Association analysis of germline mutations in CHEK2, PALB2, NBN and RECQL with the risk of ductal carcinoma in situ in Polish women

Abstract

Background

The genetic background of ductal carcinoma in situ (DCIS) has not been well explored. Previously, we reported that Polish founder mutations of BRCA1/2 confer susceptibility to DCIS. The aim of our study was to investigate the role of CHEK2, PALB2, NBN and RECQL mutations in the ethology of DCIS.

Methods

We studied 564 Polish women with DCIS for eight Polish founder alleles, including four in CHEK2 (c.1100delC, c.444 + 1G > A, del5395 and c.470T > C), two in PALB2 (c.509_510delGA and c.172_175delTTGT), one in NBN (c.657_661delACAAA) and one in RECQL (c.1667_1667 + 3delAGTA). To investigate the association of these alleles with DCIS risk, we used mutation frequencies in cancer-free controls as a reference (4000 to 4702 controls for different variants). To analyze survival, patients were followed on average for 156 months.

Results

A CHEK2 mutation (all variants combined) was associated with an increased risk of DCIS (OR = 1.7, p = 0.003). The risk was higher for CHEK2 truncating mutations (OR = 3.0, p = 0.001) than for a missense variant c.470T > C (OR = 1.5, p = 0.04). The risk was highest for carriers of CHEK2 truncating mutations with a family history of breast cancer (OR = 4.2, p = 0.01). There were no deaths reported in 52 CHEK2 mutation carriers during the follow up time. PALB2, NBN and RECQL mutations were rare among cases and were not associated with DCIS risk in Polish women.

Conclusions

Based on the current study, women with a CHEK2 mutation face an increased risk of DCIS. The presence of DCIS should be considered during surveillance of CHEK2 mutation carriers. On the other hand, DCIS patients should receive genetic counseling and testing for CHEK2 mutations.

Background

Breast cancer is the most common malignancy in women, annually diagnosed in above two million women worldwide, including over twenty thousand cases in Poland [1, 2]. The genetic basis of breast cancer has been deeply explored over the several past decades [3–7]. The two major highly penetrant genes for breast cancer, BRCA1 and BRCA2, were identified in 1994 and 1995, respectively [8, 9]. Since that, many other genes were reported to be associated with a genetic susceptibility to this cancer including PALB2, CHEK2, ATM, BARD1, BLM, NBN, XRCC2, RECQL, RAD50, RAD51C, RAD51D, TP53, PTEN, NF1, STK11 and CDH1 [10–17]. Now, identifying mutations in breast cancer susceptibility genes is used in clinical practice for risk prediction, prevention and surveillance in unaffected individuals and treatment of cancer patients [18–21].

The genetic background of invasive breast cancer has been well explored, but genetic determinants of non-invasive breast cancers have not been extensively studied [4, 22, 23]. Non-invasive breast tumors constitute about 10–15% of all breast cancers [24]. Ductal carcinoma in situ (DCIS) is the most common type of non-invasive breast cancer [25–27]. It represents about 90% of all non-invasive breast tumors [28, 29]. Most DCIS tumors are asymptomatic, identified by screening mammography [24]. DCIS is associated with a favorable prognosis, but it is a precursor of invasive ductal breast cancer, so it is important to identify risk factors for this disease as well [30, 31]. The incidence of DCIS varies significantly by a population. It has been reported that DCIS constitutes 5–25% of all breast cancer cases, and the lifetime risk of DCIS ranges between 0.5% and 3% in different countries [31]. These differences are explained by the fact that the breast screening programs vary in their recommendations and effectiveness across different populations [24].

