Table 2 Associations between variants in the BRCA2 DBD and risk of breast cancer and ovarian cancer

From: Functional evaluation and clinical classification of BRCA2 variants

 

Case

Control

OR (95% CI)f

P valuef

No. of variants

No. tested

No. of variants

No. tested

Breast cancer clinical testing cohorta

MAVE missense PStrongb

236

197,659

53

181,964

4.45 (3.30–6.13)

1.85 × 10–20

MAVE missense PStrong

 ≥99% Probability

168

197,659

32

181,964

5.38 (3.69–8.15)

7.41 × 10–17

 ≥95% Probability

201

197,659

40

181,964

5.09 (3.62–7.35)

1.71 × 10–19

MAVE missense PSMS

261

197,659

60

181,964

4.34 (3.27–5.85)

4.35 × 10–22

MAVE missense BStrong

920

197,659

1,155

181,964

0.78 (0.71–0.85)

7.14 × 10–8

MAVE missense BSMS

981

197,659

1,232

181,964

0.78 (0.71–0.85)

1.49 × 10–8

ACMG missense P/LP

210

197,659

29

181,964

6.96 (4.77–10.56)

5.87 × 10–21

ACMG missense B/LB

970

197,659

1,234

181,964

0.77 (0.70–0.83)

2.33 × 10–9

ENIGMA missense P

58

197,659

8

181,964

5.9 (3.08–12.74)

1.32 × 10–6

MAVE PSMS (no ENIGMA P)c

203

197,659

52

181,964

4.03 (2.96–5.61)

5.92 × 10-17

MAVE nonsense

205

197,659

34

181,964

5.65 (3.98–8.28)

1.23 × 10–19

Protein truncatingd

469

197,659

63

181,964

6.68 (5.19–8.74)

1.91 × 10–44

Breast cancer population-based cohorte

MAVE missense PStrong

63

81,073

26

83,247

2.49 (1.55–3.94)

4.67 × 10–5

MAVE missense PStrong

 ≥99% Probability

49

81,073

12

83,247

4.19 (2.23–7.89)

9.77 × 10–7

 ≥98% Probability

54

81,073

16

83,247

3.47 (1.98–6.06)

3.13 × 10–6

 ≥95% Probability

57

81,073

20

83,247

2.93 (1.76–4.87)

1.70 × 10–5

MAVE missense PSMS

76

81,073

42

83,247

1.86 (1.26–2.72)

1.19 × 10–3

MAVE missense BStrong

503

81,073

508

83,247

1.02 (0.90–1.15)

0.801

MAVE missense BSMS

544

81,073

551

83,247

1.01 (0.90–1.14)

0.832

ACMG missense P/LP

53

81,073

16

83,247

3.4 (1.93–6.17)

4.93 × 10–6

ACMG missense B/LB

542

81,073

554

83,247

1 (0.89–1.13)

0.952

ENIGMA missense P

26

81,073

5

83,247

5.34 (2.03–14.59)

9.52 × 10–5

MAVE nonsense

90

81,073

15

83,247

6.16 (3.58–11.15)

7.05 × 10–15

Protein truncating

179

81,073

25

83,247

7.36 (4.82–11.43)

4.87 × 10–31

Ovarian cancer clinical testing cohorta

MAVE missense PStrong

51

24,981

53

181,964

7.91 (5.32–11.79)

2.45 × 10–23

MAVE missense PStrong

 ≥99% Probability

36

24,981

32

181,964

9.93 (6.03–15.56)

3.24 × 10–19

 ≥95% Probability

44

24,981

40

181,964

9.32 (5.98–14.65)

1.05 × 10–22

MAVE missense PSMS

57

24,981

60

181,964

7.76 (5.34–11.29)

1.39 × 10–25

MAVE missense BStrong

135

24,981

1,155

181,964

0.91 (0.76–1.09)

0.362

MAVE missense BSMS

148

24,981

1,232

181,964

0.93 (0.78–1.10)

0.463

ACMG missense P/LP

43

24,981

29

181,964

11.69 (7.27–19.17)

2.28 × 10–22

ACMG missense B/LB

147

24,981

1,234

181,964

0.92 (0.77–1.09)

0.405

ENIGMA missense P

11

24,981

8

181,964

8.27 (3.47–20.03)

3.17 × 10–6

MAVE PSMS (no ENIGMA P)c

46

24,981

52

181,964

7.65 (5.05–11.59)

6.02 × 10–21

MAVE nonsense P

40

24,981

34

181,964

9.13 (5.74–14.63)

9.93 × 10–20

Protein truncating

85

24,981

63

181,964

9.35 (6.79–12.93)

1.51 × 10–40

  1. aThe clinical testing cohort included cases from patients with breast cancer and from patients with ovarian cancer (all had cancer genetic testing by Ambry Genetics), and controls from gnomAD v.4 (women), excluding the UK Biobank.
  2. bMAVE missense variants excluded variants enriched in subpopulation (>75%) or with an allele frequency of  >0.001 in any population.
  3. cMAVE missense pathogenic category variants (PStrong, PModerate and PSupporting (PSMS)) without ENIGMA-designated missense pathogenic variants.
  4. dProtein-truncating variants included frameshift and nonsense variants in the BRCA2 DBD.
  5. eCases and controls were from CARRIERS2 and BRIDGES29 studies.
  6. fFor the clinical cohort, ORs were calculated using weighted logistic regression with control populations weighted for the relative frequency of different races and ethnicities in the cases; P values were adjusted for multiple testing using the Benjamini–Hochberg method. For the population-based cohort, ORs were calculated using Fisher’s exact tests.
  7. Note that ovarian cancer included malignant epithelial tumours of the ovary and fallopian tube.
Back to article page