Proposed definition of clinical trial end points and biomarkers suggestive of disease-modifying activity
| . | Proposed definition . | Advantages . | Disadvantages and controversies . | Ref . |
|---|---|---|---|---|
| Time-to-event end points | ||||
| Thrombosis-free survival | Time from randomization (or study entry) to the date of the first major thrombosis (defined as stroke, acute coronary syndrome, transient ischemic attack, pulmonary embolism, abdominal thrombosis, deep-vein thrombosis, or peripheral arterial thrombosis; should be reported as composite and separately for arterial vs venous events) or death from any cause | Clinically relevant outcome because thrombotic events drive morbidity and mortality in PV Objective and measurable Treatment (ie, hematocrit control) has been shown to reduce risk of thrombosis | Low event rates require extended study duration Variable definition across clinical trials | 4,25,27 |
| Myelofibrosis-free survival | Time from randomization (or study entry) to the date of progression to myelofibrosis or death from any cause | Association with OS and symptom burden Objective assessment possible | Low event rates Rates depend on other factors (eg, comutations) Hematologic response may not correlate with progression to myelofibrosis | 50,100 |
| Leukemia-free survival | Time from randomization (or study entry) to the date of progression to AML or myelodysplastic syndrome or death from any cause | Association with OS Objective assessment possible | Depend on other factors (eg, comutations) Low event rates | 25 |
| OS | Time from randomization (or study entry) to the date of death from any cause | Objective assessment Improving OS as the most robust end point | Low event rates | 50 |
| Composite end points | ||||
| Progression-free survival | Time from randomization (or study entry) to the date of progression to myelofibrosis, AML or myelodysplastic syndrome, or death from any cause | Association with OS Objective assessment possible | Variable definitions across clinical trials | 25 |
| EFS | Time from randomization (or study entry) to the date of first major thrombotic event or progression to myelofibrosis, AML or myelodysplastic syndrome, or death from any cause | Captures all major contributors to PV-related morbidity and mortality Earlier readout than individual components Objective assessment possible | Variable definitions across clinical trials Not every component of the composite end point has the same implications | 25 |
| Clinical trial end points | ||||
| Possibly reflecting disease-modifying effects | ||||
| Complete hematologic response | Defined per ELN 2013 response criteria as durable (≥12 wk) peripheral blood count remission, defined as hematocrit level <45% without phlebotomies; platelet count ≤400 × 109/L, WBC count <10 × 109/L | Most commonly used end point in clinical trials Normalization of WBC and platelet count potentially reduces risk of progression to myelofibrosis and thrombosis Shorter time to trial readout | Controversy whether hematologic response correlates with risk of thrombosis | 17,21,36,81 |
| Complete response per ELN2013 criteria | Defined per ELN2013 response criteria | Composite of clinical and pathologic response Normalization of CBC associated with survival and thrombosis risk | Requires repeat bone marrow biopsy for response | 27,81 |
| Molecular response rate | Reduction in JAK2V617F allele | JAK2 allele burden is correlated with risk of venous thrombosis Molecular responses are associated with lower risk of disease progression Early readout | Requires additional prospective validation and standardization Uncertainty regarding optimal threshold Differences between arterial and venous events The threshold by which JAK2V617F VAF needs to be reduced to associate with clinical benefit remains to be defined | 16,25 |
| Clonal evolution | Acquisition of additional somatic mutations or cytogenetic abnormalities on serial bone marrow evaluations | Early readout Acquisition of certain comutations associated with higher risk of progression | Requires additional prospective validation and standardization | 100,101 |
| Symptom control | Defined per ELN2013 response criteria as reduction in MPN-SAF by ≥10 points | Patient centered Higher symptom burden is associated with quality of life in other domains and rates of depression | Unclear correlation with OS, thrombosis, and disease progression Uncertainty regarding optimal assessment tool (EQL-5D, MPN-SAF) The optimal threshold for symptom improvement is unclear | 45,81,102-104 |
| . | Proposed definition . | Advantages . | Disadvantages and controversies . | Ref . |
