The study, published in the journal Pharmacia, comes from a hematology clinic in Plovdiv and takes the perspective of the payer, Bulgaria’s National Health Insurance Fund. Researchers built Markov chain models, a standard health-economics technique that simulates patients moving between disease states year by year, to compare two futures over 20 one-year cycles. In the standard model, patients in chronic-phase CML remain on continuous tyrosine kinase inhibitor therapy for life. In the extended model, patients who achieve a stable deep molecular response, defined as MR4.5 or MR5, enter a treatment-free remission phase lasting five to seven years, with half assumed to return to treatment.

CML is driven by the BCR-ABL1 fusion gene, created when chromosomes 9 and 22 swap segments to form the Philadelphia chromosome. This abnormal protein’s runaway tyrosine kinase activity fuels uncontrolled white blood cell production. Drugs such as imatinib, marketed as Glivec, nilotinib, dasatinib, and bosutinib specifically shut down that activity, transforming a fatal malignancy into a manageable chronic condition. Since 2017, regulatory authorities have recognized that patients with sustained deep responses can safely attempt drug discontinuation under close molecular monitoring.

The modeling team extracted transition probabilities between health states from real clinical data at their center, imputing missing annual values with the median. Costs and health effects were both discounted at 3 percent per year, in line with established pharmacoeconomic practice. Health outcomes were measured in quality-adjusted life years, or QALYs, using a utility value of 0.92 for all living states, as reported in the literature. All drug costs were annualized by treatment line and expressed in Bulgarian lev. The simulations were programmed and analyzed with TreeAge Pro Healthcare, software widely used by health economists and pharmaceutical analysts.

The cohort simulation painted a sobering baseline picture. Of 151 patients in the chronic-phase cohort, 143 were projected to have died by cycle 20, including deaths from other causes, with only patients in blast crisis assumed to die from CML itself. Among the 37 patients who reached accelerated or blast phases, all were projected to die by cycle 16 under the assumed transition probabilities. These figures underscore why preventing progression through early, effective therapy remains central to CML care.

When the researchers compared cumulative 20-year costs, the savings from including treatment-free remission were modest in first-line therapy, roughly 952 lev per patient, but reached over 14,300 lev per patient in second-line therapy. Scaled to the clinic’s actual population of 110 first-line and 33 second-line patients, the projected savings were about 104,686 lev and 473,381 lev respectively. In other words, three times more money would be saved by stopping therapy in three times fewer patients, because second-line drugs are so much more expensive.

Drug-by-drug analysis reinforced the pattern. In first-line therapy, nilotinib, sold as Tasigna, emerged as the most preferred option, combining the lowest total costs under the treatment-free remission model with the largest QALY gain, 0.23, and a clearly dominant strategy marked by a negative incremental cost-effectiveness ratio. In second-line therapy, imatinib 600 mg stood out with the greatest QALY increase, 0.91, alongside significant cost reductions. Across combination strategies spanning both treatment lines, the sequence of nilotinib followed by imatinib showed the most pronounced superiority, with a full 1.00 QALY gain and categorical dominance of the treatment-free remission model.

The authors note that some scenarios showed identical accumulated QALYs with and without treatment-free remission, making an incremental cost-effectiveness ratio impossible to calculate. This reflects the assumption of equal utility across all living states. In those cases, the treatment-free remission model functions as a cost-minimizing strategy, delivering the same health outcomes at lower cost rather than a lack of benefit. Almost all combination strategies except nilotinib followed by bosutinib yielded lower total costs with equal or higher QALYs under the extended model.

The findings carry practical weight for a health system facing rising demand. Bulgaria diagnoses roughly 40 new CML patients each year, and improved survival with modern therapy means the treated population keeps growing. The study’s discussion points out that Bulgarian hematologists estimate between 20 and 25 percent of patients on tyrosine kinase therapy may already be candidates for treatment discontinuation, provided they meet strict criteria, including at least three years of therapy with a deep molecular response during the final year. The first GeneXpert molecular analysis device on the Balkan Peninsula now operates at the specialized hematology hospital in Sofia, enabling the high-tech monitoring such discontinuation requires.

The study has limitations the authors acknowledge candidly. The Markov models simplify real-world practice, assume patient homogeneity, and exclude individualized treatment approaches. Cost inputs came from a single clinic, which may limit generalizability, particularly for centers favoring generic products over far more expensive originals. Only direct drug costs were analyzed, excluding diagnostics, monitoring, and adverse event management. Even so, the categorical conclusion holds across first-line, second-line, and combination scenarios: including a treatment-free remission phase reduces total costs, preserves or improves health outcomes, and dominates continuous lifelong therapy in most drug strategies considered, offering the National Health Insurance Fund a financially and clinically superior path forward.

Subject of Research: Health economics of treatment-free remission in chronic myeloid leukemia

Article Title: Cost-effectiveness of continuous vs. treatment-free remission strategies in chronic myeloid leukemia: a Markov modeling study from a single center in Bulgaria

Article References: Goranova-Marinova, V., Dimitrova, Z., Andreevska, K., Grekova, D., Georgieva, V., Madzharov, V., Dimitrova, D., & Petkova, V. (2026). Cost-effectiveness of continuous vs. treatment-free remission strategies in chronic myeloid leukemia: a Markov modeling study from a single center in Bulgaria. Pharmacia, 73, Article e192841. https://doi.org/10.3897/pharmacia.73.e192841

Image Credits: AI Generated

DOI: 10.3897/pharmacia.73.e192841

Keywords: chronic myeloid leukemia, tyrosine kinase inhibitors, treatment-free remission, Markov model, cost-effectiveness, pharmacoeconomics, QALY, Bulgaria, BCR-ABL1, health policy, continuous, treatment-free