International clinical consensus on leukocyte adhesion deficiency-I: Modified Delphi analysis

Abstract

Background

Leukocyte adhesion deficiency-I (LAD-I) is a rare autosomal-recessive inborn error of immunity caused by mutations in ITGB2, encoding CD18, which is essential for leukocyte endothelial adhesion and tissue migration. LAD-I is predominantly characterized by frequent life-threatening infections and hyperinflammatory complications, with high rates of pediatric mortality and morbidity without allogeneic hematopoietic stem cell transplantation. Historically, diagnosis and severity classification were based on polymorphonuclear leukocyte CD18 expression. Given improved knowledge of prognostic factors and an evolving treatment landscape, contemporary guidance for severity classification and management is needed.

Objective

We sought to provide international, expert-led consensus recommendations for the diagnosis and severity classification of LAD-I; and to characterize standard of care and burden of illness for patients with severe LAD-I.

Methods

Twelve geographically diverse experts (3 steering committee members and 9 Delphi participants) were recruited to participate in a modified Delphi study. Two rounds of online questionnaires were conducted, plus a virtual workshop. Final recommendations were approved by all experts.

Results

Consensus was achieved across 26 recommendations regarding the diagnosis and treatment of patients with LAD-I. Experts agreed LAD-I severity classification should consider laboratory assessment in addition to clinical phenotype within a comprehensive framework. In the absence of symptoms, particularly in infancy, diagnosis and severity classification is based on flow cytometric assessments (CD18 and CD11a/CD11b expression) and molecular genetic testing or family history.

Conclusion

The framework developed by the Delphi panel will aid clinicians in the diagnosis, classification, and treatment of patients with LAD-I. Recommendations support best clinical practice that can lead to improved clinical outcomes.

Leukocyte adhesion deficiency-I (LAD-I) is a rare autosomal-recessive inborn error of immunity caused by loss-of-function mutations in ITGB2, encoding the β₂ integrin subunit of CD18. Deficient, absent, or dysfunctional CD18 expression prevents pairing with α integrin subunits (CD11a-d),¹ impairing adhesion to inflamed endothelia and migration to infection sites, compromising tissue homeostasis.^2,3 LAD-I affects approximately 1 in 1 million individuals globally.^4,5

Without routine newborn screening for LAD-I, flow cytometry to assess leukocyte integrin expression is typically prompted by presentation of early signs or symptoms, including marked leukocytosis and neutrophilia; umbilical cord complications (eg, delayed separation, omphalitis⁶); and/or contributory family history. Children can also present with severe/recurrent bacterial or fungal infections, recalcitrant periodontitis, and/or necrotic skin wounds/lesions absent pus formation.^3,5,6 Less frequently, diagnosis follows early genetic screening or abnormal integrin expression detected as a result of strong clinical suspicion before symptoms arise.

LAD-I severity has historically been classified by the percentage of CD18⁺ polymorphonuclear neutrophils (PMNs) in peripheral blood, with <2% indicating a severe phenotype and ≥2% but <30% indicating a moderate phenotype.⁷ Expanded flow cytometry, including CD11a/b testing,^6,8 further informs disease severity, prognosis, and treatment decisions, particularly in infants. ITGB2 sequencing aids diagnosis confirmation before pursuing potentially definitive therapies and aiding prenatal diagnosis and family planning counseling. However, no consensus currently exists on integrating immune phenotype, genotype, and clinical presentation to guide LAD-I severity classification or treatment decisions.

Best supportive care (BSC) typically comprises antimicrobial prophylaxis, infection management, and anti-inflammatory treatments, including corticosteroids;^9,10 recent evidence also supports the use of ustekinumab for its anti-inflammatory effects.¹¹ Without definitive treatment, patients with severe LAD-I often experience ≥3 severe infections annually, requiring hospitalizations, intravenous antimicrobials, and/or surgical debridement.¹² Immune dysregulation, driven by IL-23 and IL-17, contributes to chronic complications like periodontal disease,^11,13 necrotic skin lesions,^11,14 and colitis.^15,16 Despite aggressive antimicrobial and anti-inflammatory therapy, 23% to 69% of children with severe LAD-I die before age 2 without allogeneic hematopoietic stem cell transplantation (allo-HSCT),^6,8,17,18 with considerable long-term clinical and economic burden in those who survive further into childhood.¹⁹

Allo-HSCT is the only available definitive treatment, but its use is limited by donor availability, risk of graft versus host disease (GvHD), graft failure, transplant-related complications, and death.^6,20, 21, 22 Gene therapy is emerging as an alternative potential definitive treatment for patients with severe LAD-I by utilizing the patient’s hematopoietic stem cells transduced with a self-inactivating lentiviral vector containing wild-type ITGB2.23, 24, 25

Despite advances in prognostic factors, treatment, and genetic testing, most patients with severe LAD-I still present with complications associated with prior infections and hyperinflammatory conditions. An unmet need exists for practical guidance on severity classification and management of LAD-I, taking into consideration their clinical presentation and path to diagnosis. We aimed to formulate international expert-led consensus recommendations on the appropriate use of immune phenotype, ITGB2 sequencing, and clinical phenotype to guide the diagnosis and classification of moderate or severe LAD-I. Such recommendations are intended for clinicians involved in the recognition, diagnosis, and management of LAD-I, including allo-HSCT specialists, immunologists, neonatal and pediatric intensive care unit (ICU) physicians, and general practitioners, as well as health service managers and policymakers for rare pediatric immunologic disorders.

