Abstract
Background
Bullous pemphigoid research spans autoantigen and basement membrane biology, diagnosis, comorbidity, inflammatory pathways, and treatment, but its long-term conceptual development has not been mapped.
Methods
English-language articles and reviews indexed in the Science Citation Index Expanded through December 31, 2025, were retrieved from title, author-keyword, and abstract fields. Records underwent AI-assisted first-pass screening followed by human review. Bibliometrix and VOSviewer were used for parameter-tested author-keyword clustering and a full-corpus title-term sensitivity analysis.
Results
Of 3,471 screened records, 2,121 were eligible. Annual output accelerated after 2015 and exceeded 100 publications in 2022, 2023, and 2025. Author keywords were available for 55.1% of publications. The selected network contained 112 keywords, 889 links, and eight clusters. Among 924 keyword-mappable publications, autoantigens, autoantibodies, and basement membrane zone biology declined from 51.9% in 1959–1998 to 14.6% in 2021–2025. Therapeutic management and clinical outcomes increased from 7.7% to 37.8%, epidemiology and comorbidities from 1.9% to 20.8%, dipeptidyl peptidase-4 inhibitor-associated bullous pemphigoid from 0% to 12.7%, and immune checkpoint inhibitor-associated bullous pemphigoid from 0% to 10.8%. Diagnosis and related autoimmune blistering diseases peaked at 50.0% in 1999–2013 before declining to 24.3%. Inflammatory and immune effector mechanisms showed no statistically significant period variation after multiplicity adjustment. The title-term analysis recovered the major domains.
Conclusion
Bullous pemphigoid research diversified from structural, immunological, and diagnostic foundations toward treatment, clinical outcomes, comorbidity, and medication-associated disease. Although author-keyword availability was incomplete and varied substantially over time, the findings provide a longitudinal map of the field, with temporal estimates reflecting relative prominence within keyword-mappable publications.
Keywords: bullous pemphigoid, autoimmune blistering disease, treatment outcomes, immune checkpoint inhibitor, bibliometric analysis, science mapping
Introduction
Bullous pemphigoid is an acquired subepidermal autoimmune blistering disease characterized by autoantibodies directed mainly against bullous pemphigoid antigen 180 (BP180) and bullous pemphigoid antigen 230 (BP230), leading to dermoepidermal separation and blister formation. Although uncommon in the general population, it is the most frequent autoimmune blistering disease in older adults, and the burden of pruritus, erosions, and treatment-related adverse effects is substantial.1 Reported incidence has risen in recent decades. Proposed explanations include population aging, recognition of nonbullous presentations, increased diagnostic awareness, and exposure to implicated medications.2 A meta-analysis of 27 studies estimated the global incidence at 8.2 per million people, with higher rates in Europe than in Asia (10.3 vs 5.6).3 Incidence rises steeply with age. In Germany it was 19.6 per million person-years overall and 262 per million person-years among people aged 85 to 90 years.4 Bullous pemphigoid is also associated with substantial long-term mortality, with reported 1-year mortality ranging from 3.2% to 32.0% and 5-year mortality from 27.4% to 71.9%, depending on the population studied and duration of follow-up.5
The scientific understanding of bullous pemphigoid has expanded considerably. Foundational studies established the disease as an autoantibody-mediated disorder involving hemidesmosomal antigens and basement membrane disruption. Subsequent work has broadened this pathogenic framework by examining eosinophils, cytokine signaling, complement activation, type 2 inflammation, and other immune pathways.6 Diagnostic research has also progressed beyond conventional histopathology, direct immunofluorescence, and serologic testing to include newer assay platforms, biomarker studies, and noninvasive imaging approaches.7–9
Recent therapeutic studies have evaluated alternatives to systemic corticosteroids and broad immunosuppression. The Bullous Pemphigoid Steroids and Tetracyclines (BLISTER) trial showed that an initial doxycycline strategy was noninferior to prednisolone for short-term blister control and was associated with fewer severe, life-threatening, or fatal treatment-related adverse events over longer follow-up. These findings encouraged evaluation of safer initial treatment strategies for older patients.10 In 2025, the approval of dupilumab for adult patients with bullous pemphigoid marked a milestone in targeted therapy and reflects the growing clinical emphasis on steroid-sparing, mechanism-informed treatment approaches.11 Together with studies of omalizumab, rituximab,12 Janus kinase pathway inhibition,13 and other emerging agents,14 these developments illustrate the increasing orientation of bullous pemphigoid research toward treatment outcomes, safety, and individualized care.
The clinical framing of bullous pemphigoid has also expanded beyond skin-limited blistering disease. Bullous pemphigoid occurs predominantly in older adults and is frequently studied in relation to neurological disorders,15 cardiometabolic and renal comorbidity,16 medication exposure,17 malignancy-related associations,18 and immune checkpoint inhibitor-associated disease.19 These associations have practical relevance because comorbid illness, frailty, polypharmacy, and treatment-related adverse effects can shape diagnostic evaluation, treatment selection, and outcomes. Comorbidity research has placed greater emphasis on the clinical course of bullous pemphigoid in medically complex older patients.
Several narrative reviews have synthesized current knowledge on the pathogenesis, diagnosis, and management of bullous pemphigoid.20,21 These reviews provide clinically focused summaries of established evidence, but they do not quantify the broader structure, contributors, citation patterns, or long-term thematic development of the field. In the present study, a bibliometric and science-mapping approach was used to examine the macro-level evolution of bullous pemphigoid research across a screened, disease-focused Web of Science corpus. This approach complements, but does not replace, systematic reviews, scoping reviews, or mapping reviews, which are better suited to evaluating specific clinical questions, synthesizing evidence on defined interventions or outcomes, or classifying a more narrowly defined literature by substantive content. Bibliometric findings should therefore be interpreted as indicators of research activity, visibility, and conceptual organization, while recognizing that they depend on database coverage, indexing quality, and the availability of bibliographic metadata, and do not constitute direct evidence of clinical effectiveness, disease burden, or research quality.
A previous bibliometric study of pemphigoid diseases examined the 100 most-cited publications and showed that influential papers often addressed molecular mechanisms, management, and risk factors.22 However, analyses restricted to highly cited articles capture only a selective segment of the literature and cannot fully characterize field-wide publication growth, contributor patterns, author-keyword networks, or temporal changes in research priorities. A full-corpus bibliometric and science-mapping analysis may therefore provide a more comprehensive view of how bullous pemphigoid research has evolved, which thematic domains have become established, and which topics are gaining clinical and translational relevance. In addition, a bibliometric analysis has been conducted in pemphigus, a related autoimmune blistering disease. Öztekin and Öztekin analyzed 3,034 Web of Science-indexed articles identified through a title search for “pemphigus”. Their analysis identified thematic clusters involving desmoglein and desmosomal biology, autoimmunity and autoantibodies, acantholysis, treatment, rituximab, corticosteroids, quality of life, remission and relapse, and epidemiology and mortality.23
This study aimed to characterize the evolution of bullous pemphigoid-focused research from 1959 to 2025 using bibliometric indicators, author-keyword co-occurrence analysis, and science mapping with bibliometrix and VOSviewer. The specific objectives were to: (1) identify the countries/regions, institutions, journals, and publications contributing most prominently to the bullous pemphigoid literature; (2) determine the thematic domains structuring the field, as reflected by author-keyword co-occurrence patterns; and (3) assess how the relative prominence of these domains and clinically relevant topics changed over time. By mapping a screened corpus of Web of Science-indexed publications focused on bullous pemphigoid, rather than restricting the analysis to highly cited studies, this study provides a field-level view of established research domains spanning translational research and outcome-oriented care.
Materials and Methods
Study Design and Reporting Guidance
This study was designed as a bibliometric analysis of a database-defined corpus of bullous pemphigoid-focused publications indexed in Web of Science. Reporting was guided by the Preliminary Guideline for Reporting Bibliometric Reviews of the Biomedical Literature (BIBLIO).24 A completed reporting checklist is provided in Supplementary Table S1, documenting the database source, search strategy, eligibility criteria, screening procedures, counting method, network construction, clustering settings, sensitivity analyses, and data-availability statement.
Data Source and Search Strategy
A literature search was conducted on August 1, 2026, using the Science Citation Index Expanded of the Web of Science Core Collection. Web of Science was chosen because it provides standardized metadata, including citation, cited-reference, author-keyword, affiliation, and reprint-address records, that are fully compatible with bibliometrix and VOSviewer.
The dataset was restricted to English-language original articles and review articles published from the earliest indexed record through December 31, 2025. These document types were selected because they represent the primary original research and synthesis literature of the field. Non-article document types, such as editorials, letters, meeting abstracts, notes, corrections, book chapters, and retracted publications, were excluded.
