icMDS recommendations for avoiding restrictive eligibility criteria
| Category . | Recommendation . | Background . |
|---|---|---|
| Upper age limits | Upper age limits should be avoided | Several trials set upper age limits in the eligibility criteria for this analysis. The icMDS asserts that an upper age limit should be avoided given MDS is a disease mostly of older adults, and this restriction profoundly affects the generalizability of results. Not all pharmacotherapeutic outcomes that can occur in older patients with MDS can be predicted from trials with upper age restrictions. Furthermore, any exclusion of patients based solely on age is arbitrary because chronological age is not an adequate representation of biological age. A prior study evaluating a pooled patient database for patients with MDS found that patients who participated in trials were significantly younger than nonparticipants.22 Instead of imposing strict upper age limitations on trials, the icMDS encourages clinicians to use objective geriatric assessments for frailty and consideration of physiological age when assessing clinical trial candidacy for older adult patients. |
| Expected survival | Minimum expected survival should be avoided with instead prioritization of investigator judgment on whether a patient has the potential to benefit from the trial | Many studies in this analysis had eligibility criteria that required a certain minimum anticipated expected survival; the icMDS recommends avoiding this eligibility criterion. Instead, we recommend prioritizing investigator judgment of whether the patient may potentially benefit from the trial (eg, if the patient has another imminently terminal illness, then he or she is less likely to have the opportunity to benefit from a trial). Patients with MDS have limited therapeutic options resulting in a shorter life expectancy, and this generally should not prohibit them from potentially life-prolonging therapies. Indeed, post-HMA MDS has a life expectancy <6 months, but remains one of the greatest unmet treatment needs. It is well-supported in the medical community that life expectancy is an unnecessary and antiquated criterion because a judgment of life expectancy may be subjective and often incorrect. Other metrics including functional assessments may be more relevant than life expectancy when selecting qualifying study participants. We acknowledge that other terminal illnesses may preclude potential benefits from participation a clinical trial (particularly in patients with LR-MDS) and defer to investigator judgment on enrollment in these scenarios. |
| Lower limits for platelets in HR-MDS trials | Arbitrary lower limits on platelets threshold should be avoided in trials for HR-MDS | A few studies focused on HR-MDS in this review exclude patients with platelet levels lower than a certain value (most commonly <50 × 103/μL). Patients with hematologic malignancies are expected to have hematologic abnormalities at study entry, particularly in HR-MDS. Inclusion criteria for trials focused on HR-MDS should avoid arbitrary lower limits for platelets. This is excessively restrictive given that the disease process itself is characterized by cytopenias. Lower limits for platelets may be more appropriate in LR-MDS trials depending on the therapy; however, we encourage avoiding lower limits in LR-MDS particularly in therapeutics without expected myelosuppression. We encourage language for appropriate transfusion support during the trial. |
| Lower limits for hemoglobin | Arbitrary lower limits on hemoglobin thresholds should be avoided | A few studies in this review excluded patients with hemoglobin levels lower than a certain value (most commonly <9 g/dL). Patients with hematologic malignancies are expected to have hematologic abnormalities at study entry, particularly in bone marrow failure diseases such as MDS. Inclusion criteria should avoid arbitrary lower limits for hemoglobin. This is excessively restrictive given that the disease process itself is characterized by cytopenias. |
| Lower limits for absolute neutrophil count in HR-MDS trials | Arbitrary lower limits on absolute neutrophil count should be avoided in trials for HR-MDS | A few trials in this review focused on HR-MDS exclude patients with an absolute neutrophil count lower than a certain value (most commonly <500 x 103/μL). Patients with hematologic malignancies are expected to have hematologic abnormalities at study entry, particularly in HR-MDS. Inclusion criteria should avoid arbitrary lower limits for absolute neutrophil count. We acknowledge that consideration should be given to the specific therapy, and limits may be applicable to therapies with a particularly significant risk of neutropenia. However, absolute neutrophil count limits are often excessively restrictive given that the disease process itself is characterized by cytopenias. We similarly encourage avoiding lower limits in LR-MDS particularly in therapeutics without expected myelosuppression. Inclusion of patients should have language for appropriate supportive care and prophylaxis as a part of the trial. |
