Medical genetisist Minna Kankuri-Tammilehto on PRSs in predicting breast cancer risk

Dr Kankuri-Tammilehto heads the Department of Genetics and Genomics at the Turku University Hospital (TYKS) in Turku Finland. She’s a specialist in oncology and cancer genetics and has studied the prevalence of hereditary breast cancer in Finland. We asked Dr Kankuri-Tammilehto about the advantages of including information about PRS in predicting the risk for cancer, particularly breast cancer. 

 

To begin with, we learn that breast cancer cases account for two thirds of all referrals to cancer genetics at TYKS. This is a significant proportion, and has been growing during the past decade. Dr. Kankuri-Tammilehto explains that this is due to the general increase in breast cancer patients. Also, the identification of families with an increased risk of developing breast cancer has improved as new methods have become available, and the criteria for suspecting a hereditary component are more readily available and have been introduced to the training of physicians. 

 

Dr Kankuri- Tammilehto, when is there reason to suspect that a patient’s cancer could be hereditary?

In general, the most common indication is age at the cancer diagnosis. If the patient is exceptionally young or if there are many cancer cases in the family we have reason to suspect an inherited cancer. In certain cases the histological subtype can also point to a genetic component, this applies to at least renal cell carcinoma.  

 

How do you test for pathogenic genetic variants?

At the Department of Genetics and Genomics we have a process for doing this in collaboration with the units that treat the patients. If the patient and her cancer meet certain predetermined criteria, the unit treating the patient will carry out a gene panel test. Our department provides instructions for discussing the outcome of the test with the patient, since the results will also provide information about the risk of family members. 

 

In case you identify a pathogenic genetic variant as the cause of breast cancer, does this affect the treatment and counseling of the patient?

For the breast cancer patient,  we provide counseling including information about how the variant may influence treatment. 

 

The diagnosis also provides information about the risk of the patient’s relatives and we offer counseling to these individuals who have an increased risk of developing breast cancer, if they so wish. For individuals with an identified elevated risk of developing breast cancer, we provide information about measures that could decrease the risk based on the available literature. We have national recommendations for best practices for counseling. These apply to both well-understood high risk cases and also to situations where the risk is moderate. Our national guidelines ensure that the counseling of patients and relatives follows the same process everywhere in the country.  

 

How often do at-risk relatives contact you?

80% of the relatives of breast cancer patients contact us once they learn about their risk. Counselling can be offered by a healthcare professional as part of an on-site visit or via video consultation. We have also prepared video material that healthy individuals can watch to learn about their risk and available protective measures.

 

How often are you able to pin-point the genetic variant behind hereditary breast cancer? And how often are you unable to identify a causative variant? 

We still don’t know all of the genes or their variants that predispose to breast cancer, which means that the variant can’t always be identified. In these cases family history plays a big part when investigating if the cancer could be hereditary. With their consent, we then look at the medical history of the relatives of the patient. 

 

The prevalence of breast cancer cases caused by known high-risk variants (BRCA1 and BRAC2) is 10% of all high-risk hereditary cancers in the Turku area, which corresponds to the national and international prevalence. Other known high-risk variants account for 1.5-2% of cases. This means that in most cases the causative genetic variant is not identified. Counseling of relatives is then based on family medical history and prediction models, which account for moderate-risk variants. Currently individuals are instructed to return in five years with the expectation that a polygenic risk score would then be available.      

 

We then asked Dr. Kankuri-Tammilehto about her views and expectations concerning the utility of PRS in the clinic. 

 

According to recent research, a high polygenic risk score for breast cancer equals the lifetime risk of known monogenic pathogenic variants. What is your take on and view of the current research into PRSs?

This is relevant and interesting and something I also talk about in my lectures.  In my view polygenic risk score calculations are needed to improve risk predictions as part of patient care and counseling.   

 

What advantages would PRS bring to clinical practice?

I foresee several advantages:

 

Importantly, we know that breast cancer risk has a strong polygenic component and we need more data to better understand and model this. A particular example is the CHEK2 gene and its variants, which are challenging to interpret, as the breast cancer risk caused by CHEK2 variants can vary between families and even between members of the same family. In the case of CHEK2, information about the polygenic risk is therefore expected to improve risk prediction. 

 

A tool like PRS, would be very useful in understanding the risk and in planning treatment. PRS would also allow us to identify patients in high risk families and offer them routine screening. 

 

PRS could be used in designing national screening programmes for breast cancer. In Finland, women are offered their first mammogram at the age of 50 as part of the national screening programme. But for women with a high PRS for breast cancer, it might be beneficial to take part in screening five years earlier than the general population. By allowing targeted screening, PRS will benefit both health care systems and patients.

 

In my personal experience, PRS could be particularly useful in identifying individuals at high risk in families where the risk has otherwise been determined as borderline high-risk.  

 

In the long-term, PRS could also help in understanding the histological subtype and the prognosis of the cancer, but this requires more research. 

 

Are there challenges to consider?

Selecting a suitable model needs to take into account the ancestry of the patients in the area. And at the same time we need to account for diversity in our populations and have models that are not based on only Europeans, as these won’t then apply to individuals of non-European ancestry. 

The influence of the PRS also varies between cancers. In breast cancer it seems that the more variants are included the better the prediction. But the same isn’t true for prostate cancer, where about 450 variants provide good predictive power and adding more doesn’t improve this.  

 

How should PRS be incorporated into clinical practice?

In my lectures I present a scheme, where I propose the addition of a PRS model to the clinical evaluation. These models are already in use in some other countries, but in Finland we’re still looking for a suitable one. The methodology for genotyping and gene sequencing has become easy and affordable and adding these to gene panel tests when testing a patient would be feasible.   

 

You give lectures on the topic of PRS, who are your target groups and when should medical professionals receive training on PRS?

Close collaboration between clinical genetics and other specialists increases the awareness of genetic risk in specialist medical care. Often the specialists treating the patients will carry out the initial genetic test. Education on genetic risks and PRS should be available for all levels including medical school and specialist training as well in forums for medical professionals.   

 

How do you communicate the results of genetic tests to patients? Will including information about PRS complicate this? 

Already in current clinical practice the specialist in clinical genetics may have to consider and interpret the effect of more than one genetic variant. Although many specialists that treat patients can order genetic tests, the interpretation of the results is the responsibility of the specialist in clinical genetics who then ensures that the patient understands the role of the genetic risk in the disease.    

 

Unlike a risk that is based on a single pathogenic high-risk variant, polygenic risk can be influenced through life-style changes. Do physicians talk about this and how do patients receive the information?

If the patient doesn’t carry a moderate or high-risk variant, the combined effect of many minor-risk variants is indeed something that can be reduced by life-style choices. The physician treating the patient can provide information about these and typically such information is received well, as increasing awareness of life-style choices is already an integral part of health care. Carriers of moderate and high-risk variants will still receive counseling from a specialist in clinical genetics.     

 

 

Finally we asked Dr. Kankuri-Tammilehto if she was familiar with the INTERVENE Helsinki breast cancer pilot, where information about PRS is communicated to the relatives of breast cancer patients.

 

Yes, I had heard of it and I find it interesting and relevant. Providing a model that can predict risk in the Finnish population will certainly have clinical utility.