According to Polish National Cancer Registry report, in year 2022 invasive breast cancer (C50 in ICD-10 classification) was diagnosed among 21.6 thousand women (cumulative risk of 10%), while in situ breast cancer (D05 in ICD-10 classification) was diagnosed in 1.43 thousand women (cumulative risk of 0.6%). Based on the report, DCIS constitutes about 5% of breast cancers in Poland, and the lifetime risk of DCIS is only approximately 0.5% in Polish women [32]. Given that the incidence of DCIS determined by a large meta-analysis of 13 post-mortem studies in women is as high as 9%, DCIS appears to be underdiagnosed both in Poland and elsewhere [33].

In 2019, we have conducted a comprehensive analysis of breast cancer susceptibility in Poland. We detected pathogenic mutations in 512 of 1,018 (50%) families with hereditary breast cancer (HBC), and we have found that 20 founder mutations in six genes, BRCA1, BRCA2, CHEK2, PALB2, NBN and RECQL are responsible for 82% of all Polish HBC families with detected mutations [34]. It is interesting to establish whether this panel of founder mutations in BRCA1, BRCA2, CHEK2, PALB2, NBN and RECQL contribute to DCIS burden in Poland.

To answer this question, previously we investigated if BRCA1/2 mutations predispose to DCIS. We found that BRCA1/2 founder mutations are responsible for 3% of DCIS cases in Poland. BRCA2 variants conferred higher DCIS risk (OR = 11.3, p < 0.0001) than BRCA1 mutations (OR = 3.3, p = 0.01) [35]. In the current study, we established whether mutations in non-BRCA1/2 genes (from the panel of 20 Polish founder alleles) predispose to DCIS. We analyzed the frequencies of 8 founder alleles in four genes including CHEK2 (c.1100delC, c.444 + 1G > A, del5395 and c.470T > C), PALB2 (c.509_510delGA and c.172_175delTTGT), NBN (c.657_661delACAAA) and RECQL (c.1667_1667 + 3delAGTA) among 564 Polish women with DCIS and compared with mutation frequencies in Polish cancer-free controls (4000 to 4702 controls for different variants). Only CHEK2 mutations predisposed to DCIS, therefore we investigated their influence on age of DCIS diagnosis, the impact on survival among patients, and the effect of positive family history of breast cancer in CHEK2 carriers on the risk of DCIS.

Methods

Patients

We studied a series of 564 women with ductal carcinoma in situ, diagnosed between 1997 and 2019, unselected for age and family history. The patients were derived from a group of 17,086 breast cancer cases registered at the Hereditary Cancer Center in Szczecin. We excluded patients with a personal history of breast cancer or ovarian cancer before diagnosis of DCIS. The diagnosis of DCIS, with or without micro-invasion, was confirmed by a pathology report in all subjects. Each patient provided a blood sample for DNA isolation within 2 years from the date of diagnosis.

All patients signed an informed consent for genetic testing. Each patient was self- or physician-referred for genetic counseling. Clinical, demographic, and self-reported ancestry information were obtained at the time of the interview. All women were Poles. A pedigree including cancer cases in first- and second-degree relatives and their ages of diagnosis was collected for each participant. Survival data (alive or dead; if deceased, the date of death) were obtained from the Polish Ministry of the Interior and Administration in February 2020. The study was approved by the Ethics Committee of Pomeranian Medical University in Szczecin (IRB No. KB-0012/34/05/2020/Z).

The mean age of diagnosis of DCIS among the 564 women was 55.3 years (median 55.0, range 23–91 years, SD 10.8); 27.3% (154/564) women were diagnosed with DCIS before the age 50 years, 7.1% (40/564) women were diagnosed before the age 40 years. In 145 (25.7%) cases, there was a family history of breast cancer in at least one first or second degree relative (familial cases). The median follow up was 156 months. Eighteen (3.2%) patients died within the follow-up. The cases were described in more detail previously [35].

