Egyptian evidence-based consensus on clinical practice guidelines for the diagnosis and treat-to-target management of macrophage activation syndrome in children

Background Macrophage activation syndrome (MAS) is a severe life-threatening hyperinflammatory state with uncontrolled activation and proliferation of macrophages and T-lymphocytes. MAS has variable causes and risk factors. Early diagnosis and optimum management could be lifesaving. Our aim was to develop a consensus, evidence-based recommendations for the diagnosis, evaluation, and treat-to-target management of pediatric MAS. This study was carried out to achieve an Egyptian expert consensus on a treat-to-target management strategy for MAS using the Delphi technique. The multistep process strategy was used in developing a consensus, evidence-based treatment guidelines for MAS, started by developing 7 key clinical questions by a scientific committee according to the Patient/Population, Intervention, Comparison, and Outcomes (PICO) approach. The core leadership team identified pediatric rheumatology clinicians and researchers throughout Egypt. To generate evidence for MAS management, an evidence-based, systematic literature review was done. To obtain a consensus, the Delphi procedure (3 rounds) was used. Results Twenty-three expert panel participated in the 3 rounds with a response rate of 100%. A total of 19 recommendations, categorized into 2 sections (11 in the diagnosis section and 8 in management), were obtained. The agreement with the recommendations (ranks 7–9) ranged from 86.9 to 95.7%. The consensus was reached (i.e., ≥75% of respondents strongly agreed or agreed) on all the clinical standards. Algorithms for management have been also developed. Conclusion This was an expert, consensus recommendation for the diagnosis and treat to target of MAS, based on the best available evidence and expert opinion. The guidelines fill a gap in the literature as it presents a T2T approach for MAS.


Background
Macrophage activation syndrome (MAS) is also considered a type of secondary hemophagocytic lymphohistiocytosis (sHLH) associated with rheumatological diseases [1]. MAS is characterized by uncontrolled macrophage and T-lymphocyte activation and proliferation, as well as a significant increase in circulating cytokines such as IFN-gamma and GM-CSF [2]. Furthermore, its pathophysiology is still unknown, and various factors, including genetic factors, may play a role in the etiology and pathogenesis of MAS [3] demonstrating the link between MAS risk and interleukin-18. The most prevalent clinical signs of MAS include prolonged high fever, hepatosplenomegaly, neurologic impairment, and hemorrhagic abnormalities. Pancytopenia, high serum liver enzymes, triglycerides, lactate dehydrogenase, and ferritin, as well as low fibrinogen levels, are all common laboratory findings. Despite the fact that macrophage hemophagocytosis is a common finding in bone marrow examinations, this finding may be missed in the early stages of MAS [4,5]. The exact incidence of MAS is still unknown because of the wide range of clinical signs and the possibility of sustaining episodes of the disease that may pass unnoticed clinically [2]. Similar to its pathogenesis, treatment of MAS has been a matter of controversy. Parenteral treatment of high doses of corticosteroids has typically been used to treat MAS; nonetheless, occasional fatalities have been seen even among patients receiving high doses of corticosteroids. High-dose intravenous immunoglobulins, cyclophosphamide, plasma exchange, and etoposide have all been used with conflicting results [6][7][8].
Despite the fact that early diagnosis is associated with better outcomes, early detection of MAS remains a challenge. This has been attributed to the lack of a single pathognomonic trait or even a set of universal diagnostic criteria for MAS. Similarly, early management is a true life-saving competition [9]. There are few international guidelines for the management of MAS; the most recent, the Classification Criteria for Macrophage Activation Syndrome Complicating Systemic Juvenile Idiopathic Arthritis, released in 2016, focused solely on diagnosis rather than treatment. Furthermore, there are currently no Egyptian-wide, evidence-based, treat-to-target recommendations for the proper diagnosis and treatment of MAS in children. This was the motivating force for the development of this work. The objective is to provide a consensus, evidence-based recommendations for the diagnosis, evaluation, and treat-to-target management of children living with MAS. Although these guidelines were created for Egyptian children with MAS, we hope that they would be helpful to pediatric rheumatologists all around the world.

Design
The study design was developed using scientific evidence and consensus, which was based on both existing scientific evidence and clinical experience. The purpose of this multistep method was to develop a clinical gold standard for MAS treat-to-target management in children using the "Clinical, Evidence-based, Guidelines" (CEG) initiative protocol (ethical approval number: 34842/8/21). The evidence-based component of the manuscript adhered to the preferred reporting items for systematic reviews and meta-analyses criteria for systematic review publication [10]. The Egyptian College of Pediatric Rheumatology was the driving force behind the project.

