Communication Regarding Therapeutic Clinical Trial Enrollment Between Oncologists and Adolescents and Young Adults with Cancer

Introduction

Adolescent and young adult (AYA) cancer patients have low rates of enrollment in therapeutic clinical trials relative to children and older adults.^1–6 Whereas 40%–70% of children <15 years of age enroll in clinical trials,^3,7 only 7%–14% of AYAs do,^3,7,8 largely because of limited availability of trials across the AYA age range.

Prior studies have demonstrated that AYAs who enroll in clinical trials have superior survival outcomes^2,4,7 and are more likely to receive appropriate disease-directed therapy, ⁹ underscoring the importance of trial enrollment for these young patients. AYAs have not experienced the same survival gains as children and older adults have in recent decades.^10,11 Enhanced trial enrollment is an essential avenue for progress and offers opportunity to prospectively evaluate clinical and biologic features that contribute to poorer outcomes.^2,4 The National Cancer Institute's AYA Progress Review Group has defined AYAs' clinical trial enrollment as a priority research area. ¹¹

Prior study has largely focused on whether trials are available for AYA cancers. Little work has focused on the role of patient–physician communication in AYA clinical trial enrollment, despite research implicating communication as a key factor affecting enrollment among older adults, especially minorities.^12–18 We surveyed newly diagnosed AYA patients at a single large academic cancer center regarding experiences with communication surrounding clinical trial enrollment, including whether therapeutic trials were offered and whether they chose to enroll, and compared AYA reports with medical record documentation of trial discussions and enrollment. The goal of this study was to understand AYAs' experiences with trial discussions at a large academic cancer center.

Methods

We surveyed cancer patients aged 15–29 years at diagnosis, and their oncologists, at Dana-Farber Cancer Institute (DFCI), Boston, Massachusetts, from April 2014 to October 2017.^19,20 English-speaking patients were eligible if they were 1–6 weeks postdiagnosis at initial contact and oncologists provided permission for contact (permission was declined for six patients). Oncologists were asked to complete a question on patient prognosis. Medical records of participating patients were reviewed to determine whether therapeutic clinical trials were offered and whether patients enrolled.

Eligible patients were given a letter inviting participation, the survey, and an opt-out postcard. For patients <18 years, patient assent and parent/guardian consent were required; patients >18 years provided verbal consent. Surveys were offered in paper or electronic format. Nonrespondents received two subsequent contacts. Return of the baseline questionnaire was required within 12 weeks of diagnosis. Participants received a $50 gift card.

Of 303 eligible patients, 275 were approached. Of them, 218 consented and 203 completed the survey (74% eligible and approached; 67% eligible). In total, 99 oncologists completed the item on prognosis for 185 of 203 (91%) participants. Medical records of 193 of 203 participants included documented informed consent conversations about treatment at diagnosis.

Questions on trial enrollment were adapted from a previous study. ²¹ The survey explained that “Clinical trials are research studies about treatment. Treatments tested in a clinical trial may involve chemotherapy, radiation, surgery, or other drugs or therapies.” Patients were asked, “When you first started on cancer treatment, did anyone mention that enrollment in a clinical trial might be an option for you?” with response options of yes, no, and do not know. Patients who responded “yes” were asked, “Are you currently being treated as part of a clinical trial?” (yes, no, do not know). Those who responded “yes” were asked, “why did you choose to be part of a clinical trial?” Response options were “the trial offered the best treatment for my cancer”; “the trial would not help me personally, but would improve treatment of future patients”; “the trial was my only option”; “other”; and “don't know”; patients could choose all applicable responses. Those who reported that they chose not to enroll were asked why they did not, with response options of “the treatment offered on the trial had too many risks or side effects”; “I wanted to receive treatment that was proven to be the best for my type of cancer”; “I was worried about the costs of the clinical trial”; “other”; or “don't know”; patients could choose all applicable responses.

Patients were asked to report gender, race/ethnicity, education, insurance type, and level of financial dependence on parent(s)/guardian(s) before diagnosis. Age and diagnosis were extracted from medical records. Oncologists were asked “how likely you think it is that this patient will be cured of cancer,” with response categories of “extremely likely (>90% chance of cure)”; “very likely (75%–90%)”; “moderately likely (50%–74%)”; “somewhat likely (25%–49%)”; “unlikely (10%–24%)”; “very unlikely (<10%)”; or “no chance of cure.”^22–24

The study was approved by the DFCI's Institutional Review Board.

Results

Documented informed consent conversations in medical records identified 20% (n = 38/193) of patients who were offered therapeutic clinical trials and 97% (n = 188/193) who were offered standard nonprotocol treatment regimens. Female patients were more likely to be offered trials than male patients (p = 0.02, Table 1). The frequency with which trials were offered varied by diagnosis, with highest rates of trials offered to patients with leukemias and brain tumors (p = 0.004). We identified no differences by patient age (p = 0.54), race/ethnicity (0.84), medical versus pediatric oncology (p = 0.18), or insurance type (p = 0.44).

Fifty-eight percent (22/38) of patients who were offered trials enrolled; the remainder chose standard nonprotocol treatment regimens. Patients with less favorable oncologist-reported prognoses (<75% chance of cure) were more likely to enroll in trials relative to patients with more favorable prognoses (p = 0.03). The decision to enroll was not associated with patient age (p = 0.25), gender (p = 0.26), race/ethnicity (p = 0.82), education (p = 0.17), or diagnosis (p = 0.11).