Methods

Study design

To develop consensus statements/recommendations, a modified Delphi method was applied aligned with Accurate Consensus Reporting Document (aka ACCORD) for conducting and reporting formal consensus studies.²⁶ The Delphi method has become an increasingly important tool in health care research and is often used when available evidence is limited/inconsistent.27, 28, 29 Before study initiation, experts with experience in treating patients with LAD-I were invited to participate (n = 12), including a steering committee (n = 3) and Delphi participants (n = 9). Two Delphi rounds were performed followed by a consensus workshop (n = 6).

The steering committee comprised experts in pediatric immunology, infectious diseases, and cellular therapy, with over a decade of experience in diagnosis and management of patients with LAD-I and other inborn errors of immunity. All 3 also actively contribute to national and international guidelines and research (L.I.G.-G., J.H., J.W.L.). The Delphi participants comprised experts in pediatric immunology (C.B., W.I., A.G., U.A.B.)—including primary immunodeficiency disorders and leukocyte biology—pediatric hematology/oncology (S.P.), intensive care medicine (S.B.), stem cell transplantation (A.G., S.B., S.P.), and gene therapy (C.B.), reflecting the clinical specialities for which this article is relevant. A study overview is provided in Fig 1.

Fig 1.

Overview of modified Delphi study.

Delphi panel selection

Members of the Delphi panel were identified via expert-guided personal networks and snowball sampling; if experts could not participate, they were asked to recommend a suitable colleague. Invitations to participate were sent to 19 experts. The Delphi panel was selected on the basis of their publication record in LAD-I research^18,30, 31, 32 or broader publication records across comparable inborn errors of immunity.^21,33,34 To be eligible, individuals were required to have managed patients with LAD-I (personally or as part of a wider team) in the previous 5 years; there was an additional aim to recruit Delphi panelists from broad geographies. The final Delphi panel comprised 12 experts: 9 participants (from GBR, USA, DEU, IND, TUR, and ITA) and 3 experts who formed the steering committee (from the United States [J.H., J.W.L.] and Europe [L.I.G.-G.]). The steering committee supported question development for both Delphi rounds, ensuring questions were clinically relevant, accurate, and nonambiguous. The steering committee did not participate in the online Delphi rounds but contributed to content development and discussion at the consensus workshop.

Several participants were invited to attend the consensus workshop according to geographic location, clinical specialty, and availability. Of those invited to attend the consensus workshop, 3 of the participants (A.R.G., C.B., S.P.) attended alongside the steering committee. Final recommendations were approved by all experts.

Online Delphi questionnaires

A pragmatic literature review was conducted (July 2023) by performing online free-text searches across the Ovid Medline, PubMed, Google Scholar, and Google platforms. Key terms included, but were not limited to, “LAD-I burden of illness,” “LAD-I best supportive care,” “LAD-I severe disease.” First round statements and questions were drafted according to collated evidence and were then subjected to rounds of revision by the committee. The results of the pragmatic literature review are provided in this article’s Online Repository available at www.jaci-global.org.

The first Delphi round was hosted on the SmartSurvey platform;³⁵ a hyperlink to the questionnaire and instructions were distributed via email. A study overview was provided at the beginning of the questionnaire, and all participants provided their informed consent, noting that they could withdraw their responses at any time. Participants’ responses were anonymized to other participants and the study sponsor.

The first Delphi round comprised 74 questions, of which 37 were consensus statements and the remainder were alternative question types (eg, free-text explorative, multiple-choice, and ranking questions). Consensus statements and questions were stratified into 4 sections: participant information, definition of moderate and severe LAD-I, standard of care for patients with severe LAD-I, and burden of illness for patients with severe LAD-I.

First round responses were consolidated and analyzed using descriptive statistics (Table I).^27,36,37 Consensus statements were rated on 7-point Likert scales and participants were able to justify their ratings in free-text answers.

Table I.

Aggregation of Delphi responses

Likert scale response	Response stratification	Determining consensus based on:
		Agreement	Disagreement
1: Strongly disagree	1-3: Disagree	Poor/no consensus: <60% agreement	Strong/total consensus: ≥80% disagreement
2: Disagree	1-3: Disagree	Some consensus: 60-69% agreement	Good consensus: 70-79% disagreement
3: Somewhat disagree	1-3: Disagree	Some consensus: 60-69% agreement	Good consensus: 70-79% disagreement
4: Neither agree nor disagree	4: Neither agree nor disagree	Good consensus: 70-79% agreement	Some consensus: 60-69% disagreement
5: Somewhat agree	5-7: Agree	Good consensus: 70-79% agreement	Some consensus: 60-69% disagreement
6: Agree	5-7: Agree	Strong/total consensus: ≥80% agreement	Poor/no consensus: <60% disagreement
7: Strongly agree	5-7: Agree	Strong/total consensus: ≥80% agreement	Poor/no consensus: <60% disagreement

Although there is no standard definition for degree of consensus, previous studies have considered 50% to 97% agreement to be sufficient,^27,36 with 80% representing strong consensus.³⁷

First round statements with poor/no consensus or some consensus were revised according to participants’ responses, reviewed by the steering committee, and included in the second round; supplementary questions were also included. The second round Delphi comprised 73 questions (of which 16 were consensus statements), which were rated by participants using the same format as in the first round and covering the same topics. Across the two Delphi rounds, 53 consensus statements were rated by participants as described above. Those reaching consensus are presented in Table II; all other consensus statements can be found in the Online Repository.

Table II.

