Avatrombopag and lusutrombopag for thrombocytopenia in people with chronic liver disease needing an elective procedure: a systematic review and cost-effectiveness analysis

Reference checking

The bibliographies of identified research and review articles were checked for relevant studies.

Handling of citations

Identified references were downloaded into EndNote X8 [Clarivate Analytics (formerly Thomson Reuters), Philadelphia, PA, USA] bibliographic management software for further assessment and handling. Individual records in the EndNote library were tagged with searching information, such as searcher, date searched, database host, database searched, strategy name and iteration, theme and search question. This enabled the information specialist to track the origin of each individual database record and the record’s progress through the screening and review process.

Quality assurance within the search process

For all searches undertaken by the KSR information team, the main EMBASE strategy was independently peer reviewed by a second KSR information specialist. The search strategy peer review was informed by items based on the Canadian Agency for Drugs and Technologies in Health checklist. 11,12

Population

• Adults with thrombocytopenia associated with CLD needing an elective procedure.

Intervention

• Avatrombopag.

• Lusutrombopag.

Comparator

• Any comparator or none.

Outcomes

• Platelet count.

• Response rate.

• Number of platelet transfusions.

• Number of blood transfusions.

• Return to operating theatre.

• Need for rescue treatments for bleeding (referred to as ‘rescue therapy’).

• Use of concurrent treatments.

• Bleeding score.

• Mortality.

• Adverse effects of treatment.

• HRQoL.

Study design

• Randomised controlled trials (RCTs).

• Observational studies (cohort or case series) of at least 20 participants.

Study selection

Titles and abstracts identified from electronic database and other searches were independently screened by two reviewers. During this initial phase of the screening process any references that obviously did not meet the inclusion criteria were excluded. Full-paper copies of all of the remaining references were obtained. These were then independently examined in detail by two reviewers to determine whether or not they met the inclusion criteria. All papers excluded at this second stage of the screening process along with the reasons for their exclusion are listed in Table 35 (see Appendix 2). These reasons were categorised as follows:

• not relevant population (i.e. not thrombocytopenia associated with CLD needing an elective procedure)

• not relevant intervention

• not relevant outcome data (i.e. did not assess at least one of the specified outcomes or did not report relevant data or information that would allow the calculation of relevant data)

• not relevant study (i.e. not a RCT, cohort study or case series)

• insufficient study size (< 20 participants).

At both screening stages, any discrepancies between reviewers were resolved through discussion or by the intervention of a third reviewer.

A flow diagram of the numbers of studies included and excluded at each stage has been provided following guidance in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.

Data extraction

Data extraction sheets were individually designed and piloted using Microsoft Excel^® (Microsoft Corporation, Redmond, WA, USA). The extraction process was performed by two reviewers, with one checking the extraction of the other. Any discrepancies were resolved through discussion or by the intervention of a third reviewer. Studies are identified by the trial name. To avoid the duplication of data where studies (or study populations) have multiple publications, the most complete report is used as the main reference, but additional details have been extracted from the other publications as necessary. The following general information and data were extracted from each study, regardless of review topic:

• EndNote ID

• study ID or name (if reported; otherwise, surname of first author)

• year of publication

• other related publications

• study group (if reported)

• study country/countries

• recruitment dates (if relevant)

• location/setting

• study funding (public/pharmaceutical/not reported)

• study aim

• sample size

• study design

• study methods

• patient characteristics

• treatment characteristics

• results (all outcomes reported in Chapter 2, Decision problem)

• study conclusions.

Quantitative analysis and meta-analysis methods (direct ‘head-to-head’ methods)

Forest plots of effect sizes are presented for each of the main efficacy outcomes. Dichotomous outcomes (e.g. proportion of patients experiencing each type of outcome) are reported as relative risks (RRs) and odds ratios (ORs) with 95% confidence intervals (CIs).

Pooled effect sizes and 95% CIs using random-effects models are presented where two or more trials are considered to be clinically and statistically homogeneous.

The judgement of clinical homogeneity is based on the baseline characteristics of the trial populations (i.e. age, sex, aetiology of liver disease, degree of thrombocytopenia, bleeding risk and type of elective procedure). Statistical homogeneity is assessed using the I² statistic. 14 This measures the degree of inconsistency between the study results that is due to genuine heterogeneity rather than chance. The value of I² lies between 0% and 100%. For the purposes of this review, a simplified categorisation of heterogeneity is used: low (0–25%), moderate (26–75%) and high (> 75%). Studies will be considered to be sufficiently similar for the purposes of pooling only if I² < 75%. 14

Publication bias could not be assessed given that there were too few trials to use funnel plots of the point estimate plotted against the standard error (SE). 15

Indirect comparisons

Where the intervention and comparator were not compared in the same RCT (i.e. ‘head-to-head’ trials of A vs. B), but instead were separately compared with a common comparator, for example placebo, an indirect comparison of these was performed. Point estimates (with 95% CIs) were estimated using ‘indirect’ methods, for example from A versus C and B versus C, where C is a common control group (e.g. placebo). All methods are applied with consideration of the basic properties of homogeneity, similarity and consistency as reported in Dias et al. 16 All indirect comparisons are consistent with NICE methodological guidance for the conduct of direct and indirect meta-analysis,17 which includes indirect comparisons using the Bucher method. 18

Indirect meta-analysis was performed in Microsoft Excel using the Bucher method. 18 RRs and ORs with 95% CIs were calculated for each outcome and available treatment comparison.

Heterogeneity was investigated using the I² statistic for each of the pairwise comparisons. 14 If there were concerns about heterogeneity, or if any trials appeared to have results that differed substantially from the others, then one or more trials were removed in a sensitivity analysis.

Bayesian network meta-analysis

As its outputs can be directly integrated into a probabilistic cost-effectiveness analysis (CEA) framework, network meta-analysis using WinBUGs version 1.4.3 (www.mrc-bsu.cam.ac.uk/bugs/winbugs/contents.shtml) (Medical Research Council Biostatistics Unit, Cambridge, UK) was applied using a Bayesian Markov chain Monte Carlo (MCMC) approach consistent with international recommendations. This method generates a set of simulated values in the form of a posterior distribution for each of the ORs between each TPO-RA and no TPO-RA. The specification of an evidence-based baseline average risk with its SE then permits the simulation of an absolute risk for each of the three treatments, namely lusutrombopag, avatrombopag and no TPO-RA, as described in NICE Technical Support Document 2. 16 Note that the simulation from the Bayesian posterior distribution provided both statistical estimation and inference, as well as a platform for probabilistic decision-making under uncertainty. Each of the simulated absolute risks from the Bayesian MCMC was consistent and coherent and was used as an input in the CEA model to calculate the expected values of cost and quality-adjusted life-years (QALYs) using a Monte Carlo simulation.

Posterior distribution parameter estimates were obtained from 100,000 simulations after a burn-in period of 30,000 MCMC simulations, using two chains. Vague normal priors (mean 0, variance 10,000) were used for treatment effects and a vague uniform prior (0, 5) was used for the between-study standard deviation. Convergence and auto-correlation were assessed by monitoring the trace, Gelman–Rubin statistics (BGR plot) and autocorrelation plots in WinBUGS. The ORs estimated using this method were almost identical to those estimated using the Bucher method.

Degree of thrombocytopaenia

As described in Table 4, all six studies restricted patients to those with a platelet count of < 50,000/µl. ADAPT 1 and ADAPT 237 differed from the other studies in that results were published only according to the subgroups < 40,000 and 40,000–< 50,000/µl, given the variation in dose of avatrombopag according to these subgroups. Given the need to compare lusutrombopag with avatrombopag, data in these subgroups were requested of Shionogi and are presented in Chapter 3, Results, Subgroup analyses.

Disease type

As shown in Table 4, in terms of the type of CLD reported by each study, one study53 reported including a single type of disease (hepatocellular carcinoma; JapicCTI-121944), whereas five studies37,39,51,54 reported on a mixed CLD population (ADAPT-1, ADAPT-2, L-PLUS 1, L-PLUS 2, Study 202). Three studies37,51 (ADAPT-1, ADAPT-2, Study 202) reported on a CLD definition based on a model for end-stage liver disease score of ≤ 24. Two studies39,54 (L-PLUS 1, L-PLUS 2) reported on a CLD definition based on Child–Pugh class A or B; of note, the exclusion criteria reported in the L-PLUS 1 study39 implied that inclusion was based on Child–Pugh class A or B, but this was not stated explicitly. By contrast, the percentage of participants in the ADAPT trials who were in Child–Pugh class C was above zero. 37 The proportion was generally low in ADAPT-137 (i.e. no higher than 8.6% in the avatrombopag arm of the 40,000–< 50,000/µl subgroup), although it was as high as 15.2% in the placebo arm of the same subgroup in ADAPT-2. 37

Elective procedure type

Elective procedures reported in each study were quite varied (Table 5). Only one study53 reported a single type of procedure (liver radiofrequency ablation; JapicCTI-121944). The other five studies37,39,51,54 reported including mixed types of elective procedures. Only ADAPT-1 and ADAPT-2 explicitly mentioned risk of bleeding, stating that they included both ‘low-risk’ procedures, for example liver biopsy, and ‘high-risk’ procedures, for example radiofrequency ablation. Both L-PLUS 139 and L-PLUS 2,54 also according to this definition, included mixed-risk procedures, such as liver biopsy and radiofrequency ablation.

Decision rule for determining treatment dose

There appeared to be some variation in the decision rules for administering platelet transfusion prior to the elective procedure. The L-PLUS39,54 studies mandated this on the basis of a drop in platelet count below the 50,000/µl threshold, whereas this rule was not explicitly reported in the ADAPT trials. 37 However, because those eligible for the ADAPT studies37 were at ‘. . . risk of bleeding that would require a platelet transfusion, unless there was a clinically significant increase in platelet counts from baseline’, it seems likely that the same rule would be applied. There was also a difference in the decision rules for administering the intervention. In the ADAPT trials,37 all patients received avatrombopag for 5 days, whereas in the L-PLUS trials39,54 lusutrombopag was administered for between 5 and 7 days depending on platelet count (i.e. if the platelet count was at least 50 × 10⁹/l with an increase of at least 20 × 10⁹/l then no additional dose was given). The implication of this difference is that lusutrombopag was administered, on average, over a longer period than avatrombopag.

Meta-analysis

In the absence of head-to-head clinical trials of avatrombopag and lusutrombopag, the indirect comparison approach was used to assess the relative effect of these treatment interventions. On the basis of the published trials, placebo was used as the common comparator. As the dose of avatrombopag varies by platelet count, subgroup analyses were performed. Forest plots of each of the interventions compared with placebo are presented in Appendix 4.

As shown in Tables 18 and 19, the outcome on the RR scale was a little more favourable towards lusutrombopag in both outcomes that counted platelet transfusions prior to the elective procedure in all cases regardless of therapies required prior to the procedure and regardless of the subgroups. Only one statistically significant difference was identified between avatrombopag and lusutrombopag. This was in a fixed-effect analysis of the ratio of patients who required no platelet transfusion prior to elective procedure in the subgroup in which patients’ baseline platelet count was < 40,000/µl. It was in favour of lusutrombopag (RR 1.93, 95% CI 1.15 to 3.22). On the OR scale, there was no statistically significant difference in any subgroup, although there was a reversal in the point estimate to an advantage for avatrombopag in the 40,000–< 50,000/µl subgroup in terms of both outcomes.

By contrast, Table 20 shows an advantage of avatrombopag in terms of avoidance of rescue therapy, but, again, this is not statistically significant except in the fixed-effect analysis in the < 40,000/µl subgroup. On the OR scale, the value for the < 40,000/µl subgroup was not estimable and, as for the RR scale and the other outcomes, there was an advantage for avatrombopag in the 40,000–< 50,000/µl subgroup.