| No. of prior therapies | Requiring a specific number of prior therapies (such as >2 lines of therapy) should be avoided | Several trials in this analysis restrict patients based on the number of lines of prior therapies (such as >2 lines of therapy). Instead, we recommend that criteria should indicate specific prior exposure as appropriate; for example, it can be a criterion that all patients are required to have prior treatment with an HMA regardless of number of lines of therapy. The icMDS recommends against requirements for a specific number of prior therapies and instead focusing on whether or not there has been exposure to a prior class of therapy or a biosimilar, as relevant to the MDS subtype and risk. |
| Transplant candidacy | Potential candidacy for stem cell transplant should be avoided (unless study is peritransplant in design) | Many trials in this review incorporate criteria on transplant status. In April 2022, the FDA convened a panel of experts in MDS to discuss approaches to improve MDS drug development. This panel noted that a patient’s potential for HSCT may be variably interpreted, which can lead to unnecessary exclusion in clinical trials.23 The icMDS similarly advocates that transplant candidacy should not be part of trial exclusion criteria and may only be relevant when a study occurs peritransplant by design. This is of interest because imbalance in enrollment between patients who are viewed as potentially transplant eligible and transplant ineligible may impact outcomes in clinical trials. |
| Renal and hepatic laboratory values in phase 2 and 3 trials | Renal and hepatic laboratory parameters without safety signals should be avoided in later-phase trials. We recommend use of the MDRD equation or CKD-EPI formula in clinical trials for MDS | Both liver function and renal function tests were common eligibility criteria in this review, and concordance with drug safety signals was low. The icMDS stresses that laboratory parameters should be avoided in phase 2 and 3 studies once preliminary adverse events are understood in MDS and do not show safety signals. The icMDS urges concordance between renal and hepatic laboratory values in eligibility criteria and drug safety signals. For renal function, we recommend use of MDRD equation or the CKD-EPI formula instead of the C-G equation in MDS clinical trials to best represent true kidney function in older adults. For bilirubin, specific eligibility criteria for Gilbert syndrome should be defined in the protocol.24 In most cases of Gilbert syndrome, bilirubin levels are <3 mg/dL.25 |
| EF | Arbitrary EF limits that are not concordant with drug safety signals should be avoided | If an investigational therapy is not known to pose cardiac risks, arbitrary EF values should not be used to exclude patients from clinical trials. Instead of arbitrary EF limits that are not concordant with drug safety signals, trials should recommend investigator assessment of a potential participant’s risk for heart failure with a validated clinical classification system (eg, the NYHA functional classification). |
| QTc interval | If QTc prolongation is not a safety risk of the drug in early studies, then QTc limits in later-phase trials should be avoided | The icMDS agrees with recommendations from the FDA that suggests exclusion criteria on QT/QTc interval until the effects of the drug on the QT/QTc interval have been characterized.26 If QTc prolongation is not a safety risk of the drug in early studies, then QTc limits should not be imposed in later-phase studies. When QTc limits are used, we recommend use of well-characterized QT correction such as the Fridericia formula. |
| Uncontrolled hypertension | Statements on uncontrolled hypertension with arbitrary blood pressure ranges should be avoided | Several studies include uncontrolled hypertension as an exclusion criterion in this analysis and the range of blood pressures, that is, considered uncontrolled is widely variable. In the studies of drugs known to increase blood pressure, the icMDS recommends eliminating arbitrary blood pressure limitations with instead focusing on the prevention of “hypertensive urgency,” defined as a systolic blood pressure of ≥180 mm Hg or diastolic blood pressure of ≥110 mm Hg, to prevent sequelae. |