Key consensus statements and recommendations

Assessment or therapy	Statement	Participant agreement	Strength of recommendation
	Best practice for diagnosis of moderate and severe LAD-I
Molecular diagnostics for LAD-I	Severe LAD-I is defined as <2% of expected PMN CD18 expression, or ≥2% of expected PMN CD18 expression and <2% of expected PMN CD11a and/or CD11b expression.	87.5%; n = 7/8	Strong consensus
	Moderate LAD-I is defined as 2-30% of expected PMN CD18 expression.	75%; n = 6/8	Good consensus
	These laboratory assessments alone are insufficient to diagnose and distinguish between moderate and severe LAD-I in symptomatic patients. In symptomatic patients, clinical phenotype, genetic analysis of ITGB2, and/or family history of LAD-I should be considered.	87.5%; n = 7/8	Strong consensus
	In best clinical practice, PMN CD18, CD11a, and/or CD11b expression should be measured as part of diagnosing and differentiating between moderate and severe LAD-I.	100%; n = 8/8	Strong consensus
	Molecular genetic testing in conjunction with flow cytometry is ideal in confirmation of LAD-I diagnosis.	100%; n = 8/8	Strong consensus
Clinical phenotype in LAD-I	Clinical phenotype and laboratory assessments are both important. Absent clinical symptoms, newborns and infants may be diagnosed as having severe LAD-I according to laboratory assessments and genetic analysis of ITGB2 gene, or family history of LAD-I.	100%; n = 8/8	Strong consensus
	A “persisting” and “increasing” annual risk of having life-threatening infection/infection-related complication/death can be considered equally important in patients with severe LAD-I.	100%; n = 8/8	Strong consensus∗
	For patients with severe LAD-I who survive past 2 years, at minimum, their risk of life-threatening infections, infection-related complications, and death persists year after year and does not decrease in absence of definitive therapy (ie, allo-HSCT or gene therapy). In most patients, their risk of these events will increase year after year.	100%; n = 8/8	Strong consensus∗
	In patients with severe LAD-I with history of significant morbidity and who receive BSC alone (ie, in absence of definitive therapy), survival beyond 2 years of age is not marker of milder disease.	100%; n = 8/8	Strong consensus
	Best practice for treatment of severe LAD-I
Allo-HSCT	Under current clinical practice (ie, where gene therapy is not routinely available), allo-HSCT is recommended for patients with severe LAD-I. In some scenarios, allo-HSCT is not suitable: overwhelming infection or organ system disease such that patient would be unlikely to survive transplantation; family refusal of allo-HSCT; matched donor availability (although alternative [eg, haploidentical] donors are still potential definitive treatment option).	NA	Good consensus†
	For patients with severe LAD-I, allo-HSCT should be conducted as soon as possible after diagnosis, ideally in patients ≤13 months of age, as treatment outcomes are improved in younger patients; however, allo-HSCT should still be performed if patients are >13 months old.	87.5%; n = 7/8	Strong consensus
	For patients with severe LAD-I, allo-HSCT should ideally be performed either before onset of infection or after resolution of active infection. However, allo-HSCT may be performed in presence of active infections, if necessary, provided infections are controlled, patient is medically stable, and all other treatment options to resolve infection have failed.	100%; n = 8/8	Strong consensus
	For patients with LAD-I undergoing allo-HSCT, anti-inflammatory therapy (eg, ustekinumab) should be provided before transplantation to reduce cytokine-induced inflammation within bone marrow and to facilitate engraftment.	71.4%; n = 5/7	Good consensus
	If initial allo-HSCT is unsuccessful because of primary or secondary graft failure, patient should undergo second allo-HSCT.	85.7%; n = 6/7	Strong consensus
	After allo-HSCT, duration of immunosuppressive therapy can vary and is patient dependent.	71.4%; n = 5/7	Good consensus
	Follow-up procedure after allo-HSCT is case dependent; there is no universal protocol.	NA	Good consensus†
BSC	All patients with severe LAD-I require prophylactic antimicrobials (ie, antibiotics, antifungals).	100%; n = 8/8	Strong consensus
	To treat severe infections in patients with severe LAD-I, aggressive infection management should be performed in hospital setting.	100%; n = 8/8	Strong consensus
	Infections in patients with severe LAD-I often require prolonged antimicrobial treatment, including multiple courses of antimicrobials.	100%; n = 8/8	Strong consensus
	For all patients with LAD-I, anti-inflammatory therapy (eg, ustekinumab) should be prescribed as part of supportive care to treat dysregulated hyperinflammatory component of disease pathology (eg, chronic inflammation).	85.7%; n = 6/7	Strong consensus
	Steroids are prescribed in conjunction with antimicrobials if there is no improvement with antimicrobial treatment to treat inflammation associated with severe infections in patients with severe LAD-I.	75%; n = 6/8	Good consensus∗
	Granulocyte transfusions can be administered before and after allo-HSCT to treat severe infections in patients with severe LAD-I; however, some clinicians may avoid them because of potentially increased risk of transplant complications.	75%; n = 6/8	Good consensus∗
	Surgical debridement/surgical resection is often required for locoregional infections or inflammatory lesions.	87.5%; n = 7/8	Strong consensus
	Social restrictions are not recommended as long-term treatment strategy for patients with severe LAD-I.	71%; n = 5/7	Some consensus
	Burden of illness for patients with severe LAD-I
	Severe LAD-I causes significant morbidity due to infection- and inflammation-related complications associated with frequent hospitalization.	100%; n = 8/8	Strong consensus
	Severe LAD-I is associated with substantial negative impact on patient and caregiver QoL.	100%; n = 8/8	Strong consensus
	Severe LAD-I is associated with high caregiver burden.	100%; n = 8/8	Strong consensus
	There is significant burden on siblings/family members of patients with severe LAD-I who are suitable donors for allo-HSCT.	87.5%; n = 7/8	Strong consensus

NA, Not applicable.