Core team
It is made up of four professionals with backgrounds in pediatric rheumatology, inflammatory arthritis, and macrophage activation syndrome. The core team was involved in managing and coordinated the work of the team; assisted in the development of the project's scope and initial Patient/Population, Intervention, Comparison, and Outcomes (PICO) clinical questions; came to an agreement on the key questions to be included in the guidelines; nominated the expert panel; and drafted the manuscript.

Literature review team
The literature review was carried out with the help of a methodology expert, under the supervision of an experienced literature review consultant, and was based on specific research topics relating to the diagnosis and treatment of MAS. The team finished the literature search (using the PubMed/MEDLINE, EMBASE, and Cochrane databases), data abstraction, and evidence quality evaluation [11]. Following the revision, each of the literature review experts made recommendations for each section based on evidence or their own personal experience. The Oxford Centre for Evidence-based Medicine (CEBM) approach was used to establish the degree of evidence for each section [12].

Data sources and search strategies
The PICO questions (Table 1)

Study selection
Relevant studies were chosen using inclusion and exclusion criteria applied to the literature found using the search methodology.

Inclusion criteria
Articles included were systematic reviews, randomized controlled trials (RCTs), uncontrolled trials, and observational studies including cohort, case-control, and crosssectional studies.

Exclusion criteria
Conference abstracts, commentaries, editorials, and non-evidence-based personal/narrative reviews were excluded.

Expert panel
The core leadership team nominated 23 participants. The criteria for their selection included practice in the Egyptian Health System, professional knowledge, and experience in the field of pediatric rheumatology (minimum 8 years) especially including management of inflammatory arthritis and macrophage activation syndrome. The expert panel contributed in creating the scope of the project, refining the PICO questions, and voting on the recommendations by actively participating in scientific research on pediatric rheumatic disorders.

Developing the clinical care standards framework
To aid in the standardized identification of the guideline components, a structured template was created based on the answers to the structured key questions and the literature research. For each guideline component, the format in which the recommendations/information will be presented and extracted has been determined.

Delphi process
The Delphi method is a structured method for gathering vital information about a certain issue that is extensively used. It is predicated on the notion that group projections are more accurate than individual forecasts. The Delphi method's goal is to build consensus forecasts from a group of experts in a structured iterative manner. Its methodology is based on a succession of "rounds" of questionnaires sent to experts. The Delphi method generally involves the following stages: (1) A group of experts is gathered together. (2) Forecasting tasks/challenges are assigned to professionals and distributed. (3) Experts provide preliminary forecasts and justifications. In order to provide feedback, these are collated and summarized. (4) The experts receive comments, which they consider when revising their forecasts. This process can be repeated until there is a reasonable level of consensus. (5) The final forecasts are created by combining the forecasts of the experts. The key features of this method are the anonymity of participants and the controlled feedback [13][14][15].

Consensus process
To reach a consensus on the T2T (treat to target) strategy in MAS, three Delphi rounds were conducted. After the major features of the approach were determined, a discussion group worked with the scientific committee to specify the aspects that would be included in the questionnaire. The structured Delphi approach ensures that all participants' perspectives are taken into account equally, and it is especially effective for geographically diverse cities like Egypt. Online surveys were used to conduct the Delphi procedure. The Internet questionnaire's first round featured 20 issues related to MAS's T2T strategy.

Voting process
Three rounds of live online voting were held, each with a strict time limit. All members of the task force were invited to participate, and the start and end times of each round of voting were announced ahead of time. Anonymous votes were gathered and evaluated, and we handed out the unique access links. At the same time as the voting procedure, comments on rephrasing, potential ambiguity, and unidentified overlaps were received for each statement. The task force members were the only ones who could vote on the statements.

Rating
Each statement was rated between 1 and 9 with 1 being "complete disagreement" and 9 being "complete agreement. " Generally, 1-3, 4-6, and 7-9 represent disagreement, uncertainty, and agreement, respectively. It is not necessary for members to vote on all statements, and they are invited to abstain if they believe a statement is outside their area of competence. As a result, a vote of "uncertainty" indicates "inconvenience about the accuracy of the recommendation. " All statements allow for the submission of comments, which the scientific committee reviews after each round of voting. Members were also encouraged to make remarks wherever they voted a disagreement in all of the voting rounds. This will allow the panel to spot a case of statement misinterpretation and nullify the vote on that statement.

Definition of consensus
Prior to data analysis, a definition of consensus was defined. It was concluded that if at least 75% of the participants agreed (scoring 7-9) or disagreed (scores 1-3), consensus would be attained [11][12][13][14]. If a statement received a mean vote of less than 3 or a "poor" degree of agreement, it was retired. In light of the feedback, statements with an uncertainty score of 4-6 were changed. Following the second round of voting, the levels of agreement on each statement of recommendation were rated "strong" if all votes on that statement fell into the agreement bracket (7-9) [15][16][17].

