Patient-Reported Functional Outcomes in Patients With Chronic Myeloid Leukemia After Stopping Tyrosine Kinase Inhibitors

Abstract

Treatment-free remission (TFR) is a goal for patients with chronic myeloid leukemia (CML). Functional outcomes after discontinuing tyrosine kinase inhibitor (TKI) treatment have not been described. Patient-Reported Outcomes Measurement Information System (PROMIS) measures of social, physical, cognitive, and sexual function were assessed over 36 months in 172 adult patients with chronic phase CML from 14 sites at baseline (on TKI) and after discontinuation. Linear mixed-effects models described the average trajectories for each patient-reported outcome measure after discontinuation and in those who restarted TKI. Of 112 patients in TFR at 12 months, 103 (92.0%) had a 3-point or greater improvement in social function, 80 (71.4%) in social isolation, 11 (9.8%) in satisfaction with sex life, 4 (3.6%) in physical function, and no patients had a 3-point or greater improvement in cognitive function or interest in sexual activity. Patients’ scores worsened after restarting TKI. This novel information on functional outcomes in TFR can help guide patient and clinician decision making.

Tyrosine kinase inhibitors (TKIs) enable many patients with chronic myeloid leukemia (CML) to achieve similar life expectancies as the general population (1,2). Select patients can pursue a trial of TKI discontinuation, with approximately 50% remaining off therapy in treatment-free remission (TFR) (3-9). Research using patient-reported outcome measures (PROMs) provides insight into the health-related quality of life of CML patients taking TKIs (10-24), but little is known about patient function—an important consideration for treatment decision making (25-28)—during TFR or after restarting a TKI. We report results from the Life After Stopping TKIs (LAST) study on changes in function after TKI discontinuation and TKI restart.

The LAST study (NCT02269267) was a prospective longitudinal study in 14 US centers (29,30) approved by the institutional review board of each institution. Patients provided written informed consent. Eligibility criteria included adult patients with CML in chronic phase; no history of TKI resistance; taking a first-generation (imatinib) or second-generation (dasatinib, nilotinib, or bosutinib) TKI for at least 3 years with continuous documented BCR-ABL1 < 0.01% in the International Scale (ie, MR4) by real-time quantitative polymerase chain reaction (RQ-PCR) for at least 2 years. Patients restarted a TKI on loss of major molecular response (BCR-ABL1 International Scale > 0.1%).

We evaluated function using Patient-Reported Outcomes Measurement Information System (PROMIS) computerized adaptive tests administered on REDCap (31), including Ability to Participate in Social Roles and Activities (social function), Social Isolation (32), Physical Function (33), Cognitive Function (34), Satisfaction with Sex Life, and Interest in Sexual Activity (35,36). PROMIS scores are standardized and normed; moderate severity is defined as 1 or greater standard deviation (SD) worse than the US average, and severe as 2 or greater SD (37). PROMs were assessed at baseline (on TKI), monthly after TKI discontinuation for the first 6 months, at 8 and 12 months, and then every 6 months until 36 months if patients remained in TFR. If patients resumed TKI therapy, PROMs were evaluated every 3-6 months. Missing data were minimal (<5% of required assessments) (30). We considered a change of 3 points to be clinically meaningful (38). Supplementary Figure 1 (available online) shows examples of likely item responses for PROMIS scores (39).

According to our prespecified analysis plan (29), for each PROM we estimated a piecewise linear mixed-effects model with polynomial terms of time that specified 1 nonlinear trajectory after TKI discontinuation and another trajectory after TKI restart among those who resumed therapy. The models included patient-level random effects for the intercepts and linear slopes. In post hoc exploratory analyses, we examined whether age or length of time on TKI was associated with PROM trajectories. Analyses were conducted using SAS (SAS Institute, Inc) with a 2-tailed statistical significance level of .05. Analytic details are in the Supplementary Methods (available online).

PROMs were assessed in 172 patients with CML over 36 months (patient characteristics in Supplementary Table 1, available online). At baseline, patients were similar to the US population in social function and cognitive function, reported less social isolation, and fared worse in physical function and sexual function (Table 1). After TKI discontinuation, average changes were in the direction of improvement (Figure 1). Of 112 patients in TFR at 12 months, 103 (92.0%) had a 3-point or greater improvement in social function, 80 (71.4%) in social isolation, 11 (9.8%) in satisfaction with sex life, and 4 (3.6%) in physical function; no patients had a 3-point or greater improvement in cognitive function or interest in sexual activity. After TKI restart, average changes worsened across all PROMs. Time on TKI before discontinuation was not associated with PROM trajectories. For physical function and cognitive function, there were statistically significant main effects by age as well as statistically significant interactions between age and time, but these were of small magnitude so substantive findings were unchanged (Supplementary Table 2, available online).

