A rheumatoid arthritis—focused Virtual Health Assistant (aka avatar) could help save US health plans over $1 billion per year by providing the data needed to pursue an evidence-based dose reduction/discontinuation strategy for biologic DMARDs (bDMARDs). The science is strong, but is the will equally strong?
Rheumatoid arthritis (RA) is common and very expensive. According to the CDC, an estimated 0.6% of the population; about 1.5 million adults over 18 years of age have this disease.
Many of these patients are on biologic disease modifying anti-rheumatic drugs (bDMARDs), which collectively cost an estimated $7 billion in the US alone. This amount is expected to exceed $9 billion by 2020 based upon a report published by business intelligence provider GBI Research and reported on the Drug Development website.
The bDMARD class of drugs have literally revolutionized the treatment of RA, but they come at a very high cost. The comparison website GoodRx.com, which provides cost comparisons at local retail pharmacies for self-administered drugs, reports the cost for the self-administered bDMARDs to be in the range of $3000 to $3300 per month (depending on dose and specific drug). Obviously, the actual cost to the health plan will vary depending upon a variety of factors. The price of an office-administered drug (ie, infliximab) is likewise very expensive, although, again, the exact price may vary depending on the provider contract, site of administration (hospital outpatient versus physician office), quantity, and administration fees.
But the costs are not limited to just financial costs. Many of the bDMARDs come with a black box warning and numerous other side effects that affect the lives of those taking them.
Managed care companies have focused on directing patients to “preferred” or “formulary” drugs to maximize rebates and discounts and thus decrease the overall cost. They also have created prior authorization criteria to ensure appropriate utilization. But following proven dose reduction or elimination strategies for bDMARDs could dwarf the savings of current approaches.
Is it Time For a Health Plan to Promote Discontinuation or Dose Reduction?
Few physicians or health plans have attempted to actually manage the dose of the bDMARDs RA drug category over time. But this may start to change based on a growing body of evidence that demonstrates the safety and efficacy of dose reduction or discontinuation of the biologic drugs in question.
This past month, several French clinicians presented an abstract at the European League Against Rheumatism annual meeting demonstrating the ability to reduce or entirely eliminate, in a significant percentage of people, an anti-TNF bDMARD being used to control rheumatoid arthritis. In their study, TNF-blockers could be tapered for 43.2% and totally stopped for 34.1% of the patients. In their trial, only 18.2% needed to be maintained at their regular dose; obviously saving an enormous amount of money and reducing risk to the patients.
But that study was just the latest in a series of well-designed and implemented studies that have demonstrated similar results for virtually all of the current bDMARDs. In fact, a review article published in November 2014, Intensive Intervention Can Lead to a Treatment Holiday from Biological DMARDs in Patients with Rheumatoid Arthritis, by Yoshiya Tanaka, MD, PhD, and Shintaro Hirata, MD, PhD, concluded: “For early RA, approximately half of early RA patients could discontinue TNF-targeted bDMARDs without clinical flare and functional impairment after obtaining clinical remission by bDMARDs with [methotrexate].” Other patients with longstanding RA could likewise benefit from tapering or discontinuation, albeit at a lower level of success.
The authors performed a systematic literature search in mid-2014 regarding discontinuation of bDMARDs; ultimately finding 86 original articles and numerous abstracts. They excluded 67 articles for several reasons, such as lack of a description of the biologic, unspecified reasons for discontinuation, or the cessation of a biologic for a more preferable one. They added abstracts from the American College of Rheumatology and included 26 total reports in their final published article. bDMARD drugs studied in the various trials included: abatacept, adalimumab, certolizumab, etanercept, infliximab, and tocilizumab. Note, the authors did not mention any of the oral non-biologic DMARDs in this article.
The authors reviewed the symptom duration, criteria for discontinuation, follow-up period, number of patients, ratio of success and failure, duration of study and predictors of success. Their paper gave a short summary of many of the studies as well as noteworthy results. The key to success appears to be deep remission, or low disease activity (LDA) something that has defined criteria but is seldom actually measured in clinical practice in the US.
The results of these studies were astounding.
RRR Study: “71.4% of patients with deep remission were able to continue Low Disease Activity (LDA) for one year” after stopping infliximab.
BeSt Study: 56% of patients reaching LDA for 6 months while taking infliximab were able to discontinue infliximab and maintain LDA for 1 year. The findings went on to state that “more than half of patients who discontinued infliximab successfully maintained LDA for more than 8 years.”
HONOR Study: “approximately 80% of patients with deep remission were able to sustain low disease activity for one year without adalimumab.”
OPTIMA Study: LDA and remission was maintained in “66% of patients who withdrew from adalimumab treatment.”
HARD HIT Study: 44% of “patients were still in remission by 24 weeks” after stopping adalimumab.
