Moleculair onderzoek in tumorweefsel bij patiënten met (vermoeden van) kanker wordt steeds vaker aangevraagd omdat de uitkomst hiervan van belang is voor gerichte therapie (gepersonaliseerde behandeling) of voor het stellen van de diagnose. Dit onderzoek op tumorweefsel kan soms DNA-varianten aantonen die mogelijk kiembaanvarianten zijn in een gen relevant voor de klinische vraagstelling of in een gen dat los staat van de kanker van de patiënt (‘nevenbevindingen’). De kans dat een variant in een ‘niet-kankergen’ wordt gevonden is met de huidige werkwijze verwaarloosbaar. In deze richtlijn wordt een variant in een gen relevant voor de klinische vraagstelling, die mogelijk in de kiembaan aanwezig is “een aanwijzing voor een mogelijk erfelijke aanleg” genoemd. De mogelijke uitkomsten van de moleculaire tumortest – zowel de beoogde uitkomst als de aanwijzingen voor een mogelijk erfelijke aanleg – kunnen van belang zijn voor de patiënt en/of zijn/haar familieleden. Er is gebrek aan handvatten voor de zorgprofessional om de patiënt goed te informeren over de tumortest, de mogelijke uitkomsten hiervan en de terugkoppeling van de uitkomsten.

Description of studies

Below a description of each included study is provided. However, no description was given if the article was based on expert opinion (Borad (2017); Bunnik (2021); Vos (2018)).

Shirdarreh (2021) performed a narrative review to examine the knowledge and attitudes of patients and oncologists regarding primary and potential secondary findings of tumor multigene next generation sequencing in cancer patients. In total 11 studies regarding secondary findings were of interest for the current PICO question. These studies were performed in the USA (n=6), Canada (n=2), the Netherlands (n=2), and Australia (n=1). All studies included patients with cancer undergoing diagnostic tumor tests. Patients’ knowledge and attitudes were assessed by surveys and/or interviews. In total 1806 patients were studied, however 7/11 studies included fewer than 100 patients. The relative low sample size in most of the studies is a limitation.

Pujol (2019) described clinical practice considerations and schemes for managing germline findings in tumor analysis. This includes a written informed consent and a multidisciplinary approach (i.e., involving oncologist, pathologist, and geneticist in case of germline findings). Since data of a previous study population was used, no additional details are provided in the article.

Li (2020) reported ‘points to consider for reporting of germline variation in patients undergoing tumor testing’ in a statement of the ACMG (American College of Medical Genetics). This statement described the potential benefits, harms, and limitations related to reporting of germline variants identified by tumor testing.

Bijlsma (2020) performed a quantitative multicenter study to understand the preferences of patients with cancer concerning unsolicited findings (UFs) derived from whole-exome (WES) or whole-genome sequencing (WGS). Adults with cancer were included. To obtain patients’ preferences, an online questionnaire was used. To ensure that patients had enough background knowledge, two videos were provided within the questionnaire. These videos included information regarding the basic concept of genetics, potential impact of receiving information about UFs, and on four different categories of UFs (i.e., actionable, non-actionable, reproductive significance, unknown significance). Outcome regarding anxiety was assessed using the validated version of the self-reported Hospital Anxiety and Depression Scale (HADS). In total 701/845 (83%) participants completed the study. A limitation of the current study is the high education level of the participants, the fact that not all participants completed the survey, and most of the participants had no experiences with WGS.

Horiuchi (2021) performed a study to examine the incidence of secondary findings (SF) in Japanese cancer patients using WES to understand patient preferences regarding SF disclosure. WES was conducted in 2480 patients, 57% male, with a mean age of 65.7 years. The most common cancers were colorectal (28.4%), lung cancer (19.2%), and gastric cancer (12.1%). Patients had 3 times the chance to disclose the results. If the patient chose to have the results disclosed, the SF results were presented to the patient by a medical genetics doctor and/or a genetic counselor after several sessions of genetic counseling. Only registered variants in the ClinVar and HGMD database were considered.

