Diagnostic delay, SERPING1 allelic heterogeneity, and real-world lanadelumab prophylaxis in Chinese patients with hereditary angioedema due to C1 inhibitor deficiency

Background

Hereditary angioedema is a rare autosomal dominant disorder characterized by recurrent, unpredictable, non-pruritic, non-pitting swelling of the skin, gastrointestinal tract, and upper airway (1). Population-based studies estimate its prevalence at approximately 1.0–1.6 per 100,000 individuals (2–4). Although awareness of the disease has improved, clinical and genetic data from Chinese patients remain limited (5, 6).

Most cases are caused by heterozygous pathogenic variants in SERPING1, which encodes C1 inhibitor, a serine protease inhibitor involved in the regulation of complement, contact activation, coagulation, and fibrinolytic pathways (7, 8). Deficiency or dysfunction of C1 inhibitor leads to increased plasma kallikrein activity, excessive bradykinin generation, and vascular leakage (1, 9).

HAE-C1INH is classified into type 1 and type 2. Type 1 is characterized by reduced C1 inhibitor antigen and function, whereas type 2 shows normal or increased antigen level with reduced function (7–9). Type 1 is reported to be more common than type 2 in HAE-C1INH cohorts globally (5). In our cohort, 9 of 10 patients (90%) had type 1 and 1 patient (10%) had type 2, consistent with established epidemiological patterns (Table 1). Some patients exhibit typical clinical features with normal C1INH levels, a genetically distinct condition known as HAE with normal C1INH (HAE-nC1INH). A subset of HAE-nC1INH patients has been found to carry pathogenic variants in F12, PLG, ANGPT1, KNG1, MYOF, HS3ST6, CPN, DAB2IP, and other genes (9–12); however, such variants are identified in only a limited proportion of clinically suspected HAE-nC1INH cases.

Table 1. Clinical and laboratory characteristics of 10 patients with HAE-C1INH.

Reference ranges: C4, 0.10–0.40 g/L; C1INH antigen, 0.21–0.39 g/L; C1INH function, ≥ 68.0%; C1q, 159–233 mg/L.

HAE, hereditary angioedema; C1INH, C1 inhibitor; HBV, hepatitis B virus; OCD, obsessive-compulsive disorder; VAS, visual analogue scale.

More than 850 SERPING1 variants have been described, including missense, frameshift, splice-site, nonsense, and large rearrangement variants (8, 13). Most are private variants, reflecting marked allelic heterogeneity. Approximately 30%–34% arise de novo, challenging diagnosis in patients without family history (14). In patients with a strong clinical suspicion but negative conventional exon-based SERPING1 sequencing, deep intronic variants and copy-number variants may require additional methods for detection (15–18).

Due to rarity and non-specific manifestations, HAE is frequently misdiagnosed as allergic angioedema, urticaria, or acute abdomen, causing substantial diagnostic delay and morbidity (19). International surveys report mean diagnostic delays of ∼12.6 years, with 65% experiencing misdiagnosis and 24% undergoing unnecessary abdominal surgery (20). Prolonged diagnostic delays have also been observed in Asia-Pacific regions (6). Early recognition through complement testing and genetic analysis is critical (9, 21).

Lanadelumab, a fully human monoclonal antibody targeting plasma kallikrein, is recommended as first-line long-term prophylaxis (LTP) (9). The HELP trials demonstrated marked attack reduction and favorable tolerability (22–24). Real-world studies from diverse populations, including emerging Asian data, support cross-ethnic effectiveness (25–29). However, additional real-world data from Chinese patients remain limited.

We conducted this single-center retrospective observational study to characterize diagnostic delay, clinical features, complement profiles, and SERPING1 variant diversity in 10 unrelated Chinese patients with HAE-C1INH, and to describe preliminary real-world lanadelumab outcomes.