Heterogeneity

There was clinical heterogeneity in terms of invasive procedures that patients were undergoing. In both of the L-PLUS trials39,54 patients were not restricted to the elective procedure, whereas in the study by Tateishi et al. 53 only patients who were undergoing radiofrequency ablation were included. However, sensitivity analysis by exclusion of this study increased the heterogeneity in all cases. In addition, there was moderate statistical heterogeneity within each subgroup regardless of the outcome, for example for no platelet transfusion prior to the elective procedure I² = 53% and 34% in the < 40,000/µl and 40,000–< 50,000/µl subgroups, respectively (see Appendix 4). Sensitivity analysis revealed that the removal of one of the L-PLUS studies would remove this heterogeneity and reduce the I² to 0%. However, the study that needed to be removed to reduce the heterogeneity depended on the subgroup. More specifically, it was the L-PLUS 1 study39 in the < 40,000/µl subgroup and the L-PLUS 2 study54 in the 40,000–< 50,000/µl subgroup. Most importantly, this did not make any substantial change to the results.

For no rescue therapy, there was no statistical heterogeneity in the L-PLUS trials,39,54 but there was moderate heterogeneity in the < 40,000/µl subgroup. Nevertheless, given no obvious clinical difference between the ADAPT-1 and ADAPT-2 studies,37 the AG did not consider that exclusion of either was warranted. As already discussed in Results, Subgroup analyses, the lusutrombopag trials also appear to be quite different from the ADAPT trials in the much lower frequency of rescue therapy, regardless of treatment arm. This highlights that caution needs to be exercised in comparing avatrombopag with lusutrombopag.

Handling of citations

Identified references were downloaded into EndNote bibliographic management software for further assessment and handling. Individual records in the EndNote library were tagged with searching information, such as searcher, date searched, database host, database searched, strategy name and iteration, and theme or search question. This enabled the information specialist to track the origin of each individual database record and its progress through the screening and review process.

Quality assurance within the search process

For all searches undertaken by the KSR information team, the main EMBASE strategy was independently peer reviewed by a second KSR information specialist. The search strategy peer review was informed by items based on the Canadian Agency for Drugs and Technologies in Health checklist. 11,12

Identified resource use/costs studies

The systematic review of resource use/costs identified four full-text articles63–66 and three conference abstracts,29,61,62 discussing five separate studies. Two of the conference abstracts have since been published as full-text publications (the Poordad 2007 abstract61 corresponds to the Poordad et al. 201263 article and the Poordad et al. 2008 abstract62 is covered by the Poordad et al. 2011 article64), and, therefore, only the full-text publications of these studies are discussed. For the remaining conference abstract, no full-text publication was available and therefore only the content of the abstract is discussed. 67

Barnett et al. 65 conducted a study to estimate the cost of platelet transfusion for CLD patients with thrombocytopenia undergoing elective procedures in the USA. The authors developed a conceptual framework aiming to identify all direct, indirect and intangible costs of platelet transfusion. They then estimated the costs using the developed framework and cost data from the literature. The framework included the cost of generating the supply of platelets, the transfusion itself, the adverse events associated with platelet transfusion and refractoriness. The total direct cost obtained from considering all framework categories of platelet transfusion in CLD patients with thrombocytopenia scheduled to undergo an elective procedure was estimated to be in the range of US$5258–13,117. The majority of costs were attributable to the transfusion itself (US$3723–4436), followed by the cost of refractoriness (which included the opportunity cost of a delayed procedure and subsequent transfusions with human leucocyte antigen-matched platelets) (US$874–7578). A potential limitation of this study is that it is literature based, drawing cost elements from different sources with different study designs. These sources were not based on CLD patients with thrombocytopenia, as the authors could not identify published sources on this population. Therefore, the estimate may not well reflect the target population if differences exist in the costs of transfusion and the rates of related AEs and refractoriness in a CLD thrombocytopenia population in the UK. It is also noted that this study was funded by Dova Pharmaceuticals, the owner of avatrombopag.

Brown66 published a review article discussing the pharmacoeconomic analysis of thrombocytopenia in CLD. The review discussed the negative impact that thrombocytopenia and its treatment can have on costs and treatment outcomes in CLD. The impact of thrombocytopenia on patient outcomes was discussed in terms of the increased likelihood of complications during routine medical procedures as well as the cancellation, delay or prolongation of procedures, which can increase morbidity and mortality. The negative patient outcomes that can arise from platelet transfusions, such as refractoriness, infection, allergic reaction, iron overload and other transfusion reactions, were also outlined. The review also discussed the economic burden of costs associated with platelet transfusion and resulting AEs that can require further treatment and increased utilisation of health-care resources.

In a conference abstract, Poordad et al. 29 conducted a case–control study examining the economic burden of platelet transfusion in CLD patients with thrombocytopenia. A retrospective analysis was conducted in a large national US administrative claims database to examine the impact of platelet transfusion on health resource utilisation and expenditure, including hospitalisations, accident and emergency (A&E) visits and outpatient visits among CLD patients with thrombocytopenia. Data from 2012 to 2015 were used to match adult CLD patients with thrombocytopenia who received a platelet transfusion 1 : 2 based on age and sex with CLD patients with thrombocytopenia who did not receive a platelet transfusion. Among the 1173 CLD patients with thrombocytopenia included in the analysis, those with thrombocytopenia who received a platelet transfusion had a statistically significantly higher probability of having an additional outpatient office visit (1.04; p = 0.021), a non-significantly higher probability of hospitalisation (1.08; p = 0.174) and a significantly lower probability of an A&E visit (0.86; p = 0.001) than those who did not receive a platelet transfusion. Platelet transfusions were associated with significantly increased hospitalisation costs (US$25,802, 95% CI US$11,220 to US$40,660), outpatient office costs (US$3367, 95% CI US$1082 to US$5652) and total costs (US$29,717, 95% CI US$15,096 to US$44,339) and non-significantly decreased A&E costs (–US$371, 95% CI –US$1019 to US$277) compared with no transfusion.

In Poordad et al. ,64 the aim was to examine medical resource utilisation and health-care costs in HCV patients with and without thrombocytopenia from a longitudinal administrative claims database using International Classification of Diseases, Ninth Edition, Clinical Modification (ICD-9-CM), diagnosis codes. The prevalence of thrombocytopenia in HCV patients identified was found to be 3.6%, and the prevalence of thrombocytopenia in the subset of patients for whom platelet count laboratory results were available was 10.8%. HCV patients diagnosed with thrombocytopenia had a higher incidence of bleeding events (27.3% vs. 9.9%) and platelet transfusions (8.5% vs. < 1%). HCV patients diagnosed with thrombocytopenia also had a higher incidence of liver disease-related ambulatory visits (10.4% vs. 4.4%; OR 2.3, p < 0.001), emergency room visits (OR 8.6, p < 0.01) and inpatient hospital stays (OR 17.7, p < 0.01) during the year before and the year after HCV diagnosis than HCV patients without a thrombocytopenia diagnosis. HCV patients diagnosed with thrombocytopenia had significantly higher overall health-care costs (US$37,924 vs. US$12,174; p < 0.001) and liver disease-related costs (US$14,569 vs. US$4107; p < 0.001) than those without thrombocytopenia. Overall health-care and liver disease-related costs in the subset of HCV patients with complete laboratory results also found significantly higher costs among HCV patients diagnosed with thrombocytopenia than among those without thrombocytopenia (overall health-care costs US$25,482 vs. US$16,412, p < 0.001; liver disease-related costs US$23,608 vs. US$7354, p < 0.001). Where results are presented according to the two different strategies for identifying thrombocytopenia (i.e. coding identification and laboratory results), they differ quite substantially.

Poordad et al. 63 estimated the prevalence of thrombocytopenia and evaluated medical resource use and costs associated with thrombocytopenia in CLD patients. A retrospective study was performed on a longitudinal administrative claims database that included 56,445 patients with an ICD-9-CM diagnosis code for CLD in the period January 2001 to December 2003. For patients with available laboratory results, including platelet counts (35.7%), the numbers of bleeding events or platelet transfusions were also determined. The annual prevalence of thrombocytopenia among patients with CLD ranged from 3.3% to 4.1%. In comparison with patients without a thrombocytopenia diagnosis, the group of patients with a thrombocytopenia diagnosis included more males (62.6% vs. 49.4%) and experienced more anaemia (54.2% vs. 18.5%), more neutropenia (20.8% vs. 1.7%), more liver cancer (5.7% vs. 1.5%), more liver transplants (2.1% vs. < 1%) and more bleeding events (27.8% vs. 10.0%). They also received more interferon therapy (5.9% vs. 2.0%) and more platelet transfusions (8.1% vs. < 1%) and, on average, each one had more platelet count assessments (mean 3.68 vs. 2.47). Patients with a thrombocytopenia diagnosis had 2.5 times more liver disease-related ambulatory visits, 3.9 times more liver disease-related emergency room visits and 12.9 times more liver disease-related inpatient hospital stays than patients without a thrombocytopenia diagnosis. Overall medical care costs were 3.5-fold higher in patients with a thrombocytopenia diagnosis, with liver disease-related costs being 7-fold higher in patients with a thrombocytopenia diagnosis than in patients without a thrombocytopenia diagnosis. Similar results were obtained for patients with a platelet count that indicated thrombocytopenia.

In summary, the findings from the literature review that were presented above indicate that the health-care costs of patients with CLD and thrombocytopenia are substantial. Most notably, the costs of, and associated with, platelet transfusions make a relatively large contribution to those costs. This emphasises the importance of evaluating how an alternative strategy through the (additional) use of TPO-RAs compares with platelet transfusions as the current standard treatment for thrombocytopenia in patients with CLD.

Efficacy summary

Efficacy inputs in the model included the following for each treatment arm:

1. proportion of patients receiving a platelet transfusion prior to the elective invasive procedure

2. proportion of patients experiencing bleeding events following an elective invasive procedure

3. proportion of patients not receiving their elective invasive procedure during the trial period (conditional on receipt of prior platelet transfusion)

4. proportion of patients receiving rescue therapy following bleeding (conditional on receipt of prior platelet transfusion and receipt of elective invasive procedure).

For efficacy inputs 1 and 2, the proportion of patients achieving each outcome in the placebo/platelet transfusion arm was taken directly from the placebo arm of the pooled lusutrombopag clinical trials (or from L-PLUS 254 only in scenario analysis). For the lusutrombopag arm, ORs for lusutrombopag compared with placebo were estimated from the pooled trials (or from L-PLUS 254 alone in scenario analysis) and were applied to the placebo/platelet transfusion arm data. Inputs 3 and 4 were calculated as conditional probabilities in the base-case analysis using individual patient-level data from the pooled lusutrombopag trials. In a scenario analysis, these conditional probabilities could be turned off and replaced with unconditional inputs calculated using ORs, as seen for inputs 1 and 2.

In the base-case analysis, the company assumed, contrary to evidence from the lusutrombopag trials, that 100% of patients in the placebo/platelet transfusion arm would receive a platelet transfusion prior to an elective invasive procedure as a result of less intensive monitoring of platelet count prior to procedures in clinical practice. This assumption was based on clinical expert opinion. In the trials (confidential information has been removed) of placebo arm patients in the pooled trials and (confidential information has been removed) in the L-PLUS 2 trial54 received a platelet transfusion prior to surgery.

Mortality in the short-term model could occur as a result of platelet transfusion or bleeding events. The company identified two different sources for the probability of platelet transfusion-related mortality. In the base-case analysis, the company adopted values from a study by van Eerd et al. ,73 in which the base-case mortality risk associated with transfusion was estimated to be 0.3315%. The company also identified an alternative source of mortality data, from a study by Vamvakas et al. ,74 that estimated an incidence of transfusion-related death of 0.0004% from UK Serious Hazards of Transfusion (SHOT).

