I want to highlight a paper from the Journal of Clinical Oncology Practice by Cole et al##.
The research team looks at three drugs that have similar survival benefits for Chronic Myeloid Leukemia (CML). One drug, imatinib is now generic while the other two drugs are still brand name, patent protected drugs. The outcomes the researchers looked at were ER visits, inpatient hospitalizations and annual health care costs.
RESULTS:
included 1,417 receiving imatinib, 1,067 receiving dasatinib, and 647 receiving nilotinib. The 1-year risk of safety events was… higher risks among patients receiving dasatinib (RR, 1.17; 95% CI, 1.06 to 1.30) and nilotinib (RR, 1.07; 95% CI, 0.93 to 1.23) compared with those receiving imatinib. Over a median of 1.7 years, the cumulative incidence of safety events was higher among patients receiving dasatinib (HR, 1.23; 95% CI, 1.10 to 1.38) and nilotinib (HR, 1.08; 95% CI, 0.95 to 1.24) than among those receiving imatinib. One-year health care expenditures were high (median, $125,987) and were significantly higher among patients initiating second-generation TKIs compared with those receiving imatinib (difference in medians: dasatinib v imatinib, $22,393; 95% CI, $17,068 to $27,718; nilotinib v imatinib, $19,463; 95% CI, $14,689 to $24,236).
CONCLUSION:
Patients receiving imatinib had the lowest risk of hospitalization or emergency department visits and 1-year health care expenditures.
So what does all of this mean?
The first take-away is that treating CML is expensive no matter what drug is chosen. Insurance design fails when one treatment option could have an annual price of $115,000 while the other choice has a total annual cost of $140,000. Deductibles are being met in the first few weeks of treatment no matter what choice is being made. Coinsurance shuts off as maximum out of pocket limits are hit fairly quickly. Our regular tools to shift people to high value care instead of low value care are grossly ineffective at these price levels.
Secondly, imatinib leads to fewer side effects and it is cheaper than the other two drugs. It ties the other two drugs on survival and beats the other two drugs on side effects and cost. It dominates. Yet, only 45% of the treated population get the drug that is at least as good and significantly cheaper than the alternatives.
This is a case where alternative payment models for care can be very effective in shifting physician behaviors. If a CML diagnosis triggers a bundle that is designed to be big enough to readily pay for average cost of care for imatinib plus a little bit more, two things will happen. First, some physicians will change their prescribing behavior towards imantibib and away from the brand name drugs. That could make the market slightly more valuable for other generic makers to enter and drop prices significantly. Secondly, the average price of the brand name drugs will decrease to hold onto marketshare.
CML seems to be an area where there is significant volume, significant controllable price variation and meaningful clinical near substitutes. Those attributes makes CML an attractive alternative payment method target.
## https://doi.org/10.1200/JOP.19.00301
Another Scott
Very interesting. Thanks.
Yet more evidence for our friend Dean Baker.
Working Paper – The Future of the Pharmaceutical Industry (12 page .pdf)
Cheers,
Scott.
David Anderson
@Another Scott: I had read this paper about a week ago, was fascinated by it and I needed to write something NOT COVID19 for my own mental health.
I have a long, slowly cooking paper that looks at how fast people with cancer hit their OOP Maxes, and that also influenced the post.
WaterGirl
@David Anderson: We are lucky to have you!
kmax
I am a CML patient. I take brand name Gleevec (for some odd reason the generic does not work for me). I have been taking Gleevec for over 10 years.
One reason why many patients have been switched to the alternative drugs is due to developing resistance to Gleevec. I forget the percentages here, but it would not account for the high use of the alternatives. I thought the majority of patients would be on Gleevec.
Any time I see possibilities for changing how I get treatment I get nervous. For CML patients we use these drugs or die.
Can you elaborate some on what these alternative pricing models could look like?
David Anderson
@kmax: For the study, they only looked at initiation (IE people with no anti-CML agents 6 months before the first CML diagnosis shows up in a claim) so that does not look at the resistance/intolerance angle that you rightlfully raise.
An APM for a first year of CML would be the average cost of care for the first line/best in practice methods. In this case it would be imantibib as the primary drug plus supportive therapies and ancillary services plus some ER/IP utilization. On top of that average payment, the bundle would have a topper to cover a percentage of the incremental costs of the other two drugs in case someone has a specifically bad tolerance to the imantibib.
The bundled payment would go to the oncologist for total cost of care and it would be a bit more than the expected cost of care using imantibib and a bit less than the expected cost of care for the other two brand name agents. The incentive would be for the oncologist to use imantibib both as it is directly cheaper and it has fewer adverse effects and to manage the patient to stay out of the ER. At Duke, one of the ways that happens is an oncology trained primary care practice where the PCPs have oncology specific training so patients are more likely to go to them then their oncologist for moderate level concerns.
JR
A significant fraction of these patients (I believe it is around 50% in community practice) will be primary nonresponders to imatinib, and of those that respond around 30% to 40% will develop secondary resistant disease. I can think of no really good reason not to start all patients on imatinib but the truth is the majority are going to end up on something else within a few years.
kmax
@JR: I just checked a study from 2013 that I had not looked at in awhile. Its findings were about 33% of Gleevec patients had poor to no response, some of which were due to acquired resistance (2nd mutations). I had forgotten how high that number was. This accounts for some of the higher use of the second tier drugs, but not all.
I look for new studies regularly. I will see if I find something more recent about resistance
Thanks for the interesting perspective, David.
JR
@kmax:
@kmax:
Yes that’s in clinical trials. In clinical practice it has been closer to 50% primary resistance, and then of course secondary mutations develop over time.
Imatinib, while miraculous, is not that great of a drug because it has a very low barrier to resistance. If you look at the guidelines you’ll see they say that the mutations that confer resistance to imatinib are “too numerous to mention.”
Also I find the results of this study a little suspect because nilotinib had a marginally better adverse event profile than dasatinib, when the opposite is generally the case
Also note that switching drugs incurs a ton of cost due to the need for very frequent initial monitoring (both for adverse events and efficacy) when initiating the new drug.
I suspect that if you analyzed total costs across 5 years on a population basis there wouldn’t be a large difference between those who started on imatinib and those who started on second-generation agents.