Table 2.

Analysis of reported cancer history in germline BAP1 mutation carriers. This table shows the reported history, in probands and their families, of different cancer types according to predicted variant type (first five columns, associated statistical comparisons are shown in the next 5 columns) and histological appearance (indicated as whether lesions show ‘BAP-like histology’ or not, discussed in Materials and Methods section: ‘Histological review’ (last three columns)). The reported history of cancers is described in probands alone, their family members alone (first- or second-degree relatives), or within probands or family members, with the exception of cutaneous melanoma as it was the proband ascertainment characteristic and as such is only described in probands’ family members. For variant type, Fisher’s exact test (or Pearson’s chi-squared test where appropriate – marked with a ‘p’) results, odds ratios and 95% confidence intervals (calculated using the Cornfield approximation) are displayed for the association between the cancer history, BAP-like phenotypes or BAP-like histology and either any type of variant, a benign variant versus no variant, or a predicted deleterious variant versus no variant. Rows named ‘BAP-like phenotypes’ refer to the presence of at least one of the cancers previously reported to occur in the BAP1 family cancer syndrome (see Materials and Methods section: ‘Analysis of cancer phenotypes in BAP1 variant carriers’) (second to last six rows). The last row shows the distribution of individuals with BAP-like histology (52/713, as indicated in the second-to-last column) by the different BAP1 variant carrier groups. If multiple cases had been tested within the same family, only the first recruited case was retained to avoid over-estimation of cancer history (exceptions described in detail in the Results section: ‘Analysis of cancer phenotypes in BAP1 variant carriers’). ‘None’ means that no germline BAP1 mutation was detected. For the purpose of this analysis, variants that were of either ‘uncertain significance’ or were predicted to be benign by SIFT/Polyphen-2 were grouped together and called ‘benign’. The three cases with clear loss-of-function germline BAP1 mutations were grouped together with those variants that were predicted to be ‘deleterious’ by SIFT/Polyphen-2. Basal cell carcinoma (BCC). Not calculable (NC). Not applicable (NA). Cutaneous melanoma (MM)

					Analysis by type of BAP1 variant					Analysis by histological appearance
					Any variant	Benign versus no variant		Predicted deleterious versus no variant
Cancer history in the proband and family	Presence of any BAP1 variant in the proband (melanoma case)				Fisher’s exact P-value	Fisher’s exact P-value	OR (95% CI)	Fisher’s exact P-value	OR (95% CI)	BAP-like histology (in the proband’s melanoma for which central pathology review was available, total N = 713)		Fisher’s exact P-value
	N (Row %)			Total N (%)						N (Column %)