Mutation analysis

All 564 patients were tested for 8 mutations, including four in CHEK2; c.1100delC (p.Thr410Metfs), c.444 + 1G > A, del5395 (p.Thr410Metfs) and c.470T > C (p.Ile157Thr), two in PALB2: c.509_510delGA (p.Arg170Ilefs) and c.172_175delTTGT (p.Gln60Argfs), one allele of NBN: c.657_661delACAAA (p.Lys219Asnfs) and one in RECQL: c.1667_1667 + 3delAGTA (p.K555delinsMYKLIHYSFR) [36].

A large deletion of exon 9 and 10 of CHEK2 (del5395) was detected by multiplex-PCR using two primer pairs; the first pair flanked breakpoint site in intron 8 of CHEK2; the second flanked breakpoint site in intron 10. In mutation negative cases, two PCR fragments of 379 and 522 bp were amplified from the wild type allele. In mutation positive cases the forward primer of the first pair and the reverse primer of the second pair amplified an additional fragment of 450 bp, as described previously [37].

The other 7 mutations: CHEK2 c.1100delC, c.444 + 1G > A, c.470T > C; PALB2 c.509_510delGA and c.172_175delTTGT; NBN c.657_661delACAAA and RECQL c.1667_1667 + 3delAGTA were genotyped in an automated thermal cycler (LightCycler^® Real-Time PCR 480 System). The 5 µl reaction mixture contained: 1 µl (10ng) of genomic DNA, 0.125 µl TaqMan Assay 40x (Thermo Fisher Scientific, Waltham, MA, USA), 2.5 µl of reaction mix (GoTaq^®Probe qPCR Master Mix 2x, Promega), supplemented to 5 µl with 1,375 µl nuclease-free water. TaqMan assays are described in Table 1. Real-time PCR conditions were described previously [35].

Table 1.

TaqMan assays used to detect Polish founder mutations of CHEK2, PALB2, NBN and RECQL

Assay ID	HGVS	dbSNP
Pre-designed assays
C_34306823_20	NM_007194.4 (CHEK2): c.470T > C	rs17879961
C_336684713_10	NM_024675.4 (PALB2): c.509_510delGA	rs515726123
C_190888305_10	NM_024675.4 (PALB2): c.172_175delTTGT	rs180177173
C_203097466_10	NM_002485.5 (NBN): c.657_661delACAAA	rs587776650
C_324203460_10	NM_002907.4 (RECQL): c.1667_1667 + 3delAGTA	rs564485792
Custom-design assays
Not available	NM_007194.4 (CHEK2): c.1100delC	rs555607708
Not available	NM_007194.4 (CHEK2): c.444 + 1G > A	rs121908698

Statistical analysis

The prevalence of the 8 recurrent mutations was estimated in 564 DCIS patients. The odds ratios for DCIS given CHEK2, PALB2, NBN and RECQL status (positive versus negative) were calculated by two-by-two tables, using as a reference mutation frequencies in Polish cancer-free controls from our four previous studies [10, 14, 16, 38]. Additionally, we calculated odds ratios separately for each variant. Statistical significance (p-value) was assessed using Fisher exact test. All cases and controls were ethnic Poles.

To estimate all cause survival of women with DCIS, with and without a CHEK2 mutation (other mutations were not considered because were rare), we followed DCIS patients from the date of diagnosis until the date of death or February 2020, if alive. The actuarial survival between CHEK2 mutation carriers with DCIS and non-carriers with DCIS was compared, and p-value for difference was calculated using Cox regression analysis. Medians were compared using the Mann-Whitney test. Analyses were performed by MedCalc Version 22.014.

Results

In this study of 564 women with DCIS, 57 (10.1%) carried a founder non-BRCA mutation in one of the three genes including variants in CHEK2 (c.1100delC, c.444 + 1G > A, del5395 and c.470T > C), PALB2 (c.509_510delGA) and NBN (c.657_661delACAAA). CHEK2 mutations were the most common, detected in 52 (9.2%) women with DCIS. A CHEK2 truncating mutation was detected in 37 (2.5%) patients and CHEK2 c.470T > C missense variant was identified in 40 patients (7.1%). Two women carried two different CHEK2 mutations (one carried c.1100delC and c.470T > C and the second had del5395 and c.470T > C). Mutations in PALB2 and NBN were seen in 1 (0.2%) and 4 (0.7%) patients, respectively. RECQL c.1667_1667 + 3delAGTA allele was not observed among cases.