| Other malignancies | Restrictions on prior or concurrent malignancies should be avoided when the malignancy is stable and the risk of interfering with either safety or efficacy end points is low | Many studies in this analysis required patients to be 3 or 5 years from a prior cancer diagnosis. ASCO-Friends and the FDA Administration Working Group noted that patients with prior or concurrent malignancies should be permitted when the risk of the malignancy interfering with either safety or efficacy end points is low. The icMDS agrees with this statement. This is particularly relevant to MDS, because the presence of other malignancies may be directly related to the diagnosis of MDS. Indeed, therapy-related MDS (t-MDS) comprises 10% to 20% of all MDS diagnoses, because of, in part, clonal hematopoietic selection by genotoxic chemotherapy or radiotherapy. However, only 5.7% of trials for MDS include t-MDS. Patients with t-MDS, whether from therapies from other malignancies or from cytotoxic therapies such as those given for autoimmune disease, should not be excluded if other inclusion/exclusion criteria are met. The icMDS asserts that patients with concurrent malignancy should generally be included if the concurrent cancer is clinically stable and not requiring active tumor-directed therapy within the anticipated study period. This does require consideration of the nature of the particular malignancy. Current malignancies typically require more specific eligibility criteria than prior malignancies. |
| Chronic viral diseases | Broad, blanket statements on the exclusion of HIV and hepatitis without more specific criteria (as recommended per the FDA, ASCO-Friends) should be avoided | As shown by this review, eligibility criteria can disproportionately exclude potential participants from historically marginalized groups including those with chronic infections such as HIV and hepatitis. Cancer is now a leading cause of mortality in people with HIV; however, many cancer studies exclude this population. The icMDS recommends against broad, blanket statements on the exclusion of HIV and hepatitis without more specific criteria. The icMDS agrees with recommendations from ASCO-Friends and the FDA working group that patients with HIV who have low risk of AIDS-related outcomes and are compliant with standard antiretroviral therapy should be included in trials, understanding some drug interactions may impact eligibility but would be accounted for elsewhere. Eligibility criteria regarding patients known to have HIV should be evaluated in context of active therapy, CD4 count, and viral load measurements, rather than a separate criterion that may unnecessarily restrict patients with cytopenias from MDS. There can be language suggesting discussions with the medical monitor of each trial. Eligibility criteria regarding low CD4 count should be specifically associated with active HIV/AIDS rather than a separate criterion that may unnecessarily restrict patients with cytopenias from MDS. For patients with hepatitis B and C, the icMDS agrees with the criteria developed by the FDA. Active hepatitis B necessitating therapy should require the patient to be on a suppressive antiviral before the initiation of cancer therapy. Patients with a history of hepatitis C should have completed or be receiving curative antiviral treatment with a viral load that is undetectable. For incurable cancers, active hepatitis C should be included if hepatitis C viral load is stable, and the investigational agent is not suspected to exacerbate hepatitis. |
| Washout periods | This analysis showed a wide range of washout periods for the same therapeutic agents throughout MDS clinical trials. Generally, relevant clinical and laboratory parameters should be used preferentially in place of time-based washout periods to address safety considerations. This is particularly true for preceding cytotoxic chemotherapy or investigational agents with an unknown half-life. If time-based washout periods are included, pharmacological justification should be specified. Washout periods based on 5 half-lives may be appropriate for agents such as ESAs. For preceding immunotherapy, late occurring immune related adverse events (irAEs) can occur. However, in the absence of unresolved irAEs, it is recommended that the patient is eligible for trials with an understanding that late effects from prior immunotherapy may occur. An understanding of the common late occurring irAEs can be used to differentiate if effects are from prior or current therapy. |
| Category . | Recommendation . | Background . |
|---|---|---|