Seventy-seven patients (40% of 193 participants) recalled having been told that a trial might be an option. This included 21% (32/150) of patients without a documented discussion about a trial. In contrast, 11% (4/38) of patients whose records documented that a trial was offered reported that they had not been offered a trial. In total 89% (171/193) of patients were able to correctly report whether they were enrolled on a trial. However, just over one-quarter (27%, 6/22) of patients on clinical trials believed they were not on a trial. Five percent of patients not on trials believed they were enrolled (9/171) and 4% were unsure (7/171). Patients who completed surveys ≤6 weeks from diagnosis were no more likely to recognize whether a trial was offered (p = 0.34) or whether they enrolled (p = 0.79) than patients who completed surveys later.

Of patients who enrolled on trials, n = 11 reported enrolling to receive the best possible therapy for their disease; n = 9 wanted to help future patients; n = 1 reported that the trial was the only option; and n = 3 patients cited potential to avoid an aspect of standard treatment such as radiation. Of n = 16 patients who were offered trials but declined, reasons included a preference for standard therapy as the best proven therapy (n = 7); concern about risks (n = 4); concern about time commitment (n = 1); or other reasons (n = 1). Three patients were unsure about reasons.

Conclusion

Prior study on clinical trial enrollment among AYAs has largely focused on a lack of availability of trials as a key barrier to enrollment and to improved outcomes as a whole in this vulnerable group. In this study of 193 newly diagnosed AYAs at a single center, only 20% were offered therapeutic clinical trials. Unlike in a prior study, ⁶ we found no difference between pediatric and medical oncology settings. Our study did not assess the reasons that trials were not offered, such as lack of available trials, patient ineligibility, or physician decisions not to offer a trial, but these strikingly low numbers suggest a need for continued efforts to ensure that trials are available and accessible to AYAs. We also found that >40% of those offered trials chose not to enroll, highlighting communication and decision-making as potential understudied barriers to enrollment of AYAs.

Patients who are offered trials may decline to enroll for a range of reasons, reflecting individual priorities and values. Patients in our study who chose not to enroll most often cited concerns about risks and a desire to receive the best proven therapy, reflecting a risk-averse stance among patients who generally had favorable prognoses. In contrast, patients with less favorable prognoses had higher enrollment rates, perhaps reflecting the fact that, for such patients, concerns about disease outcomes may motivate patients to consider therapeutic trials in hopes of better outcomes.

Our findings also point toward concerns about communication and informed decision-making. Many patients were unable to accurately report whether a trial had been offered, and more than one-quarter of patients on clinical trials did not recall that they were enrolled. Discussions about clinical trials can be challenging for any patient; randomization and equipoise are complex concepts, ²⁵ and conversations often occur in the emotion-laden days after diagnosis. However, AYAs may need special attention in these conversations. AYAs have evolving capacity for decision-making, which requires the ability to express a choice; to understand medical information; to weigh risks and benefits; and to appreciate the relevance of choices in one's individual life. ²⁶ Evidence also suggests that, even when adolescents have these capacities, the decision-making context can affect competence. For example, adolescents may function poorly in the setting of urgent decisions, which can lead to riskier choices. Yet cancer treatment decision-making is often presented somewhat urgently, creating pressures that can impair the decisional process. Although not a focus of this study, AYAs may also be influenced by other pressures such as finances, a lack of social support during ages of transition, ambivalence about reliance on parents, and inexperience with big decisions.

Limited data exist to address these challenges. However, based on our findings and prior study, we suggest several steps in clinical trial conversations with AYAs. First, we suggest that clinicians elicit and follow AYAs' preferences about how they wish to make medical decisions. Second, clinicians should ask patients for their understanding of the decision being made to ensure that conditions for capacity are met. Third, when possible, avoid urgent or pressured decisions to improve contextual factors that help AYAs make decisions effectively. Finally, even after decisions have been made, revisit information over time to help AYAs absorb treatment information.

Our study is limited by the fact that we were not able to determine why patients were not offered trials. Future study should prospectively determine whether trials are both available and offered to enhance understanding of these issues. Cancers with excellent outcomes such as thyroid and low-stage testicular cancer often do not have available trials, and although we found differences in whether trials were offered by patient gender, this may reflect differences in available trials for select cancers with unequal gender distribution, such as breast cancer. This was a small study at a single institution. Our ability to identify factors associated with enrollment was limited. Our population had limited racial and ethnic diversity and few publicly insured patients. Because racial and ethnic minorities historically have low rates of trial enrollment,^12–18 identifying communication experiences of underrepresented minorities is of special importance for future study.

The wording of survey questions may also explain some discrepancies between patient reports and medical record documentation, because we asked patients whether anyone said a trial “might be available” for them. However, wording cannot explain other misconceptions, including a lack of knowledge about actual enrollment. Patients also completed surveys up to 12 weeks after diagnosis, which may have affected recall of early conversations. Parent factors such as education and role in the clinical trial decision were not assessed, and although our findings suggest potential communication challenges in the clinical trial decisional process, communication was not assessed directly. Finally, our study was conducted at a large cancer center. Although rates of offered trials were strikingly low, they may exceed national figures that incorporate smaller and community sites.

Nonetheless, our findings point to the multifactorial nature of low clinical trial enrollment among AYAs. Few patients are offered trials, and enrollment may be further reduced by communication and decision-making. Ultimately improvements in trial enrollment can offer a path to better outcomes for this vulnerable group. Ensuring informed decision-making is an essential step along this path.