• hereditary ovarian carcinoma;
  • BRCA1;
  • unclassified variants;
  • death;
  • recurrence


  1. Top of page
  2. Abstract
  6. Acknowledgements


The clinical relevance of BRCA1/2 alterations in ovarian carcinoma patients is debatable. Our aim was to determine factors influencing the risk of recurrence and death in ovarian carcinoma patients with BRCA pathogenic and unclassified variant mutations.


A consecutive series of 205 women with primary ovarian carcinoma were screened for mutations in BRCA1 and BRCA2 genes using a conformational sensitive gel electrophoresis and direct sequencing. Data regarding medical and familial history were collected using questionnaires. Clinical and pathologic data were extracted from medical records.


Unclassified variant mutations in BRCA1 or BRCA2 genes were found in 16 (8%) patients, and BRCA1 pathogenic mutations were found in 18 (9%) patients. No pathogenic mutation was found in BRCA2 gene. Multivariate analysis showed that BRCA1 pathogenic mutation was an independent predictor of reduced risk of relapse and death (Hazard ratios [HR] 0.52 [confidence interval {CI} 0.28–0.98] and 0.38 [CI 0.10–0.96], respectively). Unclassified variant mutation did not affect recurrence and survival (HR 0.84 [CI 0.43–1.66] and 0.94 [CI 0.48–1.82], respectively). Other factors associated with reduced risk of relapse and death were complete pathologic remission at second-look laparotomy and family history of breast and ovarian carcinoma, respectively. Recurrence and death outcomes among unclassified variant mutation carriers did not differ significantly from those in sporadic cases.


Patients with BRCA1 pathogenic mutation seem to have reduced risk of recurrence and death. These results should be interpreted with caution as they may be influenced by more intensive treatment, better response to cisplatin, and younger age of mutation carriers. Clinical relevance of BRCA1/2 unclassified variant mutations warrants further studies. Cancer 2005. © 2005 American Cancer Society.

Ovarian carcinoma is the leading cause of cancer mortality among gynecologic malignancies and the most common form of hereditary gynecologic tumor. Two recently sequenced genes, BRCA1 (1994) and BRCA2 (1995), are responsible for 80% of hereditary ovarian carcinoma cases.1, 2

Ovarian carcinoma associated with BRCA1 pathogenic mutation has a distinct clinical behavior.3–6 Namely, it is typically an advanced stage, papillary, serous, poorly differentiated adenocarcinoma, which affects younger women with family history of breast and/or ovarian carcinoma.3–7 Fifty percent of mutations occurring in BRCA1 gene account for missense alterations and intronic variations of unknown clinical relevance (see URL: It has been postulated that some unclassified variant mutations may play a role in cancer pathogenesis.8, 9 Preliminary reports show that the natural history of unclassified variant-linked ovarian tumors share features of both BRCA1-associated and sporadic ovarian carcinoma cases.7 There are no data on prognostic impact of unclassified variant mutations in ovarian carcinoma.

Prognostic impact of BRCA1 pathogenic mutation in ovarian carcinomas has been a subject of several studies over the past few years.10–17 Some studies showed significantly prolonged survival in ovarian carcinoma patients carrying BRCA1 mutation compared to sporadic carcinoma cases, whereas others did not confirm these findings in a multivariate analysis model. The aim of the current study was to add to this knowledge by determining the impact of BRCA1 pathogenic mutation on recurrence and survival in a large, consecutive series of ovarian carcinoma patients treated in a single institution. Additionally, we analyzed the prognostic relevance of BRCA1/2 unclassified variant mutations.


  1. Top of page
  2. Abstract
  6. Acknowledgements

The study group included 205 consecutive patients with primary epithelial ovarian carcinoma treated between 1997 and 2002 at the Department of Gynecology, Medical University of Gdansk, Poland. All patients agreed to undergo genetic testing and to participate in the study (n = 205). The study group was followed until the end of 2003. Disease status could be determined in 202 patients, and survival status in 186.