To ensure specificity and avoid studies mentioning bullous pemphigoid only incidentally, we omitted the Topic (TS) search field, as “Keywords Plus” frequently introduces tangential results. Instead, we restricted the query to the title (TI), author keywords (AK), and abstract (AB) fields:
TI = (“bullous pemphigoid” OR “pemphigoid, bullous”) OR AK = (“bullous pemphigoid” OR “pemphigoid, bullous”) OR AB = (“bullous pemphigoid” OR “pemphigoid, bullous”).
Artificial Intelligence-Assisted Screening and Human Review
Following bibliographic retrieval, each record underwent artificial intelligence (AI)-assisted first-pass relevance screening based on its title, abstract, and author keywords. The GPT-5.6 Sol model was accessed through the ChatGPT Work environment (OpenAI, San Francisco, CA, USA) to apply the prespecified screening workflow to the uploaded bibliographic records. Each record received a provisional classification of Include, Exclude, or Uncertain. For each classification, the model was instructed to identify the applicable eligibility criterion, provide a brief rationale with supporting text from the record, and assign a confidence level.
Publications were excluded when bullous pemphigoid appeared solely as background information, an incidental entry in a disease list, a differential diagnosis, a non-substantive comparator, shared antigen nomenclature, or a contextual reference in research primarily concerning another disease. General studies of collagen XVII, dystonin, laminin, integrins, hemidesmosomes, or basement-membrane biology were retained only when bullous pemphigoid autoimmunity, patient sera, autoantibodies, diagnosis, or pathogenesis constituted a principal or meaningfully co-primary scientific objective. The complete screening prompt is provided in Supplementary File S1, and the AI-assisted first-pass decision framework is summarized graphically in Supplementary Figure S1.
All AI-generated classifications were considered provisional and underwent human review. After the initial human review, the complete set of classification decisions underwent an AI-assisted consistency audit to identify potentially erroneous inclusions or exclusions and records for which the available bibliographic information was insufficient to determine eligibility. Records flagged during this audit were reassessed by a human investigator using the title, abstract, and author keywords, with the full text consulted when necessary. The overall screening and human-review workflow is shown in Figure 1.
Figure 1.
Flow diagram of record identification, AI-assisted screening, human review, and selection of the final bullous pemphigoid-focused corpus, 1959–2025. Document-type categories were not mutually exclusive; therefore, the category-specific counts exceed the number of unique records excluded.
Abbreviations: AI, artificial intelligence; AK, author keywords; AB, abstract; TI, title.
Bibliometric Indicators
Descriptive bibliometric indicators included annual publication output, annualized citation impact, leading countries/regions, productive institutions, leading journals, and highly cited publications. Citation data were based on Web of Science Core Collection citation counts exported on the search date. Annualized citation impact was summarized using mean total citations per year, calculated for each publication as: total Web of Science citations/(2025 − publication year + 1). This measure was used to reduce bias from longer citation accumulation time among older publications. In addition, normalized total citations were calculated by dividing each article’s total citation count by the mean total citation count of all publications in the study corpus published in the same year. Values above 1 indicated citation impact greater than the corpus-specific annual average.
Country-level publication output was attributed using the Web of Science reprint-address field (RP), which identifies corresponding-author addresses. This corresponding-author approach was selected to provide responsibility-based, article-level attribution. Each semicolon-delimited RP entry was parsed to identify its country, and country labels were standardized for spelling variants, abbreviations, and postal-address formats. For a publication with N corresponding-author entries, each entry received a fractional weight of 1/N. When multiple corresponding authors from the same country were listed, their weights were summed before aggregation by country. This fractional allocation across the countries represented in the RP field ensured that each publication contributed a total country weight of 1.0.
Institutional publication counts represented the number of articles in which each distinct Web of Science organization-enhanced institution appeared at least once. Duplicate occurrences of the same normalized institution within an article were removed; therefore, multiple authors affiliated with the same institution contributed one article count to that institution.
Author Keyword Co-Occurrence Analysis
Author-keyword co-occurrence analysis was used to examine the conceptual structure of bullous pemphigoid-focused research. Author keywords were extracted exclusively from the Web of Science author-keyword field (DE). Before analysis, the keywords were standardized using a predefined thesaurus containing synonym-merging and suppression rules. Synonym-merging rules consolidated abbreviations, alternative nomenclature, spelling variants, and singular and plural forms into canonical terms. Suppression rules removed the disease name itself and other generic or corpus-uninformative expressions that were widely distributed across the corpus but did not meaningfully distinguish thematic domains. The complete thesaurus is provided in Supplementary File S2.
The co-occurrence network was constructed in VOSviewer using full counting, such that each occurrence of a keyword and each co-occurrence between a pair of keywords contributed equally to the analysis. A co-occurrence link was established when two keywords appeared in the same publication, and link strength represented the number of publications in which the keyword pair co-occurred. Link weights were normalized using the association-strength method implemented in VOSviewer. This normalization expresses the relatedness of a keyword pair relative to the overall occurrence frequencies of the individual keywords and reduces the influence of highly frequent general terms.
Keywords were grouped using the modularity-based VOS clustering technique, with optimization performed using the smart local moving algorithm. The clustering settings were held constant across all parameter combinations: the number of random starts was set to 10, the number of iterations to 10, the random seed to 0, and the minimum cluster size to 1. In a sensitivity analysis of the selected parameter combination, increasing the number of iterations from 10 to 1,000 did not alter the cluster partition, with an adjusted Rand index (ARI) of 1.00 and normalized mutual information (NMI) of 1.00. The spatial layout was generated using the VOS mapping technique, in which keywords with stronger normalized co-occurrence relationships are positioned closer together. Cluster numbers and colors were treated as arbitrary identifiers without ordinal meaning.
A two-dimensional sensitivity analysis was conducted to select the minimum keyword-occurrence threshold and clustering resolution. Minimum occurrence thresholds of 4, 5, 6, and 7 publications were each evaluated at clustering resolutions of 0.6, 0.7, 0.8, 0.9, and 1.0, yielding 20 alternative network solutions. All settings other than the occurrence threshold and resolution parameter were held constant.
Item-level cluster assignments were exported from VOSviewer and compared externally using the ARI and normalized mutual information. Both measures are invariant to differences in the numerical labels assigned to clusters. Comparisons among resolution settings at the same occurrence threshold included all keywords retained at that threshold. Because different occurrence thresholds retained different sets of keywords, comparisons across thresholds were restricted to the keywords shared by the two networks being compared.
Parameter selection considered several complementary criteria: the number of retained keywords and links, the stability of cluster assignments across adjacent and alternative parameter settings, the presence of very small or peripheral clusters, the clinical and conceptual coherence of the resulting clusters, and the readability of the network visualization. Stability across resolutions was summarized using pairwise ARI and NMI values. The minimum occurrence threshold of six publications and a clustering resolution of 0.9 were selected for the primary analysis because this combination provided the most favorable balance between thematic coverage, partition stability, cluster interpretability, and network complexity.
Cluster labels were assigned after examining the most frequently occurring keywords, keywords with the greatest total link strength, and the broader clinical or biological relationships represented within each cluster. Labels were intended as concise interpretive summaries rather than classifications imposed before network construction.
Thematic Evolution Across Study Periods
To examine temporal thematic evolution, the corpus was divided into four successive publication periods approximately balanced by cumulative publication volume: 1959–1998 (n = 535, 25.2%), 1999–2013 (n = 526, 24.8%), 2014–2020 (n = 520, 24.5%), and 2021–2025 (n = 540, 25.5%). Volume-based rather than calendar-based partitioning was used to reduce distortion from unequal document counts across calendar intervals, given the sparse publication output of earlier decades and the marked increase in recent years.
Thematic clusters were taken directly from the author-keyword co-occurrence map generated in VOSviewer, as described in the preceding section. Temporal analyses were restricted to publications containing at least one normalized author keyword retained in the final 112-keyword network. Within this keyword-mappable subset, a publication was classified as cluster-present if it contained at least one keyword belonging to that cluster and cluster-absent otherwise. Because a single publication could contain keywords from more than one cluster, cluster membership was not mutually exclusive, and period-specific proportions therefore do not sum to 100%.
For each cluster, we report the period-specific prevalence with Wilson 95% confidence intervals, the absolute percentage-point change between consecutive periods, and the net absolute percentage-point change between the earliest period, 1959–1998, and the most recent period, 2021–2025. Cramér’s V was calculated from the Pearson statistic of the corresponding 2×4 contingency table (cluster-present versus cluster-absent by period) as a standardized descriptive measure of association between publication period and cluster prevalence.