We investigated the association of the 8 founder mutations with DCIS, using as a reference allele frequencies in cancer-free controls from our previous studies [10, 14, 16, 38]. We saw a strong association of CHEK2 mutations (all variants combined) with DCIS risk (OR = 1.7, 95%CI 1.2–2.3, p = 0.003). The odds ratio was higher (OR = 3.0, 95%CI 1.5–5.7, p = 0.001) for CHEK2 truncating mutations (c444 + 1G > A, c.1100delC or del5395) than for a missense variant c.470T > C (OR = 1.5, 95%CI 1.05–2.11, p = 0.04). PALB2, NBN and RECQL mutations were not associated with the risk of DCIS (p–values between 0.6 and 1.0), and were not further considered (Table 2).

Table 2.

Odds ratios for DCIS for Polish founder variants in CHEK2, PALB2, NBN and RECQL with corresponding p-values

Gene	Mutation	Cases N = 564		Controlsa			OR	95%CI	P-value
		Positive	%	Positive	Total	%
CHEK2	c444 + 1G > A	7	1.2	14	4346	0.3	3.9	1.3–10.3	0.007
	c.1100delC	1	0.2	7	4346	0.2	1.1	0.02–8.6	1.0
	del5395	6	1.1	16	4346	0.4	2.9	0.9–7.9	0.03
	Any truncating mutationb	14	2.5	37	4346	0.9	3.0	1.5–5.7	0.001
	c.470T > C missense variant	40	7.1	215	4346	4.9	1.5	1.05–2.1	0.04
	Any CHEK2 mutationc	52	9.2	252	4346	5.8	1.7	1.2–2.3	0.003
PALB2	c.509_510delGA	1	0.2	7	4702	0.1	1.2	0.03–9.3	0.6
	c.172_175delTTGT	0	0.0	3	4702	0.1	-	-	1.0
	Any PALB2 mutationd	1	0.2	10	4702	0.2	0.8	0.2–5.9	1.0
NBN	c.657_661delACAAA	4	0.7	22	4000	0.6	1.3	0.3–3.8	0.6
RECQL	c.1667_1667 + 3del AGTA	0	0.0	2	4702	0.04	-	-	1.0

^aMutation frequencies in 4000 to 4702 Polish cancer-free controls (for different variants) were derived from our previous studies [10, 14, 16, 38]

^bc444+1G>A or c.1100delC or del5395

^cc444+1G>A or c.1100delC or del5395 or c.470T>C

^dc.509_510delGA or c.172_175delTTGT

We also investigated the influence of CHEK2 mutations on the risk of early onset DCIS (age 50 years or below) and late onset disease (above age 50 years). The odds ratio for early onset DCIS given a CHEK2 mutation (all variants combined) was 2.4 (95%CI 1.5–3.8) and was very significant (p = 0.0004). In contrast, the odds ratio for late onset disease was 1.4 (95%CI 0.9–2.1) and it was not statistically significant (p = 0.1) (Table 3). We also checked the effect of a positive family history of breast cancer on DCIS risk among CHEK2 mutation carriers. The odds ratio for a CHEK2 mutation was higher for cases with a positive family history (OR = 2.3, 95%CI 1.3–3.9, p = 0.003) than those with no family history of breast cancer (OR = 1.4, 95%CI 0.96–2.1, p = 0.06) (Table 3). The odds ratio was the highest for women with a CHEK2 truncating mutation and family history of breast cancer (OR = 4.2, 95%CI 1.3–10.8, p = 0.01).

Table 3.

