Our population-based study shows that, adequate documentation on family history was present in only 41% of all medical records of EOC patients. This percentage is in agreement with ranges found in literature regarding patients with CRC and data required in primary care settings [17
]. In univariable analysis a slight improvement over time was seen which is encouraging however family history is still poorly recorded. This is especially true when taking into account our limited definition of an adequate history covering the majority of hereditary ovarian cancers, namely the BRCA
mutation carriers, but leaving others unattended. Previous studies on CRC defined adequate family history taking by presence of a notification of relatives with CRC only [18
]. Therefore, we feel that we used an appropriate definition for adequate family history taking in this study, but encourage clinicians to apply a much broader definition.
We believe a comprehensive adequate family history should include information on first and second degree relatives, the type of tumor they developed (especially colorectal or endometrial carcinoma) and the age at onset. Since families are getting smaller one should also note the total number of first and second-degree relatives to put it in perspective. An additional problem is that with the decreasing number of relatives family history taking is expected to be less accurate in the future. It is important to ask about second degree relatives since 50% of the mutations are paternally derived and the fathers are likely to be unaffected. Though it is known that information on second degree relatives is less reliable than information on first degree relatives [21
]. Also the absence of malignancies should be noted. In case of suspicion of hereditary cancer one should refer the patient to a specialist in the field of clinical genetics.
We defined four factors to be independently correlated with adequacy of family history taking. Having chemotherapy or surgery often requires involvement of more specialists, longer therapeutic relationships and hospitalization, leading to more opportunities to ask about family history. Regarding age, in younger patients family history taking is probably more accurate because physicians are aware of an earlier age at onset in the majority of mutation carriers. We urge specialists to keep in mind the possibility of a hereditary malignancy, even in the elderly patients. With respect to hospital type, specialized gynecologists in academic hospitals more often recorded family history. Since it is shown that treatment by a gynecologic oncologist improves outcome, more and more EOC patients will be treated by specialized gynecologists and discussed in multidisciplinary tumor boards, which eventually may improve adequacy of history taking.
With the result of the multivariable analysis the question raised if gynecologists perform worse in family history taking compared to other physicians. To answer this question we again used the database of the NCR. Family history data of all 4,858 BC and all 342 EOC patients diagnosed between 2005 and 2008 were compared. Data, extracted from medical records, included number of first-degree relatives with the same malignancy (being BC or EOC), age of the youngest relative with the same malignancy and number of first-degree relatives with another malignancy. In 30% (104/342) of all EOC patients, compared with 64% (3,103/4,858) of all BC patients, documentation on presence or absence of first-degree relatives with the same malignancy was found. 24% (81/342) of EOC patients and 20% (989/4,858) of BC patients had any documentation in their medical record about first-degree relatives with other malignancies. So family history regarding first-degree relatives with the same malignancy was taken twice more frequently in BC patients. Breast cancer is much more common compared to ovarian cancer. As a result, patients can be more aware of BC in their family and both patient and physician can be more forthcoming towards this subject. The clinician and patient can also be more aware of BC as a hereditary cancer, since there has been a lot of attention for this subject. Family history taking on first-degree relatives with another malignancy is as likely to be forgotten by surgeons in BC patients as by gynecologists in EOC patients.
Although the large sample size is an apparent strength of this study, we are aware of some limitations. It is possible that specialists did not register anything in the absence of relatives with a malignancy. In that case the family history was considered to be absent although it in fact was examined correctly. In order to determine this possible bias, we compared the data with self-administered questionnaires. Ziogas et al. [21
] showed that reliability of self-reported family history taking varies by cancer site and by degree of relative. For first-degree relatives, family history provided by patients was accurate in 83.3% for ovarian cancer. Murff et al. [22
] supported the first statement but also states that negative family history reports for ovarian cancer are less useful. Especially abdominal malignancies are reported inaccurately probably because many organs are within the abdominal cavity and it is often referred to as “abdominal cancer” [23
]. In the current study, we were unable to assess the reliability of family history taking by verifying it with a population-based registration and therefore chose a comparison with self-administered questionnaires.
Data of 147 EOC cases of mainly long-term ovarian cancer survivors were studied. In nearly 60% of the cases no agreement was found which was largely due to missing data in the EOC database in absence of any relatives with a malignancy in the questionnaires. Data from medical records were gathered in 2007 and the questionnaires were filled out by patients in 2008. It is likely that if gynecologists asked about family history, they did so in the early stages of treatment and never pursued it over time. Patients may also have developed more awareness on cancer in their family during the years since their diagnoses. As mentioned previously, these 147 cases were mainly long-term survivors. While analyzing data of these cases, 58% of these patients showed an adequate documentation in their medical records, compared to 41% in the whole EOC group. An unselected patient group with respect to survival may worsen the results in our study even more.
It is of upmost importance to have an accurate tool for the identification of hereditary cancer. The risk of EOC can currently not be reduced by screening but a prophylactic BSO offered to high-risk patients can reduce the risk significantly. BRCA mutation carriers also have an improved sensitivity to platinum chemotherapy and novel therapeutic agents such as poly (ADP-ribose) polymerase inhibitors have increased activity in these patients [24
]. High risk patients however, need to be identified firstly. Our study shows that family history taking over the last 10 years was inadequate in the majority of EOC patients. Moreover, a recent study shows that family histories change significantly over time and updates on family history every 5–10 years are recommended [25
]. Data on sensitivity of family history as a predictor of mutation carrier status are conflicting. In various studies all EOC patients both underwent family history taking and genetic testing. The proportion of BRCA gene mutation carriers having a first or second degree relative with ovarian or BC varied from 92% [26
] to 62–69% [4
]. It appears reasonable to offer genetic testing to all non-mucinous EOC patients in order to fully benefit from preventive measures like a BSO. But since genetic testing is not routinely offered in many countries and even when it is offered it is not performed in all cases, there is still a place for family history taking. Family history taking also plays an important role in the counseling of patients with a possible hereditary tumor. Moreover, patients with familial ovarian cancer but without mutation can be offered a BSO. The upcoming of electronic health record systems can be helpful by turning family history taking into a fixed item in consultations with cancer patients. Further education is needed for the physician to increase awareness of hereditary cancer since taking an adequate family history is still essential to provide high-standard care to patients with cancer and their families even in the era of genetic testing.