^∗Recommendation is based on multiple questions within Delphi questionnaire and/or discussion and agreement during consensus workshop.

^†Recommendation is based on open question in Delphi questionnaire, plus discussion and agreement during consensus workshop.

Consensus workshop

After completion of both Delphi rounds, a virtual consensus workshop was held. Discussion focused on questions with limited consensus, including referral to participant comments to understand conflicts in opinions. Questions achieving strong/good consensus and explorative questions were also presented for validation. Six consensus workshop panelists exchanged opinions, and statements were revised if necessary to align with the workshop discussions.

Results

Statements achieving strong or good consensus on aspects of best practice for the diagnosis and severity classification of moderate/severe LAD-I and treatment and burden of illness of severe LAD-I are described in Table II.

Delphi participants

Nine individuals completed the two Delphi rounds; however, one participant was excluded for not meeting eligibility criteria. Responses from the remaining 8 participants were included in the analysis. Participants were geographically situated in the United Kingdom (n = 3), United States (n = 2), Germany (n = 1), Turkey (n = 1), and India (n = 1). Most participants worked in specialist centers and had experience treating LAD-I patients with allo-HSCT (7/8, 87.5%), BSC (5/8, 62.5%), and/or gene therapy (4/8, 50%).

Best practice for diagnosis of moderate and severe LAD-I

Laboratory-based diagnostics for LAD-I

Participants agreed on existing definitions of severe (87.5% agreement; n = 7/8) and moderate LAD-I (75% agreement; n = 6/8) that were based on PMN CD18/CD11a/CD11b flow cytometric assessments, including existing thresholds. Participants advised that newborns and infants should ideally be diagnosed with severe LAD-I before clinical manifestations, and importantly, that a diagnosis in this instance should be based on flow cytometric assessments and genetic confirmation or a family history of LAD-I. In such instances, participants and panelists agreed that the health care professional should discuss potential definitive therapeutic modalities, without waiting for manifestations of LAD-I to emerge. Delphi participants and the workshop panel agreed that integrin flow cytometric assessments alone are inadequate for classifying severity in symptomatic patients; clinical phenotype should also be considered in conjunction, prioritizing disease staging by focusing on the most severe characteristics (Table II).

All participants indicated they measure CD18 and CD11a/CD11b routinely (ideally as an exact percentage), and assessments are readily available (100%; n = 8/8). Participants also reached total consensus that molecular genetic testing in conjunction with flow cytometry is preferred, but not required, to establish a LAD-I diagnosis (100%; n = 8/8). The workshop panel explained that confirmation of pathogenic mutations is not essential for a LAD-I diagnosis because a patient may possess a variant of uncertain significance that, according to in silico modeling, portends poor impact on protein function. ITGB2 sequencing is valuable in confirming LAD-I diagnosis before allo-HSCT or other definitive therapies, thus enabling prenatal diagnosis and genetic counseling.

Assessment of clinical phenotype in LAD-I

Participants agreed that LAD-I should only be considered as moderate or severe (Table II), and they strongly agreed that for patients presenting with signs and symptoms of LAD-I, clinical phenotype is as important as laboratory assessments in diagnosing and differentiating between moderate and severe LAD-I.

To assist clinicians in identifying suspected cases of symptomatic LAD-I, participants selected the 3 most common clinical manifestations experienced by patients with severe LAD-I across different age groups (Table III). Omphalitis was considered the most frequent symptom in patients aged <6 months (87.5%; n = 7/8), skin and mucosal surface infections were most common in patients aged 6 to 24 months (37.5%; n = 3/8), and both skin and mucosal infections and respiratory tract infections were most common for patients aged >24 months (37.5%; n = 3/8). These results align with published cohorts of patients with severe LAD-I, who also report the most common symptoms to be umbilical cord complications, skin infections, and lower respiratory tract infection (ie, mucosal surface infections).^6,18 The workshop panel also discussed other signs of LAD-I that likely raise clinical suspicion, prompting further assessment including immunophenotype evaluation via flow cytometry, namely leukocytosis and wounds without pus formation or that do not exhibit tissue repair, bacillus Calmette-Guérin vaccine complications,³⁸ poor wound healing, and family history

Table III.

Three most common clinical manifestations experienced by patients with severe LAD-I in different age groups according to Delphi participants

Age group	Most common clinical manifestation		Second most common clinical manifestation		Third most common clinical manifestation
	Clinical manifestation	No. (%) in agreement	Clinical manifestation	No. (%) in agreement	Clinical manifestation	No. (%) in agreement
≤6 months	Omphalitis	7 (87.5)	Skin and mucosal surface infections	5 (62.5)	Respiratory tract infections	2 (25.0)
	Skin and mucosal surface infections	1 (12.5)	Respiratory tract infections	1 (12.5)	Impaired wound healing	2 (25.0)
			Omphalitis	1 (12.5)	Sepsis	2 (25.0)
			Sepsis	1 (12.5)	Gastrointestinal inflammatory disease	1 (12.5)
					Skin and mucosal surface infections	1 (12.5)
6-24 months	Skin and mucosal surface infections	3 (37.5)	Respiratory tract infections	4 (50.0)	Skin and mucosal surface infections	3 (37.5)
	Respiratory tract infections	2 (25.0)	Skin and mucosal surface infections	2 (25.0)	Gastrointestinal inflammatory disease	2 (25.0)
	Periodontal infections	1 (12.5)	Impaired wound healing	1 (12.5)	Gastrointestinal infections	1 (12.5)
	Periodontal disease	1 (12.5)	Gastrointestinal inflammatory disease	1 (12.5)	Periodontal disease	1 (12.5)
	Omphalitis	1 (12.5)			Impaired wound healing	1 (12.5)
>24 months	Skin and mucosal surface infections	3 (37.5)	Respiratory tract infections	3 (37.5)	Skin and mucosal surface infections	2 (25.0)
	Respiratory tract infections	3 (37.5)	Impaired wound healing	2 (25.0)	Gastrointestinal inflammatory disease	2 (25.0)
	Inflammatory skin disease	1 (12.5)	Skin and mucosal surface infections	1 (12.5)	Gastrointestinal infections	1 (12.5)
	Periodontal disease	1 (12.5)	Periodontal infections	1 (12.5)	Respiratory tract infections	1 (12.5)
			Periodontal disease	1 (12.5)	Inflammatory skin disease	1 (12.5)
					Periodontal infections	1 (12.5)