As supplementary descriptive statistics, Pearson’s chi-square tests of homogeneity were used to describe whether cluster-specific proportions differed across the four periods. A test of homogeneity rather than of ordered trend was used because several clusters followed nonmonotonic patterns, for which a linear trend statistic would offset opposing changes occurring in different periods. Where any expected cell count fell below five, the P-value was obtained by Monte Carlo resampling with 10,000 replicates under a fixed random seed. Holm correction was applied across the eight cluster-specific tests. Because the dataset was defined by the database, search strategy, and eligibility criteria rather than probability sampling, P values were used to summarize between-period heterogeneity within the analyzed corpus and were interpreted together with effect sizes and confidence intervals.
Sensitivity Analysis Using Title Terms
Because author keywords were available for only 55.1% of the publications, a supplementary title-term co-occurrence analysis of the full corpus was conducted to assess whether incomplete author-keyword coverage materially affected the high-level thematic interpretation. Terms were extracted from publication titles using VOSviewer’s text-mining procedure, which applies natural language processing to identify and unify noun phrases. The thesaurus developed for the author-keyword analysis was adapted for title-term analysis by adding title-specific normalization rules and stopwords; the complete file is provided in Supplementary File S3.
After term cleaning, the network was constructed using full counting, a minimum occurrence threshold of 6, and a clustering resolution of 0.9. Because these terms were algorithmically extracted from titles rather than supplied by authors, this analysis was treated as a supportive sensitivity analysis to determine whether the broad thematic structure identified from author keywords remained evident across the full corpus.
Bibliometric Analysis and Visualization Software
Data management, descriptive bibliometric summaries, and period-specific thematic analyses were conducted using R version 4.4.0 (R Foundation for Statistical Computing, Vienna, Austria) and the bibliometrix package version 5.0.25 Network construction and visualization for co-occurrence analyses were performed using VOSviewer version 1.6.20.26
During corpus construction, ChatGPT GPT-5.6 Sol (OpenAI, San Francisco, CA, USA) was used to conduct a prespecified first-pass relevance screen based on article titles, abstracts, and author keywords. The model assigned provisional Include, Exclude, or Uncertain classifications and generated criterion-linked explanations. ChatGPT GPT-5.5 was also used during manuscript preparation to assist with language editing, organization of selected text, and troubleshooting of selected R code.
Results
Overview
The initial search yielded 3,471 records. AI-assisted first-pass screening classified 1,922 (55.4%) as Include, 965 (27.8%) as Exclude, and 584 (16.8%) as Uncertain. Following the initial human review, 1,866 (97.1%) of the AI-Include records were retained, whereas 56 (2.9%) were excluded. Among the AI-Exclude records, 884 (91.6%) were confirmed as excluded and 81 (8.4%) were reclassified as included. Of the AI-Uncertain records, 180 (30.8%) were included and 404 (69.2%) were excluded. The initial human review therefore identified 2,127 eligible publications (61.3%) and excluded 1,344 records (38.7%).
The subsequent AI-assisted consistency audit flagged 59 records for further human assessment: 27 as possible erroneous exclusions, nine as possible erroneous inclusions, and 23 as requiring full-text review. Following human adjudication, six records were reclassified from Exclude to Include and 12 from Include to Exclude; the remaining 41 flagged records retained their existing classifications. In the final decisions, 1,869 (97.2%) of the AI-Include records, 80 (8.3%) of the AI-Exclude records, and 172 (29.5%) of the AI-Uncertain records were included. The final corpus comprised 2,121 publications (61.1% of all retrieved records), while 1,350 records (38.9%) were excluded (Table 1).
Table 1.
Transition from Artificial Intelligence-Assisted First-Pass Classifications to Initial and Final Human-Review Decisions for Publications on Bullous Pemphigoid, 1959–2025
| AI First-Pass Classification | AI First-Pass Total, n (% of All Records) | Decision After Initial Human Review (% Within AI Category) | Final Decision After AI-Assisted Consistency Audit and Human Adjudication (% Within AI Category) | ||
|---|---|---|---|---|---|
| Excluded | Included | Excluded | Included | ||
| Include | 1,922 (55.4) | 56 (2.9) | 1,866 (97.1) | 53 (2.8) | 1,869 (97.2) |
| Exclude | 965 (27.8) | 884 (91.6) | 81 (8.4) | 885 (91.7) | 80 (8.3) |
| Uncertain | 584 (16.8) | 404 (69.2) | 180 (30.8) | 412 (70.5) | 172 (29.5) |
| Total | 3,471 (100.0) | 1,344 (38.7) | 2,127 (61.3) | 1,350 (38.9) | 2,121 (61.1) |
Of the 2,121 publications, 1,852 were articles and 269 were reviews. The publications appeared in 379 journals. A total of 7,847 distinct authors contributed to the corpus, with a mean of 5.8 co-authors per publication; 54 publications were single-authored. International collaboration was observed in 15.5% of publications.
Annual Publication Output and Citation Rate
Publication output increased over the study period (Figure 2A). Annual output was sparse during the first two decades after the first indexed record in 1959, remaining below 10 publications in every year through 1975. Output began to rise in the late 1970s and early 1980s, ranging from 13 to 26 publications annually between 1978 and 1985, with 20 publications in 1980. The annual count increased to 33 in 1989. From 1990 to 2008, output fluctuated between 17 and 39 publications per year, before increasing to 45 in 2009, 50 in 2011, and 55 in 2014. Growth accelerated after 2015, with annual output rising from 71 publications in 2016 to 97 in 2018. Following 81 publications in 2019 and 80 in 2020, output increased to 97 in 2021 and first exceeded 100 in 2022, with 112 publications. The annual count reached a study-period maximum of 128 in 2023, declined to 81 in 2024, and increased again to 122 in 2025.
Figure 2.
Annual publication output and annualized citation impact of bullous pemphigoid publications, 1959–2025. Panel (A) shows the number of publications by year. Panel (B) shows the mean citations per publication per year, calculated for each publication as the Web of Science Core Collection citation count divided by (2025 − publication year + 1). The shaded area represents the bootstrap 95% confidence interval for years with at least five publications. Years with fewer than five publications are shown without confidence intervals; the 1968 estimate was based on one publication and should be interpreted as an unstable small-sample value.
Mean citations per publication per year varied considerably across publication years (Figure 2B). The highest value was observed in 1968, at 6.16 citations per publication per year, although this estimate was based on a single publication. From 1969 through 2006, mean annualized citation rates generally remained between approximately 0.5 and 2.4 citations per publication per year. Rates subsequently increased, reaching 2.96 in 2008, 3.00 in 2011, 3.21 in 2012, and 4.28 in 2013. Among publication years with at least five articles, the highest mean annualized citation rate occurred in 2019, at 4.68 citations per publication per year. The corresponding rates were 3.85 in 2020, 3.14 in 2021, 3.74 in 2022, 3.07 in 2023, 3.79 in 2024, and 2.74 in 2025. The isolated peak in 1968 should be interpreted cautiously because it was derived from only one article. Although annualization adjusts for differences in publication age, estimates for recent publication years remain sensitive to short citation windows and citation lag. Annualized citation rates were calculated at the publication level and summarized by publication year; bootstrap 95% confidence intervals were omitted for years with fewer than five publications.
Countries and Regions Contributing to Bullous Pemphigoid Research
The top 10 contributing countries and regions, based on fractional attribution of corresponding-author affiliations, are shown in Table 2. The United States ranked first, with a fractional publication count of 434.3, representing 20.5% of the corpus, and also had the highest fractional total citation count at 19,491.5. Japan ranked second with 246.0 fractional publications (11.6%), followed by Germany with 204.0 (9.6%) and China with 185.5 (8.8%). Italy and the United Kingdom each accounted for 6.3% of the corpus, with fractional publication counts of 134.5 and 133.5, respectively, while France contributed 126.5 fractional publications (6.0%).
Table 2.