Chronogram of Delphi rounds
The first round took place between 12 and 15 May 2021 (4 days). The aspects about which respondents did not reach consensus in this first round were revised in view of the comments and included in the second round. The second round took place (1 week after the first round) and remained for 4 days, between 22 and 25 May 2021 (4 days). The third round took place (2 weeks after the second round) and remained for 4 days between 11 and 14 June 2021 (4 days).

Ethical aspects
This study was performed in accordance with the Helsinki Declaration. The "Clinical, Evidence-based, Guidelines" (CEG) initiative protocol was approved by the local ethical committee: ethical approval code: 34842/8/21, ethical board Tanta University. According to national standards, written ethics approval from the experts involved in this project was not required. In accordance with data protection standards, all participants were kept anonymous.

Literature research and evidence selection
In the study selection process, we found 443 potentially relevant studies by search strategy. In total, 410 were excluded by screening of title and abstracts (for duplication or the studies did not examine population or intervention of interest, did not match study design of interest, or did not report outcome measures of interest). Therefore, relevant 33 studies were included for full article review. Twenty studies were excluded as citations did not provide evidence matching a PICO. Therefore, we included 13 studies in this work ( Fig. 1).

Expert panel characteristics
The Delphi form was sent to the expert panel (n = 23), who participated in the three rounds. Respondents were drawn from different governorates and health centers across Egypt: Cairo University (26.1%), Ain Shams  In this phase, we introduced the clinical questions which will be the base and the titles of items in this guidelines work. The response rate for round 1 was 100% (23/23). Consensus was reached on the inclusion of clinical standards on 90% of the items (i.e., ≥ 75% of respondents strongly agreed or agreed). There were comments raised regarding the wording of some of the recommendations. MAS presentations with various disorders, as well as MAS and COVID-19, received more comments (excluding minor editing suggestions). The number of statements which were added after round 1 was 1 in the diagnostic criteria section, 2 in the management section, and 2 statements in the MAS and COVID-19 section.

Delphi round 2
The response rate for round 2 was 100% (23/23). Consensus was reached on the inclusion of clinical standards on 89.6% of the items (i.e., ≥ 75% of respondents strongly agreed or agreed). There were comments raised regarding the wording of some of the recommendations.
Comments (excluding minor editing suggestions) were more frequent for MAS presentations with different diseases, differential diagnosis, monitoring and follow-up, and management of resistant and severe cases section. Diversity of opinion was greatest for the item "management of resistant and severe cases. " The number of statements which were added after round 2 was one statement in the differential diagnosis section, three statements were added to the monitoring and follow-up section, and two statements were added to the MAS and COVID-19 section. Several statements were revised after round 2; most edited statements were in the management of resistant and severe cases section (5 statements), one statement was amended in each of key point, early diagnostic tools, treat-to-target strategy, and MAS and COVID-19 sections.

Delphi round 3
The response rate for round 3 was 100% (23/23). The frequency of high rank recommendation (ranks 7-9) ranged between 86.9 and 95.7%. One statement was retired for similarity with another statement. Consensus was reached (i.e., ≥75% of respondents strongly agreed or agreed) on all the clinical standards. Table 2 shows the level of evidence assigned to each statement, in accordance with the Oxford Centre for Evidence-Based Medicine (CEBM) criteria as well as mean ± standard deviation and level of agreement [11]. Agreement was unanimous (>80% agreement) for the wording of the statements.

Recommendations for the management of children with MAS
At the end of Delphi round 3, a total of nineteen recommendation items were concluded. These were categorized into 2 sections: 11 recommendations under the diagnosis and 8 under management. Table 3 presents

Early diagnostic tools
There was a high level of agreement 8.26 + 1.7 (agreement percentage: 95.65%) on the criteria for early diagnosis of MAS (level of evidence (LE) 3, strength of recommendation (SoR) B, level of agreement H). These tools are summarized in Table 4. As MAS could complicate many disorders, MAS presentations with different diseases have been assessed and presented in Table 5; also, Table 5 shows an approach to differentiate between MAS and disease activities. Table 6 shows a summary of recommendations for MAS complicating sJIA, while Fig. 2 shows an algorithm for implementation of the MAS management approach in standard clinical practice.