No. (%) in agreement indicates how many Delphi participants (out of 8) selected answer.

Definitions for recurrent, relapsed/chronic, severe, and moderate infections were developed, which achieved total consensus (100%, n = 8/8; Table IV). Without definitive therapy, participants agreed (via multiple-choice responses), in their experience, that patients with severe LAD-I typically experience 4-6 severe infections and 1-3 moderate infections per year, while patients with moderate LAD-I experience approximately 1-3 severe infections and 1-3 moderate infections per year.

Table IV.

Delphi-response definitions for different types of infections experienced by patients with LAD-I

Type of infection	Definition	Agreement	Consensus
Severe	Life-threatening infection that requires hospitalization, which can be prolonged (ie, >7 days), and which requires intravenous or intramuscular antimicrobial treatment.	100%; n = 8/8	Strong
Moderate	Infection that is not life-threatening, may require short hospital stay, is isolated to one organ system, and responds to treatment with oral and/or intravenous antimicrobial treatment.	100%; n = 8/8	Strong
Recurrent	Infection that occurs more than once after complete resolution of initial infection and can involve different organisms at same or different site of originally infected organ system.	100%; n = 8/8	Strong
Relapsed/chronic	Infection that lasts longer than expected or returns soon after resolution and involves same organism at same site of originally infected organ system.	100%; n = 8/8	Strong

Workshop panelists agreed that some severe infections can be characterized as life-threatening (that is, aligned to the Common Terminology Criteria for Adverse Events v5.0 for a grade IV infection: “life-threatening consequences; urgent intervention indicated”), especially when affecting multiple organ systems and/or requiring escalation of care (ie, care at an ICU). Panelists considered moderate infections as likely to progress, becoming severe or life-threatening, if occurring before a LAD-I diagnosis. Importantly, panelists agreed that a single life-threatening infection is sufficient to classify a patient as having severe LAD-I, irrespective of immunophenotype, and clinicians should consider definitive therapy if not considered previously.

Without definitive therapy, participants considered the annual risk of life-threatening infections, infection-related complications, and death to persist or increase in patients with severe LAD-I who survive beyond 2 years of age (Fig 2). Workshop panelists considered “persisting” and “increasing” to be of equal clinical importance for patients with severe LAD-I. Moreover, all participants (100%; n = 8/8) and panelists agreed that survival beyond 2 years of age was not indicative of milder disease.

Fig 2.

Delphi responses related to annual risk of (A) having life-threatening infection, (B) infection-related complications, and (C) death for patient with severe LAD-I who survives past 2 years of age in absence of definitive therapy (defined as allo-HSCT or gene therapy).

Diagnosing and classifying moderate or severe LAD-I

On the basis of the online Delphi results as well as the workshop panelists’ discussion and committee feedback, criteria were developed to assist in severity classification of moderate versus severe LAD-I, utilizing both clinical and laboratory elements where appropriate (Table V).^8,39,40 The criteria were developed to capture the wide variety of clinical scenarios observed globally, including the possibility of genetic testing becoming part of routine care. Participants agreed that criteria may be applied at any time during the patient’s disease course and that patients with moderate LAD-I should be routinely reassessed because their clinical phenotype can worsen over time, potentially progressing to meet the criteria for severe LAD-I.

Table V.

Algorithm for severity classification of moderate or severe LAD-I according to clinical phenotype and laboratory assessments

LAD-I classification, prognosis, and treatment∗		Clinical assessment criteria†			Laboratory assessment criteria
LAD-I classification	Diagnosis requires:	Criterion 1	Criterion 2	Criterion 3	Criterion 4	Criterion 5	Criterion 6
Severe LAD-I	Criterion 4 and any of:‡ (1 or 2) or (3 or 6) or (2 and 5)	Delayed umbilical cord separation§	≥1 life-threatening infections (CTCAE grade IV) or omphalitis (CTCAE grade III or higher) or ≥2 severe or medically significant (CTCAE grade III) LAD-I–related clinical events¶‖	Family history of LAD-I∗∗	<2% of expected PMN CD18 expression or ≥2% of expected PMN CD18 expression, and <2% of expected PMN CD11a and/or CD11b expression	2-30% of expected PMN CD18 expression, and ≥2% of expected PMN CD11a and/or CD11b expression	Mutagenic analysis of ITGB2 variants∗
Moderate LAD-I††	Criterion 5 and any of: 2 or 6	Delayed umbilical cord separation§	≥1 medically significant (CTCAE grade II) LAD-I–related clinical events¶	Family history of LAD-I∗∗	—	2-30% of expected PMN CD18 expression, and ≥2% of expected PMN CD11a and or b expression	Mutagenic analysis of ITGB2 variants∗

Severity of LAD-I is considered on spectrum and can progress. Hence, patients with moderate LAD-I should be routinely reassessed. If case fulfills multiple criteria, greater severity should be assumed.

