Editorial

Access to cancer clinical trials may not entirely level the playing field

There is compelling evidence that reducing variability in cancer care can improve patient outcomes and limit costs.

John Sweetenham, MD, FRCP, FACP
John Sweetenham

Current alternative payment models such as CMS’s Oncology Care Model include requirements for care pathway implementation and adherence, recognizing that a platform of decision support and, when appropriate, standardization of care has substantial benefits for patients with cancer.

Many would regard inclusion of patients in clinical trials as the best test of the hypothesis that consistent, protocol-driven care results in improved outcomes. There is an extensive, if somewhat controversial, body of evidence to suggest that overall outcomes for patients treated on clinical trials are better than for those treated off study. However, it is unclear how much of this apparent benefit is therapy related and how much is a function of selection bias, consistent and often more frequent patient monitoring, absence of comorbidities, and socioeconomic and other factors.

Clinical trials as an ‘equalizer’

Cancer clinical trials are changing.

When I became an oncology fellow in the 1980s, prospective, randomized phase 3 clinical trials were the gold standard for evaluating new interventions, which typically went through a classical phase 1 through phase 3 development pathway.

In many respects, that hasn’t changed — when feasible, large-scale randomized studies, depending upon the endpoints measured, provide some of the best evidence for the benefit (or lack thereof) of new cancer treatments.

That said, these studies have their own limitations, especially with the increasing recognition that significant therapeutic effects in small subgroups can be lost in the background noise of large randomized trials. The rapid emergence of highly specific targeted therapies has challenged the one-size-fits-all philosophy of big trials and has led to many innovative trial designs, as well as the inclusion of big data and real-world evidence in the evaluation process.

Real-world data are now recognized as a possible source of “surrogate” control data for clinical trials. However, the utility of real-world data for this purpose is not yet clear — robust methods for abstraction and verification of information from sources such as electronic health records or claims data are still under development but mostly are not at the level required for prospective studies. This may result in imbalances between an experimental arm and the real-world data “control.”

Although the future of large prospective trials may be uncertain, I suspect I am not alone in thinking of them as a great way to level the playing field for patients with cancer. The fact that patients are treated in a consistent manner according to the “ultimate” care pathways implies that any outcome differences can be attributed to the intervention being studied in the trial. This should eliminate any potential disparities, except for those that might have a biological basis. This does not address the issues of variable access to the trial, but at least for the patients in the study population, the trial design and conduct should be an equalizer.

There are certainly data to support this. A large, retrospective Southwest Oncology Group (SWOG) study published last year investigated whether trial participation eliminated the well-documented outcome disparities associated with living in a rural vs. urban community. Joseph Unger, PhD, MS, and colleagues evaluated outcomes of more than 36,000 patients enrolled on SWOG studies for several solid tumors and hematologic malignancies from 1986 through 2012. After adjusting for demographic differences such as age and race between the rural and urban populations, no differences in survival outcomes occurred according to place of residence.

These data, if confirmed, provide reassurance that trial participation potentially is a powerful tool for reducing cancer disparities and, more generally, support efforts to improve access to high-quality, consistent care for underserved populations.

Questioning trial fairness

My confidence in the “leveling” function of the clinical trials process was shaken recently by learning — initially on Twitter, in line with HemOnc Today’s cover story this month on social media in medicine — of a study presented at Quality Care Symposium investigating the effect of insurance status on outcomes among patients in trials.

Like the paper cited above, this report, also by Unger and colleagues, describes a retrospective study of SWOG trials conducted between 1985 and 2014, for which OS in the experimental treatment arm was superior to that of the standard arm. They identified 19 such trials that included more than 11,000 patients. Eighteen percent of the patients included in these studies had either Medicaid or no insurance.

The results of the analysis are quite shocking. The OS benefits were equal when analyzed by age, sex or race/ethnicity, but, when analyzed by insurance status, the patients with Medicaid or no insurance derived no survival benefit from the experimental arm — the survival benefit appeared to be restricted to those with adequate insurance. An analysis of PFS showed this effect begins within the first year of follow-up and appears to continue thereafter.