Top 10 Most Productive Countries and Regions Based on the Corresponding Author’s Affiliation in Articles and Reviews on Bullous Pemphigoid, 1959–2025
| Rank | Country | Fractional Publication Count (%) | MCP % | Fractional Total Citations |
|---|---|---|---|---|
| 1 | United States | 434.3 (20.5) | 12.7 | 19,491.5 |
| 2 | Japan | 246.0 (11.6) | 6.5 | 6,221.0 |
| 3 | Germany | 204.0 (9.6) | 43.6 | 9,029.0 |
| 4 | China | 185.5 (8.8) | 10.5 | 2,178.0 |
| 5 | Italy | 134.5 (6.3) | 16.0 | 3,979.0 |
| 6 | United Kingdom | 133.5 (6.3) | 11.6 | 4,707.0 |
| 7 | France | 126.5 (6.0) | 13.0 | 5,766.0 |
| 8 | Taiwan | 55.0 (2.6) | 18.2 | 946.0 |
| 9 | Israel | 50.0 (2.4) | 18.0 | 1,555.5 |
| 10 | Poland | 47.0 (2.2) | 19.2 | 887.0 |
Notes: MCP denotes multiple-country publications. The corresponding author’s country or region was identified from the Web of Science reprint author (RP) field. Country/region-level article production was calculated using fractional counting, so that when more than one corresponding-author country or region was recorded, each received a proportional share of the publication. Fractional total citations were calculated by multiplying each article’s citation count by the corresponding-author country/region weight and then summing these weighted citations across articles.
The proportion of multiple-country publications varied across the leading contributors. Germany had the highest multiple-country publication (MCP) percentage at 43.6%, substantially exceeding those of the other top-ranked countries and regions. Poland, Taiwan, and Israel had MCP percentages of 19.2%, 18.2%, and 18.0%, respectively, followed by Italy at 16.0%. Lower MCP percentages were observed for France (13.0%), the United States (12.7%), the United Kingdom (11.6%), China (10.5%), and Japan (6.5%).
Although Japan had a higher fractional publication count than Germany, Germany accumulated a substantially greater fractional citation total, at 9,029 compared with 6,221. France also accumulated more fractional citations than Italy despite having a slightly lower fractional publication count, with 5,766 compared with 3,979. The United Kingdom recorded 4,707 fractional citations from 133.5 fractional publications. These citation values were not adjusted for publication year and should therefore be interpreted as measures of cumulative fractional citation volume rather than age-normalized citation impact.
Leading Institutions
The 10 most productive institutions contributing to bullous pemphigoid research are shown in Table 3. The University of Lübeck ranked first, contributing 135 articles (6.4% of the corpus), followed by Schleswig Holstein University Hospital with 96 articles (4.5%). Hokkaido University ranked third with 60 articles (2.8%), followed by the Institut National de la Santé et de la Recherche Médicale (INSERM) with 58 articles (2.7%) and Kurume University with 57 articles (2.7%). The US Department of Veterans Affairs and the Veterans Health Administration contributed 56 (2.6%) and 53 (2.5%) articles, respectively. Assistance Publique–Hôpitaux de Paris ranked eighth with 49 articles (2.3%), followed by the Chinese Academy of Medical Sciences–Peking Union Medical College with 48 articles (2.3%) and the University of Würzburg with 47 articles (2.2%).
Table 3.
Top 10 Productive Institutions Contributing to Articles and Reviews on Bullous Pemphigoid Research, 1959–2025
| Rank | Institution, Country | Publication Count (%) |
|---|---|---|
| 1 | University of Lübeck, Germany | 135 (6.4) |
| 2 | Schleswig Holstein University Hospital, Germany | 96 (4.5) |
| 3 | Hokkaido University, Japan | 60 (2.8) |
| 4 | Institut National de la Santé et de la Recherche Médicale (INSERM), France | 58 (2.7) |
| 5 | Kurume University, Japan | 57 (2.7) |
| 6 | US Department of Veterans Affairs, United States | 56 (2.6) |
| 7 | Veterans Health Administration, United States | 53 (2.5) |
| 8 | Assistance Publique–Hôpitaux de Paris, France | 49 (2.3) |
| 9 | Chinese Academy of Medical Sciences–Peking Union Medical College, China | 48 (2.3) |
| 10 | University of Würzburg, Germany | 47 (2.2) |
Notes: Counts represent the number of articles in which an institution appeared at least once in the affiliation field. Multiple authors from the same institution within a single publication contributed one publication count.
Germany had three institutions among the top 10: the University of Lübeck, Schleswig Holstein University Hospital, and the University of Würzburg. Japan was represented by Hokkaido University and Kurume University, France by INSERM and Assistance Publique–Hôpitaux de Paris, and the United States by the US Department of Veterans Affairs and the Veterans Health Administration. China was represented by the Chinese Academy of Medical Sciences–Peking Union Medical College.
Overall, institutional output was concentrated among a relatively small group of universities, hospital systems, and national research organizations in Europe, Asia, and the United States. Counts represented the number of articles in which an institution appeared at least once in the affiliation field; repeated occurrences of the same institution within a single article were counted once. Because parent organizations, hospital systems, campuses, and affiliated units were not manually consolidated, the rankings should be interpreted as indicators of publication concentration rather than direct comparisons of institutional research performance. This caveat is especially relevant to potentially overlapping organizational entries, such as the US Department of Veterans Affairs and the Veterans Health Administration.
Leading Journals
The journals occupying the top 10 ranks for publication output in bullous pemphigoid research are shown in Table 4. The British Journal of Dermatology ranked first with 158 publications (7.4% of the corpus), followed by the Journal of Investigative Dermatology with 121 (5.7%) and the Journal of the American Academy of Dermatology with 106 (5.0%). Together, these three journals published 385 records, accounting for approximately 18.2% of the corpus. Frontiers in Immunology ranked fourth with 98 publications (4.6%), followed by Archives of Dermatology with 86 (4.0%) and the Journal of Dermatology with 75 (3.5%). Because Clinical and Experimental Dermatology and the International Journal of Dermatology were tied at rank 7 with 70 publications each, Table 4 includes 11 journals across the top 10 ranks.
Table 4.
Top 10 Journals Publishing Articles and Reviews on Bullous Pemphigoid Research, 1959–2025
| Rank | Journal | Publication Count (%) | 2025 JIF (SCIE) | JCR Category [JIF Quartile] |
|---|---|---|---|---|
| 1 | British Journal of Dermatology | 158 (7.4) | 8.2 | Dermatology [Q1] |
| 2 | Journal of Investigative Dermatology | 121 (5.7) | 7.0 | Dermatology [Q1] |
| 3 | Journal of the American Academy of Dermatology | 106 (5.0) | 12.3 | Dermatology [Q1] |
| 4 | Frontiers in Immunology | 98 (4.6) | 7.0 | Immunology [Q1] |
| 5 | Archives of Dermatology | 86 (4.0) | 4.8 (2014, latest available) | Dermatology [Q1] (2014, latest available) |
| 6 | Journal of Dermatology | 75 (3.5) | 2.3 | Dermatology [Q2] |
| 7 | Clinical and Experimental Dermatology | 70 (3.3) | 2.8 | Dermatology [Q2] |
| 7 | International Journal of Dermatology | 70 (3.3) | 4.2 | Dermatology [Q1] |
| 8 | Archives of Dermatological Research | 58 (2.7) | 2.5 | Dermatology [Q2] |
| 9 | Acta Dermato-Venereologica | 52 (2.4) | 3.8 | Dermatology [Q1] |
| 10 | Journal of the European Academy of Dermatology and Venereology | 49 (2.3) | 9.2 | Dermatology [Q1] |
Notes: Q1 and Q2 indicate first and second JIF quartiles, respectively. Journal categories, quartiles, and Journal Impact Factors were based on the 2026 Web of Science JCR. Archives of Dermatology was renamed JAMA Dermatology in January 2013. Journals with tied article counts were assigned the same rank.
Abbreviations: JCR, Journal Citation Reports; JIF, Journal Impact Factor; SCIE, Science Citation Index Expanded.
Publication activity was concentrated primarily in dermatology journals. Ten of the 11 listed journals were classified in the Dermatology category, while Frontiers in Immunology was the sole immunology-focused journal. Eight journals were ranked in the first Journal Impact Factor quartile, and three were ranked in the second quartile.
The reported Journal Impact Factors ranged from 2.3 for the Journal of Dermatology to 12.3 for the Journal of the American Academy of Dermatology. Archives of Dermatology, which contributed 86 publications (4.0%), was renamed JAMA Dermatology in January 2013. Its Journal Impact Factor of 4.8 and first-quartile (Q1) classification in Table 4 refer to 2014, the latest year for which values under the former journal title were available.
Most Cited Publications in Bullous Pemphigoid Research
The 10 most cited publications in the corpus were published between 1967 and 2013 and traced the development of bullous pemphigoid research from early immunopathological observations and antigen characterization to experimental disease models, treatment evaluation, epidemiological investigation, and comprehensive disease synthesis (Table 5). The most cited publication was the 2013 Seminar by Schmidt and Zillikens27 with 852 citations. Although the article reviewed the broader spectrum of pemphigoid diseases, it devoted its largest and most detailed section to bullous pemphigoid, addressing its epidemiology, clinical manifestations, associated diseases, target antigens, pathophysiology, diagnosis, and treatment. This was followed by the passive-transfer model developed by Liu et al28 in 1993, with 519 citations, and the randomized comparison of oral and topical corticosteroids by Joly et al29 in 2002, with 480 citations. The population-based study of incidence and mortality by Langan et al30 ranked fourth with 474 citations, while the immunoblotting study of bullous pemphigoid antigen heterogeneity by Labib et al31 ranked fifth with 452 citations.
