HOPS is a prospective, multicenter, randomized trial that aims to evaluate whether initial dosing of hydroxyurea using a novel PK-guided dosing strategy for children with SCA results in higher %HbF at 6 months compared to standard weight-based initial dosing with step-wise dose escalation. The multicenter design also allows for the validation of PK sample collection in young children with centralized PK analysis and dose selection. Participants are randomized in a 1:1 ratio to initially receive either a 20 mg/kg weight-based (Standard Arm) or a PK-guided (Alternative Arm) starting dose of hydroxyurea. Following initiation, a study-designed hydroxyurea dosing protocol will be used to escalate or reduce the initial hydroxyurea dose based on the laboratory data that will be collected through the time of the primary endpoint at month 6; the total study period is 12 months. Figure 1 outlines the schedule of events for participants in the study.
The primary objective of the study is to evaluate whether a PK-guided starting hydroxyurea dose results in a higher %HbF compared to standard weight-based (20 mg/kg) initial dosing for children with SCA. Secondary aims include careful investigation of the clinical, laboratory, and molecular determinants of the maximal hydroxyurea-induced HbF responses, as well as studies investigating changes in gene expression and regulation related to hydroxyurea starting dose and age.
Study participants are recruited from 11 pediatric sickle cell centers across the United States (Figure 2). Most study sites were selected due to their involvement in the Sickle Treatment and Outcomes Research in the Midwest (STORM) regional network, led by CCHMC and established to improve care and outcomes for individuals with SCD living in Indiana, Illinois, Michigan, Minnesota, North Dakota, Ohio, South Dakota, and Wisconsin.30 Additional study sites were included to ensure adequate enrollment and were selected based on previous collaborative relationships and investigator interest. Prior to the initiation of the study, there was significant variability in the number of patients (30 to greater than 500) and the proportion of children prescribed hydroxyurea (25-90%) at each site, mimicking the distribution of these patients across the United States and the known variability in hydroxyurea utilization. Prior to formal site selection, potential study sites completed a feasibility survey to assess patient volume, current hydroxyurea use, research capacity, and anticipated study enrollment. Subsequently, an in-person or virtual site training/initiation visit was performed, including review of study rationale and procedures as well as a comprehensive overview on the use of hydroxyurea therapy and sharing of the results from the TREAT trial. These site visits were well-received and allowed each study team to understand the rationale and strategy for dosing in the HOPS trial, which was different than the previous dosing strategies used for patients with SCA at these centers. Potential barriers to the recruitment of study participants or to the performance of study-related procedures were also reviewed to optimize the chance of smooth study success at each site.
Children with confirmed SCA (HbSS, HbSD, Hbβ0-thalassemia or other similarly severe phenotypes) who are initiating hydroxyurea therapy, following a discussion with their local clinical team, are eligible for study consideration. As each of these SCA genotypes is considered to have a similar phenotype and because we expect most enrolled participants to have HbSS disease, there will be no attempts to balance distribution of genotypes across the two study arms.
- Diagnosis of SCA (HbSS, HbSD, HbS/β0-thalassemia, or similarly severe SCA genotype)
- Age 6 months to 21 years at the time of enrollment
- Clinical decision by patient, family, and healthcare providers to initiate hydroxyurea therapy
- Current treatment with chronic, monthly blood transfusions or erythrocytapheresis. Of note, there are no restrictions regarding enrollment as to recent single blood transfusions. HbF is calculated as HbF/(HbF+HbS) to account for the presence of HbA. Children who are transitioning from chronic transfusions to hydroxyurea therapy are not eligible due to the difficulty in evaluating response with the overlap period of simultaneous transfusion therapy and hydroxyurea.
- Treatment with hydroxyurea within the past 3 months
- Hemoglobin SC disease, HbS/β+-thalassemia
- Current treatment with other investigational sickle cell medications
Recruitment and Enrollment Procedures
The study recruits patients who have decided to initiate hydroxyurea therapy based upon clinical indications and shared decision-making between the providers and the family. Informed consent is obtained by investigators or local research personnel at each study site. For children under 18 years of age, the parent or legal guardian provides written informed consent to join the study at the time of enrollment and older children sign assent as required by their local IRB. Participants who are 18 years and older sign informed consent themselves. In addition to informed consent to participate in the trial, additional consent is obtained for the storage of biological specimens for subsequent analysis.