Pinheiro (2017) performed a study to identify patient preferences regarding information and who should communicate this information. Patients who participated in the study completed a short questionnaire including several topics. Of these topics, eight were chosen for a discussion. In total 70/74 (95%) approached patients did give consent, of them 66 completed the study. The median patient age was 62 years, 30% were males. The most common cancer was lung cancer (30%), followed by colorectal cancer (26%). A limitation of the current is that it was a single center study.

Bennette (2013) performed a study to identify attributes of incidental findings that were important to patients and developed a discrete choice experiment instrument to quantify patients’ preferences. Based on a literature review and expert opinion, a set of key attributes and attribute levels was developed. Two focus groups were conducted, as well as cognitive interviews with patients who had received conventional genetic testing for familial colorectal cancer or polyposis syndromes to refine the attributes’ salience and communication. No detailed description of the study population is given in this study. Next to this limitation, the study was limited by the fact that only 12 patients were included in the focus groups and 6 in the interviews.

Stjepanovic (2018) reported patient preferences regarding return of additional germline results. A patient cohort including 1960 patients with a variety of malignancies were consented. The median age was 58 years, and 23% of the patients were males. Patients who consented to return additional findings or their delegate were contacted by the clinical genetics service, which comprised of a medical geneticist and genetic counsellor if germline medically actionable variants were covered. In total 1691 patients were eligible for tumor profiling. 1556/1691 (92%) agreed to return of additional findings. A number of 8/1596 (0.5%) patients were found with a germline medically actionable variant.

Catenacci (2015) evaluated a large cohort (retrospectively) of gastrointestinal cancer patients who had undergone routine tumor NGS for therapeutic intent. They sought to identify patients who might need follow-up for unsuspected underlying germline events, and to determine whether we could confirm the “high risk” cases post-NGS as germline carriers. The included patients were grouped into three pre-NGS risk categories; high, intermediate, low; based on age at diagnose and personal/family history. Based on NSG results, post-NGS high risk patients were contacted and if they agreed to genetic counselling, they were evaluated. In total 111 patients were analyzed. Of them 21 were classified as high post-NGS risk, and 11 of them were contacted to complete counselling and germline testing. Germline testing was completed by 7/10 patients, and in 3/10 patients confirmed to have germline mutations.

Results

Results are reported per study due to heterogeneity.

1. Patient satisfaction

Shirdarreh (2022) reported that patients had positives attitudes toward tumor NGS despite relatively little knowledge of test-related genetics concepts. However, the expectations of the patients often exceeded the reality of low clinical utility. More realistic expectations were seen by higher educated patients. These patients had also a greater genetic knowledge.

Pujol (2019) reported that clear information about possibilities and benefits of knowing germline mutations should be given by the clinician towards the patients. Additional information about preventive strategies should be given by a geneticist only to patients who require this information.

Bijlsma (2020) reported that participants with a higher quality of life were more interested in receiving UFs.

Horiuchi (2021) reported that reconfirmation of the patient’s intention before the disclosure of SFs is an important step.

Pinheiro (2017) reported that patient preferences regarding incidental findings were greater compared with physicians (56% vs. 19%). In addition, patient preferences regarding method of receiving information was (1) discussion with nurse, (2) written information.

Stjepanovic (2018) reported that the interest of patients who undergo tumor testing in germline findings is high.

Catenacci (2015) reported that patients are often “information-saturated” after a discussion of 20-40 minutes.

Expert opinion articles:

Bunnik (2021) reported that the oncologist should be able to explain key information elements in general terms to cancer patients. Upon referral patients can be informed in more detail. Ideally, patients should be prepared beforehand for the clinical and psychosocial consequences of such unsolicited findings, for themselves and for their family members, and be given the opportunity to autonomously decide whether or not to receive such unsolicited genomic information. It may be useful to offer patients written information materials or help them find educational websites or decision tools.

Borad (2017) reported that greater attention to ethical, legal, and social implications of genetic tumor testing have to be considered to ensure success.

Vos (2018) reported that patients should be able to give explicit consent about whether they would like to be informed about clinically significant unsolicited findings.

2. Anxiety

Bijlsma (2020) reported that participants with elevated levels of anxiety or depressive feelings were less inclined to receive actionable UFs.

3. Patient preferences regarding ‘return of results’

Shirdarreh (2021) reported that patients ’attitudes towards disclosure of secondary findings were variable.