Table 5.
Top 10 Most Cited Publications in the Bullous Pemphigoid Corpus, 1959–2025
| Rank | First Author, Year | Publication Title (DOI) | Total Citations | Annualized Citations (Citations/Year) | Normalized Total Citations |
|---|---|---|---|---|---|
| 1 | Schmidt E, 201327 | Pemphigoid diseases (10.1016/S0140-6736(12)61140-4) | 852 | 65.5 | 15.3 |
| 2 | Liu Z, 199328 | A passive transfer model of the organ-specific autoimmune-disease, bullous pemphigoid, using antibodies generated against the hemidesmosomal antigen, BP180 (10.1172/JCI116856) | 519 | 15.7 | 7.2 |
| 3 | Joly P, 200229 | A comparison of oral and topical corticosteroids in patients with bullous pemphigoid (10.1056/NEJMoa011592) | 480 | 20.0 | 8.2 |
| 4 | Langan SM, 200830 | Bullous pemphigoid and pemphigus vulgaris--incidence and mortality in the UK: Population based cohort study (10.1136/bmj.a180) | 474 | 26.3 | 8.9 |
| 5 | Labib RS, 198631 | Molecular heterogeneity of the bullous pemphigoid antigens as detected by immunoblotting (10.4049/jimmunol.136.4.1231) | 452 | 11.3 | 8.6 |
| 6 | Jordon RE, 196732 | Basement zone antibodies in bullous pemphigoid (10.1001/jama.1967.03120220053008) | 432 | 7.3 | 4.1 |
| 7 | Beutner EH, 196833 | Immunopathology of pemphigus and bullous pemphigoid (10.1038/jid.1968.94) | 357 | 6.2 | 1.0 |
| 8 | Bernard P, 199534 | Incidence and distribution of subepidermal autoimmune bullous skin diseases in three French regions. Bullous Diseases French Study (10.1001/archderm.1995.01690130050009) | 346 | 11.2 | 5.5 |
| 9 | Giudice GJ, 199335 | Bullous pemphigoid and herpes gestationis autoantibodies recognize a common non-collagenous site on the BP180 ectodomain (10.4049/jimmunol.151.10.5742) | 339 | 10.3 | 4.7 |
| 10 | Stanley JR, 198836 | Isolation of complementary DNA for bullous pemphigoid antigen by use of patients’ autoantibodies (10.1172/JCI113803) | 330 | 8.7 | 6.7 |
Notes: Annualized citations were calculated as total Web of Science citations/(2025 − publication year + 1); normalized total citations were calculated as publication citations/mean citations of corpus publications from the same publication year.
Abbreviation: DOI, digital object identifier.
Five of the 10 most cited publications focused primarily on immunopathology or bullous pemphigoid antigen characterization. The early studies by Jordon et al32 and Beutner et al33 established the presence and diagnostic relevance of basement membrane zone antibodies and characteristic immunopathological findings. Labib et al31 subsequently demonstrated the molecular heterogeneity of bullous pemphigoid antigens, while Giudice et al35 identified a common non-collagenous BP180 ectodomain site recognized by autoantibodies in bullous pemphigoid and herpes gestationis. Stanley et al36 extended this antigen-focused work by isolating complementary DNA encoding a bullous pemphigoid antigen using patient autoantibodies. Liu et al28 provided experimental evidence that antibodies directed against BP180 could induce disease-related tissue injury. The remaining highly cited publications included Schmidt and Zillikens’27 comprehensive review of pemphigoid diseases, which focused primarily on bullous pemphigoid, the corticosteroid treatment trial by Joly et al,29 and the epidemiological studies by Langan et al30 and Bernard et al.34
Citation metrics adjusted for publication year produced a somewhat different pattern from cumulative citation totals. The 2013 review by Schmidt and Zillikens27 had the highest annualized citation rate at 65.5 citations per year, followed by the population-based study by Langan et al30 at 26.3, the corticosteroid trial by Joly et al29 at 20.0, and the passive-transfer study by Liu et al28 at 15.7. Schmidt and Zillikens27 also had the highest normalized total citation score at 15.3. The next highest normalized scores were observed for Langan et al30 at 8.9, Labib et al31 at 8.6, and Joly et al29 at 8.2. These findings indicate that influential bullous pemphigoid research comprises foundational immunological and antigen-focused studies as well as later clinical, epidemiological, and integrative publications that achieved substantial citation impact relative to other corpus publications from the same year.
Author-Keyword Availability by Publication Period
Before interpreting the author-keyword co-occurrence and temporal cluster analyses, we examined author-keyword availability across publication periods (Supplementary Table S2). Author keywords were available for 1,169 of 2,121 records (55.1%) overall, but availability varied substantially over time. Only 77 of 535 records (14.4%) published in 1959–1998 contained author keywords, compared with 295 of 526 (56.1%) in 1999–2013, 352 of 520 (67.7%) in 2014–2020, and 445 of 540 (82.4%) in 2021–2025. Overall, 924 records (43.6%) contained at least one author keyword retained in the final 112-keyword network and were therefore considered keyword-mappable. The corresponding numbers were 52 records of 535 (9.7%), 220 of 526 (41.8%), 282 of 520 (54.2%), and 370 of 540 (68.5%) across the four successive periods. Temporal cluster proportions were calculated using these keyword-mappable publications as the denominator within each period.
Author-Keyword Co-Occurrence Network
The sensitivity analysis evaluated minimum author-keyword occurrence thresholds of 4, 5, 6, and 7 across clustering resolutions ranging from 0.6 to 1.0 (Supplementary Table S3). The four occurrence thresholds retained 167, 130, 112, and 95 keywords and produced networks containing 1,189, 988, 889, and 763 links, respectively. Among the tested thresholds, a minimum occurrence threshold of six produced the highest overall mean ARI and NMI across resolutions. At this threshold, resolution 0.9 produced eight clusters, with the smallest cluster containing four keywords, and had a mean ARI of 0.798 and mean NMI of 0.895. These values were identical to those obtained at resolution 1.0, and direct comparison confirmed that the two partitions were equivalent (ARI = 1.000; NMI = 1.000). Resolution 0.9 was therefore selected as the lowest value at which the clustering solution reached a stable plateau.
The selected network comprised 112 author keywords connected by 889 co-occurrence links and was divided into eight thematic clusters containing 23, 20, 18, 18, 13, 9, 7, and 4 keywords, respectively (Figure 3). The clusters represented (1) therapeutic management and clinical outcomes; (2) diagnosis and related autoimmune blistering diseases; (3) epidemiology and comorbidities; (4) inflammatory and immune effector mechanisms; (5) autoantigens, autoantibodies, and basement membrane zone biology; (6) immune checkpoint inhibitor–associated bullous pemphigoid; (7) COVID-19, vaccination, and emerging research fronts; and (8) dipeptidyl peptidase-4 inhibitor–associated bullous pemphigoid.
Figure 3.
Author-keyword co-occurrence network of bullous pemphigoid research. The network comprised 112 normalized author keywords occurring in at least six publications and was generated in VOSviewer using full counting, association-strength normalization, and a clustering resolution of 0.9. Node size represents keyword occurrence frequency, link thickness represents co-occurrence strength, and shorter distances indicate stronger relatedness. Colors denote eight thematic clusters: (1) therapeutic management and clinical outcomes; (2) diagnosis and related autoimmune blistering diseases; (3) epidemiology and comorbidities; (4) inflammatory and immune effector mechanisms; (5) autoantigens, autoantibodies, and basement membrane zone biology; (6) immune checkpoint inhibitor–associated bullous pemphigoid; (7) COVID-19, vaccination, and emerging research fronts; and (8) dipeptidyl peptidase-4 inhibitor–associated bullous pemphigoid.
Cluster 1 (red), therapeutic management and clinical outcomes, was the largest cluster, containing 23 keywords. Its principal treatment-related terms included dupilumab, omalizumab, rituximab, corticosteroid, biologics, methotrexate, intravenous immunoglobulin, and topical corticosteroid. The presence of mortality, prognosis, relapse, disease severity, quality of life, and the Bullous Pemphigoid Disease Area Index indicated that this cluster also encompassed treatment evaluation and patient outcomes. Psoriasis, drug-induced bullous pemphigoid, risk factors, cohort study, case-control study, meta-analysis, and systematic review reflected the broader clinical and evidence-based context of this theme.