Discussion
Macrophage activation syndrome is a life-threatening condition in childhood, which is seen most commonly in sJIA, jSLE, Kawasaki disease, and infectious diseases. It is characterized by persistent unremitting fever, hepatosplenomegaly, lymphadenopathy, marked depression of all 3 blood cell components, intravascular coagulation, impaired liver function, and central nervous system dysfunction [5]. In a population research by M H Moradinejad et al. [20], the incidence of MAS was 4.2%, whereas the incidence of sJIA, SLE, juvenile idiopathic arthritis (JIA), and polyarticular RF negative JIA was 8.2%, 16.7%, and 2.8%, respectively, with a death rate of 40%.
The 2020 Egypt census revealed that the number of population is 104,124,440, of them 33.62% are children in the age range of 0-14 years old and 18.01% in the range of 15-24 years of age [18]. Though, up till now, there is no data on the prevalence of MAS in Egypt, having over 33 million children in Egypt would give a hint on the magnitude of the problem, even if estimated at the lowest prevalence rates.  Because MAS is a potentially fulminant disorder that can evolve into a life-threatening consequence of inflammatory rheumatic disease [19], it is critical to consider it in any juvenile rheumatic disorder characterized by rapid changes in general condition and a decrease in peripheral cells.
Expert consensus recommendations for the diagnosis and treatment of MAS are expected to play an essential role, given the absence of high-level evidence for diagnosis and management of MAS. The Delphi technique has shown to be a reliable tool for achieving such consensus and selecting future research directions [21]. The Delphi technique enlists the help of a group of experts to assess the level of agreement and settle disagreements on a topic [22]. The European League Against Rheumatism/American College of Rheumatology/Pediatric Rheumatology International Trials Organization (EULAR/ACR/PRINTO) Collaborative Initiative used a Delphi approach to define classification criteria for macrophage activation syndrome (MAS) in patients with sJIA. In agreement, this task was completed in a multistep process that included a Delphi survey and Webbased processes for determining expert consensus. When using the Delphi process, consensus is reached when the percentage of people who agree or disagree is between 50 and 80% [21][22][23]. When the experts were asked about the possibilities of achieving a well-defined purpose in MAS, they came to a broad agreement. A total of 11 recommendations for diagnosis and 8 for treatment, and the percentage of agreement ranged between 86.9 and 95.7%, indicating a strong trend among the health care professionals to have a T2T approach for MAS management. A similar figure was reported in an earlier published work where an 82% consensus was achieved among 28 international experts [24].    At least one clinical criterion and at least two laboratory criteria must be present simultaneously for MAS to be diagnosed. Bone marrow aspiration is performed to search for evidence of macrophage hemophagocytosis, and it may be necessary only in doubtful cases [25]. These criteria were used to diagnose MAS in this study. Furthermore, the guidelines focused on early diagnostic techniques, with a focus on increasing increases in serum ferritin and a relative decrease in platelet count followed by a decrease in WBCs or fibrinogen levels, rather than an absolute decrease. Statements describing how to distinguish MAS from its other mimic illnesses were included to support the best diagnosis. MAS presentations in conjunction with disorders such as lupus, systemic auto-inflammation, and Kawasaki disease were also included.
Treat to target has established itself as a guiding strategy for the treatment of inflammatory arthritic conditions, and it is based on several principles: identifying a target and a tool to measure it, evaluating the target at a pre-determined time point, a commitment to change the therapy if the target is not met, and shared decisionmaking. Previous research has advocated either an evidence-based treatment method [26] or a therapy-based management algorithm [7]. In contrast, this guideline relied on treat-to-target strategy; the identified targets were to reach: fever < 38.5°C, no organometallic, no cytopenia, significant drop in serum ferritin (best is < 2000 ng/mL), triglycerides < 1.5 mmol/L, aspartate aminotransferase < 30 IU/L, and fibrinogen > 2.5 g/L. The guidelines discussed also preventive measures to lower the likelihood of developing MAS. A clear definition of resistant/irresponsive as well as severe cases has also  been identified. Monitoring and follow-up parameters, both clinical and lab, were also identified and included in this work. Such a treat-to-target approach was reported to yield superior outcomes to standard treatment protocols, and several professional organizations have endorsed it as a fundamental therapeutic strategy [27]. To our knowledge, this is the first guideline to adopt such a strategy for the treatment of MAS. Beyond the use of steroids which was highlighted in previous works [7,8,21], the algorithm proposed in this work for MAS management discussed the management of resistant and severe cases including the use of IVIG, biological treatment including IL-1 and IL-6 inhibitors, cyclosporine either mono or combined therapy. Also, the salvage role of etoposide and anti-thymocyte globulin and the promising experimental results of JAK inhibitors had been incorporated.
Given the current circumstances of the COVID-19 pandemic and possible serious conditions that appear to be linked to coronavirus disease 2019 in children [28], this guideline had a specific consideration for a rare but serious complication of COVID-19 known as a multisystem inflammatory syndrome in children (MIS-C). MIS-C resembles MAS in several manifestations; however, specific differences between both conditions have been identified. This work presented a discrete management strategy for the condition starting early diagnostic markers, clinical features, and tailored management protocol.

Conclusion
This was an expert, consensus recommendation for the diagnosis and treatment of MAS, based on the best available evidence and expert opinion. The guidelines fill a gap in the literature as it presents a T2T approach for MAS. These suggestions should make it easier to