Li (2020) reported that there is lack of evidence for the best practice to report results to patients who want them. Patients should give informed consent upfront for the possibility that results might be discovered. Positive germline test results should be returned by qualified and experienced clinicians (e.g., oncologists with genetics expertise, geneticists, and genetic counsellors).

Bijlsma (2020) reported that men would like to receive UFs more often than women. Furthermore, it was observed that one out of seven participants changed their opinion after the second video which introduced more information on the potential impact of receiving UFs and which explained the four categories. It seems important to distinguish between the categories of UF when giving information (e.g., actionable UFs, non-actional UFs, UFs of reproductive significance, UFs of unknown significance).

Bennette (2013) reported key attributes that summarize patient preferences regarding return of incidental findings; 1. Lifetime risk, 2. Treatability, 3. Severity, 4. Carrier status, 5. Drug response, 6. Total cost.

Stjepanovic (2018) reported that return of results of germline medically actionable variants in cancer predispositions genes could be done by a Genomics Tumor Board (e.g., medical oncologist, clinical molecular laboratory geneticists, genetic counsellors, and medical geneticist). The incorporation of a clinical genetics service into the clinical care is recommended for disclosure of actionable genetic variants, since previously unidentified findings present multiple challenges.

Expert opinion articles:

Bunnik (2021) suggested that when the chance of unsolicited findings is very low, opt-out options need not be actively offered.

Borad (2017) reported that it should be clear how to deal with disclosure of incidental findings that may affect individuals other than the patients. In addition, it is important to closely integrate disclosure of incidental findings with data and privacy sharing concerns. This has to be discussed with the patient.

Vos (2018) reported that unsolicited findings could only be reported if the patient has given consent. It is important to classify unsolicited findings based on their clinical relevance. Patients should be aware of DNA analysis, with its associated benefits and risks, and understand that clinically significant findings are returned.

Level of evidence of the literature

The level of evidence of the literature was not assessed, since the current summary of literature could not be assessed with the GRADE. A quality assessment tool was applied for the qualitative/quantitative research articles, please see table ‘quality of research’.

A systematic review of the literature was performed to answer the following question:

What are the benefits and harms of providing information (according to a standard protocol) about unsolicited findings, compared with standard care, by a medical specialist in patients with cancer undergoing diagnostic tumor test on the predefined outcomes?

P: patients with cancer undergoing diagnostic tumor test

I: providing information about unsolicited findings (UFs), chance/possibilities on unsolicited findings

C:  standard care (i.e., unclear and (possibly) incomplete information)

O:  patient satisfaction, anxiety, patient preferences regarding ‘return of results’

Note: In the above mentioned question we used the term unsolicited findings. Following the discussions in the working group we decided to abandon this term because unsolicited findings seemed not that suitable. We decided that a clue for a possible hereditary condition fits better. See glossary for further explanation.

Relevant outcome measures

The guideline development group considered patient satisfaction as a critical outcome measure for decision making; and anxiety and preferences as an important outcome measure for decision making.

A priori, the working group did not define the outcome measures listed above but used the definitions used in the studies.

The working group did not define a minimal clinically (patient) important difference.

Search and select (Methods)

The databases Medline (via OVID) and Embase (via Embase.com) were searched with relevant search terms from 01-01-2011 to 27-07-2021. The detailed search strategy is depicted under the tab Methods. The systematic literature search resulted in 382 hits.

Studies were selected based on the following criteria;

• patients with cancer undergoing diagnostic tumor test,
• comparison between providing information about unsolicited findings (i.e., according to protocol), and standard care (i.e., unclear and (possibly) incomplete information),
• at least one of the outcomes as defined in the PICO.

In total 45 studies were initially selected based on title and abstract screening. After reading the full text, all studies were excluded as these studies did not include a comparison (i.e., intervention/control). However, 1 review, and 11 original articles were summarized in a table, please see appendix ‘summary of findings’. This literature is described in the rows in grey. Additional literature (i.e., not included in the systematic search; n=22) was provided by the working group, which was also summarized.

Results

Only literature included in by the systematic literature search is descripted in the current summary of literature. The summary of literature includes 12 original articles (i.e., 1 narrative review, 7 qualitative/quantitative research articles, 1 guideline, 3 expert opinion articles).