In this series, there were no patients with peritoneal carcinoma. All patients had FIGO Stage IC–IV ovarian carcinoma. A total of 203 patients underwent a thorough staging laparotomy and primary cytoreduction. The goal of primary cytoreductive surgery was to dissect the maximal amount of tumor. In all patients, primary surgical excision included total hysterectomy, bilateral salpingo-oophorectomy, resection of tumor mass elsewhere in the pelvis or abdomen. and omentectomy. Two patients with FIGO Stage IC ovarian carcinoma who wished to retain fertility underwent unilateral oophorectomy, omentectomy, and staging. Optimal cytoreduction (remaining tumor < 2 cm) was achieved in 71 patients. All patients underwent adjuvant chemotherapy. Completeness of resection or cytoreduction was assessed with staging performed according to International Gynecological Cancer Society (IGSC) guidelines.16 Primary tumor sites (ovary or peritoneal) were confirmed by review of pathology reports. Surgical stage was classified according to FIGO criteria and was determined through detailed analysis of patient medical history records, surgery reports, and pathology reports.

After initial primary cytoreductive surgery, 138 patients received first-line chemotherapy of paclitaxel (175 mg/m2 in 3-hour intravenous infusion) and cisplatin (75 mg/m2), and 63 patients treated before the year 2000 received cisplatin (75 mg/m2) with cyclophosphamide (750 mg/m2).

One hundred sixty-nine patients with residual tumor following primary cytoreductive surgery, who completed 6 cycles of adjuvant chemotherapy, achieved complete clinical remission, signed informed consent, and underwent second-look laparotomy (SLL). The aim of SLL was to determine whether pathologic complete remission (pCR) was achieved. Biopsies of all suspicious lesions and staging were performed according to IGSC guidelines.16 Samples from SLL were evaluated by a pathologist and classified as negative (grossly and pathologically negative), microscopically positive (grossly negative, pathologically positive), and macroscopically positive (grossly and pathologically positive). A pCR was defined on the basis of negative histologic findings after primary cytoreductive surgery or SLL. The interval between primary surgery and SLL varied from 7 to 9 months depending on postoperative recovery and duration of first-line chemotherapy. Patients who achieved pCR were followed in outpatient clinic at regular intervals. During the first year following treatment, patients were seen every 3 months, then every 6 months, and after 5 years, patients were seen annually. Each assessment comprised complete physical examination (including breast, pelvic, and rectal examinations). CA125 was checked every 6 months. Computed tomography (CT), magnetic resonance imaging (MRI), or ultrasound (US) scans were performed only if recurrence was suspected (lesion on clinical examination, increasing level of tumor markers, or clinical symptoms).

Patients who had residual disease on SLL underwent second-line or salvage chemotherapy. If the diameter of residual changes was < 1 cm, patients were offered intraperitoneal chemotherapy at a dose of 90 mg/m2. If residual changes were > 1 cm, patients received intravenous chemotherapy including cisplatin and cyclophosphamide. Tumor recurrence was defined as elevation of CA125 level (> 35 U/mL) on two consecutive tests, the appearance of measurable lesion on clinical examination or radiologic imaging, or pathologic evidence of disease.

Clinical and pathologic data were extracted from medical records. Ovarian carcinomas in patients with no family history of cancer and no BRCA mutation or with nonpathogenic polymorphisms were defined as sporadic. Data regarding patients' medical and family history of cancer were assessed using a questionnaire. Patients with unknown data on family history of cancer due to loss of contact with their families or adoption were defined as patients with unknown family history of cancer. A family history of cancer, including breast and ovarian carcinoma, was considered to be positive if there was any first-degree relative with cancer. Information on patients' survival was extracted from medical history records and the Central Registry at Regional Oncology Center of Pomerania.

DNA Extraction and Mutation Analysis

All patients were eligible for genetic testing for BRCA1 and BRCA2 mutations. High molecular-weight DNA was extracted from the whole blood of 137 patients with a Promega Wizard Kit (Promega, Madison, WI). In 68 patients, adequate blood sample could not be obtained because of leukopenia or difficult venous access caused by damage following longstanding chemotherapy. In these patients, DNA was isolated from 50 μm sections of frozen ovarian tumor tissue using the Promega Wizard Kit. One pathogenic and two unclassified variant mutations were identified in this group, and results were confirmed using DNA isolated from paraffin-embedded healthy tissue blocks. Thus, these mutations should be considered germline.