Pharmacokinetics Studies and Determination of Hydroxyurea Dose
Once informed consent is obtained, all participants have a baseline PK visit. At this visit, participants take a single 20 mg/kg dose of liquid hydroxyurea prepared at each study site and PK microsamples are collected at 3 time points (15 minutes, 60 minutes, and 180 minutes) as described below. The liquid formulation is used for the PK studies to allow for a precise 20 mg/kg dose, but older participants are allowed to take capsules if they choose once hydroxyurea is prescribed. After the baseline visit, the participant does not start hydroxyurea until a study-determined starting dose is established and prescribed by their local provider, typically within 1-2 weeks of the PK visit. PK samples are shipped on dry ice to the central laboratory at CCHMC for measurement of hydroxyurea concentrations, determination of PK curve and area under the concentration-time curve (AUC), calculation of starting dose options for both arms, and randomization. Randomization results are blinded to the PI and staff involved in the recruitment and management of study participants until the study is complete.
Below, we describe several novel features of the study, including sparse PK sampling using microsampling devices, novel methods of measuring hydroxyurea concentrations, and determination of the optimal PK-guided dose for each individual participant.
Sparse PK Sampling
Traditional PK sampling requires collection of 1-3 mL of venous blood at many time points over several drug half-lives. This collection frequency and relatively large-volume venous blood draws over 8-12 hours is not practical in a clinical setting, particularly for infants and young children, notably the difficulties and intolerance of frequent venous blood draws in very young children and the inconvenience of having to remain in the hospital/clinic setting for a long period of time. Through the TREAT study, using historical data,23 we developed a population PK model and a sparse sampling strategy that accurately estimates hydroxyurea drug exposure using only three hydroxyurea concentrations measured at optimally designed times: 15 minutes, 60 minutes, and 3 hours after hydroxyurea administration.31 The number and timing of sample collection was selected based on known PK patters such that an accurate estimation of exposure could be made. Additionally, as we hope for this PK-guided dosing strategy to be ultimately clinically feasible, we found that the collection of 3 samples over 3 hours was acceptable to both families and feasible within a clinical setting. The TREAT cohort demonstrated the feasibility and safety of this sampling strategy in young children and older adolescent/young adults at a single center.29 The PK sampling strategy was an important feature of TREAT that resulted in high rates of enrollment with >90% of children with SCA who initiated hydroxyurea during the study period agreeing to participate.
Microsampling and Measurement of Hydroxyurea Concentrations
There is no widely established method or commercially available technique for measuring hydroxyurea concentrations in biological samples, but several new and accurate techniques have been developed.32 Our novel HPLC-based assay, requiring 0.5-1.0 mL of blood per time point, was the primary assay used in the TREAT study,29 and was a significant improvement from the previously used colorimetric assay,which required 1-2 mL per time point.33,34 We have since miniaturized the hydroxyurea assay further through the development of a highly sensitive and accurate tandem mass spectrometry-based assay (LC-MS/MS) for the quantitative measurement of hydroxyurea, requiring even smaller volumes of blood.35 Blood collection occurs using novel Volumetric Absorption Microsampling (VAMS) devices (Neoteryx, LLC, Torrance, CA), which store exactly 10 µL and samples can be collected by finger stick or heel stick, which is much preferred compared to venous sampling, for young children. Figure 3 illustrates the microsampling collection process using these VAMS devices.
To validate this novel PK microsampling method, we compared hydroxyurea concentrations and AUC measurements using the established HPLC assay to the new LC-MS/MS assay. Among 80 samples from 23 TREAT participants, both methods gave simlar hydroxyurea concentrations and AUC measurements (Figure 4A and 4B, r>0.90 for both comparisons). Importantly, this strong correlation in individual hydroxyurea concentrations also resulted in very similar recommended doses to target the desired AUC. PK-guided doses were calculated using data from both methods and demonstrated similar doses with a mean difference of -1.7 ± 2.6 mg/kg in comparing the two methods. As hydroxyurea concentrations (and thus calculated AUC) tended to be slightly higher as measured by LC-MS/MS than HPLC, there were no recommended doses calculated using LC-MS/MS values that were greater than 3 mg/kg from the HPLC calculated dose. These data, and the ease of collecting low volume samples using this technique provided confidence to use this technique as the primary method of hydroxyurea measurement for the HOPS trial.