Methods

Study design and participants

This single-center retrospective observational study included 10 consecutive unrelated index patients with confirmed HAE-C1INH who received care at the Department of Allergy, Henan Provincial People’s Hospital, from January 2022 to May 2026 (Figure 1). Eligible patients were unrelated index cases with a clinical history compatible with recurrent bradykinin-mediated angioedema and laboratory evidence of C1INH deficiency or dysfunction. HAE-C1INH was diagnosed according to the 2021 WAO/EAACI guideline (9) and the Chinese Expert Consensus (30).

Acquired C1INH deficiency was excluded based on clinical context, normal C1q levels, absence of anti-C1INH antibodies tested only in the two index patients without a family history; see below (detected by indirect enzyme-linked immunosorbent assay [ELISA]), and lack of features suggestive of lymphoproliferative or autoimmune disease. Patients with angioedema associated with chronic spontaneous urticaria, drug-induced angioedema such as ACE inhibitor-associated angioedema, or isolated allergic angioedema responsive to antihistamines were excluded. Clinical data were obtained from electronic medical records and telephone follow-up. Recorded variables included sex, age, age at symptom onset, diagnostic delay, family history, edema sites, triggers, attack frequency, attack duration, comorbidities, prior misdiagnoses, laboratory results, genetic findings, and treatments. The final follow-up was completed in May 2026. The study was approved by the Ethics Committee of Henan Provincial People’s Hospital (No. 2022-2211273107). Written informed consent for genetic testing and publication of anonymized clinical and genetic data was obtained from all participants.

Complement and C1INH measurements

Peripheral venous blood (2 mL) was collected at the time of clinical presentation. Because all index patients were confirmed clinically during acute angioedema episodes, blood samples were obtained during these acute attacks; therefore, the laboratory results reported in Table 1 represent acute-phase measurements. Most index patients were referred to and first assessed at our tertiary center during an acute attack—typically the abdominal or laryngeal event that prompted specialist referral—so that the initial confirmatory sampling was obtained during the acute phase. This reflects the referral pattern of a tertiary center rather than an absence of interval diagnosis; symptomatic first-degree relatives identified through family history were advised to undergo interval (non-acute) complement screening and cascade genetic testing. Repeat complement measurements during remission were obtained when clinically feasible to assess whether abnormalities persisted beyond the acute phase. Serum C4, C1q, and C1INH antigen levels were determined by nephelometry using the immunoturbidimetric method (scatter nephometry) (reference ranges: C4, 0.10–0.40 g/L; C1INH antigen, 0.21–0.39 g/L; C1q, 159–233 mg/L). C1INH functional activity was assessed using a chromogenic substrate assay (reference range, ≥ 68%). Anti-C1INH antibodies were tested by indirect enzyme-linked immunosorbent assay (ELISA) only in the two index patients without a family history (Patients 3 and 5), in whom acquired C1INH deficiency required exclusion despite normal C1q levels; antibody testing was not performed routinely in the eight patients with a positive family history. Although normal C1q makes acquired deficiency unlikely, C1q may remain normal in a minority of acquired cases, so antibody testing was used as a confirmatory step in these sporadic cases only. When possible, abnormal complement results were confirmed by repeat testing. Assays were performed in the hospital clinical laboratory according to standardized operating procedures. HAE-C1INH type 1 was defined by reduced C1INH antigenic level and reduced functional activity, whereas HAE-C1INH type 2 was defined by normal or elevated C1INH antigenic level with reduced functional activity.

DNA extraction and SERPING1 sequencing

HAE-C1INH was diagnosed on clinical and complement (biochemical) grounds according to the 2021 WAO/EAACI guideline; SERPING1 sequencing was performed for research and cascade-screening purposes and was not required for clinical diagnosis. Genomic DNA was extracted from 2 mL of peripheral whole blood using the DNeasy Blood & Tissue Kit (Qiagen, Hilden, Germany; Cat. No. 69504) according to the manufacturer’s instructions. Primers targeting the SERPING1 promoter region, all eight exons, and flanking intronic regions were designed using Primer3 (https://bioinfo.ut.ee/primer3-0.4.0/). PCR amplification was performed under standard conditions for 35 cycles. PCR products were purified and sequenced bidirectionally using an ABI 3500 Dx Genetic Analyzer (Applied Biosystems, Foster City, CA, USA).