Mutation scanning methods included fluorescent conformation-sensitive gel electrophoresis (F-CSGE).17, 18 The entire coding region of BRCA1, BRCA2, and 15–50 base pairs (bp) of each flanking intron were subdivided into 83 segments—33 for BRCA1 and 50 for BRCA2. All primer sequences were selected outside the Alu-repeat sequences, which are particularly abundant in the introns of BRCA1. The forward primers were fluorescently labeled with FAM, HEX, or TET. The 83 fragments were amplified in 48 monoplex or duplex PCRs (1–2 amplifiers each). The samples were subjected to electophoresis through these gels for 4.5 hours at 1680 V at 30 °C. Gels were analysed with GeneScan and Genotyper software (Applied Biosystems, Foster City, CA). Upon detection of a CSGE-variant, the segment of the gene was reamplified from the DNA sample with M 13 tail on both the reverse and forward ends (FWD: 5′-TTG TAA AAC GAC GGC CAG T; REV: 5′-GGA AAC AGC TAT GAC CAT G). Sequencing was performed with the universal primers using the ABI PRISM Big Dye Terminator Cycle (Applied Biosystems, Foster City, CA). Estimated sensitivity was 75–85%.

Statistical Analysis

Before starting this analysis, power analysis was applied to estimate the minimum sample size required for performing a survival analysis related to recurrence and death risk based on data from already published material.10 In that series, Ashkenazi women with BRCA-associated advanced ovarian carcinoma had improved disease-free interval when compared with patients with sporadic cancer (median, 49 mos and 19 mos, respectively) and had improved overall survival (median, 91 mos and 54 mos, respectively). If these data were applicable to our population, the study10 had a power of 96% to detect differences in overall survival between the women with a BRCA-associated versus sporadic tumors. We applied an alpha value of 0.05, and a two-tailed test.

All prognostic factors included in this study (for an overview, see Table 1) were related to recurrence and death risk. For both outcomes, survival curves were constructed, and the median follow-up time estimated. Data were censored at 50 months of follow-up. Univariate Cox regression analyses were performed for each prognostic factor separately, considering recurrence and death. Hazard ratios (HR) and 95% confidence intervals (CI) were estimated. To test the assumption of proportional hazards, an interaction term of a prognostic variable and a time-dependent covariate were added.19 A significant effect of that interaction term denotes the presence of a time-dependent effect and thus a violation of the proportional hazards assumption. Then, multivariate Cox regression analyses were performed including binary coding of all factors with a stepwise procedure. P values < 0.05 were considered statistically significant. All analyses were performed with program SPSS version 11.01 (SPSS, Chicago, IL).

Table 1. Clinical and Pathologic Characteristics of BRCA1 Pathogenic Mutation, BRCA1/2 Unclassified Variants Carriers and Sporadic Ovarian Cancer Patients
VariableBRCA1 pathogenic mutation carriers (n = 18) No. (%)Unclassified variants (n = 16) No. (%)Polymorphisms; sporadic cases (n = 171) No. (%)
Age of onset 
 <50 yrs15 (83)7 (44)58 (34)
 >50 yrs3 (17)9 (56)113 (66)
 I and II2 (11)3 (19)21 (12)
 III and IV16 (89)13 (81)150 (88)
 G19 (50)13 (81)157 (92)
 G39 (50)3 (19)14 (8)
 Papillary serous15 (83)11 (69)107 (63)
 Mucinous3 (17)5 (31)64 (37)
Involvement of ovaries 
 Unilateral2 (11)1 (6)46 (27)
 Bilateral16 (89)15 (94)125 (73)
SLO findings 
 No residual disease3 (17)8 (53)64 (37)
 Residual disease <1cm11 (61)4 (20)90 (53)
 Residual disease >1cm4 (22)4 (27)17 (10)
Complete pathologic remission (pCP) 
 Persistent (PR)3 (17)9 (53)60 (35)
15 (83)7 (47)111 (65)
Family history of cancer 
 Yes10 (56)5 (31)33 (19)
 No or unknown8 (44)11 (69)138 (81)
Family history of breast carcinoma 
 Yes8 (44)1 (6)12 (7)
 No or unknown10 (56)15 (94)159 (93)
Family history of ovarian carcinoma 
 Yes5 (28)2 (12)13 (8)
 No or unknown13 (72)14 (87)158 (92)