|---|---|---|---|---|
| Time-to-event end points | ||||
| Thrombosis-free survival | Time from randomization (or study entry) to the date of the first major thrombosis (defined as stroke, acute coronary syndrome, transient ischemic attack, pulmonary embolism, abdominal thrombosis, deep-vein thrombosis, or peripheral arterial thrombosis; should be reported as composite and separately for arterial vs venous events) or death from any cause | Clinically relevant outcome because thrombotic events drive morbidity and mortality in PV Objective and measurable Treatment (ie, hematocrit control) has been shown to reduce risk of thrombosis | Low event rates require extended study duration Variable definition across clinical trials | 4,25,27 |
| Myelofibrosis-free survival | Time from randomization (or study entry) to the date of progression to myelofibrosis or death from any cause | Association with OS and symptom burden Objective assessment possible | Low event rates Rates depend on other factors (eg, comutations) Hematologic response may not correlate with progression to myelofibrosis | 50,100 |
| Leukemia-free survival | Time from randomization (or study entry) to the date of progression to AML or myelodysplastic syndrome or death from any cause | Association with OS Objective assessment possible | Depend on other factors (eg, comutations) Low event rates | 25 |
| OS | Time from randomization (or study entry) to the date of death from any cause | Objective assessment Improving OS as the most robust end point | Low event rates | 50 |
| Composite end points | ||||
| Progression-free survival | Time from randomization (or study entry) to the date of progression to myelofibrosis, AML or myelodysplastic syndrome, or death from any cause | Association with OS Objective assessment possible | Variable definitions across clinical trials | 25 |
| EFS | Time from randomization (or study entry) to the date of first major thrombotic event or progression to myelofibrosis, AML or myelodysplastic syndrome, or death from any cause | Captures all major contributors to PV-related morbidity and mortality Earlier readout than individual components Objective assessment possible | Variable definitions across clinical trials Not every component of the composite end point has the same implications | 25 |
| Clinical trial end points | ||||
| Possibly reflecting disease-modifying effects | ||||
| Complete hematologic response | Defined per ELN 2013 response criteria as durable (≥12 wk) peripheral blood count remission, defined as hematocrit level <45% without phlebotomies; platelet count ≤400 × 109/L, WBC count <10 × 109/L | Most commonly used end point in clinical trials Normalization of WBC and platelet count potentially reduces risk of progression to myelofibrosis and thrombosis Shorter time to trial readout | Controversy whether hematologic response correlates with risk of thrombosis | 17,21,36,81 |
| Complete response per ELN2013 criteria | Defined per ELN2013 response criteria | Composite of clinical and pathologic response Normalization of CBC associated with survival and thrombosis risk | Requires repeat bone marrow biopsy for response | 27,81 |
| Molecular response rate | Reduction in JAK2V617F allele | JAK2 allele burden is correlated with risk of venous thrombosis Molecular responses are associated with lower risk of disease progression Early readout | Requires additional prospective validation and standardization Uncertainty regarding optimal threshold Differences between arterial and venous events The threshold by which JAK2V617F VAF needs to be reduced to associate with clinical benefit remains to be defined | 16,25 |
| Clonal evolution | Acquisition of additional somatic mutations or cytogenetic abnormalities on serial bone marrow evaluations | Early readout Acquisition of certain comutations associated with higher risk of progression | Requires additional prospective validation and standardization | 100,101 |
| Symptom control | Defined per ELN2013 response criteria as reduction in MPN-SAF by ≥10 points | Patient centered Higher symptom burden is associated with quality of life in other domains and rates of depression | Unclear correlation with OS, thrombosis, and disease progression Uncertainty regarding optimal assessment tool (EQL-5D, MPN-SAF) The optimal threshold for symptom improvement is unclear | 45,81,102-104 |
CBC, complete blood count; ELN2013, European LeukemiaNet 2013; EQL-5D, European Quality of Life–5 dimensions; MPN-SAF, Myeloproliferative Neoplasm–Symptom Assessment Form; Ref, reference; WBC, white blood cell.