Sequence data were aligned to the SERPING1 reference transcript NM_000062.3, and variant nomenclature followed Human Genome Variation Society recommendations. Primer sequences and corresponding amplicon sizes are available upon reasonable request. Because Sanger sequencing was limited to the SERPING1 promoter, coding exons, and flanking intronic regions, large deletions or duplications, deep intronic variants, and other copy-number variants may not have been detected.

Variant annotation and classification

Variants were annotated using ClinVar, LOVD, HGMD, and gnomAD v4.1. In silico prediction tools included REVEL for missense variants, MutationTaster/AutoPVS1 for frameshift variants, and SIFT Indel for in-frame deletions. Variant classification was independently performed by two investigators according to ACMG/AMP criteria (31), and discrepancies were resolved by consensus. Variants of uncertain significance were interpreted in the context of clinical phenotype, complement results, family history, and segregation data when available.

Lanadelumab prophylaxis and outcomes

Seven patients received lanadelumab long-term prophylaxis after shared decision-making. The initial dose was 300 mg subcutaneously every 2 weeks (q2w). Dosing intervals were individualized and extended in selected patients meeting either of the following criteria sustained over ≥6 months: (1) attack-free status, defined as zero angioedema episodes during the preceding 6-month observation period; or (2) well-controlled disease, defined as an AECT score ≥10 with no more than one mild breakthrough attack per year that did not require acute rescue treatment. Extension was not protocol-driven but individualized based on clinical judgment and patient-specific factors. Post-treatment AAR was calculated as the number of attacks during lanadelumab prophylaxis divided by the duration of lanadelumab exposure in years.

Clinical outcomes included: (1) AAR, defined as the number of attacks per year at baseline, based on the 12 months before lanadelumab initiation, and during the prophylaxis period, annualized according to the duration of lanadelumab exposure; (2) disease control, assessed by the AECT (score range, 0–16; ≥10 indicating well-controlled disease) at baseline, based on the 3-month period before treatment, and at the most recent 3-month follow-up; and (3) safety outcomes, including breakthrough attacks, injection-site reactions, serious adverse events, thromboembolic events, and treatment discontinuation. Breakthrough attacks were defined as angioedema episodes occurring during lanadelumab prophylaxis, irrespective of whether rescue medication was used. During the study period (January 2022–May 2026), multiple on-demand treatments were available in China for acute HAE-C1INH attacks, including: (1) icatibant (30 mg subcutaneous injection); (2) fresh-frozen plasma (FFP); and (3) attenuated androgens (danazol) in selected cases. As documented in Table 1, Patients 1 had previously received FFP. Patient 9 received subcutaneous icatibant 30 mg for an acute abdominal attack.

Statistical analysis

Statistical analyses were conducted using SPSS version 20.0 (IBM Corporation, Armonk, NY, USA). Normality was assessed with the Shapiro–Wilk test. Normally distributed continuous variables are expressed as mean ± standard deviation (SD), and non-normally distributed variables are expressed as median (interquartile range [IQR]). Categorical variables are presented as counts and percentages. Given the small sample size and paired design, pre- and post-treatment AAR and AECT scores were compared using the exact two-sided Wilcoxon signed-rank test. No formal power calculation was performed because of the exploratory nature of the study and the limited sample size. No adjustment for multiple comparisons was applied. P-values were interpreted descriptively rather than as confirmatory evidence.

Results

Baseline clinical characteristics

Baseline demographic and clinical data are summarized in Table 1. The cohort included 6 females and 4 males. Mean age at last follow-up was 32.8 ± 9.9 years, ranging from 18 to 46 years. Nine patients had HAE-C1INH type 1 and one had type 2. Eight patients had a positive family history, whereas Patients 3 and 5 had no known family history.