CTCAE, Common Terminology Criteria for Adverse Events v5.

^∗Confirmatory genetic diagnosis (criterion 6) of LAD-I is best practice but not mandatory, regardless of other criteria, before allo-HSCT, but it is mandatory before gene therapy. Genetic evaluation confirms presence of confirmed/suspected pathogenic variants, or variants of uncertain significance considered deleterious according to stepwise ABC system³⁹ and American College of Medical Genetics and Genomics⁴⁰ classification guidelines. Genetic testing alongside flow cytometry assays for PMN CD18, CD11a, and/or CD11b expression is ideal but not mandatory for LAD-I diagnosis.

^†Clinical phenotype should include clinical suspicion plus confirmed manifestations. In absence of clinical symptoms, newborns may be diagnosed as having severe LAD-I according to PMN expression criteria and genetic analysis of ITGB2 gene, or family history of LAD-I.

^‡Presenting with criteria 4 and 6 alone is sufficient for diagnosis of severe LAD-I; definitive therapy options should be discussed as soon as possible after diagnosis.

^§Delayed is defined as umbilical cord separation after 21 days of life.⁸

^¶LAD-I clinical events include either bacterial or fungal infection, or inflammatory events such as pyoderma gangrenosum or periodontitis, within any consecutive 18-24-month period.

^‖If criteria 2 and 5 are present, ≥3 (as opposed to ≥2) severe or medically significant (CTCAE grade III) LAD-I–related clinical events need to be present as part of criterion 2 to warrant diagnosis of severe LAD-I.

^∗∗Biological/blood relative only.

^††Combinations of (criteria 1 + 5) or (criteria 3 + 5) alone are insufficient for diagnosis of moderate LAD-I. However, delayed umbilical cord separation and/or family history of LAD-I, in conjunction with 2-30% PMN CD18 expression, should raise clinical suspicion, leading to close monitoring of patient by specialist for clinical events indicative of LAD-I. In these instances, genetic evaluation is suggested.

Best practices for treatment of severe LAD-I

Allo-HSCT

Participants agreed that allo-HSCT should be conducted as soon as possible after severe LAD-I diagnosis, ideally at ≤13 months of age for improved transplant outcomes (100% agreement; n = 8/8), in line with a retrospective analysis of registry data.²¹ However, there was strong consensus that patients >13 months of age are not precluded from receiving definitive treatment (87.5% agreement; n = 7/8).

Only 50% of participants (n = 4/8) agreed allo-HSCT should be performed within 3 months of diagnosis; this timing was considered aspirational, and many external factors influence timing of allo-HSCT. Participants agreed that approximately 2 to 6 months are required to prepare and stabilize a patient with severe LAD-I before allo-HSCT, noting variability of patients’ symptoms and progression, and the time to find a donor. Participants agreed that allo-HSCT should ideally be performed either before onset of infection or after active infections are resolved (100%; n = 8/8). However, if necessary, allo-HSCT may be performed in the presence of active infections if infections are controlled, the patient is medically stable, and all other treatment options to resolve infections have failed (100%; n = 8/8). Most participants would consider using anti-inflammatory therapy (eg, ustekinumab) before allo-HSCT to reduce cytokine-induced inflammation with a goal to facilitate engraftment (71.4% agreement; n = 5/7).

Participants reached strong consensus that patients should undergo second allo-HSCT after primary or secondary graft failure (85.7% agreement; n = 6/7). Few participants had managed patients with severe LAD-I who experienced unsuccessful second allo-HSCT.

Most participants agreed (75%; n = 6/8) that allo-HSCT is not always suitable, including situations in which a patient has an overwhelming infection, an organ system disease reducing likelihood of survival, family refusal, and no suitably matched donor options or donor refusal. Participants who disagreed (United States, n = 1; Great Britain, n = 1) stated that mismatched unrelated donors could be used in cases where there were no matched donor options. Workshop panelists agreed and suggested that alternative donors (eg, haploidentical) may still enable definitive treatment, although the availability of this therapy is limited to centers with expertise and/or research protocols investigating the use of alternative donors. Older patients yet to undergo allo-HSCT could still receive a transplant to improve survival and quality of life (QoL) as opposed to continued BSC.

On the potential impact of gene therapy on allo-HSCT use, participant responses indicated that gene therapy may potentially enable a shorter time to treatment compared with allo-HSCT by relieving reliance on donor availability. Participants indicated that in instances where two potentially definitive treatment options are available, shared decision-making with the patient/caregiver is recommended to consider their preferences, as well as issues around timing. Participants specified that gene therapy, once approved and if available, would be primarily considered for patients with severe disease for whom allo-HSCT donor options would increase risk of GvHD (ie, only haploidentical or other mismatched donors are available), or in those patients unlikely or unable to tolerate the preconditioning regimens for allo-HSCT. The Delphi panel had limited experience of managing patients with moderate LAD-I; however, they discussed that the use of a potentially definitive treatment (allo-HSCT or gene therapy) in moderate LAD-I would be considered on a case-by-case basis, and it would depend on patient/caregiver preferences, availability and adherence to BSC, the patient’s QoL, risk of GvHD, and ability to tolerate allo-HSCT preconditioning. Participants stated that additional evidence on the use of definitive treatments in moderate LAD-I would be useful.