  1. Top of page
  2. Abstract
  6. Acknowledgements

Detailed characteristics of BRCA1 pathogenic, BRCA1/2 unclassified variant carriers, and sporadic cases are illustrated in Tables 1 and 2. Pathogenic mutations in the BRCA1 gene were found in 18 (9%) patients, and unclassified variant mutations in BRCA1/BRCA2 genes were found in 13 (8%) patients. In the remaining 154 patients, no BRCA mutation was identified. Recurrence was established in 93 (46%) of 202 assessable patients, and 97 (52%) of 186 patients died (Table 3). Of those deceased, 93 deaths were cancer-related deaths, and 4 patients died without apparent recurrence.

Table 2. Type of Location of BRCA Mutation
MutationBRCA geneExonNo. of casesPercentage frequency n = 34Type of alterationPredicted defect
  1. F: frameshift; M: missense; UV: unclassified variant.

4956 A>G1ex16B716UVSer>Gly
4486 G>T2ex11J715UVAsp>Tyr
185insC1ex224FStop 39
300 T>G1ex512MCys61Gly
3405 C>T1ex1112StopGly>stop
IVS 22+51ex2212Splice–siteStop 1803
3111 G>A2ex11D12UVGln>Gln
5075 A>G1ex16B12UVMet>Ile
Table 3. Patient Characteristics Related to Recurrence and Death Rates (Frequencies); Univariate Cox Regression Analysis (Hazard Ratios and 95% Confidence Intervals)
VariableHR (% frequency)RecurrenceDeath
93/202 46%97/186 52%
CI (range)HR (% frequency)CI (range)
Mutation status 
 Sporadic59/154 (38)175/139 (54)1
 Unclassified variants10/16 (62)0.84 (0.43–1.66)10/16 (63)0.94 (0.48–1.82)
 BRCA1/ mutation16/18 (89)0.58 (0.31–1.10)7/18 (39)0.39 (0.18–0.84)
Age of onset 
 <5043/77 (56)130/74 (41)1
 ≥5050/121 (41)1.03 (0.67–1.56)67/112 (60)0.63 (0.41–0.97)
 I and II11/26 (42)18/25 (32)1
 III72/158 (46)1.18 (0.63–2.24)79/148 (53)1.85 (0.89–3.82)
 IV10/14 (71)1.65 (0.69–3.94)10/13 (77)2.67 (1.05–6.79)
 128/66 (42)132/59 (541
 248/107 (45)1.06 (0.66–1.69)47/103 (46)0.90 (0.57–1.41)
 317/25 (68)1.08 (0.58–1.99)18/24 (75)1.24 (0.69–2.21)
 Papillary serous60/128 (47)165/122 (53)1
 Nonpapillary serous30/70 (47)0.88 (0.57–1.35)32/64 (50)1.04 (0.68–1.59)
Ovarian involvement 
 Unilateral17/48 (35)122/46 (48)1
 Bilateral76/150 (51)1.10 (0.64–1.87)75/140 (54)1.01 (0.63–1.62)
First line chemotherapy regimen 
 Paclitaxel + cisplatin65/135 (48)160/124 (48)1
 Cisplatin + cyclophosphamide28/63 (44)1.01 (0.64–1.57)36/61 (59)1.14 (0.75–1.74)
Residual changes on second look laparotomy 
 Absent22/74 (30)133/68 (49)1
 <1cm56/99 (56)1.67 (1.02–2.75)45/94 (48)1.09 (0.69–1.71)
 >1cm15/25 (60)1.59 (0.82–3.07)18/23 (78)1.95 (1.10–3.48)
Complete pathologic remission 
 Yes20/71 (28)132/64 (50)1
 No73/127 (58)1.78 (1.08–2.94)64/121 (53)1.18 (0.77–1.81)
Route of second line chemotherapy 
 Intraperitoneal34/64 (53)127/60 (45)1
 Intravenous40/65 (62)1.04 (0.66–1.63)37/62 (60)0.48 (0.29–0.80)
Complete clinical remission after second line chemotherapy 
 Yes61/98 (62)137/87 (43)1
 No13/31 (42)0.65 (0.43–1.26)26/34 (77)0.46 (0.22–0.95)
Family history of cancer 
 No63/153 (41)179/139 (57)1
 Yes30/45 (67)0.78 (0.43–1.42)18/47 (38)0.40 (0.18–0.86)
Family history of breast cancer 
 No81/178 (46)189/165 (54)1
 Yes12/20 (60)0.65 (0.43–1.26)8/21 (38)0.46 (0.22–0.95)
Family history of ovarian cancer 1 1
 No80/179 (45)0.84 (0.43–1.66)90/166 (54)0.94 (0.48–1.82)
 Yes13/19 (68)0.58 (0.31–1.10)7/20 (35)0.39 (0.18–0.84)