In the base-case analysis, bleeding-related mortality was taken from a study by Takaki et al. ,75 which estimated that the rate of death from either major or minor bleeding following radiofrequency ablation (RFA) was 0.83%. Two alternative sources of estimates of bleeding-related mortality were included in the model. Lo et al. 76 estimated a mortality rate of 6% from upper gastrointestinal haemorrhage and oesophageal variceal bleeds (assumed to be a major bleed) and Triantos and Kalafateli77 estimated a 20% mortality rate from acute variceal bleeding (assumed major bleed).

Chronic liver disease-related mortality was incorporated into the long-term model to estimate lifetime QALYs for those patients surviving the short-term model. In the base-case model, data were used from a systematic review by D’Amico et al. ,78 with 1-year survival estimated at 84%.

The model included AEs relating to the treatment and to platelet transfusion. SAEs that were possibly or probably related to the drug were included in the model. Thrombus-related AEs are particularly relevant to TPO-Ras; therefore, any severe thrombus-related events in any of the three lusutrombopag trials79–81 (3 mg dose) were included in the model. In its submission, the company states that comprehensive data for all platelet transfusion-specific AEs were not available. Therefore, data for platelet transfusion AEs were taken from the van Eerd et al. 73 study, which reports the incidence of AEs per unit of fresh-frozen plasma transfused. 73

Health-related quality-of-life summary

Health-related quality-of-life data were not collected in the trials. The base-case analysis adopted a baseline utility value of 0.544 in both treatment groups, estimated for patients with CLD/cirrhosis. This utility value is from a study by Sullivan et al. 82 that provides EuroQol-5 Dimensions (EQ-5D) index scores for a wide variety of chronic conditions based on UK community preferences (using US-based panel survey data). One-off disutilities were included in the model for platelet transfusions, bleeding events, rescue therapy and AEs. In the base-case analysis, a disutility of 0.1 for patients experiencing serious platelet transfusion-related AEs was applied for one model cycle (4 weeks). This value was taken from TA293,83 a previous NICE appraisal of eltrombopag for thrombocytopenic purpura. In the base-case analysis, the company assumed the same disutility for rescue therapy as for platelet transfusion, stating that clinical experts advised that platelet transfusion would be most common in clinical practice.

Utilities summary

Disutilities for bleeds were also identified from the literature. The literature provided separate disutilities for bleeds classified as major and those classified as minor. The company assumed that all bleeds were major, stating that no studies were identified that reported the proportion of bleeds classified as major or minor following an elective invasive procedure in this population, and that minor bleeds would be expected to have a minor impact on costs and QALYs. Therefore, a disutility associated with a major bleeding event of 0.397 for a duration of 1 week was adopted from Jugrin et al. 84 For thrombus-related AEs, the company incorporated a disutility of 0.029, applied over 1 week, estimated by Jugrin et al. 84 for related thrombotic events (index deep-vein thrombosis and index pulmonary embolism).

The baseline utility value for CLD/cirrhosis patients adopted in the short-term model was also used to calculate QALYs throughout the long-term model. Utility values were adjusted to incorporate the natural decline in utility observed with ageing using the Ara and Wailoo85 equation to generate utility multipliers by age and sex.

Costs summary

The drug acquisition cost of (confidential information has been removed) for 7 days of 3 mg of lusutrombopag was included in the model. As lusutrombopag is an oral medication, no administration costs were required. The base-case cost of platelet transfusion was based on the TA293 appraisal83 of eltrombopag. In the eltrombopag appraisal, this cost was assumed to comprise the cost of blood transfusion (weighted average cost of £57.72 in 2011/12, code 821 blood transfusion) and the cost of 2 units of platelets (2 × £230.393 in 2011/12), which resulted in a cost per transfusion of £517.28 in 2011/12. The company used expert opinion to inform the average number of units of platelets that would be received per transfusion. The expert stated that most often platelet transfusions would contain either 2 or 4 units and, therefore, it was assumed that an average of 3 units of platelets would be received per transfusion. This resulted in a base-case cost of £812.61 (inflated to 2017/18), which included both administration and platelet acquisition. Two alternative costs of platelet transfusion were included in the model. One alternative was based on NHS Reference Costs 2017–1886 for single plasma exchange or other intravenous blood transfusion. Here it was assumed that a single transfusion was sufficient to transfuse the required number of units of platelets, which resulted in a cost per transfusion of £517.28. The final option was based on a poster by Varney and Guest,87 which estimated the cost per unit of adult platelet concentrate to be £347 in 2002/3, resulting in a cost per transfusion of £1493.21 (inflated to 2017/18).

The costs associated with treating transfusion-related complications were based on the costs of complications from fresh-frozen plasma transfusion, reported in van Eerd et al. 73 The cost of managing portal vein thrombosis (PVT) in lusutrombopag patients was assumed to be £958.95, based on NHS Reference Costs 2017–1886 for percutaneous transluminal, embolectomy or thrombolysis, of blood vessel, with a CC (complication) score of 0–4 in a day-case setting. The same cost of one platelet transfusion was assumed for all rescue therapies.

All patients in both treatment arms were assumed to have received an elective invasive procedure and to incur the relevant costs. Although the short-term model allowed for the possibility of delaying the procedure beyond the 35-day cycle, all patients were assumed to receive their procedure at some point. Base-case procedural costs were estimated using the pooled proportion of patients receiving each procedure in the three trials and the relevant NHS Reference Costs 2017–1886 in the elective inpatient setting. In the base-case analysis, the company included a sunk cost for cancelled or delayed procedures, assuming that there may not be enough time to reallocate a pre-assigned clinician or hospital bed to another patient procedure, thus wasting clinician time. A sunk cost of £566.05 for delayed elective invasive procedures was included, which was based on a study based on an NHS reference cost that the company had stated had been removed from subsequent years’ NHS reference costs. 88

Patient population

The patient population considered is CLD patients with severe thrombocytopenia (i.e. a platelet count of < 50,000/µl) who are scheduled to undergo an elective invasive procedure.

The patient population is divided into two subgroups:

1. patients with a platelet count of < 40,000/µl

2. patients with a platelet count of 40,000–< 50,000/µl.

This immediate division of the population into platelet count subgroups is necessitated by the fact that each of these subgroups receives a different dose of avatrombopag, as described below. Therefore, it is not possible to conduct a direct comparison between lusutrombopag and avatrombopag without this subgroup separation.

Interventions

Lusutrombopag is administered orally once per day at a dose of 3 mg for up to 7 days, with the first dose taken a minimum of 9 days prior to the scheduled procedure. 8

Avatrombopag for patients with a platelet count of < 40,000/µl is administered orally once per day at a dose of 60 mg (three tablets of 20 mg), with the first dose administered 10–13 days prior to the scheduled procedure and the regimen for 5 days (i.e. the procedure is scheduled 5–8 days after the last dose). For patients with a platelet count of 40,000–< 50,000/µl, the administration and timing of avatrombopag are the same, but the dose is reduced to 40 mg (two tablets of 20 mg).

Standard of care entails patients being given a platelet transfusion if their platelet count fails to reach ≥ 50,000/µl on the day of the scheduled procedure.

Model structure

The AG model is based on the structure for lusutrombopag submitted by Shionogi. Similar to that model, the AG model combines a short-term decision tree considering costs and QALYs over a 35-day period (matching the time horizon of all trials, as shown in Table 7), during which severely thrombocytopenic CLD patients are scheduled to undergo an elective invasive procedure. Those patients alive at the end of the short-term model enter the long-term Markov model, which assesses QALYs and mortality over a lifetime time horizon of 50 years. The AG short-term decision tree model has the following chance nodes:

• receiving/not receiving platelet transfusion (taken from the avatrombopag and lusutrombopag trials)

• receiving/not receiving the elective invasive procedure within the 35-day study period

• rescue therapy/no rescue therapy (taken from avatrombopag and lusutrombopag trials)

• death/no death due to platelet transfusion, surgery or rescue therapy (taken from the literature).

The structure of the AG short-term decision tree model, shown in Figure 4, differs in several ways from that of the original Shionogi model discussed in Review of the lusutrombopag submission. In the Shionogi model, a chance node for death due to platelet transfusion was placed directly after the receipt of transfusion before the chance node for undergoing an elective invasive procedure. In the AG model, both mortality due to platelet transfusion prior to elective invasive procedure and mortality due to surgical complications were considered after the chance nodes for undergoing surgery and requiring rescue therapy.

FIGURE 4.

Structure of the short-term decision tree model. Data from AG model.

The Shionogi model also allowed for the probability of delays to scheduled procedures and modelled the potential impact of delays on quality of life and mortality and the additional costs that may be incurred as a result of such delays. Additional costs resulting from surgery delays included a possible additional platelet transfusion, as well as sunk costs resulting from last-minute delays leading to wasted surgeon and surgical theatre time. The AG did not feel that the inclusion of a sunk cost was necessary, as surgical theatre slots would usually be filled by other procedures and surgeons could effectively fill their time with other tasks. In addition, the fact that Shionogi identified a sunk cost unit cost from the NHS Reference Costs from 2009/10 but this was subsequently removed from the reference costs suggests that it is no longer considered an appropriate cost to include in a model. The Shionogi model also contained a chance node for death due to surgery delay. However, this was assumed to carry a probability of 0 in the base-case analysis and was removed by the AG.

The Shionogi model structure contained a separate chance node for bleeding events and a subsequent chance node for the requirement of rescue therapy. However, the AG had concerns regarding this structure and the data it was based on. The AG was unable to trace back the numbers used to calculate bleeding event efficacy to the lusutrombopag trials’ CSRs. 79–81 On clarification request, the company provided data on the number of bleeding events in each trial and treatment group. 56,57 However, in contrast to how it was implemented in the original Shionogi submission model, these numbers did not suggest that lusutrombopag substantially reduced the odds of bleeding. In addition, these conditional probabilities were not available for avatrombopag. The small number of World Health Organization grade 2 bleeding events and the rescue events seen in the trials led to concerns about the confidence that can be placed in conditional probabilities based on such data. Therefore, the AG felt that bleeding events were better modelled as a surgical complication rather than as a separate event. Therefore, bleeding events and their impact on the mortality and quality of life of patients were modelled as a surgical-related AE and a source of mortality. The chance node for requiring rescue therapy was retained.

The long-term Markov model presented by Shionogi was utilised without changes in the AG model. In the long-term model, data from the literature regarding CLD-related mortality and utility values were used to estimate the number of QALYs that would accrue over the expected remaining life of the patient with a cycle length of 1 year. QALYs in the long-term model were discounted at a rate of 3.5%. No cost discounting was incorporated as costs are included only in the short-term model, in which discounting is inappropriate.

Assessment group input parameters

Probabilistic sensitivity analysis and scenario analyses

Given the parametric uncertainty surrounding the input parameters utilised in the model, probabilistic sensitivity analysis, consisting of 2000 iterations, was run to test parameter uncertainty in the model. All parameters except drug prices, drug doses and discount rates were included in the probabilistic sensitivity analysis (see Appendix 6). As is standard practice, appropriate distributions were fitted to included parameters. Beta distributions were used for probabilities, proportions, risks and utilities, gamma distributions were used for costs, beta tree was used for Child–Pugh categories and normal distributions were used for age and the number of ATDs per transfusion. Where SEs were unknown, they were estimated as 20% of the mean value. For efficacy parameters obtained from WinBUGS, probabilistic values were drawn from CODA (convergence diagnostic and output analysis) output. Cost-effectiveness planes and cost-effectiveness acceptability curves will be provided to examine the uncertainty related to the decision.