| Upper age limits | Upper age limits should be avoided | Several trials set upper age limits in the eligibility criteria for this analysis. The icMDS asserts that an upper age limit should be avoided given MDS is a disease mostly of older adults, and this restriction profoundly affects the generalizability of results. Not all pharmacotherapeutic outcomes that can occur in older patients with MDS can be predicted from trials with upper age restrictions. Furthermore, any exclusion of patients based solely on age is arbitrary because chronological age is not an adequate representation of biological age. A prior study evaluating a pooled patient database for patients with MDS found that patients who participated in trials were significantly younger than nonparticipants.22 Instead of imposing strict upper age limitations on trials, the icMDS encourages clinicians to use objective geriatric assessments for frailty and consideration of physiological age when assessing clinical trial candidacy for older adult patients. |
| Expected survival | Minimum expected survival should be avoided with instead prioritization of investigator judgment on whether a patient has the potential to benefit from the trial | Many studies in this analysis had eligibility criteria that required a certain minimum anticipated expected survival; the icMDS recommends avoiding this eligibility criterion. Instead, we recommend prioritizing investigator judgment of whether the patient may potentially benefit from the trial (eg, if the patient has another imminently terminal illness, then he or she is less likely to have the opportunity to benefit from a trial). Patients with MDS have limited therapeutic options resulting in a shorter life expectancy, and this generally should not prohibit them from potentially life-prolonging therapies. Indeed, post-HMA MDS has a life expectancy <6 months, but remains one of the greatest unmet treatment needs. It is well-supported in the medical community that life expectancy is an unnecessary and antiquated criterion because a judgment of life expectancy may be subjective and often incorrect. Other metrics including functional assessments may be more relevant than life expectancy when selecting qualifying study participants. We acknowledge that other terminal illnesses may preclude potential benefits from participation a clinical trial (particularly in patients with LR-MDS) and defer to investigator judgment on enrollment in these scenarios. |
| Lower limits for platelets in HR-MDS trials | Arbitrary lower limits on platelets threshold should be avoided in trials for HR-MDS | A few studies focused on HR-MDS in this review exclude patients with platelet levels lower than a certain value (most commonly <50 × 103/μL). Patients with hematologic malignancies are expected to have hematologic abnormalities at study entry, particularly in HR-MDS. Inclusion criteria for trials focused on HR-MDS should avoid arbitrary lower limits for platelets. This is excessively restrictive given that the disease process itself is characterized by cytopenias. Lower limits for platelets may be more appropriate in LR-MDS trials depending on the therapy; however, we encourage avoiding lower limits in LR-MDS particularly in therapeutics without expected myelosuppression. We encourage language for appropriate transfusion support during the trial. |
| Lower limits for hemoglobin | Arbitrary lower limits on hemoglobin thresholds should be avoided | A few studies in this review excluded patients with hemoglobin levels lower than a certain value (most commonly <9 g/dL). Patients with hematologic malignancies are expected to have hematologic abnormalities at study entry, particularly in bone marrow failure diseases such as MDS. Inclusion criteria should avoid arbitrary lower limits for hemoglobin. This is excessively restrictive given that the disease process itself is characterized by cytopenias. |
| Lower limits for absolute neutrophil count in HR-MDS trials | Arbitrary lower limits on absolute neutrophil count should be avoided in trials for HR-MDS | A few trials in this review focused on HR-MDS exclude patients with an absolute neutrophil count lower than a certain value (most commonly <500 x 103/μL). Patients with hematologic malignancies are expected to have hematologic abnormalities at study entry, particularly in HR-MDS. Inclusion criteria should avoid arbitrary lower limits for absolute neutrophil count. We acknowledge that consideration should be given to the specific therapy, and limits may be applicable to therapies with a particularly significant risk of neutropenia. However, absolute neutrophil count limits are often excessively restrictive given that the disease process itself is characterized by cytopenias. We similarly encourage avoiding lower limits in LR-MDS particularly in therapeutics without expected myelosuppression. Inclusion of patients should have language for appropriate supportive care and prophylaxis as a part of the trial. |
