As somatic genomic testing becomes increasingly adopted as part of standard care for MBC, there is a critical need to assess its impact on patients outside of implications for clinical care. In this study, patients received identical genomic testing (Foundation Medicine) with results delivered immediately after progression providing a consistent setting for evaluation of patient understanding and perception of genomic testing.
There is growing awareness of differences in genetic testing comprehension, as has been documented previously in the scope of germline genomic testing and race,28 and in this study patient understanding of somatic genomic testing showed substantial variability. We used a non-validated 7-question metric to assess genetic knowledge and this demonstrated a wide range of baseline understanding of genetics, with a range 0% to 100% of questions answered correctly in both pre- and post-test. This suggests that providers cannot assume any baseline knowledge of genetics when offering somatic genomic tumor testing. In terms of sociodemographic patient factors, participants who identified as non-white and participants with lower annual incomes had significantly lower baseline genetic comprehension, however, the limited sample size prevented evaluation of possible confounders, such as literacy, numeracy, and other metrics of socioeconomic status. It is important to note that education was not associated with baseline genetic knowledge, possibly skewed by the fact that the observed cohort was predominantly well-educated, limiting our ability to detect an association.
We hypothesized that patients’ genetic knowledge may improve over their time on a prospective clinical trial of somatic genomic testing. There was no formal educational intervention as part of this study but during the informed consent process patients received information on the nature of genetics and genetic testing, the information it could yield, and how that information could impact care. However, our data do not demonstrate any significant change in genetic comprehension from pre- to post-study. While objective genetic comprehension remained stable, patients’ self-assessed ability to explain genetics’ relationship with health to others increased. This suggests a potentially inflated self-appraised ability and understanding, an area explored in the perception of patients’ health and in students’ assessment of knowledge, but not before in the context of patient health education to our knowledge.29 There is evidence that information seeking can falsely increase a patient’s subjective knowledge within the domain of genetic testing and we did see that most patients (85%) sought information after their genomic testing.30 It is important for physicians to be aware of this phenomenon because it may influence patients concealing health illiteracy.31 Collectively, these findings reinforce the need for interventions to educate patients about their care or research studies and to target health literacy, particularly in the informed consent process for genetic/genomic testing.
To investigate patient understanding of genomic testing, we evaluated exploratory questions regarding patient motivations, expectations, perceived risks, and perceived benefits. Notably, the majority of patients mistakenly believed that the NGS test results would explain how they could change their own behavior to lower their risk of disease. This demonstration of advanced cancer patients’ unrealistic expectations of somatic genomic testing corroborates and further informs recent study findings that indicated over 90% of patients believed the NGS test would tell them if they had a higher risk for noncancerous diseases (e.g. Alzheimer’s Disease) or would test for viruses like Human Immunodeficiency Virus despite the informed consent process explicitly stating these would not be accounted for.32
In cancer care, there has been a growing focus on mental health, because of the preliminary documentation of associations between depression and worse survival outcomes in women with advanced breast cancer 16. Although still disputed, it has been hypothesized that psychological distress, including anxiety and depression, may elicit disease progression by interfering with the hypothalamic pituitary adrenal axis.33 While further study is warranted to explore the relationship between psychological health and clinical outcomes, we began to explore the impact of tumor genomic testing on emotional wellbeing. Our cohort of participants had a slightly higher average CES-D depression score (13.6) at baseline in comparison to another trial evaluating depressive-symptoms alone in MBC, which had an average CES-D score of 11.1.16 A recent meta-analysis assessed depression in patients with breast cancer worldwide and found that depression was prevalent in 32.2% of patients, similar to our study population with a prevalence of 38% with a CES-D score reflective of depression. In terms of anxiety, 47% of women displayed anxiety in our cohort.34 In comparison to other advanced cancer types, such as patients with late stage ovarian cancer with a mean BAI score of 16.88–19 at baseline, our cohort had less severe anxiety on average (mean score of 11.3). However, our study cohort had a slightly greater proportion of patients who experienced at least mild anxiety (47%) in comparison to patients with late stage ovarian cancer (44%).35 Interestingly, anxiety and depression positively correlated in our study, indicating that psychosocial interventions for MBC should be comprehensive and not only focus on depression or anxiety. Patients’ mental health status remained rather stable over the duration of the study.
This is the first study to our knowledge to evaluate MBC patients’ trust in their doctor using the validated TPS metric. Patients’ trust in their medical oncologist remained stable over the duration of the study with an average TPS score of 48.4–49.4 (pre- and post-survey, respectively). This is higher than other documented TPS study scores in rheumatic disease and primary care patients which ranged between 41.9 and 45.7, respectively.36,37 The stability of the TPS score was interesting given our prior finding that patients lost confidence in their cancer treatment success after undergoing the genomic testing.20
Self-efficacy in cancer care, defined as patients’ confidence in maneuvering through their cancer care and relationship with their medical oncologist, trended downward from before to after genomic testing. One possible explanation of this finding is that the physicians unintentionally overstated the ability of the NGS test to impact treatment and patients subsequently became disappointed with the results. Within the context of CASE-cancer, this suggests that patients lose confidence in their perceived capacity to navigate their cancer care from before to after the NGS testing. Because increased self-efficacy, amongst other patient psychological factors and communication styles with their physician, is correlated with better disease outcomes, future study of ways to preserve patient self-efficacy is needed.38 One potential method to preserve patient self-efficacy is through increased patient education and empowering supportive interventions.38–40
As genomic testing becomes increasingly integrated into clinical care, it is important to evaluate how the translation of this valuable information holistically affects patients. While other studies are beginning to evaluate patient comprehension and perceptions of genomic testing in other disease groups,41 this exploration in cancer care has primarily focused on germline genetic testing.42–44 These studies still support the need for greater attention on the translation of genomic testing into clinical care, as similar concerns about patient misconceptions about results and what the tests offer are apparent.41
Our study does have limitations. There was a relatively limited sample size, partly due to a lower response rate of 58% (58/100). The study still offers unique insight into critical areas regarding the translation of somatic genomic testing into clinical care because of its longitudinal design and utilization of multiple metrics. It is important to note that our use of non-validated metrics to evaluate patient comprehension of genetics, subjective perception of genetic knowledge, or patient perceived risks and benefits of genetic testing does limit our study. Further, the single-center design also supports the need for further exploration of these areas.
In conclusion, this is the first study to longitudinally evaluate the role of depression, anxiety, trust in physician, self-efficacy, objective genetic knowledge, subjective assessment of genetic knowledge, motivations and expectations in regards to somatic genomic testing in a single advanced cancer population. Further exploration of these areas in a larger study are warranted and will hopefully inform how to best support patients who are undergoing tumor genomic testing by better explaining the role of NGS in cancer care through the creation of patient-centered education initiatives.