Cluster 2 (green), diagnosis and related autoimmune blistering diseases, contained 20 keywords. It combined diagnostic methods, including enzyme-linked immunosorbent assay, immunofluorescence, direct immunofluorescence, indirect immunofluorescence, and immunoblotting, with related or differential diagnoses such as pemphigus, pemphigus vulgaris, epidermolysis bullosa acquisita, mucous membrane pemphigoid, linear IgA bullous dermatosis, pemphigoid gestationis, and dermatitis herpetiformis. BP230 and anti-BP180 antibody were also included, reflecting the role of serological and tissue-based markers in distinguishing autoimmune blistering diseases.
Cluster 3 (blue), epidemiology and comorbidities, comprised 18 keywords. Its most prominent terms included comorbidity, epidemiology, pruritus, neurological disorder, dementia, stroke, incidence, Parkinson disease, multiple sclerosis, malignancy, and cancer. Additional terms such as elderly, eosinophilia, atopic dermatitis, blister, autoimmune, and Mendelian randomization indicated interest in associated conditions, susceptible populations, and possible epidemiologic or causal relationships.
Cluster 4 (yellow), inflammatory and immune effector mechanisms, also contained 18 keywords. Eosinophil, cytokines, immunoglobulin E, inflammation, complement, immunoglobulin G, blister fluid, T-helper 2, neutrophils, T cells, mast cells, basophils, eotaxin, interleukin-4, and interleukin-5 characterized cellular and humoral mechanisms involved in tissue injury and blister formation. Keratinocyte, C4d, and immunohistochemistry further reflected investigation of local inflammatory responses and immune deposition.
Cluster 5 (purple), autoantigens, autoantibodies, and basement membrane zone biology, contained 13 keywords and occupied a central position in the network. BP180 was the most frequent keyword in the entire map, accompanied by autoimmunity, autoantibody, hemidesmosome, autoantigen, basement membrane, basement membrane zone, collagen, epitope, antibody, immunoglobulin G subclass, dermal-epidermal junction, and mouse model. This cluster represented the structural and immunological foundations of bullous pemphigoid pathogenesis.
Cluster 6 (cyan), immune checkpoint inhibitor–associated bullous pemphigoid, comprised nine keywords: immune checkpoint inhibitors, immunotherapy, pembrolizumab, nivolumab, melanoma, immune-related adverse events, pharmacovigilance, immunosuppression, and adverse effects. Its composition identified a distinct research theme concerning bullous pemphigoid arising during anticancer immunotherapy and its clinical management.
Cluster 7 (orange), coronavirus disease 2019 (COVID-19), vaccination, and emerging research fronts, contained seven keywords. COVID-19, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), and vaccine formed the principal subgroup, while pathogenesis, BP180-NC16A, Janus kinase inhibitors, and infant represented smaller or more recent lines of investigation. The mixed composition of this cluster suggests that it captured newer topics that had not yet developed into larger, fully separated thematic domains.
Cluster 8 (brown), dipeptidyl peptidase-4 inhibitor–associated bullous pemphigoid, was the smallest and most narrowly defined cluster. It consisted of dipeptidyl peptidase-4 (DPP-4) inhibitors, diabetes mellitus, vildagliptin, and DPP-4. Its separation from the broader therapeutic and epidemiologic clusters indicated that drug-associated bullous pemphigoid related to this medication class had become a distinct research theme.
Temporal Evolution of Author-Keyword Themes
Among the 924 keyword-mappable publications, 52 were published in 1959–1998, 220 in 1999–2013, 282 in 2014–2020, and 370 in 2021–2025. The prevalence of most thematic clusters changed significantly across publication periods (Figure 4). Because cluster membership was nonexclusive, publications could contribute to more than one thematic cluster.
Figure 4.
Temporal evolution of eight author-keyword thematic clusters in bullous pemphigoid research, 1959–2025. Points represent the proportion of keyword-mappable publications containing at least one author keyword assigned to each thematic cluster within four publication periods: 1959–1998, 1999–2013, 2014–2020, and 2021–2025. Because cluster membership was nonexclusive, a publication could contribute to more than one cluster. Shaded bands and error bars represent Wilson 95% confidence intervals, and numbers above the points indicate the corresponding prevalence percentages. “pp” denotes the absolute percentage-point change between 1959–1998 and 2021–2025, with upward and downward triangles indicating increases and decreases, respectively. Cramér’s V summarizes the association between publication period and membership in each cluster. P values were derived from chi-square tests and adjusted for multiple comparisons using the Holm method.
The largest temporal change occurred in autoantigens, autoantibodies, and basement membrane zone biology, which declined consistently from 51.9% in 1959–1998 to 46.4% in 1999–2013, 29.1% in 2014–2020, and 14.6% in 2021–2025. This represented a net decrease of 37.3 percentage points and the strongest association with publication period among the eight clusters (Cramér’s V = 0.30; Holm-adjusted P < 0.001).
In contrast, therapeutic management and clinical outcomes increased steadily from 7.7% in the earliest period to 12.7%, 25.9%, and 37.8% in successive periods, corresponding to a net increase of 30.1 percentage points (V = 0.24; Holm-adjusted P < 0.001). Epidemiology and comorbidities also expanded substantially, rising from 1.9% to 12.3% and 20.9% before remaining nearly unchanged at 20.8% in 2021–2025. The net increase was 18.9 percentage points (V = 0.14; Holm-adjusted P = 0.002).
Diagnosis and related autoimmune blistering diseases followed a nonmonotonic pattern. Its prevalence increased from 30.8% in 1959–1998 to a peak of 50.0% in 1999–2013, then declined to 33.0% in 2014–2020 and 24.3% in 2021–2025. Despite a modest net decrease of 6.4 percentage points between the first and last periods, the overall association with publication period was significant (V = 0.21; Holm-adjusted P < 0.001).
Three more recent themes increased progressively. Immune checkpoint inhibitor–associated bullous pemphigoid was absent in 1959–1998 but increased from 2.7% in 1999–2013 to 6.4% in 2014–2020 and 10.8% in 2021–2025, yielding a net increase of 10.8 percentage points (V = 0.14; Holm-adjusted P = 0.002). COVID-19, vaccination, and emerging research fronts increased from 1.9% to 2.3%, 4.3%, and 8.1%, corresponding to a net increase of 6.2 percentage points (V = 0.11; Holm-adjusted P = 0.018). Dipeptidyl peptidase-4 inhibitor–associated bullous pemphigoid was absent in the earliest period and remained uncommon in 1999–2013 (0.9%) but increased to 8.5% in 2014–2020 and 12.7% in 2021–2025. Its net increase of 12.7 percentage points was significantly associated with publication period (V = 0.18; Holm-adjusted P < 0.001).
The prevalence of inflammatory and immune effector mechanisms decreased from 30.8% in 1959–1998 to 19.1% in 1999–2013, increased slightly to 22.3% in 2014–2020, and then declined to 18.4% in 2021–2025. Although this corresponded to a net decrease of 12.4 percentage points, the variation across periods was not statistically significant after adjustment for multiple comparisons (V = 0.08; Holm-adjusted P = 0.154).
Sensitivity Analysis Using Title Terms
The sensitivity analysis retained 80 normalized title terms occurring in at least six publications, yielding a network of 210 co-occurrence links and nine clusters at a resolution of 0.9. Frequently occurring terms included dipeptidyl peptidase, inhibitor, risk, dermatitis, mortality, management, BP180, eosinophil, and dupilumab. The network recovered the principal domains identified in the author-keyword analysis, including structural and serologic research, inflammatory and immune effector mechanisms, therapeutic management, DPP-4 inhibitor–associated bullous pemphigoid, mortality and prognosis, immune checkpoint inhibitor–related research, and COVID-19 and vaccination (Supplementary Figure S2).
Although the title-term network produced nine clusters rather than the eight identified from author keywords, the major substantive domains were broadly concordant. Some treatment, diagnostic, pediatric, and immune checkpoint inhibitor–related terms were distributed across multiple clusters, indicating that the two analyses did not yield a one-to-one cluster correspondence. Nevertheless, the title-term analysis supported the overall thematic structure identified in the primary author-keyword analysis.
Discussion
Growth, Global Contributions, and Publication Structure
This bibliometric analysis of 2,121 articles and reviews provides a field-level account of bullous pemphigoid-focused research published from 1959 through 2025. Annual output accelerated after 2015, exceeded 100 publications in 2022, 2023, and 2025, and reached a study-period maximum of 128 in 2023. The decline in 2024 followed by renewed output in 2025 cautions against interpreting a single-year change as a sustained reversal. Annualized citation rates were higher for several publication years in the late 2010s but fluctuated in the most recent years. Publication volume and citation visibility measure different dimensions of research activity, and recent-year estimates remain sensitive to short citation windows and citation lag despite adjustment for publication age.