Given the structural uncertainty surrounding the input parameters utilised in the model, the AG conducted a series of scenario analyses for various efficacy, mortality, safety, cost and utility parameters. These scenario analyses are listed below and explained in more detail in the following section:

1. drug prices

2. number of ATDs per platelet transfusion

3. cost of platelet transfusion

4. cost of rescue therapy

5. inclusion of grade 2 bleeding AEs

6. probability of requiring platelet transfusion, estimated from international trials only

7. efficacy model input parameters derived from fixed-effect meta-analysis models

8. literature source for long-term Child–Pugh grade-specific mortality

9. under-reporting factor for SHOT data platelet transfusion-specific mortality

10. alternative literature source for surgery-related mortality

11. alternative literature source for baseline CLD utility

12. alternative literature source for bleeding disutility

13. alternative literature source for PVT disutility

14. alternative literature source for transfusion-related AE disutilities

15. alternative values for elective invasive procedure delay disutility and duration.

Assessment group base-case deterministic results

The base-case deterministic model results from the AG model are shown in Table 29. The price of avatrombopag for both subgroups is assumed to be (confidential information has been removed), equal to the price of lusutrombopag.

In both subgroups, no TPO-RA incurred the lowest costs and fewest QALYs. In the < 40,000/µl subgroup, lusutrombopag is the next cheapest option, with an incremental cost compared with no TPO-RA of £592 and incremental QALYs of 0.00017 (which is equivalent to a gain of 1.5 quality-adjusted life-hours), resulting in a deterministic incremental cost-effectiveness ratio (ICER) of around £3,400,000. Avatrombopag 60 mg is the most expensive option in this subgroup but incurs a lower QALY gain than lusutrombopag, with an incremental QALY of –0.000079. Avatrombopag 60 mg is therefore dominated by lusutrombopag in the < 40,000/µl subgroup. In the 40,000–< 50,000/µl subgroup, lusutrombopag is the cheapest option after no TPO-RA, with an incremental cost of £624 and an incremental QALY of 0.000000007, resulting in an ICER of > £84,000,000,000 compared with no TPO-RA. Avatrombopag 40 mg is the most expensive option in this subgroup but provides a higher QALY gain, with an incremental QALY gain of 0.00041 over lusutrombopag. This results in an ICER of £21,947 for avatrombopag 40 mg compared with lusutrombopag. However, it should be noted that the incremental QALYs are extremely small, and in both subgroups all treatments resulted in almost identical QALYs.

The disaggregated cost results in Table 30 show that, although the costs of platelet transfusion, AE management and rescue therapy are higher for no TPO-RA than for lusutrombopag and avatrombopag (except for AE costs in the 40,000–< 50,000/µl subgroup), the combined difference in cost is still substantially lower than the drug costs for lusutrombopag and avatrombopag. This results in incremental costs of > £500 for both treatments compared with no TPO-RA. In the face of such small incremental QALYs, this incremental cost has a large impact on the ICER. In both subgroups, the dominance of one treatment over the other is mostly due to the differences in the QALY decrements as a result of bleeding, which lead to small but important differences in the total QALYs (Table 31).

Probabilistic sensitivity analysis results

The probabilistic results in Table 32 for the < 40,000/µl subgroup follow the same pattern as that of the deterministic results. Lusutrombopag is more expensive than no TPO-RA by £600 [i.e. (confidential information has been removed) more expensive] and more effective by 0.0001 QALYs, resulting in an ICER of approximately £4,000,000. Avatrombopag 60 mg is slightly more expensive and slightly less effective than lusutrombopag and is therefore dominated. In the 40,000–< 50,000/µl subgroup, no TPO-RA is again the cheapest option. Lusutrombopag is the next cheapest and most effective option, with an incremental cost of £626 and incremental QALYs of 0.0004. Avatrombopag 40 mg is £10 more expensive than lusutrombopag and –0.00054 QALYs less effective and is therefore dominated by lusutrombopag.

The cost-effectiveness planes (Figures 5 and 6) for both subgroups show that, for the majority of iterations, both treatments are more costly and more effective than no TPO-RA. However, each diagram also shows that a substantial proportion of iterations fall in the north-west quadrant, where the treatments are more expensive but less effective than no TPO-RA. This can be seen most prominently for avatrombopag in the 40,000–< 50,000/µl subgroup, for which it appears that approximately half of the iterations suggest that avatrombopag is less effective than no TPO-RA (orange points). This indicates that, given the uncertainties in the model, the treatments should be regarded as having equivalent effectiveness in terms of QALYs.

FIGURE 5.

Cost-effectiveness plane for subgroup: platelet count of < 40,000/µl (lusutrombopag and avatrombopag 60 mg vs. placebo).

FIGURE 6.

Cost-effectiveness plane for subgroup: platelet count of 40,000–< 50,000/µl (lusutrombopag and avatrombopag 40 mg vs. placebo).

The cost-effectiveness acceptability curves in turn (Figures 7 and 8) show that, for all threshold ICERs up to £100,000, no TPO-RA has 100% probability of being the most cost-effective treatment.

FIGURE 7.

Cost-effectiveness acceptability curve for platelet count of < 40,000/µl.

FIGURE 8.

Cost-effectiveness acceptability curve for platelet count of 40,000–< 50,000/µl.

Scenario analysis results

Given the uncertainty surrounding the input parameters utilised in the model, the AG conducted a series of scenario analyses using various efficacy, mortality, safety, cost and utility parameters. These scenario analyses are listed below and the results of each are provided in Operational validation efforts on the assessment group model.

1. drug prices

2. number of ATDs per platelet transfusion

3. cost of platelet transfusion

4. cost of rescue therapy

5. inclusion of grade 2 bleeding AEs

6. probability of requiring platelet transfusion, estimated from international trials only

7. cost of elective invasive procedure taken from international trials only

8. literature source for long-term Child–Pugh grade-specific CLD mortality

9. under-reporting factor for SHOT data platelet transfusion-specific mortality

10. alternative method for calculating surgery-related mortality

11. alternative literature source for baseline CLD utility

12. alternative literature source for bleeding disutility

13. alternative literature source for PVT disutility

14. alternative literature source for transfusion-related AE disutilities

15. alternative values for elective invasive procedure delay disutility and duration

16. cost of elective invasive procedure cancellation

17. proportion of patients requiring platelet transfusion hospitalised the day before elective invasive procedure.

The description and the results of these analyses are presented in Appendix 7. Among these scenarios, only the first three (i.e. using different drug prices, different number of ATDs per platelet transfusion and different platelet transfusion costs) had a substantial impact on the incremental results.

Operational validation efforts on the assessment group model

The AG conducted the following validation efforts:

• comparing the clinical outcomes of the AG economic model with those of clinical trials

• comparing the economic and health outcomes of the AG economic model and the Shionogi economic model.

Search strategy

1. (avatrombopag or doptelet or AKR 501 or AKR501 or AS 1670542 or AS1670542 or E 5501 or E5501 or oralE 5501 or oralE5501 or YM 477 or YM477 or 570406-98-3 or 677007-74-8).af. (33)

2. (lusutrombopag or mulpleta or S 888711 or S888711 or 1110766-97-6).af. (14)

3. or/1-2 (46)

4. exp Thrombocytopenia/ (45,457)

5. (thrombocytopeni$ or thrombocytopaeni$ or thrombopeni$ or thrombopaeni$ or macrothrombocytopeni$ or macrothrombocytopaeni$).ti,ab,ot,hw. (69,081)

6. ((11q or 11q23) adj3 (disorder$ or syndrome$ or delet$ or jacobsen)).ti,ab,ot,hw. (574)

7. (jacobsen adj3 syndrome$).ti,ab,ot,hw. (129)

8. paris trousseau.ti,ab,ot,hw. (30)

9. kasabach merritt.ti,ab,ot,hw. (704)

10. (hemangioma or haemangioma).ti,ab,ot,hw. (32,339)

11. (thrombotic adj2 (microangiopath$ or micro angiopath$)).ti,ab,ot,hw. (3354)

12. (hemolytic uremic or haemolytic uremic).ti,ab,ot,hw. (7663)

13. gasser$.ti,ab,ot,hw. (1689)

14. HELLP Syndrome/ (1709)

15. (HELLP adj2 syndrome$).ti,ab,ot,hw. (2561)

16. ((hemolysis or haemolysis) adj2 liver adj2 platelet$).ti,ab,ot,hw. (7)

17. May Hegglin.ti,ab,ot,hw. (221)

18. ((haemolytic or hemolytic) adj2 (anaemi$ or anemi$) adj2 (microangiopathic or micro angiopathic)).ti,ab,ot,hw. (1411)

19. moschcowitz.ti,ab,ot,hw. (107)

20. werlhof.ti,ab,ot,hw. (120)

21. Wiskott-Aldrich Syndrome/ (1428)

22. (wiskott and Aldrich).ti,ab,ot,hw. (3312)

23. (immunodeficiency 2 or immunodeficiency2 or Imd2).ti,ab,ot,hw. (44)

24. ((platelet$ or thrombocyte$) adj3 (defici$ or reduc$ or low or lower or lowest or few or fewer or fewest or decrease or decreases or decreased or defective or destruc$ or destroy$)).ti,ab,ot,hw. (22,231)

25. or/4-24 (132,417)

26. exp Liver Diseases/ (521,414)

27. ((liver$ or hepat$ or intrahepat$) adj2 (disease$ or disorder$ or lesion$)).ti,ab,ot,hw. (163,004)

28. (cirrhosis or cirrhoses or cirrhotic).ti,ab,ot,hw. (123,945)

29. (chronic adj3 destructive cholangitis).ti,ab,ot,hw. (98)

30. ((fibrosis or fibroses or scar$) adj3 (liver$ or hepat$)).ti,ab,ot,hw. (23,356)

31. ((hepatitis or hepatopath$) adj3 (chronic or acute or persistent or long stand$ or long term or recurr$)).ti,ab,ot,hw. (76,827)

32. ((liver$ or hepat$ or intrahepat$) adj3 inflam$).ti,ab,ot,hw. (13,126)

33. (haemochromatosis or hemochromatosis or bronze$ diabet$ or recklinghausen applebaum or siderochromatosis).ti,ab,ot,hw. (10,335)

34. primary biliary cholangitis.ti,ab,ot,hw. (552)

35. ((liver$ or hepat$ or intrahepat$) adj3 carcinoma$).ti,ab,ot,hw. (110,103)

36. (hepatocarcinoma or hepatoma$).ti,ab,ot,hw. (30,671)

37. or/26-36 (614,221)

38. 25 and 37 (9693)

39. Receptors, Thrombopoietin/ (1355)

40. ((thrombopoietin$ or c-Mpl) adj3 (agonist$ or agent$ or mimetic$ or receptor$)).ti,ab,ot,hw. (1939)

41. (eltrombopag or promacta or revolade or SB 497115 or SB497115 or 496775-61-2).ti,ab,ot,hw,rn. (631)

42. (romiplostim or nplate or remiplistim or amg 531 or amg531 or 267639-76-9).ti,ab,ot,hw,rn. (521)

43. promegapoietin.ti,ab,ot,hw,rn. (12)

44. Platelet Transfusion/ (6808)

45. ((platelet$ or thrombocyt$) adj3 (transfus$ or infus$ or administ$)).ti,ab,ot,hw. (12,351)

46. Splenectomy/ (21,173)

47. (splenectom$ or (spleen adj3 (resect$ or remov$ or surg$))).ti,ab,ot,hw. (30,967)

48. Splenic Artery/ and Embolization, Therapeutic/ (667)

49. ((spleen or splenic or eria lienalis or lienal) adj3 (embolisation or embolization or embolism or embolus or thrombus or embolotherap$ or therap$ occlus$)).ti,ab,ot,hw. (999)

50. Megakaryocytes/ (7273)

51. ((megakaryocyte$ or karyocyte$) adj3 (stimul$ or maturat$ or produc$)).ti,ab,ot,hw. (1186)

52. Thrombopoiesis/ (848)

53. (thrombopoiesi$ or thrombocytopoies$ or megakaryocytopoies$).ti,ab,ot,hw. (2678)

54. ((platelet$ or thrombocyt$) adj3 (produc$ or formation or stimulat$)).ti,ab,ot,hw. (155,25)

55. Portasystemic Shunt, Transjugular Intrahepatic/ (2365)

56. (transjugular intrahepatic portosystemic shunt$ or transjugular intrahepatic porto systemic shunt$ or transjugular intrahepatic portacaval shunt$ or transjugular intrahepatic porta systemic shunt$ or transjugular intrahepatic portasystemic shunt$ or transjugular intrahepatic shunt$ or transjugular intrahepatic stent$ or TIPS or TIPSS).ti,ab,ot,hw. (29,852)

57. or/39-56 (96,920)

58. 38 and 57 (897)

59. 3 or 58 (919)

60. exp animals/ not humans/ (4,540,224)

61. 59 not 60 (894).

