Expert Panel Tackles Merits of End Points in Castrate-Resistant Prostate Cancer: Impact on Treatment and Coverage Decisions

Wayne Kuznar

AVBCC Special Feature, Prostate Cancer


This special feature is supported by funding from Janssen Pharmaceutical Companies of Johnson & Johnson.

Nearly 12% of American men will be diagnosed with prostate cancer at some point during their lives. Prostate cancer may remain indolent, but it often requires treatment, as evidenced by the more than 26,000 deaths attributable to this disease annually. Although the 5-year survival is high for men with localized prostate cancer, this drops to 29% in men with metastatic prostate cancer. Even with treatment, prostate cancer progresses in more than 33% of men. The median time to metastatic disease is approximately 8 to 10 years after the detection of biochemical recurrence.

The definition of nonmetastatic castrate-resistant prostate cancer (CRPC) is no evidence of metastatic disease on imaging; a castrate level of testosterone; and rising levels of serum prostate-specific antigen (PSA), despite the castrate level of testosterone. Periodic imaging and measurement of serum PSA levels are important to detect metastatic disease as early as possible to facilitate treatment.

Current treatment modalities for metastatic CRPC include abiraterone acetate (Zytiga) and enzalutamide (Xtandi). Among the newest therapies for CRPC are 2 new-generation androgen receptor antagonists—apalutamide (Erleada) and darolutamide.

Apalutamide was just approved by the FDA in February 2018 as the first treatment of nonmetastatic CRPC; it is also the first FDA approval that was based on the new end point of metastasis-free survival. Darolutamide is still in development and has received a fast track designation by the FDA.

At the 2017 Association for Value-Based Cancer Care Summit, an expert panel discussed the merits of different end points used in clinical trials of treatments for CRPC and their value to clinicians, patients, and payers. The panel also deliberated on the utility of value-based metrics on treatment and coverage decisions. The following is a highlight of the panel discussion.

Moderator:
What study end points are important when evaluating new therapies for prostate cancer and for metastatic prostate cancer? Is the time from initiation of a therapy to metastasis a viable end point?

Participant:
From a regulatory standpoint, FDA approval has traditionally been awarded based on the prolongation of survival, but lately, progression-free survival, combined with imaging and PSA end points, has bled into how we look at outcomes. The next end point that was considered was metastasis-free survival, which leads nicely into patient quality-of-life issues that create the classic physician–patient decision-making end point.

Thus, payers have put considerable effort into health economic outcome reporting data, because payers need to understand that if clinicians will invoke a therapy that is expensive, they should know the value to patients’ quality of life and the effect on downstream costs, such as emergency department visits, skeletal-related adverse events, and inpatient hospitalization for conversion from nonmetastatic CRPC to symptomatic M1 CRPC. Many men present with M1 CRPC, and we need to do a better job in the clinician community to avoid that presentation.

Participant:
From a clinician’s standpoint, what we can do to delay the progression from nonmetastatic CRPC to metastatic CRPC would clearly not be opposed by payers. Disease progression is a very important end point: Are we slowing the progression of disease, and are we accruing more months of quality of life?

Participant:
When I was with a large health insurance organization, we would look at a test, such as Oncotype DX, which is not inexpensive, and how we could apply it and its return on investment. Using this test reveals that 33% of patients with breast cancer are at low risk and don’t require treatment, and there’s a large cohort of patients for whom treatment will dramatically improve their quality of life. There’s a group in the middle in whom we have to weigh personal factors about treatment or nontreatment.

Regarding CRPC, an insurer would pay for a good predictive test to do the risk stratification. If the patient is at high risk or at medium risk, with 1 or 2 additional nuggets of information, such as young age or a good health status, then payers will cover any therapy that prolongs the time to metastasis.

Participant:
In the CRPC space, there’s a tremendous pushback among payers and the FDA regarding the prevention of metastasis. The disconnect between hormone-sensitive prostate cancer and CRPC is huge. One type is clearly a more aggressive disease. Once the patient becomes castrate-resistant, he is in a higher-risk stratum.

Taking it a step further, the concept of M0 disease is a big phantom. It’s not the M0 disease, but rather our imaging studies are inadequate to find the metastasis. So it really is M1 disease, but we have failed to admit that we cannot see it.

