Karl Kilgore, PhD, Avalere Health, An Inovalon Company, Washington, DC, discusses study results which showed greater health care resource utilization (HCRU) and costs with CAR-T therapy than transplantation for patients with relapsed/refractory diffuse large B-cell lymphoma (DLBCL).
Transcript
Hi, I'm Karl Kilgore, Senior Research Scientist at Avalere Health, an Inovalon Company, in Washington, DC.
This is a study that we presented at ASH 2020 entitled "Cost and Healthcare Utilization and Relapsed/Refractory Diffuse Large B-Cell Lymphoma: A Real-World Analysis of Medicare Beneficiaries Receiving Chimeric Antigen Receptor T-Cell (CAR-T), Autologous and Allogeneic Hematopoietic Stem Cell Transplants."
In this study, we used the 100% Medicare fee-for-service claims data to pull a sample of patients who received one of the 3 primary treatments for patients with relapsed/refractory DLBCL. These patients typically are considered for autologous hematopoietic stem cell transplant, allogeneic transplants, and CAR-T therapy.
The time period from which these data were drawn was from 2017 to 2019. The CAR-T patients represent the very first real-world experience with CAR T-cell therapy in the Medicare population. There were approximately 200 patients.
Each of the groups, we matched the patients who received CAR-T autoSCT and alloSCT using propensity score match on demographic characteristics and some clinical characteristics, specifically age, gender, race, region of the country ECOG performance status, and Charlson-Deyo Comorbidity Index. We matched the autoHCT and the alloHCT patients separately to the same CAR-T cohort.
The patient groups were very similar after the match. The median age across all groups was 69 to 70. They were predominantly White, non-dual eligible status. The only notable difference post-match was the mean Charlson score was slightly higher for the autoHCT payers than for the alloHCT payers. Not surprising, alloHCT is a little bit harder on the patient.
We looked at healthcare utilization and costs 6 months prior to the index procedure compared to 6 months after the procedure. Health care utilization, we looked in particular inpatient utilization, both in terms of length of stay.
Prevalence of inpatient went down for all groups from pre- to post-. The inpatient prevalence was higher pre- for the auto group, and it was higher with both time periods for the allo group.
Similar trends were seen for length of stay, except that the average length of stay for the alloHCT group was actually 5 days on average higher in the post-period than in the pre-period.
While most utilization and cost statistics decreased for all the groups, this was one exception where inpatient length of stay increased for the alloHCT group. We also looked at emergency department utilization, which went down similarly from pre- to post- for all groups.
Total health care costs decreased for all groups from pre- to post-. Costs were higher in the pre-period for autoHCT compared to CAR-T, and they dropped to about equal levels for alloHCT costs were higher pre- and post- compared to CAR-T.
Now, the utilization and cost figures that I shared with you were the 6 months pre- and the 6 months post- the index procedure itself. That is, they do not include cost for the index procedure itself. We stratified costs by whether the patient was in a clinical trial or was treated in an inpatient prospective payment system-exempt cancer hospital.
In all, very briefly, the cost of all procedures carried significant costs with them, CAR-T with higher and also had a higher standard deviation that was more variation in the CAR-T costs. In terms of utilization during the stay, all of these patients in the study receive their index procedure in the inpatient setting.
In our data set, we had a small number of patients that received CAR-T and autoHCT in the outpatient setting, but the numbers were too low to report. In terms of utilization—so this was all inpatient—the length of stay was longer for both HCT groups than for CAR-T.
The likelihood of an ICU transfer was greater for CAR-T for more of the CAR-T patients had an ICU stay as part of their procedure stay than HCT, but the HCT patients who had a stay in the ICU spent a significant number of days longer in the ICU than CAR-T patients did.
Quick summary then of the finding, procedure costs were greater for CAR-T than for HCT, but when you take into account the period prior to the procedure and the cost of treatment after the procedure, you get significant cost offsets, which are important to take into account for treatment planners, for policymakers in determining the treatment course for the elderly, relapsed/refractory DLBCL patient.
A couple of caveats to this is a descriptive study while we matched patients using propensity score match on patient characteristics and clinical characteristics at baseline period, the 6 months prior to treatment. We were not able to match patients on the prior treatments that they had received because as we went further back in time, our N got smaller.
I will add that the data suggest that the autoHCT patients, for most of them, this was their second-line treatment compared to CAR-T, where most of these patients had 2, 3, 4, 5 prior lines of chemotherapy prior to CAR-T.
I want to add that these data reflect very early experience with CAR-T. It is in fact, the very first year of experience, of real-world experience, with CAR-T. If we continue to pursue this line of reasoning, we may well see some differences in costs and ICU utilization as providers become more familiar with this new procedure. Thank you.