MEDLINE 805.

MEDLINE Epub Ahead of Print 18.

MEDLINE In-Process & Other Non-Indexed Citations 71.

MEDLINE Daily Update 0.

Search strategy

1. #41 (#39 AND #40) 255

2. #40 pubstatusaheadofprint OR publisher[sb] OR pubmednotmedline[sb] 3,121,488

3. #39 (#4 OR #38) 3451

4. #38 (#26 AND #37) 3428

5. #37 (#27 OR #28 OR #29 OR #30 OR #31 OR #32 OR #33 OR #34 OR #35 OR #36) 176,154

6. #36 “ Portasystemic Shunt, Transjugular Intrahepatic”[Mesh] OR “transjugular intrahepatic portosystemic shunt”[tiab] OR “transjugular intrahepatic porto systemic shunt”[tiab] OR “transjugular intrahepatic portacaval shunt”[tiab] OR “transjugular intrahepatic porta systemic shunt”[tiab] OR “transjugular intrahepatic portasystemic shunt”[tiab] OR “transjugular intrahepatic shunt”[tiab OR “transjugular intrahepatic stent*”[tiab] OR TIPS[tiab] OR TIPSS[tiab] 29,035

7. #35 (platelet*[tiab] OR thrombocyt*[tiab]) AND (produc*[tiab] OR formation[tiab] OR stimulat*[tiab]) 71,046

8. #34 “ Thrombopoiesis”[Mesh] OR thrombopoiesi*[tiab] OR thrombocytopoies*[tiab] OR megakaryocytopoies*[tiab] 2712

9. #33 “ Megakaryocytes”[Mesh] OR (megakaryocyte*[tiab] OR karyocyte*[tiab]) AND (stimul*[tiab] OR maturat*[tiab] OR produc*[tiab]) 4666

10. #32 (spleen[tiab] OR splenic[tiab] OR “eria lienalis”[tiab] OR lineal[tiab]) AND (embolisation[tiab] OR embolization[tiab] OR embolism[tiab] OR embolus[tiab] OR thrombus[tiab] OR embolotherap*[tiab] OR “therapautic occlusion”[tiab]) 2234

11. #31 “ Splenic Artery”[Mesh] AND “Embolization, Therapeutic”[Mesh] 683

12. #30 “ Splenectomy”[Mesh] OR splenectom*[tiab] OR (spleen[tiab] AND (resect*[tiab] OR remov*[tiab] OR surg*[tiab])) 38,387

13. #29 “ Platelet Transfusion”[Mesh] OR ((platelet*[tiab] OR thrombocyt*[tiab]) AND (transfus*[tiab] OR infus*[tiab] OR administ*[tiab])) 47,154

14. #28 eltrombopag[tiab] OR promacta[tiab] OR revolade[tiab] OR “SB 497115”[tiab] OR SB497115[tiab] OR romiplostim[tiab] OR nplate[tiab] OR remiplistim[tiab] OR “amg 531”[tiab] OR amg531[tiab] OR promegapoietin[tiab] 825

15. #27 “ Receptors, Thrombopoietin”[Mesh] OR (thrombopoietin*[tiab] OR c-Mpl[tiab]) AND (agonist*[tiab] OR agent*[tiab] OR mimetic*[tiab] OR receptor*[tiab]) 1980

16. #26 (#15 AND #25) 11,827

17. #25 (#16 OR #17 OR #18 OR #19 OR #20 OR #21 OR #22 OR #23 OR #24) 649,767

18. #24 (liver*[tiab] OR hepatic[tiab] OR intrahepatic[tiab]) AND carcinoma*[tiab] 75,099

19. #23 haemochromatosis[tiab] OR hemochromatosis[tiab] OR “bronze diabetes”[tiab] OR “bronze diabetic”[tiab] OR “recklinghausen applebaum”[tiab] OR siderochromatosis[tiab] OR “primary biliary cholangitis”[tiab] OR hepatocarcinoma[tiab] OR hepatoma*[tiab] 40,197

20. #22 (liver*[tiab] OR hepatic[tiab] OR intrahepatic[tiab]) AND inflam*[tiab] 57,427

21. #21 (hepatitis[tiab] OR hepatopath*[tiab]) AND (chronic[tiab] OR acute[tiab] OR persistent[tiab] OR “long standing”[tiab] OR “long term”[tiab] OR recurr*[tiab]) 91,895

22. #20 (fibrosis[tiab] OR fibroses[tiab] OR scar*[tiab]) AND (liver*[tiab] OR hepatic[tiab]) 40,403

23. #19 chronic[tiab] AND “destructive cholangitis”[tiab] 118

24. #18 cirrhosis[tiab] OR cirrhosis[tiab] OR cirrhotic[tiab] 95,558

25. #17 “ liver disease”[tiab] OR “liver diseases”[tiab] OR “hepatic disease”[tiab] OR “hepatic diseases”[tiab] OR “intrahepatic disease”[tiab] OR “intrahepatic diseases”[tiab] OR “liver disorder”[tiab] OR “liver disorders”[tiab] OR “hepatic disorder”[tiab] OR “hepatic disorders”[tiab] OR “intrahepatic disorder”[tiab] OR “intrahepatic disorders”[tiab] OR “liver lesion”[tiab] OR “liver lesions”[tiab] OR “hepatic lesion”[tiab] OR “hepatic lesions”[tiab] OR “intrahepatic lesion”[tiab] OR “intrahepatic lesions”[tiab] 108,675

26. #16 “ Liver Diseases”[Mesh] 521,434

27. #15 (#5 OR #6 OR #7 OR #8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #14) 188,201

28. #14 (platelet*[tiab] OR thrombocyte*[tiab]) AND (defici*[tiab] OR reduc*[tiab] OR low[tiab] OR lower[tiab] OR lowest[tiab] OR few[tiab] OR fewer[tiab] OR fewest[tiab] OR decrease[tiab] OR decreases[tiab] OR decreased[tiab] OR defective[tiab] OR destruc*[tiab] OR destroy*[tiab]) 99,513

29. #13 “ immunodeficiency 2” OR immunodeficiency2 OR Imd2 46

30. #12 Moschcowitz[tiab] OR werlhof[tiab] OR “Wiskott-Aldrich Syndrome”[Mesh] OR (wiskott[tiab] AND Aldrich[tiab]) 2664

31. #11 (haemolytic[tiab] OR hemolytic[tiab]) AND (anaemi*[tiab] OR anemi*[tiab]) AND (microangiopath*[tiab]) 1765

32. #10 (hemolysis[tiab] OR haemolysis[tiab]) AND liver[tiab] AND platelet*[tiab] 1247

33. #9 “ HELLP Syndrome”[Mesh] OR “HELLP syndrome” OR “HELLP syndromes” 2583

34. #8 (thrombotic[tiab] AND microangiopath*[tiab]) OR “hemolytic uremic” OR “haemolytic uremic” OR gasser*[tiab] 12,074

35. #7 “ jacobsen syndrome” OR “paris trousseau” OR “kasabach merritt” OR “May Hegglin” OR hemangioma[tiab] OR haemangioma[tiab] 17,717

36. #6 (11q[tiab] OR 11q23[tiab]) AND (disorder*[tiab] OR syndrome*[tiab] OR delet*[tiab] OR Jacobsen[tiab]) 1605

37. #5 “ Thrombocytopenia”[Mesh] OR thrombocytopeni*[tiab] OR thrombocytopaeni*[tiab] OR thrombopeni*[tiab] OR thrombopaeni*[tiab] OR macrothrombocytopeni*[tiab] OR macrothrombocytopaeni*[tiab] 73,938

38. #4 (#2 OR #3) 47

39. #3 lusutrombopag OR mulpleta OR “S 888711” OR S888711 14

40. #2 avatrombopag OR doptelet OR “AKR 501” OR AKR501 OR “AS 1670542” OR AS1670542 OR “E 5501” OR E5501 OR “oralE 5501” OR oralE5501 OR “YM 477” OR YM477 34.

Search strategy

1. avatrombopag/ (64)

2. (avatrombopag or doptelet or AKR 501 or AKR501 or AS 1670542 or AS1670542 or E 5501 or E5501 or oralE 5501 or oralE5501 or YM 477 or YM477 or 570406-98-3 or 677007-74-8).af. (135)

3. lusutrombopag/ (33)

4. (lusutrombopag or mulpleta or S 888711 or S888711 or 1110766-97-6).af. (33)

5. or/1-4 (163)

6. exp thrombocytopenia/ (157,171)

7. (thrombocytopeni$ or thrombocytopaeni$ or thrombopeni$ or thrombopaeni$ or macrothrombocytopeni$ or macrothrombocytopaeni$).ti,ab,ot. (87,986)

8. ((11q or 11q23) adj3 (disorder$ or syndrome$ or delet$ or jacobsen)).ti,ab,ot. (1015)

9. (jacobsen adj3 syndrome$).ti,ab,ot. (187)

10. paris trousseau.ti,ab,ot. (49)

11. kasabach merritt.ti,ab,ot. (793)

12. (hemangioma or haemangioma).ti,ab,ot. (18,275)

13. (thrombotic adj2 (microangiopath$ or micro angiopath$)).ti,ab,ot. (5177)

14. (hemolytic uremic or haemolytic uremic).ti,ab,ot. (7454)

15. gasser$.ti,ab,ot. (1885)

16. (HELLP adj2 syndrome$).ti,ab,ot. (3305)

17. ((hemolysis or haemolysis) adj2 liver adj2 platelet$).ti,ab,ot. (11)

18. May Hegglin.ti,ab,ot. (262)

19. ((haemolytic or hemolytic) adj2 (anaemi$ or anemi$) adj2 (microangiopathic or micro angiopathic)).ti,ab,ot. (2048)

20. moschcowitz.ti,ab,ot. (93)

21. werlhof.ti,ab,ot. (55)

22. (wiskott and aldrich).ti,ab,ot. (2815)

23. (immunodeficiency 2 or immunodeficiency2 or Imd2).ti,ab,ot. (71)

24. ((platelet$ or thrombocyte$) adj3 (defici$ or reduc$ or low or lower or lowest or few or fewer or fewest or decrease or decreases or decreased or defective or destruc$ or destroy$)).ti,ab,ot. (33,439)

25. or/6-24 (221,567)

26. chronic liver disease/ or liver disease/ or liver cirrhosis/ or liver fibrosis/ or chronic hepatitis/ (244,905)

27. ((liver$ or hepat$ or intrahepat$) adj2 (disease$ or disorder$ or lesion$)).ti,ab,ot. (170,572)

28. (cirrhosis or cirrhoses or cirrhotic).ti,ab,ot. (134,378)

29. ((chronic adj3 nonsuppurative destructive cholangitis) or (chronic adj3 non suppurative destructive cholangitis)).ti,ab,ot. (126)

30. ((fibrosis or fibroses or scar$) adj3 (liver$ or hepat$)).ti,ab,ot. (38,165)

31. ((hepatitis or hepatopath$) adj3 (chronic or acute or persistent or long stand$ or long term or recurr$)).ti,ab,ot. (93,566)