| No. of prior therapies | Requiring a specific number of prior therapies (such as >2 lines of therapy) should be avoided | Several trials in this analysis restrict patients based on the number of lines of prior therapies (such as >2 lines of therapy). Instead, we recommend that criteria should indicate specific prior exposure as appropriate; for example, it can be a criterion that all patients are required to have prior treatment with an HMA regardless of number of lines of therapy. The icMDS recommends against requirements for a specific number of prior therapies and instead focusing on whether or not there has been exposure to a prior class of therapy or a biosimilar, as relevant to the MDS subtype and risk. |
| Transplant candidacy | Potential candidacy for stem cell transplant should be avoided (unless study is peritransplant in design) | Many trials in this review incorporate criteria on transplant status. In April 2022, the FDA convened a panel of experts in MDS to discuss approaches to improve MDS drug development. This panel noted that a patient’s potential for HSCT may be variably interpreted, which can lead to unnecessary exclusion in clinical trials.23 The icMDS similarly advocates that transplant candidacy should not be part of trial exclusion criteria and may only be relevant when a study occurs peritransplant by design. This is of interest because imbalance in enrollment between patients who are viewed as potentially transplant eligible and transplant ineligible may impact outcomes in clinical trials. |
| Renal and hepatic laboratory values in phase 2 and 3 trials | Renal and hepatic laboratory parameters without safety signals should be avoided in later-phase trials. We recommend use of the MDRD equation or CKD-EPI formula in clinical trials for MDS | Both liver function and renal function tests were common eligibility criteria in this review, and concordance with drug safety signals was low. The icMDS stresses that laboratory parameters should be avoided in phase 2 and 3 studies once preliminary adverse events are understood in MDS and do not show safety signals. The icMDS urges concordance between renal and hepatic laboratory values in eligibility criteria and drug safety signals. For renal function, we recommend use of MDRD equation or the CKD-EPI formula instead of the C-G equation in MDS clinical trials to best represent true kidney function in older adults. For bilirubin, specific eligibility criteria for Gilbert syndrome should be defined in the protocol.24 In most cases of Gilbert syndrome, bilirubin levels are <3 mg/dL.25 |
| EF | Arbitrary EF limits that are not concordant with drug safety signals should be avoided | If an investigational therapy is not known to pose cardiac risks, arbitrary EF values should not be used to exclude patients from clinical trials. Instead of arbitrary EF limits that are not concordant with drug safety signals, trials should recommend investigator assessment of a potential participant’s risk for heart failure with a validated clinical classification system (eg, the NYHA functional classification). |
| QTc interval | If QTc prolongation is not a safety risk of the drug in early studies, then QTc limits in later-phase trials should be avoided | The icMDS agrees with recommendations from the FDA that suggests exclusion criteria on QT/QTc interval until the effects of the drug on the QT/QTc interval have been characterized.26 If QTc prolongation is not a safety risk of the drug in early studies, then QTc limits should not be imposed in later-phase studies. When QTc limits are used, we recommend use of well-characterized QT correction such as the Fridericia formula. |
| Uncontrolled hypertension | Statements on uncontrolled hypertension with arbitrary blood pressure ranges should be avoided | Several studies include uncontrolled hypertension as an exclusion criterion in this analysis and the range of blood pressures, that is, considered uncontrolled is widely variable. In the studies of drugs known to increase blood pressure, the icMDS recommends eliminating arbitrary blood pressure limitations with instead focusing on the prevention of “hypertensive urgency,” defined as a systolic blood pressure of ≥180 mm Hg or diastolic blood pressure of ≥110 mm Hg, to prevent sequelae. |