1. Reduced etanercept plus MTX
2. MTX with placebo
3. Double placebo.
PRIZE Study: Methotrexate (MTX)-naïve, early RA patients who were initially treated with MTX and etanercept achieved disease activity remission in 70% of patients. The patients were then randomized to a double-blind, 39-week trial of either:
Sustained remission was seen in 63.5 % of those who maintained reduced etanercept plus MTX, 38.5% of those maintained on MTX plus placebo and 23.1% of those maintained on double placebo. The authors went on to state, “There was no significant radiographic progression in any treatment group.”
EMPIRE Study: 110 DMARD naïve patients with early inflammatory arthritis were randomized to receive MTX plus etanercept or MTX plus placebo. After either reaching “no swollen or tender joints” or reaching 52 weeks, all injections were stopped. In those who had initially been treated with etanercept plus MTX, 57.7% remained in remission and LDA after ceasing etanercept for the next 26 weeks.
ACT-RAY Study: 556 patients with established RA with mean duration of over 8 years who inadequately responded to MTX were randomized to receive either 8mg/kg TCZ plus MTX or 8mg/kg of TCZ plus placebo. After ceasing therapy, 14% did not flare.
DREAM Study: In 187 patients who showed LDA on TCZ (for varying periods of time) and then were discontinued: 13.4 % remained LDA after discontinuing TCZ and 9.1% remained in drug-free remission at 52 weeks. According to the authors, most responded promptly to restart of TCZ.
Cessation or Dose Reduction Strategy Is Effective
Overall this review article stated: Taken together, these recent studies indicate that “30-79% of early RA patients could discontinue bDMARDs without clinical flare and functional impairment after reduction of disease activity…”
But, Dose Reduction Strategy Requires Good Data
The authors made a point that close monitoring is essential. Basically disease activity must be measured on a regular basis both prior to and after initiating a dose reduction strategy. First to see who might be a candidate for dose reduction or cessation and then to measure response to dose reduction to provide data on who needs to be restarted on their bDMARD.
It would be virtually unheard of for this to occur in the US, as US physicians do not actually measure disease activity on a regular basis. In fact, none of the studies in this systematic review article were performed in the US. This author speculates that it is because we lack data, something that is actually rather easy to collect—if we have a virtual health assistant acting as the collection agent.
An article on the Arthritis Foundation website discusses how rheumatologist Nathan Wei, MD, occasionally uses the DAS28 in his practice. “Some days I’m a DAS user, some days I’m not.” He said he spends more time asking patients, "How are you doing?"
The article went on to say he would then “consider what the patient says about his or her condition, the results of lab tests, and an occasional imaging exam (usually an MRI) to make treatment decisions.”
A Virtual Health Assistant Can Produce Data to Allow Discontinuation of bDMARDs in RA Patients Through Intelligent Engagement
A recent article reflected on how an avatar, more properly called a virtual health assistant (VHA), could collect this data from patients.
A VHA, using advanced artificial intelligence and natural language procession and residing on a smartphone, could easily engage in a conversation with a person with rheumatoid arthritis. An earlier AJMC series of articles described how a VHA can literally talk to a person. This capability is already possible and is where the human-computer interface is rapidly heading.
A VHA could first and foremost inquire whether the prescribed medication was taken as directed. A VHA can enter into a dialogue about barriers to adherence (ie, financial issues, adverse events, treatment expectations, prior authorization issues, etc) and actually assist in solving the issue. A VHA can contact the pharmacy for a refill, and remind the patient of an upcoming appointment—the capabilities are virtually endless.
The VHA could, on a prescheduled basis, ask the patient to record each painful and swollen joint by touching the smart phone (or tablet) screen at the appropriate joint on an anatomic diagram. In addition it could perform a clinical assessment such as the RAPID3 scale (recommended by the American College of Rheumatology guidelines); a score that asks questions about limitations in performing activities of daily living, sleep difficulties, and overall health.
The device can also be programmed to answer literally tens of thousands of questions a person might ask concerning rheumatoid arthritis, allowing the person to make sense of his or her disease as well as help the person deal with ambivalence and resistance.
A VHA can also be programmed to monitor overall health and encourage healthy lifestyle choices such as smoking cessation, weight control, healthy diet, and recommended exercise. A VHA could also pop up with videos of stretching and strengthening exercises, especially important for those with joint disease.
Health plans have entered the era of big data. But they have to find a way to create a return on investment for collecting and analyzing big data. A VHA can easily create this return on investment with just one drug class: the bDMARDs for RA. Think about what it would mean to your health plan to shave 25% or more from the bDMARD drug class. For a regional health plan, it could mean tens of millions in savings. For our entire country, the savings could be measured in billions!
In addition, patients would be exposed to fewer side effects and physicians could use VHA generated data for clinical decisions instead of relying on lab tests and MRIs as Dr Wei earlier stated. In summary, everyone would benefit.
RA is the perfect case study for using a VHA as a data collection tool because the data is currently not available. The data can provide incredible insight into the lives of those suffering with this terrible disease. The data can lead to actionable decisions by physicians, and the savings could be enormous. The science is strong, but is the will equally strong?