32. ((liver$ or hepat$ or intrahepat$) adj3 inflam$).ti,ab,ot. (20,905)

33. (haemochromatosis or hemochromatosis or bronze$ diabet$ or recklinghausen applebaum or siderochromatosis).ti,ab,ot. (9700)

34. primary biliary cholangitis.ti,ab,ot. (1046)

35. liver cell carcinoma/ (136,789)

36. ((liver$ or hepat$ or intrahepat$) adj3 carcinoma$).ti,ab,ot. (122,282)

37. (hepatocarcinoma or hepatoma$).ti,ab,ot. (35,186)

38. or/26-37 (532,951)

39. 25 and 38 (13,778)

40. thrombopoietin receptor/ (1769)

41. ((thrombopoietin$ or c-Mpl) adj3 (agonist$ or agent$ or mimetic$ or receptor$)).ti,ab,ot. (2199)

42. eltrombopag/ (1783)

43. (eltrombopag or promacta or revolade or SB 497115 or SB497115 or 496775-61-2).ti,ab,ot,hw,rn,tn. (1834)

44. romiplostim/ (1552)

45. (romiplostim or nplate or remiplistim or amg 531 or amg531 or 267639-76-9).ti,ab,ot,hw,rn,tn. (1698)

46. promegapoietin.ti,ab,ot,hw,rn,tn,dj. (25)

47. thrombocyte transfusion/ (17,075)

48. ((platelet$ or thrombocyt$) adj3 (transfus$ or infus$ or administ$)).ti,ab,ot. (13,882)

49. splenectomy/ (32,248)

50. (splenectom$ or (spleen adj2 (resect$ or remov$ or surg$))).ti,ab,ot. (27,238)

51. spleen artery/ and exp artificial embolism/ (457)

52. ((spleen or splenic or eria lienalis or lienal) adj3 (embolisation or embolization or embolism or embolus or thrombus or embolotherap$ or therap$ occlus$)).ti,ab,ot. (1536)

53. megakaryocyte/ and (stimulation/ or cell maturation/) (1079)

54. ((megakaryocyte$ or karyocyte$) adj3 (stimul$ or maturat$ or produc$)).ti,ab,ot. (1555)

55. thrombocytopoiesis/ (4137)

56. (thrombopoiesi$ or thrombocytopoies$ or megakaryocytopoies$).ti,ab,ot. (2708)

57. ((platelet$ or thrombocyt$) adj3 (produc$ or formation or stimulat$)).ti,ab,ot. (20,991)

58. transjugular intrahepatic portosystemic shunt/ (3426)

59. (transjugular intrahepatic portosystemic shunt$ or transjugular intrahepatic porto systemic shunt$ or transjugular intrahepatic portacaval shunt$ or transjugular intrahepatic porta systemic shunt$ or transjugular intrahepatic portasystemic shunt$ or transjugular intrahepatic shunt$ or transjugular intrahepatic stent$ or TIPS).ti,ab,ot. (35,802)

60. or/40-59 (124,052)

61. 39 and 60 (1558)

62. 5 or 61 (1651)

63. animal/ or animal experiment/ (3,692,962)

64. (rat or rats or mouse or mice or murine or rodent or rodents or hamster or hamsters or pig or pigs or porcine or rabbit or rabbits or animal or animals or dogs or dog or cats or cow or bovine or sheep or ovine or monkey or monkeys).ti,ab,ot. (4,424,329)

65. 63 or 64 (5,722,776)

66. exp human/ or human experiment/ (19,263,219)

67. 65 not (65 and 66) (4,428,740)

68. 62 not 67 (1614).

Search strategy

1. #1 avatrombopag or doptelet or “AKR 501” or AKR501 or “AS 1670542” or AS1670542 or “E 5501” or E5501 or “oralE 5501” or oralE5501 or “YM 477” or YM477 47

2. #2 lusutrombopag or mulpleta or “S 888711” or S888711 11

3. #3 #1 or #2 58

4. #4 MeSH descriptor: [Thrombocytopenia] explode all trees 1121

5. #5 (thrombocytopeni* or thrombocytopaeni* or thrombopeni* or thrombopaeni* or macrothrombocytopeni* or macrothrombocytopaeni*):ti,ab,kw 7871

6. #6 ((11q or 11q23) NEAR/3 (disorder* or syndrome* or delet* or jacobsen)):ti,ab,kw 42

7. #7 (jacobsen NEAR/3 syndrome*):ti,ab,kw 0

8. #8 “paris trousseau” 2

9. #9 “kasabach merritt” 4

10. #10 (hemangioma or haemangioma):ti,ab,kw 298

11. #11 (thrombotic NEAR/2 (microangiopath* or micro angiopath*)):ti,ab,kw 70

12. #12 (hemolytic uremic or haemolytic uremic) 135

13. #13 (gasser*):ti,ab,kw 100

14. #14 MeSH descriptor: [HELLP Syndrome] this term only 45

15. #15 (HELLP NEAR/2 syndrome*):ti,ab,kw 130

16. #16 ((hemolysis or haemolysis) NEAR/3 platelet*):ti,ab,kw 9

17. #17 “May Hegglin” 0

18. #18 ((haemolytic or hemolytic) NEAR/2 (anaemi* or anemi*) NEAR/2 (microangiopathic or micro angiopathic)):ti,ab,kw 16

19. #19 (moschcowitz):ti,ab,kw 1

20. #20 (werlhof):ti,ab,kw 0

21. #21 MeSH descriptor: [Wiskott-Aldrich Syndrome] this term only 6

22. #22 (wiskott and aldrich):ti,ab,kw 24

23. #23 (“immunodeficiency 2” or immunodeficiency2 or Imd2):ti,ab,kw 1

24. #24 ((platelet* or thrombocyte*) NEAR/3 (defici* or reduc* or low or lower or lowest or few or fewer or fewest or decrease or decreases or decreased or defective or destruc* or destroy*)):ti,ab,kw 2416

25. #25 #4 OR #5 OR #6 OR #7 OR #8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR #22 OR #23 OR #24 10,523

26. #26 MeSH descriptor: [Liver Diseases] explode all trees 13,186

27. #27 ((liver* or hepat* or intrahepat*) NEAR/2 (disease* or disorder* or lesion*)):ti,ab,kw 7716

28. #28 (cirrhosis or cirrhoses or cirrhotic):ti,ab,kw 8338

29. #29 (chronic NEAR/3 destructive cholangitis):ti,ab,kw 1

30. #30 ((fibrosis or fibroses) NEAR/3 (liver* or hepat*)):ti,ab,kw 1583

31. #31 ((hepatitis or hepatopath*) NEAR/3 (chronic or acute or persistent or long stand* or long term or recurr*)):ti,ab,kw 9152

32. #32 ((liver or hepat* or intrahepat*) NEAR/3 inflam*):ti,ab,kw 663

33. #33 (haemochromatosis or hemochromatosis or bronze* diabet* or recklinghausen applebaum or siderochromatosis):ti,ab,kw 96

34. #34 primary biliary cholangitis:ti,ab,kw 287

35. #35 ((liver* or hepat* or intrahepat*) NEAR/3 carcinoma*):ti,ab,kw 3866

36. #36 (hepatocarcinoma or hepatoma*):ti,ab,kw 172

37. #37 #26 OR #27 OR #28 OR #29 OR #30 OR #31 OR #32 OR #33 OR #34 OR #35 OR #36 27,420

38. #38 #25 and #37 787

39. #39 MeSH descriptor: [Receptors, Thrombopoietin] this term only 45

40. #40 ((thrombopoietin* or c-Mpl or mpl) NEAR/3 (agonist* or agent* or mimetic* or receptor*)):ti,ab,kw 196

41. #41 (eltrombopag or promacta or revolade or “SB 497115” or SB497115):ti,ab,kw 198

42. #42 (romiplostim or nplate or remiplistim or “amg 531” or amg531):ti,ab,kw 157

43. #43 promegapoietin 0

44. #44 MeSH descriptor: [Platelet Transfusion] this term only 300

45. #45 ((platelet* or thrombocyt*) NEAR/3 (transfus* or infus* or administ*)):ti,ab,kw 3034

46. #46 MeSH descriptor: [Splenectomy] this term only 176

47. #47 (splenectom* or (spleen NEAR/2 (resect* or remov* or surg*))):ti,ab,kw 617

48. #48 MeSH descriptor: [Splenic Artery] this term only 18

49. #49 ((spleen or splenic or eria lienalis or lienal) NEAR/3 (embolisation or embolization or embolism or embolus or thrombus or embolotherap* or “therap* occlus*”)):ti,ab,kw 38

50. #50 MeSH descriptor: [Megakaryocytes] this term only 28

51. #51 ((megakaryocyte* or karyocyte*) NEAR/3 (stimul* or maturat* or produc*)):ti,ab,kw 27

52. #52 MeSH descriptor: [Thrombopoiesis] this term only 8

53. #53 (thrombopoiesi* or thrombocytopoies* or megakaryocytopoies*):ti,ab,kw 89

54. #54 ((platelet* or thrombocyt*) NEAR/3 (produc* or formation or stimulat*)):ti,ab,kw 848

55. #55 MeSH descriptor: [Portasystemic Shunt, Transjugular Intrahepatic] this term only 94

56. #56 (“transjugular intrahepatic portosystemic shunt*” or “transjugular intrahepatic porto systemic shunt*” or “transjugular intrahepatic portacaval shunt*” or “transjugular intrahepatic porta systemic shunt*” or “transjugular intrahepatic portasystemic shunt*” or “transjugular intrahepatic shunt*” or “transjugular intrahepatic stent*” or TIPS or TIPSS):ti,ab,kw 1028

57. #57 #39 OR #40 OR #41 OR #42 OR #43 OR #44 OR #45 OR #46 OR #47 OR #48 OR #49 OR #50 OR #51 OR #52 OR #53 OR #54 OR #55 OR #56 5620

58. #58 #38 and #57 110

59. #59 #3 or #58 146.

Cochrane Database of Systematic Reviews 8.

Cochrane Central Register of Controlled Trials 138.

Search strategy

Database last updated 24 January 2019, 13:06.

Search strategy

Title/Abstract: avatrombopag OR doptelet OR “AKR 501” OR AKR501 OR “AS 1670542” OR AS1670542 OR “E 5501” OR E5501 OR “oralE 5501” OR oralE5501 OR “YM 477” OR YM477 OR lusutrombopag OR mulpleta OR “S 888711” OR S888711

OR

Title/Abstract: (thrombocytopeni* OR thrombocytopaeni* OR thrombopeni* OR thrombopaeni* OR macrothrombocytopeni* OR macrothrombocytopaeni*) AND (“liver* disease*” OR “hepatic disease*” OR “liver* disorder*” OR “hepatic disorder*” OR “liver* lesion*” OR “hepatic lesion*” OR cirrho* OR fibros* OR “liver* carcinoma*” OR “hepatic carcinoma*”)

OR

Title/Abstract: ((platelet* OR thrombocyte*) AND (defici* OR reduc* OR low OR lower OR lowest OR few OR fewer OR fewest OR decrease OR decreases OR decreased OR defective OR destruc* OR destroy*)) AND (“liver* disease*” OR “hepatic disease*” OR “liver* disorder*” OR “hepatic disorder*” OR “liver* lesion*” OR “hepatic lesion*” OR cirrho* OR fibros* OR “liver* carcinoma*” OR “hepatic carcinoma*”).

Records retrieved: 212.