Table 1.

Baseline PROs, changes at 6 and 12 months for patients in TFR, and changes at 6 and 12 months for patients who restart TKI

PROMIS measurea	Baseline scores		Predicted mean change in TFRb, (95% CI)		Clinically meaningful improvementc			Predicted mean change after restarting TKIb, (95% CI)
					≥3-point improvement, % (95% CI)		≥5-point improvement, % (95% CI)
	Mean (SD)	% with moderate or severe impairment	6 mo	12 mo	6 mo	12 mo	12 mo	6 mo	12 mo
Ability to participate in social roles and activities	51.1 (9.6)	12.3	3.4 (2.7 to 4.1)	4.7 (3.7 to 5.6)	72.1 (64.3 to 79.8)	92.0 (86.9 to 97.0)	35.7 (26.8 to 44.6)	−1.6 (−2.6 to −0.6)	−2.6 (−4.2 to 0.0)
Social isolation	43.5 (8.9)	2.9	−2.6 (−3.2 to −2.0)	−3.4 (−4.2 to −2.6)	19.4 (12.6 to 26.2)	71.4 (63.0 to 79.8)	5.4 (1.2 to 9.5)	2.4 (1.3 to 3.5)	2.4 (1.2 to 3.7)
Physical function	47.3 (8.3)	19.9	0.1 (−0.1 to 0.3)	0.3 (−0.1 to 0.7)	0.0	3.6 (0.1 to 7.0)	3.6 (0.1 to 7.0)	−1.5 (−2.5 to −0.6)	−1.6 (−2.8 to −0.4)
Cognitive function	49.3 (10.1)	18.0	1.1 (0.6 to 1.5)	1.8 (1.0 to 2.6)	0.8 (0.0 to 2.2)	0.0	0.0	−2.6 (−3.9 to −1.4)	−2.5 (−4.0 to −1.0)
Sexual satisfaction	44.5 (8.6)	26.8	1.3 (0.7 to 2)	1.7 (0.8 to 2.5)	0.0	9.8 (4.3 to 15.3)	0.0	−0.4 (−0.9 to 0.2)	−0.7 (−1.7 to 0.3)
Interest in sexual activity	39.7 (10.3)	52.4	0.1 (−0.1 to 0.4)	0.3 (−0.1 to 0.7)	0.0	0.0	0.0	−0.4 (−0.9 to 0.1)	−0.8 (−1.9 to 0.3)

^a All measures except for social isolation equate higher scores with improved functioning. A score of 50 corresponds to the average in the US general population, with a standard deviation (SD) ±10. CI = confidence interval; PROs = patient-reported outcomes; TFR = treatment-free remission; TKI = tyrosine kinase inhibitor.

^b Predicted mean change based on mixed models.

^c Model-estimated proportion of patients in TFR who reached a 3- or 5-point or greater change by 6 and 12 months.

Figure 1.

Model-predicted mean changes in Patient-Reported Outcomes Measurement Information System (PROMIS) functional domains. (Left) After tyrosine kinase inhibitor (TKI) discontinuation. (Right) After restarting TKIs. Higher scores indicate better health. Trajectories (black line) are shown with 95% confidence intervals (gray lines), and the distributions of patient scores over time are shown as box plots.

Social function, an integral component of well-being for patients with cancer (40,41), had the largest improvement during TFR. The better than average scores on social isolation may be related to increased social support given to those with cancer (14) and to those who participate in clinical trials (42). Conceivably, the reduction in symptoms, especially fatigue and diarrhea (30), may have helped patients to better participate in social activities, leading to the improvement in function.

Sexual function is an area not previously studied in patients with CML (13-17,43,44), although sexual dysfunction after cancer treatment is common and associated with impaired health-related quality of life (45). Although sexual satisfaction improved for a minority of patients in TFR, discontinuation was not associated with improvement in sexual interest. Given the dearth of information about sexual side effects of TKIs, future investigations should explore these findings.

Prior studies of physical function among patients with CML on TKIs have had mixed results (10,11,18,22), which may help explain the negligible improvement we observed during TFR. Likewise, prior research on cognition in CML patients shows inconsistent findings (19,20,22-24,46,47), though it is of utmost importance to patients with CML (48). Our study is the first, to our knowledge, to evaluate cognition in TFR, and only a small percentage of patients experienced notable improvement. This is a subject in need of further study, perhaps using cognitive testing.

The TFR population is unique among patients with CML, and prior research suggests a relationship between achieving a deep molecular response and PROM outcomes (17). It is plausible that patients eligible for discontinuation were better able to tolerate TKIs, an area not previously studied. There may also be other factors that contribute to who attempts TFR, because not all eligible patients choose to do so (49).