Probability of 3-year survival in the entire group was 40% (95% CI; range, 31–49%; Fig. 1A). The probability of 3-year survival in BRCA1 pathogenic mutation group was 77% (95% CI; range, 54–100%), and the median survival was 45 months (95% CI; range, 38–52 mos). Among unclassified variant mutation carriers, probability of 3-year survival was 30% (95% CI; range, 22–44%), and the median survival was 30 months (95% CI; range,26–34 mos). In sporadic cancer group, probability of 3-year survival was 31% (95% CI; range, 20–42%), and the median was 28 months (95% CI; range, 25–31 mos; Fig. 1C). Figure 1D shows overall survival in months for patients with a BRCA1 mutation, for patients with an unclassified variant mutation, and for sporadic cases.

thumbnail image

Figure 1. (A) Overall survival in the entire group. (B) Disease-free survival according to BRCA1 mutation (follow-up time in mos). log rank: 2.84; df = 1; P = 0.092. Non-BRCA1-group is unbroken line; BRCA group is broken line. (C) Overall survival according to BRCA1 mutation (follow-up time in mos). log rank: 5.52; df = 1; P = 0.019. Non-BRCA1-group is unbroken line; BRCA group is broken line. (D) Overall survival in months for patients with a BRCA1 mutation, for patients with an unclassified variant mutation, and for sporadic cases.

Download figure to PowerPoint

In univariate analysis, the presence of residual changes on SLL and the absence of pCR were related to an increased risk of recurrence (respectively, HR 1.67 [1.02–2.75] and 1.78 [1.08–2.94; Table 3]). A BRCA1 pathogenic mutation, age of onset < 50 years, and a family history of breast and/or ovarian carcinoma were related to a decreased risk of death, whereas high FIGO grade and the presence of residual changes on SLL were related to an increased risk of death (Table 3).

In the Cox regression model, there was no violation of the assumption of proportional hazards relating to the two outcomes and any prognostic factor (results not presented). pCR on SLL was an independent favorable factor influencing the risk of recurrence, whereas BRCA1 pathogenic mutation status, family history of breast and ovarian carcinoma, and low or intermediate tumor grade were independent predictors of reduced death risk (Table 4). Unclassified variant mutation status did not influence recurrence or death rates in ovarian carcinoma patients.

Table 4. Relative Risk of Recurrence and Death: Multivariate Modal (Hazard Ratios and 95% CI)
VariableRecurrence (CI range)Death (CI range)
BRCA1 mutation status 
 Yes0.52 (0.28–0.98)0.38 (0.15–0.96)
Grading 3 
 Yes 2.47 (1.38–4.44)
Complete pathologic remission 
 No2.05 (1.23–3.44) 
Family history of cancer 
 Yes 0.43 (0.24–0.76)

Family history of breast and ovarian carcinoma in first-degree relatives was the strongest predictor of BRCA1 mutation (P = 0.000; P = 0.002, respectively). Tumors of the colon, prostate, stomach, esophagus, lung, and pancreas, as well as malignant melanoma were not associated with BRCA1 pathogenic mutation or unclassified variant (n = 16). Five ovarian carcinoma patients had family history of pancreatic carcinoma and malignant melanoma of whom none harbored BRCA1/2 mutation or unclassified variant. This group will be screened for p16 (Leiden factor) mutation.21


  1. Top of page
  2. Abstract
  6. Acknowledgements

The current study showed that BRCA1 pathogenic mutation, pCR or intermediate tumor, grade, and family history of breast and ovarian carcinoma were independent favorable predictors of recurrence and/or death among ovarian carcinoma patients. The unclassified variant mutation did not affect recurrence or survival.

