Evidence indicates that light chain (AL) amyloidosis is caused by deposition of abnormally folded light chains in organs like the heart, kidney, liver, and nerves. These abnormal light chains are secreted by plasma cells; and often lead to involvement of more than one organ in a given patient. Organ damage is the main driver of morbidity and the leading cause of early death in such patients. Treatment in patients with AL amyloidosis is targeted toward the underlying plasma cell clone. The objective of treatment is the rapid reduction of circulating light chains to stop further organ damage and allow for gradual degradation of the organ amyloid deposits, resulting in organ improvement. While both hematologic and organ responses are important in AL amyloidosis, there has been no model to concurrently assess both hematologic response and individual organ responses for a given patient and compare these across patients receiving different treatments. Surbhi Sidana, MD, and colleagues sought to fill this gap, to develop a system that can be used in the clinic to assess the impact of therapy, as well as provide a framework for regulatory purposes for drug approvals.
In this study, published in Blood Cancer Journal, we developed a composite model that can allow for both hematologic and organ response assessment in patients with AL amyloidosis. To develop this model, we assessed outcomes with treatment in two large independent cohorts of patients with newly diagnosed AL amyloidosis. The testing cohort included 473 Mayo Clinic patients, and the validation cohort included 575 patients from the Amyloidosis Center in Pavia, Italy. The majority of patients in both cohorts had cardiac involvement, and about one-half had involvement of more than one major organ (heart, kidney, liver). Response assessment was conducted at 6 months from the start of treatment. As there was no existing model to evaluate multiple organ responses simultaneously, we first developed a combined parameter to assess organ response across multiple organs. Using existing criteria for individual organ responses, organ response was classified as one of the following:
- All organ response – response in all of the involved and evaluable organs (heart, kidney, liver).
- Mixed organ response – response in at least one of the organs
- No organ response.
The rates of combined 6-month organ response in the Mayo and Pavia cohorts were:
- All organ response – 26% and 21%, respectively.
- Mixed organ response – 4% and 18%, respectively.
- No organ response – 60% and 61%, respectively.
Each patient was assigned a score for their combined organ and hematologic responses. Scores for hematologic response varied from 0-3, with 0 representing complete response, 1 representing very good partial response, 2 representing partial response, and 3 representing no response. Scores for organ response varied from 0-2, with 0 representing all organ response, 1 representing mixed response, and 2 representing no response; lower scores indicated better response. We estimated survival outcomes of patients with different scores (1-5) relative to a score of 0, which indicated complete organ and hematologic response. Using these scores, we developed a composite model with which patients could be divided into two main groups based on similar outcomes for survival. Group 1 included those with scores of 0-3, and group 2 included those with scores of 4-5. In the Mayo cohort, group 1 had significantly better survival outcome compared with group 2 (median overall survival, not reached vs. 34 months) with a hazard ratio (HR) of 3.4 (95% confidence interval [CI]: 2.5-4.6). This model was then validated in the Pavia cohort, in which median survival in group 1 was 87 months, compared with 23 months in group 2 (HR, 2.8 (95% CI, 2.2-3.5).
The composite model performed better for predicting survival than using hematologic response or organ response alone. As this model was predictive of survival, it may be used as a surrogate endpoint for treatment in clinical trials, allowing for rapid assessment of treatment outcomes. This model can also allow us to compare outcomes of patients with different treatments. Future studies using this model are needed to evaluate the utility of changing treatment based on early composite response assessment.