The chief limitation is this was a single-arm study that does not permit causal attributions to TFR and restarting TKI. Furthermore, there may be other variables that we did not evaluate that influenced PROM trajectories, for example, medical comorbidities. Finally, we used thresholds for interpreting the magnitude of change based on methodological work in other conditions.

The LAST study provides novel insight into the patient experience during TFR and highlights the decline in function after restarting TKI, a critical time for patients who attempt TFR. TFR is a primary goal of CML treatment by national guidelines (50,51). Information on patient function in TFR and TKI restart can help inform patient and clinician decision-making.

Funding

This work was supported by the National Cancer Institute at the National Institutes of Health (grant R01 CA184798 to E.A. and K.F.).

Notes

Role of the funder: The content is solely the responsibility of the authors. The funder had no role in the study design; collection, analysis, or interpretation of data; writing of the manuscript.

Disclosures: K.S.: Research support from the American Society of Hematology. E.A.: personal fees from Novartis, Bristol Myers Squibb, and Takeda and research support from Novartis and Takeda. J.C.: Grants (to his institution) and personal fees from Novartis, Pfizer, Sun Pharma and Takeda; and grants from Bristol Myers Squibb during the conduct of the study. M.D.: Received grants from the National Institutes of Health as a subcontract with the Medical College of Wisconsin during the conduct of the study; personal fees and research funding from Blueprint, Takeda, Novartis, and Incyte; personal fees from Medscape, Fusion Pharma, Sangamo, and DiserSol; and research funding from SPARC, Leukemia & Lymphoma Society, and Pfizer outside the submitted work. V.K.: Personal fees from Novartis and Pfizer outside the submitted work. R.L.: Consultant or advisor to Agios, Amgen, Ariad/Takeda, Astellas, Celgene/Bristol Myers Squibb, CVS/Caremark, Epizyme, MorphoSys, and Novartis, and has received clinical research support to his institution from Astellas, Celgene, Cellectis, Daiichi Sankyo, Forty Seven/Gilead, Novartis, Rafael Pharmaceuticals, and royalties from UpToDate. M.M: Received personal fees and research funding from Novartis Oncology and Bristol Myers Squibb; personal fees from Takeda and Pfizer; and research funding from Sun Pharma/SPARC outside the submitted work. V.O.: Research funding from Pfizer, Inc; consulting for Takeda, Bristol Myers Squibb; education outreach for American Society of Clinical Oncology Advantage. J.P-I.: Grants from the National Institutes of Health during the conduct of the study; and personal fees from Janssen, AbbVie, AstraZeneca, Novartis, Pfizer, and Takeda outside the submitted work. J.R.: Grants from the National Cancer Institute during the conduct of the study; and personal fees from Novartis, Bristol Myers Squibb, Takeda, Amgen, Cepheid, Genentech outside the submitted work. N.S.: received funding for the conduct of clinical research during the conduct of this study. R.S.: Speakers Bureau, Consultant PharmEssential, Consultant AbbVie. K.F.: Grants to her institution from NIH, Novartis, Incyte, and Jazz, and consulting fees from ReFocus for work unrelated to this manuscript. L.L., K.W., C.S., J.T.: no disclosures.

Author contributions: Conceptualization: K.F., K.W., E.A. Designed the analysis: K.F., L.L., K.W. Methodology: K.F., L.L., K.W. Data curation: K.F., L.L., K.W. Formal analysis: L.L., K.W. Project administration: E.A., K.F, J.C., M.D., V.K., R.L., M.M, V.O., J.P.-I., J.R., C.S., N.S., R.S., J.T. Funding acquisition: E.A., K.F. Supervision: K.F., K.W., E.A., J.C., M.D., V.K., R.L., M.M, V.O., J.P.-I., J.R., C.S., N.S., R.S., J.T. Writing—original draft: K.S., K.F. Writing—reviewing and editing: K.S., K.F., E.A., L.L., K.W., J.C., M.D., V.K., R.L., M.M, V.O., J.P.-I., J.R., C.S., N.S., R.S., J.T.

Acknowledgements: We would like to acknowledge the contributions of Ellen K. Ritchie of Weill Medical College of Cornell University and Martha Wadleigh of the Dana-Farber Cancer Institute, as well as the contributions of H. Jean Khoury, MD to the conceptualization and operationalization of this research prior to his death.

Disclaimer: The content is solely the responsibility of the authors and does not represent the official views of the National Cancer Institute.

Data Availability

The data underlying this article cannot be shared publicly due to the privacy of individuals that participated in the study. The data will be shared on reasonable request to the corresponding author.