Effect of Preventive Hormonal Therapy on Breast Density: A Systematic Qualitative Review

Table 3

Aromatase inhibitors: characteristics of the reviewed studies (n: 7).

Authors/year	Endpoint	Design	High risk definition	Population/number of cases	Age	Method of BD measurement/time	Mammograms readers	Reproducibility	BD baseline
Letrozole (LET)
Cigler et al. 2010 [51]	Effect of letrozole on breast density	A multicenter, randomized, double-blind, placebo-controlled trial (2 : 1 ratio of LET (2.5 mg daily) or placebo)	NA	Healthy postmenopausal women with or without prior BC and with an estimated >25% BD on baseline mammogram/a total of 120 patients would be required; 67 women were randomized and 50 were included in the final analysis (LET: 31, placebo: 19)^1	Median: 57.4 y; LET: 57.3 y; placebo: 58.1 y	Visually according to Boyd classification and BIRADs systems at baseline and 12 months and by a computer-assisted thresholding program at baseline, 12 and 24 months	A single observer blinded^1a	NA	Mean baseline BD: 39.6% (95% CI 32.3–47.0) for women on LET and 40.0% (95% CI 32.4–47.7) on placebo
Vachon et al. 2007 [52]	Effect of letrozole on breast density	Subset analysis of a prospective randomized trial	NA	Women with early-onset BC after 5 yrs of tamoxifen/204 women were potentially eligible; 104 women met all study criteria: LET: 𝑛=56 (54%); placebo: 𝑛=48 (46%)	Median age (range): placebo: 64.7 y (49–80); LET: 64.6 y (30–84)	On digitized mammograms using a computer-assisted thresholding program at baseline and 9–15 months	A trained programmer	Intraobserver variation no greater than 10% on over 500 images	Placebo/LET 0–9%: 12 (26%)/15 (28%); 10–24%: 20 (43%)/21 (39%); 25–49%: 12 (26%)/15 (28%); 50–74%: 3 (6%)/3 (6%); 75+%: 0 (0%)/0 (0%). Mean (SD): placebo (𝑛=33): 20% (14.1); LET (𝑛=35): 18.5% (14.5)
Smith et al. 2012 [53]	Effect of letrozole on breast density	Prospective, single-arm study	LCIS, ADH, ALH, or DCIS, or BRCA-1 mutation or a Gail 5 y of BC > 8%	Postmenopausal women with risk factors for developing BC/20 enrolled, 16 evaluable	Median: 58.9 y	On digitized mammograms using a computer-assisted thresholding software at baseline, 6 and 12 months	An experienced reader blinded to time sequence	NA	Range: 1.049% to 64.1425% (median: 27.70%)^2
Fabian et al.; 2007 [54]	Effect of letrozole on biomarkers of BC risk including breast density	Phase II biomarker based pilot prospective, single-arm study	(1) Gail 5 y of BC ≥ 1.67%; (2) prior ADH, ALH, or LCIS; (3) prior treated contralateral DCIS; (4) known BRCA1/2 mutation	Postmenopausal women with risk factors for developing BC and who were on HRT/42 women^3	Median 50 y (range 39–68)	On digitized mammograms using a computer-assisted thresholding software at baseline and 6 months	A single reviewer blinded to the time sequence of mammograms	NA	Median BD at baseline for the entire group: 31.5%
Mousa et al.; 2008 [55]	Effect of letrozole on breast density	Preliminary retrospective case control study	NA	Postmenopausal women who were on HRT/56 women: cases (𝑛=28) received low-dose HRT daily + LET 2.5 mg and controls: HRT alone (𝑛=28). BD assessed for 18 women in cases and 22 in controls^4	For 18 cases and 22 controls: mean ± SD: cases: 63.06 y ± 6.99; controls: 65.45 y ± 8.28	Visually according to BI-RADs system and by a computer-assisted thresholding program at initial time and after 24 months of treatment	One radiologist blinded to time sequence and treatment arm	Correlation coefficient for the intraobserver reliability: 0.92 for the visual scores of BD	Total IPI for 1st mammogram: mean ± SD: cases (𝑛=18): 2.1×10^10±2.6×10^10; controls (𝑛=22): 1.5 × 10^10 ± 2.4 × 10^10
Anastrozole (ANAS)
Prowell et al.; 2011 [56]	Effect of anastrozole on breast density	Prospective, single-arm study	NA	Postmenopausal women with BC with aromatase inhibitor^5as their only systemic therapy/54 patients; 50 assessed for breast density measurement, 43 patients had both baseline and 12-month mammograms	Mean age (range): 62.5 y (47–79)	On digitized mammograms using a computer-assisted thresholding program at baseline, 6 months and 12 months	A single observer blinded to time sequence	NA	Median BD at baseline (range): 13.4% (0.1–66.2)
Exemestane (EXE)
Cigler et al. 2011 [57]	Effect of exemestane on breast density	Prospective double-blind randomized trial (1 : 1 ratio)	NA	Healthy postmenopausal women/98 women were randomized (49 to EXE 25 mg; 49 to PLAC) and 65 had BD data at baseline and 12 months, and 43 at baseline and 24 months	Median: 56.9 y median: EXE: 56.9 y, Placebo: 56.8 y	Visually according to Boyd classification and BI-RADs system at baseline and on digitized mammograms using a computer-assisted method at baseline, 6, 12, and 24 months	A trained physician blinded^1a	NA	Mean baseline BD: 33.9% (95% CI from 27.6 to 40.2) for women on EXE (𝑛=38) and 36.5% (95% CI from 33.3 to 42.8) for women on placebo (𝑛=34)

1Only 43 women (27 on letrozole and 16 on placebo) had both baseline and 24 months. ^1aBlinded to any clinical data corresponding to each mammogram. LCIS: lobular carcinoma in situ, ADH: atypical ductal hyperplasia, ALH: atypical lobular hyperplasia, and DCIS: ductal carcinoma in situ. ²Personal calculation. HRT: hormonal replacement therapy. ³Thirty-five of 42 subjects were taking estrogen alone (various preparations) and seven estrogen + progesterone. Fifty-nine percent were on transdermal preparations. ⁴Study participants: all received letrozole (Femara) 2.5 mg orally 3 times weekly with the exception of 2 women who were given anastrozole (Arimidex) 1 mg/day because of headaches with letrozole. IPI: computerized calculation of integrated pixel intensity. ⁵Anastrozole 1 mg daily.