Breast cancer screening,body mass index and prognosis benefit

Beeken et al hypothesized that screening compliance would be lower for obese individuals than for those with healthy weight status, but their study ¹ found no association between weight status and breast cancer screening in England. The study, one of the few to investigate body mass index (BMI) and breast cancer screening, was conducted in a country with high breast cancer screening compliance. ² In southern Italy, conversely, there is low screening compliance, causing delays in breast cancer diagnosis and also increased mortality.^3–5

The positive association between BMI and increased incidence and mortality is well recognized, but there are few studies on the association between BMI, mode of breast cancer detection, and prognosis. ⁶ We undertook a study to verify whether there were differences in breast cancer prognosis between screen-detected v.s. symptomatic women, according to BMI status.

We analyzed data from a clinical series of 448 women diagnosed with incident, histologically-confirmed breast cancer at the G. Pascale Foundation, National Cancer Institute of Naples. Detailed eligibility criteria are reported elsewhere. ⁷ BMI was categorized as: normal weight (<25 kg/m²), overweight (25–29 kg/m²) and obese (≥30 kg/m²). Tumours were considered screen-detected if suspicious findings were first detected by breast imaging within the routine national screening programme. Women were followed for a median of 4.6 years and rates of recurrence were calculated among screened and symptomatic patients by dividing the number of events by the total person-time at risk. Unadjusted relative risks (rate ratios) were obtained by dividing the event rate in the screened group by that in symptomatic group. Multivariate associations were estimated using the Cox regression model. Relative risks (hazard ratios) and 95% confidence intervals (CIs) were calculated, adjusting for influential factors. ⁸

We found no association comparing screen-detected and symptomatic women with BMI categories using the Chi-Square test (p = 0.24) (Table 1). Breast cancer recurrence was diagnosed in 12 women in the screen-detected group (11.2%) and 81 women in the symptomatic group (26%) (Table 2). The rate of recurrence was 22 per 1000 person-years for screen-detected (95% CI 11-39) and 64 per 1000 person-years for symptomatic women (95% CI 51-80). The unadjusted relative risk for screen-detected women was 0.37 (95% CI 0.21-0.66). Adjusting for age, education, lymph-nodes, tumour size, Ki-67, and receptor status using a Cox model yielded a relative risk of 0.46 (95% CI 0.23-0.91). The relative risk was lower when restricting to a BMI <25: adjusted risk of recurrence associated with screen-detected women was 0.07 (95% CI 0.13-0.73). In the other BMI categories, adjusted relative risks associated with screen-detected women were closer to 1.

Table 1.

Characteristics of study participants by mode of breast cancer detection.

Descriptive Characteristics	Mode of BC detection
	Symptomatic		Screen-Detected		p-value
	No.	%	No.	%
TOTAL	334	74.5	107	25.5
Body mass index					0.24
<25	132	41.4	48	44.9
25–29	105	32.9	40	37.4
≥30	82	25.7	19	17.8

Table 2.

Breast cancer recurrence in symptomatic and screen-detected women and stratification according to BMI.

	Person-years*	No. of events	Rate x1000 person-years (95% CI)**	Unadjusted HR (95% CI)	p†	Adjusted HR (95% CI)	p‡
DFS
Symptomatic	1268	81 (6.3%)	64 (51–80)	1 (referent)		1 (referent)
Screen-detected	542	12 (2.2%)	22 (11–39)	0.37 (0.21–0.66)	0.001	0.46 (0.23–0.91)	0.03
by BMI
BMI <25
Symptomatic	447	27 (6%)	60 (40–70)	1 (referent)		1 (referent)
Screen-detected	234	1 (0.4%)	4 (0.1–24)	0.07 (0.009–0.48)	0.008	0.09 (0.13–0.73)	0.02
BMI 25 – 29
Symptomatic	419	27 (6.4%)	65 (42–94)	1 (referent)		1 (referent)
Screen-detected	182	7 (3.8%)	38 (15–79)	0.58 (0.18–1.42)	0.5	1.32 (0.51–3.39)	0.6
BMI ≥30
Symptomatic	352	24 (6.8%)	68 (44–101)	1 (referent)		1 (referent)
Screen-detected	92	4 (4.3%)	43 (12–111)	0.63 (0.22–1.82)	0.4	0.43 (0.09–2.08)	0.3

*

Rates were calculated from the reference date (ie. date of diagnosis)

**

CI = confidence interval

†

Crude hazard ratios (HR) and 95% confidence intervals (CIs)

‡

HR and 95% CIs were calculated adjusting for age, education, lymph-nodes, T size, Ki67 and receptor status.

Our findings confirmed that screen-detected women have better prognosis and, similarly to Beeken et al, there was no association between weight status and breast cancer screening. Our study highlighted the potential impact of weight on breast cancer prognosis. Heavier women seemed to have more aggressive cancers and this was evident among the screen-detected group. This unexpected finding may be partly due to the association with higher BMI and metabolic syndrome, both negative factors for breast cancer risk and prognosis, ⁹ and partly explained by the strong positive correlation between BMI and absolute breast density, a biomarker of breast cancer risk that has been suggested as the gold standard breast density measurement. ¹⁰ Because the numbers in our study were small, further investigations are needed.