BSC

Strong consensus was achieved on BSC practices for patients with severe LAD-I. Participants emphasized that BSC consists of symptom management and surveillance, antimicrobial treatment, vaccinations, and good hygiene practices. Prophylactic antimicrobials should be administered but may not prevent infection, with severe infections requiring prolonged antimicrobial treatment and in-hospital management (100% agreement; n = 8/8). Strong consensus was achieved on conducting surgical debridement/surgical resection for locoregional infections or inflammatory lesions (87.5% agreement; n = 7/8) and ustekinumab to treat hyperinflammatory symptoms (85.7% agreement; n = 6/7).

In a series of exploratory questions, participants stated that steroids can treat inflammation in conjunction with antimicrobials if there is no improvement with antimicrobial treatment alone. Granulocyte transfusions can be administered before and after transplantation to treat severe infections in patients with severe LAD-I, though some participants did not advocate their use owing to high rates of associated alloimmunization, which can adversely affect patients that undergo subsequent allo-HSCT.⁴¹ Participants disagreed on immunoglobulin replacement therapy; some participants prescribed this routinely, while others stated that it is infrequent.

The workshop panel agreed that social restrictions should not be a long-term treatment strategy and should only be recommended for patients undergoing allo-HSCT in accordance with standard practices.

Burden of illness for patients with severe LAD-I

All participants agreed that, despite BSC, severe LAD-I has a substantial negative effect on patient and caregiver QoL, a high caregiver burden, and significant morbidity caused by infection- and inflammation-related complications (100% agreement; n = 8/8). A significant burden exists for siblings/family members of patients with severe LAD-I who are suitable donors for allo-HSCT (87.5% agreement; n = 7/8).

Participants reached good consensus that a patient with severe LAD-I typically experiences 4 infection-related emergency department visits per year (75% agreement; n = 6/8) and total consensus that approximately 75% of infection-related emergency department visits required hospitalization (100% agreement; n = 8/8). A typical patient may spend over 2 weeks (approximately 17 days) as an inpatient per infection-related hospitalization (100% agreement; n = 8/8). There was variable consensus when discussing ICU admissions, in part the result of local protocol divergence. Workshop panelists agreed that “patients with severe LAD-I will have ≥1 infection-related ICU admission per year” lasting approximately 5 days.

Discussion

LAD-I is a rare inborn error of immunity with only ∼300 cases published before 2018, for which disease severity was classified mostly by PMN CD18 expression.⁶ As knowledge of the natural history of disease and underlying disease mechanism has evolved, other features have become important to further characterize the immune phenotype and its role in assessing disease prognosis. This modified Delphi study forms a comprehensive severity classification framework that is reflective of current clinical practice. The study participants concluded that in presymptomatic infants, LAD-I diagnosis and severity can be determined by laboratory assessments alone, whereas in symptomatic patients, clinical phenotype should be considered alongside laboratory assessments. The proposed framework focuses on early LAD-I identification and severity classification, enabling timely treatment considerations before substantial disease complications accumulate. The comprehensive criteria developed for the diagnosis and classification of patients with moderate or severe LAD-I (Table V) expand on the established precedent,^42,43 providing relevant definitions which incorporate both clinical and molecular features. These criteria enable consideration of scenarios where clinical phenotype may drive diagnosis while also accounting for scenarios where patients have yet to present an indicative clinical phenotype but are still considered to have severe LAD-I. For example, newborns and young infants may be diagnosed by immunophenotype and confirmatory molecular genetic testing, particularly given the increasing availability of genetic testing. Similarly, there are published reports of patients with splice-site mutations who show a delayed onset of clinical symptoms,¹⁸ highlighting that severity can change over time. Patients with moderate LAD-I should be monitored and have disease severity regularly reassessed. Potentially definitive therapeutic modalities should be considered in all diagnosed patients classified as having severe LAD-I.

Our research highlights the importance of shared decision-making in patients where treatment options are less clear—that is, for patients with a moderate LAD-I immunophenotype. It should be noted that although treatment of this subpopulation was discussed at length during the study, we do not provide explicit recommendations for the management of patients with moderate disease. This is because the Delphi participants only reported limited direct experience with moderate LAD-I without a severe phenotype. This is in part due to fewer patients with moderate LAD-I in the absence of a severe phenotype being seen in expert transplant centers,^6,18 and future research pertaining to the moderate LAD-I population is warranted. Nevertheless, these criteria will aid clinicians in accurate diagnosis and classification of patients with LAD-I and ensure expedient referral for patient-centered treatment, including definitive therapies. Furthermore, given that our proposed definition for severe LAD-I includes those with an immunophenotype typical of moderate LAD-I if the clinical phenotype is considered severe, our recommendations can be considered relevant to patients with a moderate immunophenotype but severe clinical morbidity.

Importantly, a delayed diagnosis of severe LAD-I leads to elevated risk of mortality^6,8,17,18 and accumulation of infection- and inflammation-related disease complications that may ultimately lead to sepsis and organ damage. While survival past infancy may be indicative of a milder phenotype in other diseases, good/strong consensus indicates that the risk of developing severe infections, infection- or inflammatory-related complications, and mortality persists or increases with age in the absence of definitive treatment.¹² This study also addresses the previously inadequate definition of BSC for patients with severe LAD-I and the paucity of evidence-based standard of care for patients with severe LAD-I. The recommendations we provide for BSC can assist with the management of LAD-I–associated infections and inflammation before definitive therapy.⁴⁴ Recommendations for allo-HSCT also guide clinicians toward optimal timing and preparative strategies for allo-HSCT and further characterize scenarios where allo-HSCT is unsuitable. As part of allo-HSCT preparative strategies, the use of anti-inflammatory therapy (eg, ustekinumab) is recommended to mitigate cytokine-driven inflammation in the bone marrow and support engraftment. However, this study did not assess access to ustekinumab, and variability in availability may have influenced the level of consensus for this recommendation.