Odds ratios for DCIS for CHEK2 mutations by age of diagnosis and by family history of breast cancer

Group	Total No.	CHEK2 c444 + 1G > A		CHEK2 1100delC		CHEK2 del5395		Any CHEK2 truncating mutationb		CHEK2 c.470T > C missense variant		Any CHEK2 mutationc		OR	95%CI	P-value
		No.	%	No.	%	No.	%	No.	%	No.	%	No.	%
All DCIS patients	564	7	1.2	1	0.2	6	1.1	14	2.5	40	7.1	52	9.2	1.6	1.2–2.3	0.003
Age ≤ 50 years	186	3	1.6	1	0.5	0	0.0	4	2.2	20	10.8	23	12.4	2.3	1.4–3.6	0.0008
Age > 50 years	378	4	1.1	0	0.0	6	1.6	10	2.6	20	5.3	29	7.7	1.3	0.9-2.0	0.14
Cases with FHa	145	3	2.0	1	0.7	1	0.7	5	3.4	13	9.0	18	12.4	2.3	1.3–3.9	0.003
Cases without FHa	419	4	1.0	0	0.0	5	1.2	9	2.1	27	6.4	34	8.1	1.4	0.96–2.1	0.06
Cancer-free controls	4346							37	0.9	215	4.9	252	5.8	ref	ref	ref

^aFH - positive family history of breast cancer in first- or second-degree relative

^bc444 + 1G > A or c.1100delC or del5395

^cc444 + 1G > A or c.1100delC or del5395 or c.470T > C

Finally, we explored the impact of CHEK2 mutations on survival of DCIS patients. The median follow-up among 564 women with DCIS was 156 months (range 1–270, SD 62.7), including 166 months (range 18–258, SD 54.6) for CHEK2 mutation carriers versus 156 (range 1–270, SD 63.4) for non-carriers (p = 0.3, Mann-Whitney test). There were no deaths in 52 mutation carriers (0%) versus 18 deaths in 512 non-carriers (3.5%) during the follow up time (p = 0.4). The 10-year survival was 100% for carriers compared to 97% for non-carriers. The age-adjusted p-value for mortality associated with a CHEK2 mutation was 0.9 (Cox regression analysis). There were no deaths among mutation carriers, so the hazard ratio (carriers vs. non-carriers) was not estimated.

Discussion

In Poland, we have initiated a program to detect DNA mutations which predispose to DCIS. We focused on Polish founder mutations in BRCA1, BRCA2, PALB2, CHEK2, NBN and RECQL (18 alleles), which account for one-half of Polish breast cancer families [34]. Previously, we reported that Polish founder mutations of BRCA1/2 confer susceptibility to DCIS, and are responsible for 3% of DCIS cases in Poland [35]. In the current study, we focused on 8 recurrent mutations in four non-BRCA1/2 genes (from the panel of 20 Polish alleles) including CHEK2 (c.1100delC, c.444 + 1G > A, del5395 and c.470T > C), PALB2 (c.509_510delGA, c.172_175delTTGT), NBN (c.657_661delACAAA) and RECQL (c.1667_1667 + 3delAGTA).

We found a strong association of CHEK2 mutations with an increased risk of DCIS. CHEK2 founder mutations are present with a frequency of 6% in a genetically homogenous population of Poland [34]. According to our current study, CHEK2 mutations (all alleles together) confer a 1.7-fold increased risk of DCIS. Based on these figures, we estimate that CHEK2 founder mutations account for 4% of DCIS cases in Poland (population attributable risk estimate). We consider that CHEK2 mutations may be important cause of DCIS in other homogenous populations, in which founder alleles of this gene are present, i.e. the Netherlands, Finland and Eastern European countries [39, 40]. Our results suggest that CHEK2 truncating mutations confer a 3-fold (p = 0.001) increased risk of DCIS and a missense c.470T > C variant confers lower risk (OR = 1.5, p = 0.04).