Mean age at symptom onset was 21.7 ± 11.1 years. Median age at onset was 20.5 years, with a range of 7–39 years. Mean diagnostic delay was 9.4 ± 7.0 years, and the longest delay was 23 years.

All patients had recurrent peripheral edema. Seven patients had abdominal attacks, and 6 patients had facial/laryngeal involvement. Attacks were non-pitting, non-pruritic, and asymmetric. Most episodes resolved spontaneously within 2–7 days.

Reported triggers included emotional stress, fatigue, cold exposure, menstruation, minor trauma, and prolonged standing. Three patients had no identifiable trigger.

Before appropriate diagnosis or treatment, all patients had received oral antihistamines without benefit. One patient had been treated for chronic urticaria. Five patients with abdominal symptoms had been hospitalized for suspected pancreatitis, appendicitis, or gastroenteritis. Two received fresh-frozen plasma with clinical improvement. Patient 10 had severe laryngeal edema requiring emergency tracheotomy and had previously received fresh-frozen plasma with clinical improvement.

Comorbidities included obsessive-compulsive disorder, chronic hepatitis B, allergic asthma, anemia, drug allergy to amoxicillin, and pulmonary tuberculosis. Based on available clinical records and annualized estimates before treatment, the mean attack frequency was 13.4 ± 8.3 attacks/year. The mean untreated attack duration was 3.4 ± 1.3 days, and the mean peak VAS score was 6.9 ± 2.5.

Complement and C1INH results

Serum C4 was reduced in all 10 patients during attacks, with a mean value of 0.049 ± 0.019 g/L (reference 0.10–0.40 g/L). Because all patients were confirmed during acute attacks, the C4 values reported in Table 1 represent acute-phase measurements. To address whether complement reduction was transient (acute consumption artifact) or persistent, repeat complement measurements during remission were obtained when clinically feasible. Remission-phase sampling was available for 6 patients (Patients 1, 2, 4, 6, 8, and 10). C4 remained reduced during remission (mean 0.061 ± 0.015 g/L), and the acute-phase versus remission-phase difference was not statistically significant (p = 0.78). Among the 9 patients with type 1 HAE-C1INH, mean C1 inhibitor antigen was 0.069 ± 0.029 g/L (reference 0.21–0.39 g/L), and median functional activity was 11.0% (IQR 4.3–13.5%; reference ≥ 68%). Patient 4, who had type 2 HAE-C1INH, had increased C1 inhibitor antigen of 0.70 g/L and reduced function of 6.3%. Serum C1q was within the normal range in all patients (range 163–222 mg/L; reference 159–233 mg/L), consistent with hereditary rather than acquired C1INH deficiency (Table 1).

SERPING1 variants

Sanger sequencing identified 10 distinct heterozygous SERPING1 variants, each observed in a single family (Table 2).

Table 2. SERPING1 variants identified in 10 unrelated patients with HAE-C1INH.

ACMG/AMP evidence codes were assigned based on population frequency, predicted variant effect, previous reports, phenotype specificity, and segregation data when available.

The variant spectrum included:

Frameshift variants (n = 3): c.933_934insGGAA (p.Pro312Glyfs17), c.6dup (p.Ser3Leufs17), and c.1358_1359insCTGG (p.Val454Trpfs*20).

Missense variants (n = 6): c.1067T > A (p.Val356Glu), c.1034G > A (p.Gly345Glu), c.1396C > T (p.Arg466Cys), c.1346T > A (p.Leu449Gln), c.1483G > A (p.Val495Ile), and c.1121T > C (p.Leu374Pro).

In-frame deletion (n = 1): c.816_818delCAA (p.Asn272del).

All variants were absent from or extremely rare in gnomAD v4.1. According to ACMG/AMP criteria, c.1396C > T was classified as pathogenic; seven variants were classified as likely pathogenic; and two, c.1483G > A and c.1121T > C, were classified as variants of uncertain significance (VUS).

