18F-FDG PET–Derived Tumor Blood Flow Changes After 1 Cycle of Neoadjuvant Chemotherapy Predicts Outcome in Triple-Negative Breast Cancer

Article Figures & Data

Download figure
Open in new tab
Download powerpoint
FIGURE 1.
Bivariate scatterplots of tumor BF response (ΔBF) and tumor metabolism response (ΔSUV_max).
Download figure
Open in new tab
Download powerpoint
FIGURE 2.
Kaplan–Meier curves showing OS according to menopausal status, histologic response, tumor BF response (ΔBF cutoff = −30%), and association of pathologic response and ΔBF. P values were determined with log-rank test.

TABLE 1

Factors Affecting Survival

	OS^*			Disease-free survival^*
Variable	Events	HR (95% CI)	P	Events	HR (95% CI)	P
Menopausal status
No	1/28	1		2/28	1
Yes	6/18	9.7 (1.1–84)	0.035	6/18	5.4 (1.0–28)	0.040
Glucose metabolic response (ΔSUV_max cutoff = −50%)
Good metabolic response	2/20	1		2/20	1
Poor metabolic response	5/20	3.1 (0.6–16)	NS	5/20	3.0 (0.6–16)	NS
Unknown	0/6			1/6
BF response (ΔBF cutoff = −30%)
Good BF response	1/20	1		1/20	1
Poor BF response	6/20	9.6 (1.1–84)	0.037	6/20	9.2 (1.1–80)	0.040
Unknown	0/6			1/6
Type of surgery
Breast-conserving	3/30	1		4/30	1
Mastectomy	4/16	4.7 (1.0–22)	0.046	4/16	3.3 (0.8–14)	NS
Histologic response
pCR	0/20	NC	—	1/20	1
Non-pCR	7/26			7/26	6.7 (0.8–57)	NS
Histologic response + BF response (cutoff = −30%)
pCR	0/20	NC	—	1/20	1
Non-pCR and good BF response	1/10			1/10	1.4 (0.3–20)	NS
Non-pCR and poor BF response	6/13			6/13	13 (1.5–116)	0.017
Unknown	0/3			0/3

↵* Univariate Cox proportional hazards model (patients with unknown variables were not included in the model).
Tumor size (>5 cm), American Joint Committee on Cancer staging, lymph node involvement, history of pregnancy, Scarf–Bloom–Richardson grading, tumor architectural differentiation, nuclear pleomorphism, number of mitosis, SUV_max1, BF₁, and the SUV_max1/BF₁ ratio were not associated with OS. Data in parentheses are 95% CIs.
HR = hazard ratio; NS = nonsignificant (P > 0.05); NC = not calculable.