In the current study, we observed strong association of CHEK2 mutations (all variants combined) with cases diagnosed at age 50 or below (OR = 2.4, p = 0.0004), but not with cases diagnosed above age 50 (OR = 1.4, p = 0.1). Therefore, CHEK2 mutations may predispose to early onset DCIS. However, the difference in the two odds ratios was not statistically significant, and other studies did not report association between CHEK2 and early onset DCIS, so further studies are needed in this regard.

Our study suggests that the risk of DCIS in a woman with a CHEK2 mutation is not determined solely by the presence of the mutation; penetrance is also dependent on her family history of cancer. We observed that the DCIS risk for a woman with a CHEK2 mutation and a family history of breast cancer was greater (OR = 2.3, p = 0.003) than that of a CHEK2 mutation carrier who has no family history of breast cancer (OR = 1.4, p = 0.06). Of note, the risk was highest for patients with a CHEK2 truncating mutation and a positive family history of breast cancer (OR = 4.2, p = 0.01). These data support the model that both family history and a CHEK2 mutation are risk factors for breast cancer that operate in concert, and that the CHEK2 mutation modifies the risk of cancer imparted by other cancer genes [41–45].

We also analysed all cause survival of DCIS patients with a CHEK2 mutation. Our data suggest the survival among CHEK2 carriers with DCIS is good, and it is similar to that of non-carriers with DCIS. The 10-year mortality was 0% for carriers compared to 3% for non-carriers. Of note, Elshof et al. reported that 10-year breast cancer mortality among DCIS patients is 2.3% for women at age < 50 years, and 1.4% for those > 50 years at diagnosis [46]. According to Narod et al. at 20 years, the mortality among DCIS patients is 3.3% [47].

Other studies support association between CHEK2 and DCIS [48–52]. The Breast Cancer Association Consortium reported the risk of 2.2 for DCIS in association with CHEK2 truncating variants, 0.8% of 4187 DCIS patients carried a CHEK2 truncating mutation [48]. Evans analyzed 311 women with DCIS, 4 (1.29%) had CHEK2 c.1100delC mutation [49]. Petridis et al., studied DCIS patients diagnosed before age 50 and detected CHEK2 pathogenic mutations in 2.4% of 655 DCIS cases, OR for a CHEK2 mutation was 8.0 (95%CI 2.9–22.1) [50]. Mundt et al. found that CHEK2 truncating variants confer 1.8-fold increased risk of DCIS, and CHEK2 c.470T > C missense variant increases the risk by 1.3-fold (p = 0.02) [51]. Huang et al. showed that CHEK2 truncating variants are associated with a moderate risk of DCIS (OR = 3.3, 95%CI 2.31–4.52) [52].

The aggregate evidence suggest that PALB2 mutations may also predispose to DCIS, but this evidence is based on a few studies. Petridis et al. reported OR of 14.9 (95%CI 1.8–123.9) for DCIS given a PALB2 mutation [50]. Huang et al. found that PALB2 mutations confer 2.6-fold increased risk of DCIS (95%CI 1.26–5.12) [52]. The Breast Cancer Association Consortium also reported a moderate risk of 2.5 for DCIS in association with PALB2 mutations [48]. We saw no association of PALB2, NBN and RECQL mutations with DCIS, but these mutations were rare in Polish women with DCIS, and our study is inconclusive in this regard. Other studies (than our study) have not investigated the role of NBN and RECQL in the etiology of DCIS.