Co-segregation information was available for 8 of the 10 families (Table 2), providing supportive evidence for variant interpretation in several families.

Long-term prophylaxis with lanadelumab

Seven patients (Patients 2, 4, 5, 6, 7, 8, and 10) received lanadelumab for LTP. Specifically, four patients (Patients 5, 7, 8, and 10; 57%) achieved complete attack-free status and had intervals extended to q6w (Patients 5, 7, 8) or q4w (Patient 10). Two additional patients (Patients 2 and 4; 29%) achieved well-controlled disease with only sporadic mild breakthrough attacks (AAR <1 per year) and had intervals extended to q8w (Patient 2) and q10w (Patient 4). Patient 6 (14%), despite achieving AECT ≥13 (well-controlled), experienced one mild abdominal breakthrough attack. Overall, interval extension ranged from q4w to q10w at the most recent follow-up (Table 3).

Table 3. Lanadelumab prophylaxis regimens and clinical outcomes in seven patients with HAE-C1INH.

Lanadelumab was initiated at 300 mg subcutaneously every 2 weeks in all treated patients. Dose intervals were subsequently extended according to individual disease control after shared decision-making. “None” indicates no breakthrough attacks during lanadelumab prophylaxis. q2w, every 2 weeks; q4w, every 4 weeks; q6w, every 6 weeks; q8w, every 8 weeks; q10w, every 10 weeks.

Effect on attack frequency and disease control

Changes in AAR and AECT scores are summarized in Table 4.

Table 4. Changes in annualized attack rate and Angioedema Control Test scores before and after lanadelumab prophylaxis.

Data are presented as individual patient values or median (IQR). Baseline AAR refers to the 12 months before lanadelumab initiation, and post-treatment AAR was annualized according to each patient’s duration of lanadelumab exposure. Baseline AECT refers to the 3 months before lanadelumab initiation, and post-treatment AECT refers to the most recent 3 months. Paired baseline and post-treatment values were compared using the exact two-sided Wilcoxon signed-rank test. P-values were interpreted descriptively because of the exploratory nature of the study and limited sample size. AAR, annualized attack rate; AECT, angioedema control test; HAE, hereditary angioedema; HAE-C1INH, hereditary angioedema due to C1 inhibitor deficiency; IQR, interquartile range.

Among the 7 patients receiving lanadelumab, median AAR decreased from 11 (IQR, 7–23) to 0 (IQR, 0–1) attacks/year (p = 0.016). Median AECT score increased from 3 (IQR, 2–4) to 13 (IQR, 13–13) (p = 0.016). Four of seven patients (57%) remained completely attack-free during the observation period. Three patients experienced mild breakthrough attacks during lanadelumab prophylaxis: Patient 2 experienced multiple mild attacks, all of which resolved spontaneously; Patient 4 experienced one mild breakthrough attack with hip swelling, which resolved within 2 h; and Patient 6 experienced one mild abdominal attack, which resolved within 1 h. No breakthrough attacks required acute treatment. All seven patients achieved AECT scores ≥10 during treatment, indicating well-controlled disease.

Safety

Lanadelumab was well tolerated. No serious adverse events, thromboembolic events, or treatment discontinuations occurred. No clinically significant injection-site reactions were recorded.

Duration of lanadelumab exposure varied across the cohort according to treatment initiation dates. The seven patients receiving lanadelumab had a median follow-up duration of 18 months (range, 12–26 months) as of May 2026. Specifically, Patient 2 had the longest exposure (26 months; 2.2 years), Patient 4 had 30 months (2.5 years), Patient 5 had 25 months (2.1 years), Patient 6 had 23 months (1.9 years), Patient 7 had 20 months (1.7 years), Patient 8 had 19 months (1.6 years), and Patient 10 had 12 months of follow-up. All patients remained on treatment at the time of manuscript preparation, with no discontinuations due to adverse events or loss of efficacy.