The current study confirms previous data on improved outcome in ovarian carcinoma patients carrying BRCA mutations.4, 6, 10, 15 Additionally, we investigated the prognostic impact of unclassified BRCA variants. To our knowledge, this is the first published study addressing this issue. The lack of clinical significance of BRCA unclassified variants is of substantial importance, as this type of alteration could be identified in up to 15% of ovarian carcinoma patients who undergo full sequence evaluation of the BRCA1 and BRCA2 genes.20

The role of BRCA unclassified variant mutation in ovarian carcinoma pathogenesis has not yet been determined. Unclassified variant mutations in BRCA1 or BRCA2 genes may be critically involved in BRCA function and consequently associated with a high penetrance and predisposition to cancer. This variant may also be a modifier of BRCA1, BRCA2, or a related gene function with similar impact on cancer occurrence as protein-truncating mutation. However, unclassified variants may merely represent a polymorphic form of the gene, with no impact on cancer risk or prognosis. In our study, 60% of BRCA1/2 unclassified variant carriers had family history of breast and ovarian carcinoma. Therefore, we strongly suspect that unclassified variant mutations might have predisposed this group to cancer occurrence. Further, our previous preliminary report showed that the natural history of ovarian carcinoma related to unclassified variant mutations shared features of both BRCA1-associated and sporadic ovarian carcinoma cases.7 Future molecular and linkage studies on BRCA unclassified variant alterations are warranted to establish their role in ovarian carcinoma pathogenesis.

A limitation of this study is the small number of unclassified variant and pathogenic mutation cases (16 and 18, respectively of a total of 202 patients). However, based on the large survival differences between patients with and without BRCA1 mutation,10 we expected to have sufficient statistical power to detect a difference between the two groups. Further, all patients in this series were treated in a single institution with standardized treatment, thus the impact of confounding factors was reduced.

Results of our study should be interpreted cautiously. First, patients with BRCA1 pathogenic mutation more often required second-line chemotherapy because of residual disease on SLL, resulting from higher initial FIGO stage compared to unclassified variant carriers and sporadic cases. Second, BRCA1 pathogenic mutation carriers respond better to cisplatin, which in this study was the main chemotherapy agent (administered intraperitoneally or intravenously). Thus, postponed recurrence and death might have occurred due to more intensive treatment and better response to cisplatin in ovarian carcinoma patients carrying BRCA1 pathogenic mutation. In vitro studies show that BRCA1 defective cells of breast or ovarian tumors demonstrate a specific chemosensitivity profile dependent on BRCA1 expression.22, 23, 24, 25 Conversely, longer survival in advanced stage ovarian carcinoma in BRCA1 mutation carriers may be related to greater penetration in younger women. Age of onset among BRCA1/BRCA2 unclassified variant carriers did not differ from sporadic cases, therefore younger age might have been an underlying survival advantage in the group of patients carrying BRCA1 pathogenic mutation.

The results of studies addressing the impact of BRCA1/BRCA2 pathogenic mutation status on survival in patients with ovarian carcinoma are contradictory (Table 5).4–15 Three studies4, 6, 10 showed longer survival among BRCA1/BRCA2 pathogenic mutation carriers, three studies11, 13, 15 did not show survival differences, and one study14 showed significantly worse survival in patients with BRCA1/BRCA2 pathogenic mutation. Only two studies included disease-free survival in their analyses.6, 10

Table 5. Prognostic Impact of BRCA Alterations in Published Series of Ovarian Carcinoma
PopulationaNo. patients with ovarian carcinomaNo. carriersClinical and pathologic characteristicsAlterations testedDisease-free survival (BRCA carriers vs. sporadic cases)Overall survival (BRCA carriers vs. sporadic cases)
  • a

    See numbered references in Reference section of this article.