The causes of this finding are unclear, although the authors cite limited access to supportive services or low adherence to protocol-directed follow-up due to financial constraints as possible contributors.

At the same meeting, Huey and colleagues from The University of Texas MD Anderson Cancer Center reported on data that explored the financial impact of phase 1 trial participation. They demonstrated that the out-of-pocket and nonmedical costs of trial participation were higher among those with low household income — further evidence that even for patients enrolled in a study, the financial effects are relatively higher for those with limited resources. The out-of-pocket costs were often higher than expected, suggesting possible room for improvement in financial counseling of these patients. The study does not report outcomes, but the findings raise questions about the fairness of the clinical trials process.

The findings of these studies are preliminary and need to be confirmed. If the benefits of large randomized trial participation are restricted to those with good insurance, this is truly very disturbing.

In the short term, it raises issues of trial design and interpretation, as well as questions regarding the informed consent process — the description of potential benefit for an insured and underinsured patient might have to be framed very differently. In the longer term, these data support the need for policy changes to eliminate income and insurance coverage as barriers to clinical trial enrollment.

Further extension of coverage for patients on trials would eliminate another source of cancer disparity and, if the results of this recent study are confirmed, have the potential to improve survival and prevent relapse among underinsured patients — a possible source of cost savings.

References:

Unger JM, et al. JAMA Netw Open. 2018;doi:10.1001/jamanetworkopen.2018.1235.

The following were presented at Quality Care Symposium; Sept. 6-7, 2019; San Diego:

Huey R, et al. Abstract 8.

Unger JM, et al. Abstract 119.

For more information:

John Sweetenham, MD, FRCP, FACP, is HemOnc Today’s Chief Medical Editor for Hematology. He also is associate director for clinical affairs at Harold C. Simmons Comprehensive Cancer Center at UT Southwestern Medical Center. He can be reached at john.sweetenham@utsouthwestern.edu.

Disclosure: Sweetenham reports no relevant financial disclosures.

There is compelling evidence that reducing variability in cancer care can improve patient outcomes and limit costs.

John Sweetenham, MD, FRCP, FACP
John Sweetenham

Current alternative payment models such as CMS’s Oncology Care Model include requirements for care pathway implementation and adherence, recognizing that a platform of decision support and, when appropriate, standardization of care has substantial benefits for patients with cancer.

Many would regard inclusion of patients in clinical trials as the best test of the hypothesis that consistent, protocol-driven care results in improved outcomes. There is an extensive, if somewhat controversial, body of evidence to suggest that overall outcomes for patients treated on clinical trials are better than for those treated off study. However, it is unclear how much of this apparent benefit is therapy related and how much is a function of selection bias, consistent and often more frequent patient monitoring, absence of comorbidities, and socioeconomic and other factors.

Clinical trials as an ‘equalizer’

Cancer clinical trials are changing.

When I became an oncology fellow in the 1980s, prospective, randomized phase 3 clinical trials were the gold standard for evaluating new interventions, which typically went through a classical phase 1 through phase 3 development pathway.

In many respects, that hasn’t changed — when feasible, large-scale randomized studies, depending upon the endpoints measured, provide some of the best evidence for the benefit (or lack thereof) of new cancer treatments.

That said, these studies have their own limitations, especially with the increasing recognition that significant therapeutic effects in small subgroups can be lost in the background noise of large randomized trials. The rapid emergence of highly specific targeted therapies has challenged the one-size-fits-all philosophy of big trials and has led to many innovative trial designs, as well as the inclusion of big data and real-world evidence in the evaluation process.

Real-world data are now recognized as a possible source of “surrogate” control data for clinical trials. However, the utility of real-world data for this purpose is not yet clear — robust methods for abstraction and verification of information from sources such as electronic health records or claims data are still under development but mostly are not at the level required for prospective studies. This may result in imbalances between an experimental arm and the real-world data “control.”