Acute treatment with icatibant

Patient 9 received subcutaneous icatibant 30 mg for an acute abdominal attack. Symptom relief was achieved within 30 min, and the attack resolved completely within 2 h.

Patients not receiving long-term prophylaxis

Three patients did not receive LTP. Patient 1, who had comorbid obsessive-compulsive disorder, experienced reduced attack frequency after optimization of psychotropic medications and elected to defer LTP. Patient 3 had infrequent, mild attacks and declined prophylaxis after shared decision-making. Patient 9 declined both attenuated androgens and lanadelumab but retained access to on-demand icatibant for acute attacks.

Discussion

In this single-center retrospective observational study of 10 unrelated Chinese index patients with HAE-C1INH, we observed long diagnostic delays, frequent misdiagnosis before confirmation, marked heterogeneity of SERPING1 variants, and improved disease control among patients receiving lanadelumab prophylaxis.

Diagnostic delay and misdiagnosis

The mean diagnostic delay in this cohort was 9.4 years, with the longest delay reaching 23 years. This finding is consistent with previous Chinese and international reports showing substantial delays in the diagnosis of HAE-C1INH (20, 32). In the present series, abdominal attacks were often mistaken for pancreatitis, appendicitis, or gastroenteritis, whereas cutaneous swelling was misdiagnosed as allergic angioedema or chronic urticaria. One patient required emergency tracheostomy before HAE was diagnosed.

These findings indicate that HAE-C1INH should be considered in patients with recurrent angioedema without wheals, particularly when attacks are non-pruritic, poorly responsive to antihistamines or corticosteroids, or accompanied by recurrent abdominal pain or laryngeal symptoms. Earlier testing of serum C4 and C1 inhibitor antigen and function in emergency, gastroenterology, dermatology, otolaryngology, and allergy settings may help reduce diagnostic delay (9, 21, 30).

Clinical phenotype and laboratory diagnosis

The clinical manifestations in this cohort were broadly consistent with previous reports of HAE-C1INH (6, 33, 34). All patients had peripheral edema, 70% had abdominal attacks, and 60% had facial or laryngeal involvement. The relatively high proportion of laryngeal involvement may reflect referral bias to a tertiary center, but it also underlines the need for timely diagnosis and access to on-demand treatment, as untreated laryngeal attacks can be fatal (35).

Reported triggers included emotional stress, fatigue, cold exposure, menstruation, minor trauma, and prolonged standing, consistent with known precipitating factors in HAE-C1INH (36). The role of hormonal factors as triggers could not be assessed in this small cohort.

All patients had reduced C4 levels during attacks. C4 remains a useful first-line screening test, but normal C4 outside attacks does not exclude HAE-C1INH when clinical suspicion is high (9, 37, 38). Therefore, C1 inhibitor antigen and functional testing remain essential for diagnosis. Normal C1q levels in all patients supported hereditary rather than acquired C1 inhibitor deficiency (9, 39). Consistent with this, anti–C1INH antibody testing was reserved for the two sporadic cases without a family history and was not performed routinely, because a positive family history together with normal C1q made acquired deficiency clinically improbable in the remaining patients.

SERPING1 variant spectrum

Ten distinct heterozygous SERPING1 variants were identified in 10 unrelated families, with no recurrent variant. This finding is consistent with the marked allelic heterogeneity of HAE-C1INH (8, 13, 14). The variant spectrum included frameshift variants, missense variants, and one in-frame deletion.

Frameshift variants are expected to cause loss of function, most likely through nonsense-mediated mRNA decay or production of truncated non-functional protein (7, 8). The c.1396C > T variant in Patient 4 was associated with HAE-C1INH type 2, consistent with previous genotype–phenotype observations for variants affecting the reactive center loop (8, 40). The c.1067T > A variant identified in Patient 2 has also been reported in a Chinese patient with type 1 HAE-C1INH (41). Other missense variants were interpreted according to population frequency, in silico predictions, phenotype compatibility, and available segregation evidence.