Research contributions were unevenly distributed across countries and institutions. The United States accounted for the largest fractionally attributed corresponding-author output, followed by Japan and Germany, whereas Germany had the highest proportion of multiple-country publications among the leading contributors. The institutional ranking showed a different pattern: the University of Lübeck and Schleswig Holstein University Hospital occupied the first two positions, followed by Hokkaido University, INSERM, and Kurume University. The US Department of Veterans Affairs and the Veterans Health Administration also appeared among the top 10 institutions. These findings suggest concentration in several specialized European and Asian centers alongside a more distributed contribution from the United States. This interpretation requires caution because countries were fractionally credited through corresponding-author affiliations, whereas each institution received one full publication count whenever it appeared in an article; the two rankings are complementary and are not directly comparable measures of performance.
The journal distribution remained strongly centered in dermatology. Because two journals were tied at rank 7, the top 10 ranks comprised 11 journals; 10 were classified in Dermatology, and Frontiers in Immunology was the sole immunology-focused outlet. The British Journal of Dermatology, Journal of Investigative Dermatology, and Journal of the American Academy of Dermatology together published 18.2% of the corpus, indicating substantial concentration in major specialist journals while leaving most publications distributed across a wide range of sources. The position of Frontiers in Immunology also reflects a visible immunology-oriented publication stream. The present analysis did not examine changes in journal access models over time and therefore cannot determine whether open-access publishing independently influenced the field’s growth.
The citation profile includes both foundational mechanistic studies and later clinical research. Foundational publications on basement membrane zone antibodies, bullous pemphigoid antigens, hemidesmosomal proteins, and passive-transfer models remained highly cited, while later treatment trials, population-based studies, and integrative reviews also achieved substantial cumulative and publication-year-normalized citation impact. The continuing visibility of both groups indicates that mechanistic immunodermatology and clinically oriented investigation remain jointly important to the field’s knowledge structure.
Shifts in Thematic Prominence in Bullous Pemphigoid Research
Previous bibliometric work on highly cited pemphigoid publications identified molecular mechanisms, management, and risk factors as major subjects, while a separate study of pemphigus documented parallel structural, immunological, therapeutic, and outcome-related themes.22,23 The present full-corpus analysis extends these observations by identifying eight author-keyword domains and quantifying their prevalence across four publication-volume-balanced periods. The selected network comprised 112 keywords connected by 889 links, and the temporal analysis included 924 publications with at least one retained keyword.
The largest long-term change occurred in the cluster representing autoantigens, autoantibodies, and basement membrane zone biology. Its prevalence declined from 51.9% in 1959–1998 to 46.4% in 1999–2013, 29.1% in 2014–2020, and 14.6% in 2021–2025, a net decrease of 37.3 percentage points and the strongest association with publication period among the eight clusters (Cramér’s V = 0.30). This pattern indicates a declining relative share within a rapidly expanding literature; it does not indicate disappearance or declining scientific value. BP180 remained the most frequent keyword and occupied a central network position. Research on BP180, BP230, hemidesmosomes, epitopes, and the dermal-epidermal junction established the mechanistic framework on which later diagnostic and therapeutic studies depended.
Diagnosis and related autoimmune blistering diseases followed a nonmonotonic trajectory. The cluster increased from 30.8% in 1959–1998 to 50.0% in 1999–2013, then declined to 33.0% in 2014–2020 and 24.3% in 2021–2025 (Cramér’s V = 0.21). Because this cluster combined direct and indirect immunofluorescence, enzyme-linked immunosorbent assay, immunoblotting, BP230, anti-BP180 antibody, and related autoimmune blistering diseases, it should not be interpreted as a measure of serologic diagnosis alone. The peak in 1999–2013 is compatible with a phase in which antigen-specific assays and differential diagnostic frameworks became established. The later decline reflects reduced relative prominence after diagnostic consolidation and thematic diversification.
Therapeutic management and clinical outcomes showed the largest increase, rising from 7.7% in 1959–1998 to 12.7%, 25.9%, and 37.8% in successive periods, a net increase of 30.1 percentage points (Cramér’s V = 0.24). The cluster combined established treatments with dupilumab, omalizumab, rituximab, intravenous immunoglobulin, methotrexate, and corticosteroids, as well as mortality, prognosis, relapse, disease severity, quality of life, and the Bullous Pemphigoid Disease Area Index. This expansion is consistent with the field’s increasing emphasis on treatment effectiveness, safety, steroid-sparing strategies, and patient-centered outcomes. The BLISTER trial provided evidence for a safer initial doxycycline strategy in selected patients, and the 2025 approval of dupilumab marked a regulatory milestone for targeted therapy.10,11,37 A single-arm meta-analysis reported a pooled complete-response proportion of 68% for dupilumab, but the predominantly retrospective and uncontrolled evidence does not establish comparative effectiveness.38
Epidemiology and comorbidities also became a major domain, increasing from 1.9% in the earliest period to 12.3% in 1999–2013 and 20.9% in 2014–2020, then remaining essentially unchanged at 20.8% in 2021–2025. The net increase was 18.9 percentage points (Cramér’s V = 0.14). The near-identical prevalence in the final two periods suggests that epidemiology and comorbidity research had become an established component of the field rather than continuing to accelerate. The cluster’s neurological terms, including dementia, stroke, Parkinson disease, and multiple sclerosis, are consistent with an expanding clinical focus on multimorbidity in older patients. Systematic reviews and cohort studies have linked bullous pemphigoid with neurological disease and have identified dementia, stroke, heart disease, and diabetes as adverse prognostic factors for mortality.16,39,40
DPP-4 inhibitor-associated bullous pemphigoid formed a separate four-keyword cluster and increased from 0% in 1959–1998 to 0.9% in 1999–2013, 8.5% in 2014–2020, and 12.7% in 2021–2025 (Cramér’s V = 0.18). Its separation from the broader epidemiology and treatment clusters indicates that research centered on DPP-4 inhibitors, diabetes mellitus, vildagliptin, and DPP-4 had developed sufficient internal co-occurrence to form a distinct topic.
Immune checkpoint inhibitor-associated bullous pemphigoid also emerged as a separate domain, increasing from 0% in the earliest period to 2.7%, 6.4%, and 10.8% in successive periods (Cramér’s V = 0.14). The cluster included immune checkpoint inhibitors, pembrolizumab, nivolumab, melanoma, immune-related adverse events, pharmacovigilance, immunosuppression, and adverse effects. This growth is consistent with the expanding literature on the presentation, diagnosis, treatment, and oncologic implications of bullous pemphigoid arising during immunotherapy.19,41 Clinical guidance now recognizes bullous pemphigoid within the spectrum of blistering dermatological immune-related adverse events.42 Its network separation makes this treatment-associated entity visible as a coherent publication theme.
Inflammatory and immune effector mechanisms showed a fluctuating rather than steadily directional pattern. Prevalence decreased from 30.8% in 1959–1998 to 19.1% in 1999–2013, increased to 22.3% in 2014–2020, and declined to 18.4% in 2021–2025. The overall variation was not statistically significant after Holm correction (Cramér’s V = 0.08; adjusted P = 0.154). The continued presence of eosinophils, cytokines, immunoglobulin E, complement, T-helper 2 pathways, neutrophils, mast cells, basophils, interleukin-4, and interleukin-5 indicates sustained mechanistic investigation despite the absence of a clear monotonic temporal trend.
The cluster labelled COVID-19, vaccination, and emerging research fronts increased from 1.9% to 2.3%, 4.3%, and 8.1% across the four periods. This cluster was intentionally interpreted cautiously because it combined COVID-19, SARS-CoV-2, and vaccine with pathogenesis, BP180-NC16A, Janus kinase inhibitors, and infant. Its early-period prevalence arose from the non-COVID terms and must not be interpreted as evidence of pre-pandemic COVID-19 research. The cluster is better viewed as a small grouping of recent or incompletely separated research fronts than as a single biologically coherent domain.
Comparison with the pemphigus bibliometric literature suggests both shared and disease-specific patterns.23 Both fields retain strong foundations in autoantigen and autoantibody research while devoting increasing attention to treatment, remission or relapse, quality of life, epidemiology, and mortality. Bullous pemphigoid research showed additional emphasis on hemidesmosomal biology, neurological comorbidity, DPP-4 inhibitor-associated disease, immune checkpoint inhibitor-associated disease, and emerging targeted therapies. The comparison remains descriptive because the studies differed in corpus construction, relevance screening, keyword harmonization, and network settings.