For HOPS, PK samples are collected in duplicate using the described sparse sampling strategy (samples collected at 15 minutes, 60 minutes, and 180 minutes following the hydroxyurea dose) and the VAMS devices as shown in Figure 3.
Determination of PK-guided dose
As was done in the TREAT study, hydroxyurea concentrations are incorporated into the previously described population PK-model using MW/Pharm (Mediware, Prague, Czech Republic).36 To determine the PK-guided dose, each participant’s absorption profile is used to determine the dose that would achieve target AUC of 115 mg*h/L. This AUC target was determined through the analysis of PK studies performed in a cohort of children with SCA from the Hydroxyurea Study of Long-term Effects (HUSTLE, NCT00305175) after they reached a clinically-determined maximum tolerated dose.23,31 The optimal PK-guided dose is calculated and recorded for all participants, but only those participants randomized to the Alternative Arm would initiate hydroxyurea at this dose. At the conclusion of the study, the PK-guided dose calculated for the Standard Arm will be compared to the dose achieved through the dose escalation process. The PK-model allows for determination of a specific dose (in mg) that approximates an AUC of 115 mg*h/L, but at times this dose is not convenient using common dosage forms (100 mg/mL liquid or 500 mg capsules). For participants who choose to take liquid hydroxyurea, the recommended starting doses (for both arms) are rounded to the nearest 20 mg (0.2mL). For older participants who choose to take hydroxyurea capsules, a daily dose is selected that best approximates the recommended dose. This at times requires different doses on different days. For example, if a dose of 750 mg is recommended, the participant would alternate taking one (500 mg) and two (1000 mg) capsules each day for an average daily dose of 750 mg. Hydroxyurea will only be started if there are no baseline cytopenias as defined in the toxicity criteria in Table 1.
Randomization and blinding
Participants who complete their baseline PK visit are randomized in a 1:1 ratio to receive either hydroxyurea using a starting dose of 20 mg/kg/day (Standard Arm) or an individualized, PK-guided dose (Alternative Arm). All participants have both a standard (20 mg/kg) and PK-guided starting dose calculated and entered into a locked Research Electronic Data Capture (REDCap) form visible only to select members of the Data Coordinating Center, after which randomization occurs. The random allocation sequence is generated using the REDCap randomization module. The randomization procedure is stratified by age (age< 2 and age>2 years). This stratification for age is performed to increase the likelihood of having age balance in each treatment arm due to the fact that the primary endpoint (%HbF) is typically higher in children less than 2 years than in older children.6 In the current era, guided by the 2014 NHLBI guidelines, most children with SCA are at least offered hydroxyurea and many begin taking hydroxyurea at a young age. We thus anticipate that the enrollment age for HOPS will be young, though likely not quite as young as the TREAT cohort given that there simply are not many older patients who have not yet been offered or started on hydroxyurea. The stratification for age is included to ensure each arm is balanced in terms of older and younger participants. Randomization is performed using a truncated binomial rule with permuted blocks, each 4 in size, within each stratum. The probability of assigning either treatment within each permuted block will be ½, until one of the two treatments has been assigned twice; all subsequent patients within the block receive the remaining treatment. This ensures that within each block as well as at the end of the study the treatment assignment is balanced. The Data Coordinating Center manager and REDCap data specialist, who are not involved in participant screening, enrollment, or assessment, are the only people with knowledge of the study arm. At this time, the local study team is informed that randomization has occurred and starting dose is available. The starting dose is provided as an absolute (mg) dose and entered into the REDCap study database.
The study is designed with the intent of a double-blind design, but is not formally labeled a “double-blind” trial due to the fact that hydroxyurea is used as an open-label study medication and the study team will know if the mg/kg dose is notably different than 20 mg/kg. Despite this possibility, the study arm is not explicitly provided to the provider or the family, and the same procedures are used for dose escalation or reduction throughout the remainder of the study. Additionally, because there are some patients on the Alternative (PK-guided) Arm who may have a dose that is very close to or the same as the Standard Arm dose of 20 mg/kg dose, it is not always possible to know the study arm assignment. Accordingly, we do not anticipate that lack of formal blinding will create bias in the treatment or outcomes of enrolled participants.