Search strategy

1. (avatrombopag or doptelet or AKR 501 or AKR501 or AS 1670542 or AS1670542 or E 5501 or E5501 or oralE 5501 or oralE5501 or YM 477 or YM477 or 570406-98-3) 2

2. (lusutrombopag or mulpleta or S 888711 or S888711 or 1110766-97-6) 0

3. 1 OR #2 2

4. MeSH DESCRIPTOR Thrombocytopenia EXPLODE ALL TREES 107

5. (thrombocytopeni* or thrombocytopaeni* or thrombopeni* or thrombopaeni* or macrothrombocytopeni* or macrothrombocytopaeni*) 369

6. (11q or 11q23) 0

7. (jacobsen near3 syndrome*) 0

8. (paris trousseau) 0

9. (kasabach merritt) 1

10. (hemangioma or haemangioma) 34

11. (thrombotic near2 (microangiopath* or micro angiopath*)) 0

12. (hemolytic uremic or haemolytic uremic) 14

13. (gasser*) 4

14. MeSH DESCRIPTOR HELLP Syndrome EXPLODE ALL TREES 5

15. (HELLP near2 syndrome*) 11

16. ((hemolysis or haemolysis) near2 liver near2 platelet*) 2

17. (May Hegglin) 0

18. ((haemolytic or hemolytic) near (anaemi* or anemi*)) 18

19. (microangiopath* near thrombotic) 0

20. (moschcowitz or werlhof) 0

21. MeSH DESCRIPTOR Wiskott-Aldrich Syndrome EXPLODE ALL TREES 0

22. (wiskott and Aldrich) 5

23. (immunodeficiency 2 or immunodeficiency2 or Imd2) 1

24. ((platelet* or thrombocyte*) near3 (defici* or reduc* or low or lower or lowest or few or fewer or fewest or decrease or decreases or decreased or defective or destruc* or destroy*)) 24

25. #4 OR #5 OR #6 OR #7 OR #8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR #22 OR #23 OR #24 467

26. MeSH DESCRIPTOR Liver Diseases EXPLODE ALL TREES 1983

27. ((liver or hepat* or intrahepat*) near (disease* or disorder* or lesion*)) 723

28. (cirrhosis or cirrhoses or cirrhotic) 643

29. (chronic near3 cholangitis) 1

30. ((fibrosis or fibroses or scar*) near3 (liver* or hepat*)) 49

31. ((hepatitis or hepatopath*) near3 (chronic or acute or persistent or long stand* or long term or recurr*)) 547

32. ((liver* or hepat* or intrahepat*) near3 inflam*) 20

33. (haemochromatosis or hemochromatosis or bronze* diabet* or recklinghausen applebaum or siderochromatosis) 37

34. (primary biliary cholangitis) 1

35. ((liver* or hepat* or intrahepat*) near3 carcinoma*) 516

36. (hepatocarcinoma or hepatoma*) 14

37. #26 OR #27 OR #28 OR #29 OR #30 OR #31 OR #32 OR #33 OR #34 OR #35 OR #36 2427

38. #25 AND #37 36

39. #3 OR #38 37.

Database of Abstracts of Reviews of Effects  19.

Health Technology Assessment database 7.

NHS Economic Evaluation Database 11.

Search strategy

Search strategy

Search strategy

((MH:c15.378.140.855 OR MH:c15.378.100.100.970 OR thrombocytopeni* OR thrombocytopaeni* OR thrombopeni* OR thrombopaeni* OR macrothrombocytopeni* OR macrothrombocytopaeni* OR trombocitopeni* OR ((platelet* OR thrombocyte*) AND (defici* OR reduc* OR low OR lower OR lowest OR few OR fewer OR fewest OR decrease OR decreases OR decreased OR defective OR destruc* OR destroy*))) AND (MH:C06.552 or “liver disease” OR “liver diseases” OR “hepatic disease” OR “hepatic diseases” OR “intrahepatic disease” OR “intrahepatic diseases” OR “liver disorder” OR “liver disorders” OR “hepatic disorder” OR “hepatic disorders” OR “intrahepatic disorder” OR “intrahepatic disorders” OR “liver lesion” OR “liver lesions” OR “hepatic lesion” OR “hepatic lesions” OR “intrahepatic lesion” OR “intrahepatic lesions” OR hepatopatias OR cirrhosis OR cirrhoses OR cirrhotic OR cirrose OR cirrosis OR ((liver$ OR hepatic OR intrahepatic) AND carcinoma$))) OR (avatrombopag OR doptelet OR “AKR 501” OR akr501 OR “AS 1670542” OR as1670542 OR “E 5501” OR e5501 OR “oralE 5501” OR orale5501 OR “YM 477” OR ym477 OR lusutrombopag OR mulpleta OR “S 888711” OR s888711).

Search limited to non-MEDLINE databases:

• LILACS (89)

• IBECS (45)

• BINACIS (13)

• CUMED (4)

• MedCarib (4)

• LIS -Health Information Locator (1)

• Index Psychology – Theses (1).

Search strategy

1. (avatrombopag or doptelet or AKR 501 or AKR501 or AS 1670542 or AS1670542 or E 5501 or E5501 or oralE 5501 or oralE5501 or YM 477 or YM477).af. (15)

2. (lusutrombopag or mulpleta or S 888711 or S888711 or 1110766-97-6).af. (10)

3. 1 or 2 (25)

4. exp thrombocytopenia/ (19,173)

5. (thrombocytopeni$ or thrombocytopaeni$ or thrombopeni$ or thrombopaeni$ or macrothrombocytopeni$ or macrothrombocytopaeni$).ti,ab,hw. (18,543)

6. ((11q or 11q23) adj3 (disorder$ or syndrome$ or delet$ or jacobsen)).ti,ab,hw. (132)

7. (jacobsen adj3 syndrome$).ti,ab,hw. (41)

8. (paris trousseau or kasabach merritt or hemangioma or haemangioma).ti,ab,hw. (2487)

9. (thrombotic adj2 (microangiopath$ or micro angiopath$)).ti,ab,hw. (1515)

10. (hemolytic uremic or haemolytic uremic or gasser$).ti,ab,hw. (643)

11. hellp syndrome/ (410)

12. (HELLP adj2 syndrome$).ti,ab,hw. (415)

13. ((hemolysis or haemolysis) adj2 liver adj2 platelet$).ti,ab,hw. (0)

14. May Hegglin.ti,ab,hw. (10)

15. ((haemolytic or hemolytic) adj2 (anaemi$ or anemi$) adj2 (microangiopathic or micro angiopathic)).ti,ab,hw. (77)

16. (moschcowitz or werlhof or (wiskott and Aldrich)).ti,ab,hw. (468)

17. wiskott-aldrich syndrome/ (460)

18. (immunodeficiency 2 or immunodeficiency2 or Imd2).ti,ab,hw. (0)

19. ((platelet$ or thrombocyte$) adj3 (defici$ or reduc$ or low or lower or lowest or few or fewer or fewest or decrease or decreases or decreased or defective or destruc$ or destroy$)).ti,ab,hw. (1916)

20. or/4-19 (24,421)

21. exp Liver Diseases/ (70,505)

22. ((liver$ or hepat$ or intrahepat$) adj2 (disease$ or disorder$ or lesion$)).ti,ab,hw. (27,653)

23. (cirrhosis or cirrhoses or cirrhotic).ti,ab,hw. (14,624)

24. (chronic adj3 destructive cholangitis).ti,ab,hw. (3)

25. ((fibrosis or fibroses or scar$) adj3 (liver$ or hepat$)).ti,ab,hw. (4585)

26. ((hepatitis or hepatopath$) adj3 (chronic or acute or persistent or long stand$ or long term or recurr$)).ti,ab,hw. (8107)

27. ((liver$ or hepat$ or intrahepat$) adj3 inflam$).ti,ab,hw. (1780)

28. (haemochromatosis or hemochromatosis or bronze$ diabet$ or recklinghausen applebaum or siderochromatosis).ti,ab,hw. (1151)

29. primary biliary cholangitis.ti,ab,hw. (230)

30. ((liver$ or hepat$ or intrahepat$) adj3 carcinoma$).ti,ab,hw. (13,730)

31. (hepatocarcinoma or hepatoma$).ti,ab,hw. (900)

32. or/21-31 (89,117)

33. 20 and 32 (2415)

34. thrombopoietin/ (1145)

35. ((thrombopoietin$ or c-Mpl) adj3 (agonist$ or agent$ or mimetic$ or receptor$)).ti,ab,hw. (206)

36. (eltrombopag or promacta or revolade or SB 497115 or SB497115 or 496775-61-2).ti,ab,hw. (279)

37. (romiplostim or nplate or remiplistim or amg 531 or amg531 or 267639-76-9).ti,ab,hw. (256)

38. promegapoietin.ti,ab,hw. (0)

39. ((platelet$ or thrombocyt$) adj3 (transfus$ or infus$ or administ$)).ti,ab,hw. (896)

40. (splenectom$ or (spleen adj3 (resect$ or remov$ or surg$))).ti,ab,hw. (1139)

41. ((spleen or splenic or eria lienalis or lienal) adj3 (embolisation or embolization or embolism or embolus or thrombus or embolotherap$ or therap$ occlus$)).ti,ab,hw. (141)

42. megakaryocytes/ (2226)

43. ((megakaryocyte$ or karyocyte$) adj3 (stimul$ or maturat$ or produc$)).ti,ab,hw. (72)

44. (thrombopoiesi$ or thrombocytopoies$ or megakaryocytopoies$).ti,ab,hw. (114)

45. ((platelet$ or thrombocyt$) adj3 (produc$ or formation or stimulat$)).ti,ab,hw. (944)

46. (transjugular intrahepatic portosystemic shunt$ or transjugular intrahepatic porto systemic shunt$ or transjugular intrahepatic portacaval shunt$ or transjugular intrahepatic porta systemic shunt$ or transjugular intrahepatic portasystemic shunt$ or transjugular intrahepatic shunt$ or transjugular intrahepatic stent$ or TIPS or TIPSS).ti,ab,hw. (2278)

47. or/34-46 (8073)

48. 33 and 47 (221)

49. 3 or 48 (227).

Search strategy

(avatrombopag OR doptelet OR lusutrombopag OR mulpleta) OR ((thrombocytopeni* OR thrombocytopaeni* OR thrombopeni* OR thrombopaeni* OR macrothrombocytopeni* OR macrothrombocytopaeni* OR (platelet* OR thrombocyte*) AND (defici* OR reduc* OR low OR lower OR lowest OR few OR fewer OR fewest OR decrease OR decreases OR decreased OR defective OR destruc* OR destroy*)) AND (“liver disease*” OR “hepatic disease*” OR “liver disorder*” OR “hepatic disorder*” OR “liver lesion*” OR “hepatic lesion*” OR cirrhosis OR cirrhosis OR cirrhotic OR “liver* carcinoma*” OR “hepatic carcinoma*”)).

Records retrieved: 40.

Search strategy

IDEAS search interface.

(avatrombopag | doptelet | lusutrombopag | mulpleta | thrombocytopenia | thrombocytopenic | thrombocytopaenia | thrombocytopaenic | thrombopenia | thrombopenic | thrombopaenia | thrombopaenic).

Records retrieved: 14.

Search strategy

(avatrombopag OR doptelet OR “AKR 501” OR AKR501 OR “AS 1670542” OR AS1670542 OR “E 5501” OR E5501 OR “oralE 5501” OR oralE5501 OR “YM 477” OR YM477 OR lusutrombopag OR mulpleta OR “S 888711” OR S888711) OR ((thrombocytopenia OR thrombocytopenic OR thrombocytopaenia OR thrombocytopaenic OR thrombopenia OR thrombopenic OR thrombopaenia OR thrombopaenic OR macrothrombocytopenia OR macrothrombocytopenic OR macrothrombocytopaenia OR macrothrombocytopaenic) AND (liver OR hepatic OR intrahepatic OR cirrhosis OR cirrhoses OR cirrhotic)).

319 studies found.

