Exploiting evolutionary principles to prolong tumor control in preclinical models of breast cancer

Abstract

Conventional cancer treatment strategies assume that maximum patient benefit is achieved through maximum killing of tumor cells. However, by eliminating the therapy-sensitive population, this strategy accelerates emergence of resistant clones that proliferate unopposed by competitors-an evolutionary phenomenon termed "competitive release." We present an evolution-guided treatment strategy designed to maintain a stable population of chemosensitive cells that limit proliferation of resistant clones by exploiting the fitness cost of the resistant phenotype. We treated MDA-MB-231/luc triple-negative and MCF7 estrogen receptor-positive (ER(+)) breast cancers growing orthotopically in a mouse mammary fat pad with paclitaxel, using algorithms linked to tumor response monitored by magnetic resonance imaging. We found that initial control required more intensive therapy with regular application of drug to deflect the exponential tumor growth curve onto a plateau. Dose-skipping algorithms during this phase were less successful than variable dosing algorithms. However, once initial tumor control was achieved, it was maintained with progressively smaller drug doses. In 60 to 80% of animals, continued decline in tumor size permitted intervals as long as several weeks in which no treatment was necessary. Magnetic resonance images and histological analysis of tumors controlled by adaptive therapy demonstrated increased vascular density and less necrosis, suggesting that vascular normalization resulting from enforced stabilization of tumor volume may contribute to ongoing tumor control with lower drug doses. Our study demonstrates that an evolution-based therapeutic strategy using an available chemotherapeutic drug and conventional clinical imaging can prolong the progression-free survival in different preclinical models of breast cancer.

Fig. 1. MRI volumetric data for the different treatments applied to MDA-MB-231 preclinical models

(A to D) Different cohorts of mice injected at different times. Control (Ctrl) animals did not receive any chemotherapy. The mice in the ST group received the standard high-dose treatment [paclitaxel, 20 mg/kg intraperitoneally, twice a week for 2.5 weeks]. AT groups received different doses depending on the algorithm (AT-1 or AT-2; see Materials and Methods). Data are means ± SD. More detailed results, including ADC values and volumes for individual tumors, are available in fig. S1. Arrows indicate when the treatment in the ST group was stopped.

Fig. 2. MRI volumetric data for the different treatments applied to MCF7 preclinical models

Ctrl animals did not receive any chemotherapy. The mice in the ST group received the standard high-dose treatment (paclitaxel, 20 mg/kg intraperitoneally, twice a week for 2.5 weeks). AT groups received different doses depending on the algorithm (AT-1 or AT-2; see Materials and Methods). Data are means ± SD. More detailed results, including ADC values and volumes for individual tumors, are available in fig. S1. Arrow indicates when the treatment was stopped in the ST mice.

Fig. 3. Survival of animals treated by different therapeutic algorithms

The Kaplan-Meier plots show animals with tumor volumes smaller than 1000 mm³ after treatment according to the standard (STD) or adaptive (AT-1 and AT-2) strategies. In the legend, n corresponds to the number of animals under each treatment algorithm. (A and B) Mice injected with MDA-MB-231 (A) and MCF7 (B) cell lines. Statistics were calculated with the Mantel-Cox test.

Fig. 4. Total delivered dose of paclitaxel with each algorithm

Total delivered dose of paclitaxel for MDA-MB-231 cohort D (upper graph) and MCF7 cohort (lower graph) is shown as the cumulative dose of paclitaxel for each day after MRI monitoring had started. Blue corresponds to mice treated under AT-1; green, mice treated under AT-2; and red, animals under ST treatment. Note that mouse#9 and mouse#13 in the MDA-MB-231 group received two sessions of ST therapy to demonstrate that the cancer cells that recur after standard therapy are drug-resistant. Cumulative dose graphs for MDA-MB-231 cohorts A to C are shown in fig. S2.

Fig. 5. Mean percentage of necrotic tissue in tumors under different treatment algorithms as determined by DW MRI

The percentages of tissue volume that is necrotic are represented relative to the first day of treatment. The data summarize results for MDA-MB-231 and MCF7, in the left and right graphs, respectively, under different therapies: standard (ST), AT-1, and AT-2. After an initial increase, the necrotic volume decreases in the AT-1 (black diamonds) cohort and increases in the ST and AT-2 cohorts. Results are means ± SE.

Fig. 6. Changes in tumor flow and perfusion under different treatments

Bar graphs show the changes in area under the curve (AUC) for blood flow and perfusion measurements derived from DCE MRI before (Beginning), during (Middle), and after (End) therapy relative to pretreatment baseline. In this case, only Ctrl, ST, and AT-1 therapies were monitored. The numbers in parentheses correspond to the number of mice analyzed in each group. Results are means ± SE.

Fig. 7. Histology of tumors under different therapeutic regimens

The graphs show H&E, CD31, and SMA analysis of the indicated tumors under different therapeutic regimens, with MDA-MB-231 in the top row and MCF7 in the bottom row. Tumors treated according to AT-1 developed less necrosis (detected by H&E staining) than those on AT-2, and greater vascular density (CD31 staining) in the case of MDA-MB-231 tumors. In MCF7 tumors, necrosis did not correlate with the stability of the tumors, but increased vascular density was observed in tumors that were treated with AT-2. The AT-1 cohort had a higher microvessel density (CD31) than the Ctrl or ST groups but did not achieve statistical significance. However, the AT-1 group did have significantly higher vessel functionality demonstrated by the SMA staining. Significance testing was performed by Student's t test. P values are indicated above a line showing each comparison.