Study-directed Hydroxyurea Dosing
The primary objective of HOPS is to compare PK-guided dosing to traditional, weight-based dosing. After selection of the starting dose, all participants, regardless of starting dose, are monitored and have dosage adjusted in the same way. The dose may be adjusted every 8 weeks based on laboratory values to target moderate myelosuppression. The maximum daily dose of hydroxyurea on the HOPS protocol will not exceed 35 mg/kg/day. The dose adjustment and toxicity criteria (Table 1) were decided upon through a consensus of study investigators and are less conservative than are used by most centers or previously published settings, tolerating lower absolute neutrophil, absolute reticulocyte, and platelet counts to optimize clinical benefits while still maintaining patient safety. This consensus decision was based on clinical experience that severe myelosuppression, even with the higher doses used in the TREAT trial, are uncommon with hydroxyurea therapy. With each study visit, participants’ prescribed dose and laboratory values are reviewed by their clinical provider who determines one of four options: 1) continue to prescribe hydroxyurea at the same dose, 2) escalate the prescribed dose, 3) adjust the prescribed dose to account for weight gain, or 4) temporarily hold and/or decrease the prescribed dose. Table 1 summarizes the dose adjustment and toxicity criteria. The study also created a HOPS Dosing Calculator, available on the study website to assist prescribers and to reduce the potential for variation in dosing regimens across study sites (Figure 5). While the calculator is designed to guide dosing decisions, dosing decisions can rely upon clinical discretion at any time throughout the study.Medication adherence is encouraged and patients/caregivers self-report their adherence in person at study visits, by telephone, or if the family agrees, through an electronic REDCap survey sent automatically on a monthly basis by text message or e-mail. We recognize that suboptimal adherence is the primary barrier to effective hydroxyurea therapy, but we aim for this study to mimic “real-world” circumstances and purposefully did not include excessive efforts to address medication adherence.
The primary endpoint is the mean %HbF at 6 months, and participants who are randomized to either the Standard or Alternative Arm are included in the primary endpoint analysis according to the intent to treat principle. We hypothesize that the %HbF will be at least 5 percentage points higher (e.g. HbF of 25% compared to HbF of 30%) in children who initiate hydroxyurea at the PK-guided dose compared to those who start at a standard, 20 mg/kg dose. While there is no true perfect biomarker that predicts the morbidity and mortality of SCA, %HbF was chosen as the primary endpoint as it is the most well-established protective factor that is able to prevent polymerization of HbS and the subsequent complications of SCA. While traditional therapy often achieves modest levels of HbF, we aim to maximize the HbF response beyond the 30% level that has been postulated to be necessary to truly prevent HbS polymerization and RBC sickling. Although safety is not a primary study endpoint, we will carefully collect and analyze the frequency and severity of both laboratory and clinical adverse events, particularly cytopenias that may be the result of hydroxyurea dosing. We will also investigate the clinical, laboratory, and molecular determinants of the maximum hydroxyurea-induced HbF responses. Parameters such as age, sex, baseline (pre-treatment) %HbF, ANC, ARC, and the number of alpha-globin genes present will be analyzed, along with the optimal hydroxyurea dose, and selected PK (AUC, Cmax, t1/2) and PD (%HbF, Hb, MCV, ARC, ANC) variables. We also will analyze the epigenomic signature and gene expression patterns of study participants receiving hydroxyurea therapy after they have achieved the optimal dose, with the goal of elucidating the underlying mechanisms that determine optimal hydroxyurea dose and ultimate HbF response.
Sample Size and Statistical Analysis Plan
A total of 116 patients will be enrolled (58 per arm). The difference in the mean %HbF at six months will be compared between treatments using a one-sided Welch’s two sample t-test. Welch’s t-test will be used due to evidence in the preliminary data that the standard deviations in the two treatment arms may different. A one-sided test is used due to the hypothesis, based on preliminary data, that the PK-guided dosing will have a higher HbF% than the standard, weight-based dosing arm. Significance (and superiority for the primary endpoint) will be assessed at the 0.05 level of significance. The power calculation is based on two separate concluded studies. For data representative of the Standard Arm, the SWITCH trial yielded a mean %HbF and SD of 29.1% and 6.7% post treatment, respectively.37 For data representative of the Alternative Arm, the TREAT trial yielded a mean %HbF and SD of 34.7% and 9.9% post treatment, respectively.29 For the purpose of power calculation we used a conservative estimate of 5% for the mean %HbF difference and 7% and 10% for the SD for the Standard and Alternative Arm, respectively. Based on the preliminary data a sample size of 104 (52 per arm) evaluable patients will have an approximate power of 0.90 and will control the one-sided error rate. Anticipating a 10% drop out rate, we will randomize 116 (58 per arm) patients to one of the two treatment arms. For the determination of predictors of %HbF response, multivariate linear-regression models will be used to identify independent predictors of ‘%HbF response. A p-value of less than 0.05 will be considered statistically significant.