Jewish107122 BRCA1 12 BRCA2Age FIGO Histology Grading Surgical cytoreduction Response to chemotherapy SurvivalFounder mutationsBRCA1/2 carriers: longer (median 49 vs. 19 mos)BRCA1/2 carriers: longer median survival (91 vs. 54 mos)
USA1125059 BRCA1 dysfunctionsAge FIGO Histology Grading Tumor site Residual disease Preoperative CA125 SurvivalBRCA1 germline (founder mutations) BRCA1 somatic BRCA1 silent mRNA (promoter hypermethylation)Not analyzedNo difference (median 49 vs. 42 mos)
USA425053 BRCA1Age FIGO Histology Grading Survival22 coding exons and intronic splice donor and splice acceptor BRCA1 regionsNot analyzedBRCA1 carriers: longer median survival (77 vs. 29 mos)
UK1327062 BRCA1/2FIGO Histology Grading SurvivalFounder mutationNot analyzedNo difference (median BRCA1 21 mos, BRCA2 16 mos, sporadic 20 mos)
Jewish693388 BRCA1/2Age FIGO Histology Grading Surgical outcome Response to chemotherapy SurvivalFounder mutationsBRCA1/2 14 mos Sporadic: 7 mosBRCA 1/2 carriers: longer
Norway1424936 BRCA1 8 BRCA2Age FIGO Histology Grading SurvivalFounder mutationsNot analyzedBRCA1/2 carriers: shorter 5-year survival (63% vs. 91%)
Jewish15896171 BRCA1 58 BRCA2Age FIGOFounder mutationsNot analyzedBRCA1/2 carriers: longer median survival (53 vs.
   Histology Grading Survival  38 mos)
Poland (current series)20218 BRCA1 16 unclassified variants BRCA1/2Age FIGO Histology Grading CPR Residual changes on SLL Treatment regimen Recurrence Survival Family history of cancerScreening of the whole length of BRC1/2 and 15–30 flanking introns BRCA1 pathogenic: longer median survival; unclassified variants: no difference. (BRCA1/2 45 mos; unclassified 30 mos; sporadic 29 mos)

In the current study, unclassified variant mutations did not have an impact on recurrence and survival. This may be a result of unclassified variant itself or clinical characteristics of these tumors, which share features of both BRCA1-associated and sporadic ovarian carcinoma cases.7 However, the number of detected cases with unclassified variant might well have been to small to detect minor survival differences. Hence, further studies of the outcome of this subset should be continued on larger cohorts.

A remarkable finding in this series is increased survival in patients with a family history of breast or ovarian carcinoma. This effect is comparable to that in patients with a BRCA1 mutation. An explanation may be that some patients with a family history of cancer carry an undetected BRCA1 mutation. Indeed, the sensitivity of genetic testing for this mutation is in the range of 65–75%.17, 18 Additionally, in almost 33% of patients, DNA was isolated from tumor tissue, a technique considered less sensitive than sequencing using lymphocytes as a source. A second explanation is that this group of patients may carry still unknown, genetic alteration comparable to BRCA1.

In conclusion, in our study BRCA1 pathogenic mutation was associated with prolonged recurrence and survival, whereas unclassified variant mutations had no detectable prognostic relevance. Favorable prognostic impact of BRCA1 pathogenic mutation should be interpreted cautiously though, as this effect may be related to different treatment outcomes rather than differences in natural tumor behavior. Increased survival in familiar ovarian carcinoma patients may be related to other genetic factors modifying the outcome. Clinical relevance of BRCA1/2 unclassified variants, both in terms of cancer pathogenesis and clinical behavior, warrants further studies on larger cohorts of patients.


  1. Top of page
  2. Abstract
  6. Acknowledgements

The authors thank Jeannelle Kraan (Department of Human Genetics, Leiden University Medical Center) for her excellent technical assistance.


  1. Top of page
  2. Abstract
  6. Acknowledgements