Two variants, c.1483G > A and c.1121T > C, were classified as variants of uncertain significance. Both variants were detected in patients with typical clinical features and compatible complement abnormalities. However, they should not be considered disease-causing without additional segregation or functional evidence. Because functional validation was unavailable and segregation evidence was limited, these variants should be interpreted cautiously and re-evaluated when additional genetic or functional evidence becomes available.

These findings also highlight the genetic heterogeneity of HAE-C1INH and the potential value of comprehensive genetic evaluation in selected patients, particularly sporadic cases and families requiring cascade screening. Broader testing may be considered when conventional exon-based sequencing is negative or when clinical and genetic findings are discordant (15–18).

Lanadelumab prophylaxis

Lanadelumab prophylaxis was associated with a marked reduction in attack frequency and improved disease control in this cohort. Four of the seven treated patients remained attack-free during follow-up, and the remaining three experienced only mild breakthrough attacks during lanadelumab prophylaxis, none of which required acute treatment. No serious adverse events or treatment discontinuations were recorded.

These observations are consistent with the HELP trial program and real-world studies showing that lanadelumab substantially reduces attack frequency and improves disease control in patients with HAE-C1INH (22–29). The present findings add to the limited real-world evidence on lanadelumab use in Chinese patients (27). In this cohort, good disease control was maintained in several patients after individualized extension of dosing intervals, which ranged from every 4 to every 10 weeks at the last follow-up. However, interval extension was individualized rather than protocol-driven, and treatment duration varied among patients. Therefore, these findings should be interpreted as preliminary and should not be considered evidence supporting a specific dose-extension schedule. Prospective studies with predefined criteria for interval adjustment are needed.

During the study period (January 2022–May 2026), the therapeutic landscape for HAE-C1INH in China evolved substantially. Lanadelumab became available and represented the targeted first-line long-term prophylaxis option recommended per international guidelines (9, 21, 30). Before lanadelumab became available (before ∼2021), management in China relied primarily on non-targeted strategies, including FFP for acute attacks and attenuated androgens (danazol) for long-term prophylaxis; icatibant became available in some major centers but with limited accessibility and higher cost.

In the wider Chinese context, HAE remains substantially underdiagnosed. Although population-based prevalence estimates of ∼1.0–1.6 per 100,000 would predict roughly 14,000–22,000 affected individuals nationwide, only a few hundred to around one to two thousand patients have been formally identified, indicating a large diagnostic gap. Expertise and diagnostic capacity are concentrated in a small number of tertiary referral centers in major cities—most prominently Peking Union Medical College Hospital, which coordinates the national HAE registry and the Chinese HAE patient organization—together with a limited number of provincial referral centers such as ours. Complement and C1INH functional testing, on-demand icatibant, and lanadelumab are largely confined to these urban centers, whereas patients in rural and lower-tier settings face limited testing availability, longer referral pathways, and reduced access to targeted therapy. Reimbursement is expanding but remains geographically uneven. Strengthening the national registry, broadening cascade screening of affected families, and decentralizing diagnostic testing are therefore key priorities for reducing diagnostic delay and access inequity in China.

As of mid-2026, lanadelumab has become increasingly accessible through medical insurance coverage in major urban centers, though significant regional disparities in availability and reimbursement persist. For acute on-demand management, icatibant is now available as the targeted bradykinin B2 receptor antagonist in select hospitals, but remains more limited in distribution than FFP; FFP continues to be the most widely accessible acute treatment option across both urban and rural regions and hospital tiers. Attenuated androgens continue to be used in some centers as an alternative long-term prophylaxis option, typically where lanadelumab is unavailable, rather than as an add-on to lanadelumab. In our cohort, no patient received attenuated androgens in combination with lanadelumab; combining attenuated androgens with lanadelumab is not consistent with current clinical practice and would not be an appropriate treatment approach, as guideline-recommended targeted prophylaxis is intended to replace, rather than supplement, non-targeted androgen therapy.