Data Collection Methods and Data Management
CCHMC served as the Data Coordinating Center and the Data Management Team. Data is collected using standardized paper case report forms (CRFs), which is subsequently entered into a secured REDCap database hosted at Cincinnati Children’s Hospital Medical Center. REDCap is a secure, web-based software platform designed to support data capture for research studies, providing 1) an intuitive interface for validated data capture; 2) audit trails for tracking data manipulation and export procedures; 3) automated export procedures for seamless data downloads to common statistical packages; and 4) procedures for data integration and interoperability with external sources.38,39 The REDCap data capture system or equivalent electronic data capture system provides a platform for study-wide oversight of safety monitoring activities with automated notifications within the protocol context. Notifications are based on key study functions including adverse events serious adverse events tracking and CRF completion. These notifications and reporting tools are used to ensure timely communication between the study sites, protocol management, and coordinating center staff, and that reporting requirements are met in all instances. All study data and biospecimens for the proposed studies will be collected directly from the study participants themselves, through a guardian for younger children in the case of questionnaires, or through the medical record for clinically obtained exams. Upon informed consent, participants are issued a study ID number and a numerical study identifier to be used throughout the remainder of the study to ensure confidentiality.
Regulatory and compliance study monitoring of HOPS clinical trials is a continuous, ongoing review of the conduct of this trial to ensure that it is conducted, documented, and reported in accordance with the Institutional Review Board (IRB) approved protocol; the International Conference of Harmonisation (ICH) Good Clinical Practice Guidelines (GCPs); institutional policies; and applicable regulatory requirements.
Study monitoring includes monitoring for central elements defined as:
- Data related to primary study endpoints;
- Appropriateness of consent documentation;
- Protocol eligibility;
- Protocol compliance;
- Timeliness of data entry, including the reporting of adverse events (AE/AR) and serious adverse events (SAE/SSAR) as events are reported;
- Documentation of response assessment measures;
- Essential regulatory documentation; and
- Site investigator supervision of overall conduct of the study.
Regulatory and compliance monitoring includes a combination of on-site visits (scheduled to occur at least annually or more frequently based on enrollment, the degree of risk or severity of monitoring findings, and other study management issues) along with off-site ongoing (remote) efforts. The study has a specific Data Safety Monitoring Plan that includes a designated Medical Monitor unaffiliated with the study who regularly reviews safety data and all adverse and serious adverse events.
Each clinical site uses locally approved informed consent documents, based upon the HOPS Model Consent Form (Appendix). Once a potentially eligible patient has been identified and approached, a consent conference takes place with the patient and family to discuss the study and explain the study procedures. The protocol is carefully described including the purpose of the study, risks and benefits of study participation, the treatment and assessments, and the patients’ rights and responsibilities if they enroll in the study, including the ability to withdraw from the study at any time. At a minimum, one parent or the legally authorized representative(s) is included in the conference. When the consent is signed, a consent process note must be written in the medical record or the research chart. The original consent document is filed in the participant’s research record and a signed copy of the consent document should be provided to the participant and/or LAR or as directed by the local ethical review board. The Informed Consent CRF is completed in REDCap at this time as well. In addition to the study consent, there is an optional consent to store blood for future research purposes. The willingness or unwillingness to sign this optional consent does not affect the ability of the child to receive the full breadth of study treatment and procedures.
The following list provides the scenarios in which a HOPS participant may be removed from the protocol.
- Initiation of chronic transfusions at any time after enrollment on the study
- Parental or family decision to withdraw from the study
- Participant decides to no longer participate in the study
- Investigators may discontinue any participant at their discretion, if in their professional opinion, the participant’s health, safety, and/or well-being is threatened by continued participation in the study.
- Participant decides not to initiate or to discontinue hydroxyurea treatment permanently
- Participants who are not able to complete necessary study procedures. Specifically, if a participant has a three-month (90 day) period without a complete blood count during the first six months of treatment, the participant may be removed from the study.
If more than 10% of participants in either arm are removed from the study prior to the completion of six months of therapy as designed by the sample size estimates, we will recruit new participants to reach the required sample size of 104 participants.