	None (N = 1868)	Benign (N = 100)	Predicted deleterious (N = 9)							No (N = 661)	Yes (N = 52)
Renal cancer (in the proband)	5 (71.4)	2 (28.6)	0 (0)	7 (100.0)	0.08	0.045	7.6 (0.7–47.0)	1	0 (0–174.3)	3 (0.5)	0 (0)	1
Renal cancer (in the family)	48 (90.6)	4 (7.5)	1 (1.9)	53 (100.0)	0.1	0.3	1.6 (0.4–4.4)	0.2	4.7 (0.1–36.5)	12 (1.8)	1 (1.9)	1
Renal cancer (in the proband or family)	53 (88.3)	6 (10.0)	1 (1.7)	60 (100.0)	0.06	0.07 p	2.2 (0.7–5.3)	0.2	4.3 (0.1–32.9)	15 (2.3)	1 (1.9)	1
BCC (in the proband)	261 (93.9)	14 (5.0)	3 (1.1)	278 (100.0)	0.3	0.99 p	1.0 (0.5–1.8)	0.1	3.1 (0.5–14.5)	99 (15.0)	8 (15.4)	0.9 p
BCC (in the family)	114 (93.4)	5 (4.1)	3 (2.5)	122 (100.0)	0.02	0.7 p	0.8 (0.3–2.0)	0.02	7.7 (1.2–36.5)	33 (5.0)	1 (1.9)	0.5
BCC (in the proband or family)	346 (94.5)	17 (4.7)	3 (0.8)	366 (100.0)	0.4	0.7 p	0.9 (0.5–1.6)	0.4	2.2 (0.4–10.4)	119 (18.0)	8 (15.4)	0.6 p
Meningioma (in the proband)	4 (100.0)	0 (0)	0 (0)	4 (100.0)	1	1	0 (0–18.1)	1	0 (0–219.4)	2 (0.3)	0 (0)	1
Meningioma (in the family)	0 (0)	0 (0)	1 (100.0)	1 (100.0)	0.005	NC	NC	0.005	NC	0 (0)	1 (1.9)	0.07
Meningioma (in the proband or family)	4 (80.0)	0 (0)	1 (20.0)	5 (100.0)	0.02	1	0 (0–18.1)	0.02	58.3 (1.1–670.5)	2 (0.3)	1 (1.9)	0.2
Mesothelioma (in the proband)	0 (0)	0 (0)	1 (100.0)	1 (100.0)	0.005	NC	NC	0.005	NC	1 (0.2)	0 (0)	1
Mesothelioma (in the family)	4 (66.7)	0 (0)	2 (33.3)	6 (100.0)	0.0004	1	0 (0–18.1)	0.0003	133.1 (10.1–1084.2)	1 (0.2)	2 (3.9)	0.02
Mesothelioma (in the proband or family)	4 (57.1)	0 (0)	3 (42.9)	7 (100.0)	0.000003	1	0 (0–18.1)	0.000003	233 (26.7–1660.1)	2 (0.3)	2 (3.9)	0.03
Cutaneous melanoma (in the family)	146 (93.0)	8 (5.1)	3 (1.9)	157 (100.0)	0.049	0.95 p	1.03 (0.4–2.2)	0.03	5.9 (0.9–27.9)	47 (7.1)	4 (7.7)	0.8
Ocular melanoma (in the proband)	0 (0)	0 (0)	0 (0)	0 (0)	NC	NC	NC	NC	NC	0 (0)	0 (0)	NC
Ocular melanoma (in the family)	2 (100.0)	0 (0)	0 (0)	2 (100.0)	1	1	0 (0–36.1)	1	0 (0–439.1)	1 (0.2)	0 (0)	1
Ocular melanoma (in the proband or family)	2 (100.0)	0 (0)	0 (0)	2 (100.0)	1	1	0 (0–36.1)	1	0 (0–439.1)	1 (0.2)	0 (0)	1
BAP-like phenotype (in the proband)	267 (93.4)	16 (5.6)	3 (1.0)	286 (100.0)	0.2	0.6 p	1.1 (0.6–2.0)	0.1	3.0 (0.5–14.1)	102 (15.4)	8 (15.4)	1 p
BAP-like phenotype (in the family)	166 (93.3)	8 (4.5)	4 (2.2)	178 (100.0)	0.01	0.8 p	0.9 (0.4–1.9)	0.006	8.2 (1.6–38.4)	46 (7.0)	3 (5.8)	1
BAP-like phenotype MM (in the family)	289 (93.5)	16 (5.2)	4 (1.3)	309 (100.0)	0.08	0.9 p	1.04 (0.6–1.8)	0.04	4.4 (0.9–20.4)	85 (12.9)	5 (9.6)	0.5 p
BAP-like phenotype (in the proband or family)	397 (94.1)	21 (5.0)	4 (0.9)	422 (100.0)	0.2	0.95 p	0.98 (0.6–1.6)	0.1	3.0 (0.6–13.8)	132 (20.0)	10 (19.2)	0.9 p
BAP-like phenotype MM (in the proband or family)	496 (94.1)	27 (5.1)	4 (0.8)	527 (100.0)	0.5	0.9 p	1.02 (0.6–1.6)	0.3	2.2 (0.4–10.3)	163 (24.7)	12 (23.1)	0.8 p
BAP-like histology present (in the proband’s melanoma) for which central pathology review was available (N = 713)	49 (94.2)	1 (1.9)	2 (3.9)	52 (100.0)	0.1	0.7	0.4 (0.01–2.6)	0.1	3.6 (0.4–19.8)	NA	NA	NA