Advanced search option

Drugs@FDA searched

Search strategy

Search strategy

(avatrombopag OR doptelet OR lusutrombopag OR mulpleta) OR ((thrombocytopeni* OR thrombocytopaeni* OR thrombopeni* OR thrombopaeni*) AND (liver* OR hepat*) AND (thrombopoietin* receptor* OR thrombopoietin* agonist* OR thrombopoietin* agent* OR eltrombopag OR promacta OR revolade OR romiplostim OR nplate OR platelet transfus* OR platelet infus* OR platelet admin* OR thrombocyt* transf* OR thrombocyt* infus* OR thrombocyt* admin* OR splenectom* OR spleen resect* OR spleen remov* OR spleen surger* OR emboli* OR thrombus OR megakaryocyte* OR karyocyte* OR thrombopoiesis OR thrombocytopoies OR megakaryocytopoies OR platelet produc* OR thrombocyt* produc* OR platelet forma* OR thrombocyt* forma* OR platelet stimul* OR thrombocyt* stimul*)).

Records retrieved: 37.

Search strategy

(avatrombopag OR doptelet OR lusutrombopag OR mulpleta) OR ((thrombocytopeni* OR thrombocytopaeni* OR thrombopeni* OR thrombopaeni*) AND ((thrombopoietin* NEAR receptor*) OR (thrombopoietin* NEAR agonist*) OR (thrombopoietin* NEAR agent*) OR eltrombopag OR promacta OR revolade OR romiplostim OR nplate OR (platelet NEAR transfus*) OR (platelet NEAR infus*) OR (platelet NEAR admin*) OR (thrombocyt* NEAR transf*) OR (thrombocyt* NEAR infus*) OR (thrombocyt* NEAR admin*) OR splenectom* OR (spleen NEAR resect*) OR (spleen NEAR remov*) OR (spleen NEAR surger*) OR emboli* OR thrombus OR megakaryocyte* OR karyocyte* OR thrombopoiesis OR thrombocytopoies OR megakaryocytopoies OR (platelet NEAR produc*) OR (thrombocyt* NEAR produc*) OR (platelet NEAR forma*) OR (thrombocyt* NEAR forma*) OR (platelet NEAR stimul*) OR (thrombocyt* NEAR stimul*)).

Records retrieved: 41.

Search strategy

1. Keyword: avatrombopag

2. Keyword: doptelet

3. Keyword: lusutrombopag

4. Keyword: mulpleta

5. Keyword: thrombocytopeni* liver* thrombopoietin*

6. Keyword: thrombocytopaeni* liver* thrombopoietin*

7. Keyword: thrombopeni* liver* thrombopoietin*

8. Keyword: thrombopaeni* liver* thrombopoietin*

9. Keyword: thrombocytopeni* hepati* thrombopoietin*

10. Keyword: thrombocytopaeni* hepati* thrombopoietin*

11. Keyword: thrombopeni* hepati* thrombopoietin*

12. Keyword: thrombopaeni* hepati* thrombopoietin*

13. Keyword: thrombocytopeni* liver* eltrombopag

14. Keyword: thrombocytopaeni* liver* eltrombopag

15. Keyword: thrombopeni* liver* eltrombopag

16. Keyword: thrombopaeni* liver* eltrombopag

17. Keyword: thrombocytopeni* hepati* eltrombopag

18. Keyword: thrombocytopaeni* hepati* eltrombopag

19. Keyword: thrombopeni* hepati* eltrombopag

20. Keyword: thrombopaeni* hepati* eltrombopag

21. Keyword: thrombocytopeni* liver* romiplostim

22. Keyword: thrombocytopaeni* liver* romiplostim

23. Keyword: thrombopeni* liver* romiplostim

24. Keyword: thrombopaeni* liver* romiplostim

25. Keyword: thrombocytopeni* hepati* romiplostim

26. Keyword: thrombocytopaeni* hepati* romiplostim

27. Keyword: thrombopeni* hepati* romiplostim

28. Keyword: thrombopaeni* hepati* romiplostim

29. Keyword: thrombocytopeni* liver* “platelet transfus*"

30. Keyword: thrombocytopaeni* liver* “platelet transfus*"

31. Keyword: thrombopeni* liver* “platelet transfus*"

32. Keyword: thrombopaeni* liver* “platelet transfus*"

33. Keyword: thrombocytopeni* hepati* “platelet transfus*"

34. Keyword: thrombocytopaeni* hepati* “platelet transfus*"

35. Keyword: thrombopeni* hepati* “platelet transfus*"

36. Keyword: thrombopaeni* hepati* “platelet transfus*"

37. Keyword: thrombocytopeni* liver* splenectom*

38. Keyword: thrombocytopaeni* liver* splenectom*

39. Keyword: thrombopeni* liver* splenectom*

40. Keyword: thrombopaeni* liver* splenectom*

41. Keyword: thrombocytopeni* hepati* splenectom*

42. Keyword: thrombocytopaeni* hepati* splenectom*

43. Keyword: thrombopeni* hepati* splenectom*

44. Keyword: thrombopaeni* hepati* splenectom*

45. Keyword: thrombocytopeni* liver* “splenic emboli*"

46. Keyword: thrombocytopaeni* liver* “splenic emboli*"

47. Keyword: thrombopeni* liver* “splenic emboli*"

48. Keyword: thrombopaeni* liver* “splenic emboli*"

49. Keyword: thrombocytopeni* hepati* “splenic emboli*"

50. Keyword: thrombocytopaeni* hepati* “splenic emboli*"

51. Keyword: thrombopeni* hepati* “splenic emboli*"

52. Keyword: thrombopaeni* hepati* “splenic emboli*"

53. Keyword: thrombocytopeni* liver* megakaryocyte*

54. Keyword: thrombocytopaeni* liver* megakaryocyte*

55. Keyword: thrombopeni* liver* megakaryocyte*

56. Keyword: thrombopaeni* liver* megakaryocyte*

57. Keyword: thrombocytopeni* hepati* megakaryocyte*

58. Keyword: thrombocytopaeni* hepati* megakaryocyte*

59. Keyword: thrombopeni* hepati* megakaryocyte*

60. Keyword: thrombopaeni* hepati* megakaryocyte*.

Records retrieved: 90.

Search strategy

1. quality-adjusted life years/ or quality of life/ (179,815)

2. (sf36 or sf 36 or sf-36 or short form 36 or shortform 36 or sf thirtysix or sf thirty six or shortform thirtysix or shortform thirty six or short form thirty six or short form thirtysix or short form thirty six).ti,ab,ot. (23,334)

3. (sf6 or sf 6 or sf-6 or short form 6 or shortform 6 or sf six or sfsix or shortform six or short form six).ti,ab,ot. (1938)

4. (sf12 or sf 12 or sf-12 or short form 12 or shortform 12 or sf twelve or sftwelve or shortform twelve or short form twelve).ti,ab,ot. (5044)

5. (sf6D or sf 6D or sf-6D or short form 6D or shortform 6D or sf six D or sfsixD or shortform six D or short form six D).ti,ab,ot. (745)

6. (sf20 or sf 20 or sf-20 or short form 20 or shortform 20 or sf twenty or sftwenty or shortform twenty or short form twenty).ti,ab,ot. (386)

7. (sf8 or sf 8 or sf-8 or short form 8 or shortform 8 or sf eight or sfeight or shortform eight or short form eight).ti,ab,ot. (488)

8. "health related quality of life".ti,ab,ot. (37,648)

9. (Quality adjusted life or Quality-adjusted-life).ti,ab,ot. (11,042)

10. "assessment of quality of life".ti,ab,ot. (1664)

11. (euroqol or euro qol or eq5d or eq 5d).ti,ab,ot. (9022)

12. (hql or hrql or hqol or h qol or hrqol or hr qol).ti,ab,ot. (17,843)

13. (hye or hyes).ti,ab,ot. (63)

14. health$ year$ equivalent$.ti,ab,ot. (40)

15. (hui or hui1 or hui2 or hui3 or hui4 or hui-4 or hui-1 or hui-2 or hui-3).ti,ab,ot. (1339)

16. (quality time or qwb or quality of well being or “quality of wellbeing” or “index of wellbeing” or “index of well being”).ti,ab,ot,hw. (817)

17. (Disability adjusted life or Disability-adjusted life or health adjusted life or health-adjusted life or “years of healthy life” or healthy years equivalent or “years of potential life lost” or “years of health life lost”).ti,ab,ot. (3371)

18. (QALY$ or DALY$ or HALY$ or YHL or HYES or YPLL or YHLL or qald$ or qale$ or qtime$ or AQoL$).ti,ab,ot. (12,572)

19. (timetradeoff or time tradeoff or time trade-off or time trade off or TTO or Standard gamble$ or “willingness to pay”).ti,ab,ot. (6642)

20. 15d.ti,ab,ot. (1625)

21. (HSUV$ or health state$ value$ or health state$ preference$ or HSPV$).ti,ab,ot. (373)

22. (utilit$ adj3 (“quality of life” or valu$ or scor$ or measur$ or health or life or estimat$ or elicit$ or disease$)).ti,ab,ot. (10,844)

23. (utilities or disutili$).ti,ab,ot. (6548)

24. (CLDQ or Chronic Liver Disease Questionnaire$).ti,ab,ot,hw. (161)

25. (LDSI or Liver Disease Symptom Index$).ti,ab,ot,hw. (18)

26. (LDQOL or Liver Disease Quality of Life Questionnaire$).ti,ab,ot,hw. (26)

27. (EORTC QLQ-HCC18 or EORTC QLQ-LMC21).ti,ab,ot,hw. (13)

28. (PLD-Q or Polycystic Liver Disease Questionnaire$).ti,ab,ot,hw. (5)

29. or/1-28 (228,242)

30. animals/ not (animals/ and humans/) (4,507,390)

31. 29 not 30 (226,165)

32. letter.pt. (1,013,622)

33. editorial.pt. (479,604)

34. historical article.pt. (349,760)

35. or/32-34 (1,824,832)

36. 31 not 35 (217,667)

37. exp Thrombocytopenia/ (45,457)

38. (thrombocytopeni$ or thrombocytopaeni$ or thrombopeni$ or thrombopaeni$ or macrothrombocytopeni$ or macrothrombocytopaeni$).ti,ab,ot,hw. (69,081)

39. ((11q or 11q23) adj3 (disorder$ or syndrome$ or delet$ or jacobsen)).ti,ab,ot,hw. (574)

40. (jacobsen adj3 syndrome$).ti,ab,ot,hw. (129)

41. paris trousseau.ti,ab,ot,hw. (30)

42. kasabach merritt.ti,ab,ot,hw. (704)

43. (hemangioma or haemangioma).ti,ab,ot,hw. (32,339)

44. (thrombotic adj2 (microangiopath$ or micro angiopath$)).ti,ab,ot,hw. (3354)

45. (hemolytic uremic or haemolytic uremic).ti,ab,ot,hw. (7663)

46. gasser$.ti,ab,ot,hw. (1689)

47. HELLP Syndrome/ (1709)

48. (HELLP adj2 syndrome$).ti,ab,ot,hw. (2561)

49. ((hemolysis or haemolysis) adj2 liver adj2 platelet$).ti,ab,ot,hw. (7)

50. May Hegglin.ti,ab,ot,hw. (221)

51. ((haemolytic or hemolytic) adj2 (anaemi$ or anemi$) adj2 (microangiopathic or micro angiopathic)).ti,ab,ot,hw. (1411)

52. moschcowitz.ti,ab,ot,hw. (107)

53. werlhof.ti,ab,ot,hw. (120)

54. Wiskott-Aldrich Syndrome/ (1428)

55. (wiskott and Aldrich).ti,ab,ot,hw. (3312)

56. (immunodeficiency 2 or immunodeficiency2 or Imd2).ti,ab,ot,hw. (44)

57. ((platelet$ or thrombocyte$) adj3 (defici$ or reduc$ or low or lower or lowest or few or fewer or fewest or decrease or decreases or decreased or defective or destruc$ or destroy$)).ti,ab,ot,hw. (22,231)

58. or/37-57 (132,417)

59. 36 and 58 (595).