In this single-center cohort (based at a tertiary teaching hospital in a provincial capital), lanadelumab was accessible and selected as the long-term prophylaxis strategy in 7 of 10 patients, with demonstrable efficacy and safety benefits. Three patients declined long-term prophylaxis for various reasons: Patient 1 experienced reduced attack frequency after optimization of psychotropic medications for comorbid obsessive-compulsive disorder and elected to defer LTP; Patient 3 had infrequent, mild attacks and declined prophylaxis after shared decision-making; and Patient 9 declined both attenuated androgens and lanadelumab but retained access to on-demand icatibant for acute attacks. These findings support the expanding use of targeted biological therapies in Chinese patients with HAE-C1INH, consistent with international guidelines recommending early intervention for high-risk patients. However, they also underscore the need for further expansion of treatment options, improved awareness among healthcare providers, and enhanced healthcare system readiness to support timely diagnosis and equitable access to advanced therapies across urban and rural regions and hospital tiers. Access barriers and regional heterogeneity remain significant challenges to optimizing HAE management in China.

Acute treatment and psychological factors

One patient received icatibant for an acute abdominal attack and had symptom relief within 30 min and complete resolution within 2 h, consistent with the known efficacy of icatibant as a bradykinin B2 receptor antagonist for acute HAE attacks (42).

Patient 1 reported fewer attacks after optimization of treatment for obsessive-compulsive disorder. Emotional stress is a recognized trigger of HAE attacks (36), and psychosocial burden is common in patients with HAE (43, 44). This single observation does not establish causality, but it suggests that psychological assessment and stress management may be useful components of care in selected patients.

A previous Chinese report has described lanadelumab prophylaxis, regional genetic diagnosis, and individualized management of HAE-C1INH (45). The present study extends these observations by integrating diagnostic delay, complement phenotyping, SERPING1 variant classification, and real-world individualized lanadelumab prophylaxis outcomes in 10 unrelated index patients. This combined clinical-genetic-treatment perspective may provide additional insights into HAE-C1INH recognition and management in China.

Limitations

This study has limitations. The small sample size and single-center retrospective design limit generalizability and may have introduced recall bias, particularly because some pre-treatment attack data were obtained through telephone follow-up. Follow-up duration and lanadelumab dosing intervals varied, and interval extension was individualized rather than protocol-driven; therefore, off-label extension beyond approved schedules should be interpreted cautiously. In addition, comparisons of AAR and AECT scores were exploratory because only seven patients received lanadelumab, and p-values should be considered descriptive. Genetic analysis was limited to Sanger sequencing of the SERPING1 promoter, coding exons, and flanking intronic regions, so large deletions or duplications, deep intronic variants, and variants in other HAE-associated genes may have been missed. Segregation analysis was incomplete in some families, and functional validation was unavailable for VUS. Despite these limitations, our findings provide additional real-world evidence on the diagnosis, SERPING1 allelic heterogeneity, and individualized lanadelumab prophylaxis of HAE-C1INH in Chinese patients.

Conclusion

This single-center retrospective observational study suggests that Chinese patients with HAE-C1INH may experience long diagnostic delays and frequent misdiagnosis before confirmation. Ten distinct SERPING1 variants were identified in 10 unrelated families, supporting the marked genetic heterogeneity of the disease. Lanadelumab prophylaxis was associated with reduced attack frequency, improved AECT scores, and good tolerability in this small real-world cohort. Earlier complement testing, appropriate genetic analysis, and access to both on-demand and prophylactic therapies may improve the care of Chinese patients with HAE-C1INH. Larger prospective multicenter studies are needed to clarify genotype–phenotype correlations and optimize individualized long-term prophylaxis strategies.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Enza D'Auria, University of Milan, Italy

Reviewed by: Mauro Cancian, University Hospital of Padua, Italy

Michihiro Hide, Hiroshima City Hospital Organization, Japan