Although the future of large prospective trials may be uncertain, I suspect I am not alone in thinking of them as a great way to level the playing field for patients with cancer. The fact that patients are treated in a consistent manner according to the “ultimate” care pathways implies that any outcome differences can be attributed to the intervention being studied in the trial. This should eliminate any potential disparities, except for those that might have a biological basis. This does not address the issues of variable access to the trial, but at least for the patients in the study population, the trial design and conduct should be an equalizer.

PAGE BREAK

There are certainly data to support this. A large, retrospective Southwest Oncology Group (SWOG) study published last year investigated whether trial participation eliminated the well-documented outcome disparities associated with living in a rural vs. urban community. Joseph Unger, PhD, MS, and colleagues evaluated outcomes of more than 36,000 patients enrolled on SWOG studies for several solid tumors and hematologic malignancies from 1986 through 2012. After adjusting for demographic differences such as age and race between the rural and urban populations, no differences in survival outcomes occurred according to place of residence.

These data, if confirmed, provide reassurance that trial participation potentially is a powerful tool for reducing cancer disparities and, more generally, support efforts to improve access to high-quality, consistent care for underserved populations.

Questioning trial fairness

My confidence in the “leveling” function of the clinical trials process was shaken recently by learning — initially on Twitter, in line with HemOnc Today’s cover story this month on social media in medicine — of a study presented at Quality Care Symposium investigating the effect of insurance status on outcomes among patients in trials.

Like the paper cited above, this report, also by Unger and colleagues, describes a retrospective study of SWOG trials conducted between 1985 and 2014, for which OS in the experimental treatment arm was superior to that of the standard arm. They identified 19 such trials that included more than 11,000 patients. Eighteen percent of the patients included in these studies had either Medicaid or no insurance.

The results of the analysis are quite shocking. The OS benefits were equal when analyzed by age, sex or race/ethnicity, but, when analyzed by insurance status, the patients with Medicaid or no insurance derived no survival benefit from the experimental arm — the survival benefit appeared to be restricted to those with adequate insurance. An analysis of PFS showed this effect begins within the first year of follow-up and appears to continue thereafter.

The causes of this finding are unclear, although the authors cite limited access to supportive services or low adherence to protocol-directed follow-up due to financial constraints as possible contributors.

At the same meeting, Huey and colleagues from The University of Texas MD Anderson Cancer Center reported on data that explored the financial impact of phase 1 trial participation. They demonstrated that the out-of-pocket and nonmedical costs of trial participation were higher among those with low household income — further evidence that even for patients enrolled in a study, the financial effects are relatively higher for those with limited resources. The out-of-pocket costs were often higher than expected, suggesting possible room for improvement in financial counseling of these patients. The study does not report outcomes, but the findings raise questions about the fairness of the clinical trials process.

PAGE BREAK

The findings of these studies are preliminary and need to be confirmed. If the benefits of large randomized trial participation are restricted to those with good insurance, this is truly very disturbing.

In the short term, it raises issues of trial design and interpretation, as well as questions regarding the informed consent process — the description of potential benefit for an insured and underinsured patient might have to be framed very differently. In the longer term, these data support the need for policy changes to eliminate income and insurance coverage as barriers to clinical trial enrollment.

Further extension of coverage for patients on trials would eliminate another source of cancer disparity and, if the results of this recent study are confirmed, have the potential to improve survival and prevent relapse among underinsured patients — a possible source of cost savings.

References:

Unger JM, et al. JAMA Netw Open. 2018;doi:10.1001/jamanetworkopen.2018.1235.

The following were presented at Quality Care Symposium; Sept. 6-7, 2019; San Diego:

Huey R, et al. Abstract 8.

Unger JM, et al. Abstract 119.

For more information:

John Sweetenham, MD, FRCP, FACP, is HemOnc Today’s Chief Medical Editor for Hematology. He also is associate director for clinical affairs at Harold C. Simmons Comprehensive Cancer Center at UT Southwestern Medical Center. He can be reached at john.sweetenham@utsouthwestern.edu.

Disclosure: Sweetenham reports no relevant financial disclosures.