Despite substantial progress in the practice of allo-HSCT,^20,21,32 participants recognized the potential for gene therapy as another definitive option to treat severe LAD-I, given current matched donor limitations and GvHD risks. The risks of preconditioning–associated infections are also lower in patients who receive gene therapy compared with allo-HSCT.⁴⁵ While it is not possible to perform randomized controlled trials in this patient population, and long-term supportive data are limited, a long-term follow-up study of lentiviral-based gene therapy for severe LAD-I is currently ongoing (NCT06282432).^24,45,46 Though we acknowledge that equitable access to emerging therapies such as gene therapy is often limited, treatment access was not within the study’s scope. By including emergent practices where they are available, this study may provide support for policy change and improved access for emerging, effective therapies.

The consensus statements/recommendations were developed through robust methodology aligned with best practice.^26,28,29 Key aspects incorporated in the study design were participant anonymity, iteration of Delphi rounds, feedback, and statistical stability of consensus.^28,29 A Delphi first round typically comprises open-ended questions for idea generation,^47,48 which was not conducted; that is why we refer to the design as “modified Delphi.” This study was sponsored by a pharmaceutical company. Delphi panelists were selected according to their expertise and were not employees of the sponsoring company. The sponsor had an active role in panelist selection and question development but was not involved in the analysis or interpretation of data; nor did they influence discussion or decisions made by the panel during the Delphi process. All Delphi panelists disclosed potential conflicts of interest, and responses were anonymized and analyzed independently to minimize bias.

This study is subject to certain limitations. First, patients with LAD-I and their caregivers were not represented in the Delphi, so we are unable to confirm that recommendations pertaining to the burden of illness accurately reflect patients’/caregivers’ experience. Expert elicitation on the likely clinical manifestations of LAD-I was used to complement published evidence. Although quantitative data from large database or registry studies would provide a more robust approach, the rarity of LAD-I and lack of disease-specific ICD-10 coding makes such comprehensive analyses impractical. Expert elicitation provides critical context regarding the significance of clinical manifestations; it can capture aspects not routinely documented in retrospective records, such as educational needs and nuanced patient management considerations. Our estimates for health care resource utilization align with the previous literature for nontransplanted patients with severe LAD-I;¹² however, we acknowledge that only long-term, real-world evidence can provide robust estimates. Delphi studies typically include 15 to 20 participants,²⁸ although there is no minimum size. The small number of participants included in our study reflects the small number of eligible participants with extensive experience in the diagnosis and management of LAD-I, given its rarity. Nevertheless, a panel of 5 experts may be considered sufficient to control for chance agreement.²⁸ Finally, consensus statements were drafted for conciseness, thereby potentially omitting certain nuances of clinical management.

Conclusion

This study is the first to gather international clinical expert opinion on the diagnosis, standard of care, and burden of illness among patients with severe LAD-I. Consolidated recommendations establish best clinical practice, enabling improvement of clinical outcomes for patients with severe LAD-I. Provision of clinically relevant diagnostic and severity classification criteria and contemporary views on prognosis and treatment practices are the first steps to raising awareness of LAD-I (in both moderate and severe forms) in primary care clinicians and specialists, accelerating patient referral and timely provision of definitive treatment. Because this is the first consensus panel on severe LAD-I, further exercises may be required in the future as knowledge grows and potential new treatment modalities are introduced.

Clinical implication.

Clinically relevant criteria and evidence-based assessments improve awareness of LAD-I among primary care physicians and specialists, expediting patient referral and timely provision of definitive therapy.

Disclosure statement

Supported by Rocket Pharmaceuticals (Cranbury, NJ), which provided support for the modified Delphi process (including providing honoraria to the steering committee and Delphi participants) as well as the medical writing and publication costs for this study. Rocket Pharmaceuticals actively participated in the development of this report but was not involved with data collection or management. The financial support provided to authors of this study was provided at fair market value to cover participants’ time only; it was not to influence responses provided by Delphi participants or consensus workshop panel participants.

Disclosure of potential conflict of interest: J. Heimall is speaker for CSL Behring; receives research support from ADMA, CSL Behring, Regeneron, and Sumitomo; is a consultant for CSL Behring and Sumitomo; and receives author royalties from UpToDate. C. Booth reports educational honoraria from Chiesi Farmaceutici; ad hoc consulting for Rocket Pharmaceuticals and Ensoma; and is principal investigator for clinical trials sponsored by Rocket Pharmaceuticals. M. Chitty-Lopez is an employee of Rocket Pharmaceuticals. M. Bailey is a former employee of Rocket Pharmaceuticals. P. Carter, E. Matthews, and J. Lee are employees of HEOR, which received funding from Rocket Pharmaceuticals for this study. S. Prockop receives support for the conduct of clinical trials through Boston Children’s Hospital from Pierre Fabre and Cabaletta; is consultant for Atara Biotherapeutics, Ensoma, HEOM, Pierre Fabre, DSMB Stanford University, and NYBC; holds equity shares in Regatta Biotherapy (a nonexcluded organization); and is inventor of intellectual property related to development of a third-party viral specific T cells program, with all rights assigned to Memorial Sloan Kettering Cancer Center. J. W. Leiding is shareholder at bluebird bio; is consultant for Sobi, Grifols, Amgen, and Rocket Pharmaceuticals; and is speaker for Grifols. The rest of the authors declare that they have no relevant conflicts of interest.