In the current and our previous study [35] we described mutation frequencies in BRCA1/2, PALB2, CHEK2, NBN and RECQL in Polish women with DCIS. It is interesting to compare whether the mutation frequencies observed in our DCIS cohort differ from those in the general breast cancer population in Poland. For BRCA1, recurrent mutations were detected in 1.2% of DCIS cases compared to 3.6% in invasive breast cancer cases. This difference may be explained by the fact that BRCA1 mutations predispose much strongly to invasive breast cancers than to DCIS [53]. They predispose in particular to medullar and ductal G3 tumors (usually without DCIS component), and BRCA1-associated breast cancers are commonly triple-negative, that is they have no expression of estrogen receptor (ER), progesterone receptor (PR) and human epidermal growth factor receptor 2 (HER2). Regarding CHEK2, truncating mutations were identified with similar frequency in DCIS cases (2.5%) and invasive tumors (2.9%). Also, CHEK2 c.470T > C missense variant had similar prevalence (7.1%) in DCIS cases and invasive breast cancer cases (7.7%) [53]. These data suggest that CHEK2 mutations confer similar risk of DCIS and invasive breast cancer. This is consistent with the evidence that breast tumors which arise in carriers of CHEK2 mutations seem to be similar to those of breast cancers in the population at large. PALB2 mutations were detected in 0.2% of DCIS cases compared to 0.9% of invasive breast cancer cases [14]. This suggests that PALB2-mutation carriers are more likely to be diagnosed with invasive breast cancer than with DCIS. For NBN and RECQL, the mutation frequencies were similar in DCIS cases and women with invasive breast cancer (0.7% vs. 1.0%, and 0.0% vs. 0.2%, respectively) [16, 54], and the number of mutation carriers was low (in DCIS), so no clear conclusions can be drawn. For BRCA2, we were not able to perform the comparison because BRCA2 mutation frequency is not established in Polish women with unselected breast cancer. Of note, the above data are based on relatively low number of carriers among Polish DCIS cases in individual genes and further studies are needed.

The major reason why we focused our study on inherited predisposition to ductal carcinoma in situ is that women in Poland receive this specific diagnosis, and therefore we need to know whether or not DCIS may be caused by specific mutations in breast cancer susceptibility genes (and to what extent). This knowledge is important in clinical practice for several reasons. It provides useful information what mutations predispose to DCIS, and thus what genes should be screened in DCIS patients. Furthermore, if a mutation is found (i.e. in BRCA1/2 or CHEK2) then specific surveillance protocols may be recommended for the mutation carriers (usually for breast and other cancers, because most breast cancer susceptibility genes are associated with a multi-site cancer predisposition). Finally, women with non-invasive breast tumors (in contrast to those with invasive cancers) are not commonly referred to outpatient genetic clinics in Poland [55]. Our evidence that mutations in BRCA1/2 and CHEK2 predispose Polish women to DCIS, should encourage clinicians to refer patients with DCIS for genetic counselling and testing more commonly.

Conclusions

Based on the current study, women with a CHEK2 mutation face an increased risk of DCIS, in particular in the presence of a positive family history of breast cancer. The presence of DCIS should be considered during surveillance of CHEK2 mutation carriers, in particular when calcifications on mammography, a hypoechoic area on ultrasound or non-mass enhancement on breast MRI are seen. On the other hand, DCIS patients should receive genetic counseling and testing for CHEK2 mutations.

Abbreviations

DCIS

Ductal carcinoma in situ

BC

Breast cancer

IBC

Invasive breast cancer

HBC

Hereditary breast cancer

OR

Odds ratio

P

Probability value

Authors’ contributions

Author Contributions: Conceptualization, S.F. and C.C.; methodology, S.F., W.K. and C.C.; software, S.F., D.W., H.R.; validation, S.F., J.L. and C.C.; formal analysis S.F., I.F., C.C.; investigation, S.F., I.F., M.C.,; resources, A.J., M.L., K.S., T.H., T.D., J.G., C.C.; data curation, S.F., M.Cz., K.G.; writing—original draft preparation, S.F.; writing—review and editing, C.C. and S.F.; visualization, S.F., I.F., M.C.; supervision, C.C.

Funding

This research received no external funding.

Data availability

The datasets generated and/or analysed during the current study are not publicly available due the fact that the main research data supporting the results of this study are included in Tables 2 and 3 but are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.