Participant:
If a patient in his mid- or late 80s has a few comorbidities, even if he is androgen-sensitive, the threshold for starting androgen-deprivation therapy (ADT) for him is high. By contrast, if a patient who has received ADT comes to my practice and he is otherwise tolerating his ADT well, and has a PSA doubling time well above 10, I’m not inclined to prescribe treatment for him.

Meanwhile, if he is an 83-year-old patient and his mother and father lived into their late 90s, I’m going to follow him closely, because I do not want to miss that window of opportunity for potential therapy. If you believe that delaying metastasis-free survival by not just a few weeks or months but by years is important, it’s going to be difficult not to institute therapy, regardless of the patient’s age.

As payers, if we have patients who can obtain a drug that is going to markedly delay metastatic disease burden—how could we make a compelling argument to not pay for the drug?

Participant:
We don’t have overall survival data with the use of existing therapies for CRPC, but if we are delaying disease progression without affecting overall survival, I believe that is a compelling reason for society to pay for the treatment, because of the morbidity and quality-of-life issues with metastatic disease. Many things in healthcare don’t have an impact on overall survival, yet we devote our resources to them.

Participant:
I talk to patients about the 3 “P’s”: prolonging survival, preserving quality of life, and preventing therapy complications. There’s a fourth P, prevention, which is preventing economic toxicity. For years, we never talked about economic toxicity, but it’s front and center now.

Moderator:
Metastasis-free survival appears to resonate as an end point. In thinking of value, is it important for payers to see that a drug has the potential to reduce overall healthcare expenditures, at least within a certain window?

Participant:
As a payer, it is always interesting and of some value to have that information, but by and large, a reduction in healthcare expenditures is not something that adds to the binary decision to cover or not to cover a drug.

Participant:
Newly diagnosed prostate cancer is where we will have the maximum impact on disease progression, and where true cost-savings will accrue initially. We want to try to manage the cost versus the best outcomes, which is exactly what MACRA (Medicare Access and CHIP Reauthorization Act of 2015) has called on us to do. In our situation, we have used genomic testing in patients with newly diagnosed prostate cancer for risk stratification and for determining which patients are appropriate for active surveillance.

The rates of active surveillance vary across the country, to as low as 13% in the South; we think that percentage should be 40% to 60%. There’s a huge delta to be gained with patients with newly diagnosed prostate cancer, hence the cost-savings without diminishing the quality of care. We are pairing up with an industry partner to develop a model for advanced prostate cancer.

Participant:
As a nonclinician, I want to ask whether the fear of reimbursement affects treatment decisions, or is the choice tied to potential outcomes?

Participant:
Reimbursement never enters into treatment decisions. Each clinician has his or her own experiences; the clinician has seen side effects and bad outcomes tied to the comorbidities that a patient brings to the table. That combination forges the decision. There are patients I’d normally treat with enzalutamide, but they are already dragging, and I may not think that they would tolerate the fatigue.

Participant:
The factors that enter into treatment decisions include burden of disease, volume of disease, location of disease, comorbidities, symptomatology, and spectrum. However, maybe the ability to pay can have an impact on treatment decisions, as well the proximity to the clinic.

One of the things that organized medicine has been trying to do is to create clinics of excellence, because not everybody should be caring for patients with advanced prostate cancer. Even within a practice, there has to be better collaboration to promote specialization, because you cannot dabble in advanced cancer treatments.

Busy medical oncologists are dealing with breast, lung, colon, and marrow cancers mainly, and prostate, bladder, and kidney cancers become the afterthought. We are promoting specialization at our level, with interactions with government and interactions with our own member practices.

The amount of new information, biomarkers, predictive markers, and diagnostic imaging is mind-boggling, let alone the therapeutics, and they all have significant cost implications. So it really has to be, “get to the right person.”

Moderator:
How important are quality-of-life metrics as secondary end points to the clinician community and to payers? How important are they in clinical trials?

Participant:
Nobody dies from rising PSA levels. What people die from is a critical amount of tumor burden, and location of the tumor burden that results in complications of that tumor burden, whether it’s pain related, orthopedic related, or inanition cachexia. But a rising PSA level has been very predictive in multiple studies. That’s why we have clinical trials, such as SPARTAN, ARAMIS, and PROSPER, because tumor burden has been a huge unmet need.