| Other malignancies | Restrictions on prior or concurrent malignancies should be avoided when the malignancy is stable and the risk of interfering with either safety or efficacy end points is low | Many studies in this analysis required patients to be 3 or 5 years from a prior cancer diagnosis. ASCO-Friends and the FDA Administration Working Group noted that patients with prior or concurrent malignancies should be permitted when the risk of the malignancy interfering with either safety or efficacy end points is low. The icMDS agrees with this statement. This is particularly relevant to MDS, because the presence of other malignancies may be directly related to the diagnosis of MDS. Indeed, therapy-related MDS (t-MDS) comprises 10% to 20% of all MDS diagnoses, because of, in part, clonal hematopoietic selection by genotoxic chemotherapy or radiotherapy. However, only 5.7% of trials for MDS include t-MDS. Patients with t-MDS, whether from therapies from other malignancies or from cytotoxic therapies such as those given for autoimmune disease, should not be excluded if other inclusion/exclusion criteria are met. The icMDS asserts that patients with concurrent malignancy should generally be included if the concurrent cancer is clinically stable and not requiring active tumor-directed therapy within the anticipated study period. This does require consideration of the nature of the particular malignancy. Current malignancies typically require more specific eligibility criteria than prior malignancies. |
| Chronic viral diseases | Broad, blanket statements on the exclusion of HIV and hepatitis without more specific criteria (as recommended per the FDA, ASCO-Friends) should be avoided | As shown by this review, eligibility criteria can disproportionately exclude potential participants from historically marginalized groups including those with chronic infections such as HIV and hepatitis. Cancer is now a leading cause of mortality in people with HIV; however, many cancer studies exclude this population. The icMDS recommends against broad, blanket statements on the exclusion of HIV and hepatitis without more specific criteria. The icMDS agrees with recommendations from ASCO-Friends and the FDA working group that patients with HIV who have low risk of AIDS-related outcomes and are compliant with standard antiretroviral therapy should be included in trials, understanding some drug interactions may impact eligibility but would be accounted for elsewhere. Eligibility criteria regarding patients known to have HIV should be evaluated in context of active therapy, CD4 count, and viral load measurements, rather than a separate criterion that may unnecessarily restrict patients with cytopenias from MDS. There can be language suggesting discussions with the medical monitor of each trial. Eligibility criteria regarding low CD4 count should be specifically associated with active HIV/AIDS rather than a separate criterion that may unnecessarily restrict patients with cytopenias from MDS. For patients with hepatitis B and C, the icMDS agrees with the criteria developed by the FDA. Active hepatitis B necessitating therapy should require the patient to be on a suppressive antiviral before the initiation of cancer therapy. Patients with a history of hepatitis C should have completed or be receiving curative antiviral treatment with a viral load that is undetectable. For incurable cancers, active hepatitis C should be included if hepatitis C viral load is stable, and the investigational agent is not suspected to exacerbate hepatitis. |
| Washout periods | This analysis showed a wide range of washout periods for the same therapeutic agents throughout MDS clinical trials. Generally, relevant clinical and laboratory parameters should be used preferentially in place of time-based washout periods to address safety considerations. This is particularly true for preceding cytotoxic chemotherapy or investigational agents with an unknown half-life. If time-based washout periods are included, pharmacological justification should be specified. Washout periods based on 5 half-lives may be appropriate for agents such as ESAs. For preceding immunotherapy, late occurring immune related adverse events (irAEs) can occur. However, in the absence of unresolved irAEs, it is recommended that the patient is eligible for trials with an understanding that late effects from prior immunotherapy may occur. An understanding of the common late occurring irAEs can be used to differentiate if effects are from prior or current therapy. |
ASCO, American Society of Clinical Oncology; C-G, Cockcroft-Gault; CKD-EPI, Chronic Kidney Disease Epidemiology Collaboration; ESAs, erythropoiesis-stimulating agents; HMA, hypomethylating agent; HSCT, hematopoietic stem cell transplant; MDRD, Modification of Diet in Renal Disease; NYHA, New York Heart Association.