which to test perioperative ketorolac.