When we talk about prolonging survival, right there is also preserving the quality of life. An oral-based therapy that prevents metastases and keeps patients from requiring taxane-based chemotherapy—which carries a risk for serious adverse events, or analgesics, or therapy that preserves the ability to carry out daily activities—is a favorable quality-of-life metric that resonates with patients.

Moderator:
Do these secondary end points tip the scales at all in favor of one drug over another when multiple drugs are available, or are you predominantly looking at the primary end point when you make treatment decisions?

Participant:
If we are talking about drugs covered by Medicare Part D, you can make a drug nonformulary and choose one drug over another. What tips that decision?

If you have essentially similar outcomes, but one drug may have many more side effects and affects quality of life more than the other drug, that would tip the scale to the drug with fewer side effects. Ultimately, the treatment decision is heavily influenced by the price of the drug.

Moderator:
After approval, we know that these drugs are coming to market, but there may be a lag time before they are included in the National Comprehensive Cancer Network (NCCN) recommendations, which payers rely on. Would the consideration be on a one-off basis, or on an individual basis, before you draft medical necessity requirements?

Participant:
Generally, we know that even if a drug has not made it to the drug compendia yet, there is enough literature support to develop coverage policies at the release of the drug. We don’t like appeals any more than physicians do.

Moderator:
Let’s say we have 3 oral drugs for M1 CRPC, and each of the 3 drugs has slight variations in safety and tolerability for specific patient populations, but their phase 3 clinical trial data are fairly comparable in terms of their survival and radiographic progression-free survival. You now are in a position to authorize payment. Of the 3 oral drugs, 2 are very close in price, maybe differ by 5% to 10%. The price of the third drug is 30% lower. How does that affect your coverage decision?

Participant:
If there was that much of a price differential, we would clearly choose the lower-cost drug, based on the parameters you set up. Then it comes down to the individual clinician choice, and that’s where I get involved with an appeal, if the drug was denied initially.

Participant:
As payers, what we did was we looked at the various markets and engaged companies that use clinical pathways that meet our standards for quality and operational efficiency, and we obtained an agreement from the practicing clinicians that they would use these clinical pathways. We tried to select clinical pathways that did not disrupt the office workflow.

If clinicians were 80% compliant with the most cost-effective drug choices, given that the other options are equally good in terms of efficacy and safety, then they would not have to obtain approval to use the drugs. Once clinicians started rising above that exception level, that affected their savings bonus on the efficiency end, which accounted for hospitalization rates and emergency department visits. We tried to bypass the friction of medical director appeal and grievance, and gave the clinicians the most autonomy, while holding them to a population health standard.

Moderator:
How did you determine what was cost-effective?

Participant:
We looked at the NCCN guidelines, efficacy, safety profile, comorbidities, and cost. If the drugs are roughly equivalent in efficacy and safety, then we can create clinical pathways with general cost parameters, but we can add enough wiggle room so that people can do business without having friction in their workflow, or without denying patients the therapy that best fits their specific needs.

Moderator:
Do you think value frameworks are going to take on increased importance in consideration of drug coverage on the payer side? How about on the clinician side, in larger health systems?

Participant:
Value frameworks will work in the larger integrated delivery health systems as opposed to the small community hospitals, because larger delivery health systems are currently using these types of standards and metrics. One of my clients is a delivery health system with $8.4 billion in revenue, and they have been using these types of value-based metrics for years. In smaller groups, in which 10 or 20 doctors practice, these metrics are not necessarily used.

We will see a lot of accountable care organizations (ACOs) dissipating when studies show that ACOs are not efficient or cost-beneficial to anybody. We will go back to the standard fee-for-service reimbursement model that we have right now.

Participant:
The reason why value frameworks have been accepted in the bigger health systems is that these systems get to share in the cost-savings. That’s why incentives such as alternative payment methods are so attractive.

Participant:
Urology is one of the specialties in which the patient population will not contribute to cost-sharing benefits in the long-run.

Participant:
A collaboration between the payer and the provider is optimal. A third entity sometimes makes things more difficult; although its value framework can be used as a template or a guideline to begin the discussion, ultimately, it comes down to a mutual agreement between the payer and the clinician. That’s the ideal world.