These temporal findings require cautious interpretation because author-keyword availability increased from 14.4% in 1959‒1998 to 82.4% in 2021‒2025, and the proportion of all records that were keyword-mappable increased from 9.7% to 68.5%. Zero or low prevalence in early periods cannot establish that a topic was historically absent. The full-corpus title-term network recovered the main structural, diagnostic, inflammatory, therapeutic, mortality, medication-associated, immune checkpoint inhibitor, and COVID-19-related domains, supporting the broad organization of the author-keyword map. Because the title-term analysis was not used to estimate period-specific prevalence, it does not independently validate the magnitude of the temporal trajectories.
Implications for Research
The principal practical value of this analysis lies in identifying where publication activity has concentrated and how the field’s conceptual balance has changed. Therapeutic management and clinical outcomes accounted for 37.8% of keyword-mappable publications in 2021–2025, while DPP-4 inhibitor-associated and immune checkpoint inhibitor-associated bullous pemphigoid accounted for 12.7% and 10.8%, respectively. These patterns support continued evaluation of comparative effectiveness, durability of response, long-term safety, steroid-sparing effects, treatment burden, and outcomes in older adults with multimorbidity. A Cochrane review of bullous pemphigoid interventions similarly points to the importance of balancing disease control against corticosteroid-related harm.43
The persistence of epidemiology and comorbidities at approximately one fifth of keyword-mappable publications in the two most recent periods suggests a sustained research attention to characterize prognosis in clinically heterogeneous populations. Future studies could examine whether relapse, mortality, and treatment tolerance differ by neurological disease, frailty, diabetes, DPP-4 inhibitor exposure, malignancy, or immune checkpoint inhibitor exposure. Medication-associated forms should be analyzed separately where possible because their timing, management constraints, and competing clinical risks may differ.
The continued presence of structural, diagnostic, and inflammatory domains also supports translational work linking molecular findings with clinically applicable biomarkers and treatment-response outcomes. The co-occurrence map cannot determine whether these areas are sufficiently integrated, so claims of a translational gap would require direct study of citation or article-level connections. The present findings are best used to guide more focused systematic reviews, cohort studies, and comparative trials rather than to infer effectiveness or causal mechanisms from publication patterns.
Strengths and Limitations
A principal strength of this study was the construction of a systematically screened, bullous pemphigoid-focused corpus from all records retrieved through the Web of Science search. Each record underwent AI-assisted first-pass classification using prespecified eligibility criteria and a complete prompt reproduced in Supplementary File S1, thereby allowing the screening procedure and decision rules to be examined and replicated. The entire corpus was subsequently reviewed by a human investigator, and all AI-generated classifications remained provisional. After the initial human review, the complete set of eligibility decisions underwent an AI-assisted consistency audit to identify potentially erroneous inclusions or exclusions and records for which the available bibliographic information was insufficient to determine eligibility. Every flagged record was reassessed by a human investigator using the title, abstract, and author keywords, with the full text consulted when necessary. Final eligibility decisions were based exclusively on human adjudication. This staged workflow, together with disclosure of the screening prompt, eligibility criteria, and transitions between AI classifications and human decisions, increased the transparency, auditability, and procedural reproducibility of corpus construction while reducing the inclusion of publications in which bullous pemphigoid appeared only as background information, incidental terminology, a differential diagnosis, or shared antigen nomenclature.
A second strength was the systematic sensitivity analysis of the author-keyword network parameters. Rather than selecting a single occurrence threshold and clustering resolution without comparison, the analysis evaluated multiple combinations of minimum keyword occurrence thresholds and clustering resolutions. Cluster agreement was quantified using the ARI and NMI, while thematic coherence, small-cluster formation, network coverage, and visualization readability were also considered. These procedures reduced dependence on an arbitrary parameter choice and provided an explicit basis for selecting the final network.
A third strength was the supplementary title-term analysis conducted across the full corpus. Because author-keyword availability was incomplete and varied substantially over time, terms were independently extracted from publication titles using VOSviewer’s text-mining procedure. The recovery of the principal thematic domains in the full-corpus analysis provided additional evidence that the broad conceptual structure identified from author keywords was not materially altered by incomplete keyword coverage.
A fourth strength was the use of thematic clusters, rather than isolated keywords, as the units of temporal analysis. The same network-derived cluster assignments were applied across all periods, and publications represented a cluster when they contained at least one retained keyword assigned to that domain. This approach reduced the influence of fluctuations in individual terms, terminology changes, and related synonyms. It also supported interpretation at the level of broader research domains while allowing publications to contribute to multiple clusters. Reporting period-specific prevalence, percentage-point changes, Cramér’s V, confidence intervals, and multiplicity-adjusted tests further supported interpretation of the magnitude, direction, and consistency of temporal differences.
Several limitations should be considered. First, the analysis was restricted to English-language articles and reviews indexed in Web of Science and therefore reflects that database’s journal coverage, citation tracking, affiliation coding, and indexing practices. Relevant publications absent from the database, published in other languages, or incompletely indexed were not represented. The language restriction may have affected country and institutional rankings, journal distributions, and the apparent prevalence of themes.
Second, author keywords were available for 55.1% of the publications and were less common in earlier decades. Temporal comparisons therefore describe a changing and potentially selective subset, and the earliest estimates are particularly uncertain. The supplementary title-term analysis used the full corpus and supported the broad thematic structure, but title terms are not equivalent to author keywords. In addition, the 2021–2025 period overlapped with the COVID-19 pandemic and its aftermath; the study cannot separate pandemic-related changes in research activity or publication timing from other temporal influences.
Third, thematic clusters depended on keyword normalization, the minimum occurrence threshold, and the selected clustering resolution. The sensitivity analyses showed stability across several parameter combinations, but they did not test alternative clustering algorithms or all possible normalization decisions. Some clusters were conceptually heterogeneous. In particular, the diagnosis cluster combined methods with related autoimmune blistering diseases, and the COVID-19 cluster included several other emerging topics. Conversely, the separation of DPP-4 inhibitor-associated and immune checkpoint inhibitor-associated bullous pemphigoid should not be treated as proof of etiologic or biological independence. Cluster labels were assigned by the authors and should be interpreted as summaries of co-occurrence patterns rather than fixed classifications.
Fourth, bibliometric indicators measure publication activity and citation visibility rather than methodological quality, certainty of evidence, clinical effectiveness, or disease burden. Bibliometric measures can assist research assessment but may be misleading when interpreted as stand-alone indicators of scientific quality.44 Cumulative citations favor older publications, whereas annualized and publication-year-normalized measures cannot fully account for different citation trajectories or the short citation windows of recent studies. Journal-level indicators likewise cannot establish the importance of an individual article.
Finally, country and institutional productivity used different attribution rules. Country output was fractionally allocated across corresponding-author entries, whereas each distinct institution received one full article count when it appeared in a publication. Related universities, hospitals, campuses, parent organizations, and health systems were not manually consolidated. Country and institutional rankings should therefore be interpreted as complementary descriptions of corresponding-author attribution and institutional appearances, not as directly comparable measures of research performance.
Conclusion
Bullous pemphigoid research expanded and diversified from 1959 to 2025 while remaining centered in dermatology. Among keyword-mappable publications, autoantigens, autoantibodies, and basement membrane zone biology decreased from 51.9% in 1959–1998 to 14.6% in 2021–2025. Over the same periods, therapeutic management and clinical outcomes increased from 7.7% to 37.8%, epidemiology and comorbidities from 1.9% to 20.8%, DPP-4 inhibitor-associated bullous pemphigoid from 0% to 12.7%, and immune checkpoint inhibitor-associated bullous pemphigoid from 0% to 10.8%. Diagnosis and related autoimmune blistering diseases peaked at 50.0% in 1999–2013 before declining to 24.3%, whereas inflammatory and immune effector mechanisms showed no statistically significant variation across periods after multiplicity adjustment.
These patterns indicate a shift in relative research attention from structural and diagnostic foundations toward treatment, patient outcomes, comorbidity, and medication-associated disease, while mechanistic and diagnostic investigation remained integral to the field. The full-corpus title-term analysis recovered the principal domains and supported the broad thematic structure identified from author keywords. These findings provide a basis for more focused evidence synthesis concerning biologic and pathway-directed therapies, outcomes in older adults with multimorbidity, and DPP-4 inhibitor-associated and immune checkpoint inhibitor-associated disease.
These priorities were inferred from publication patterns and should not be interpreted as direct evidence of research quality, causal effects, clinical benefit, or unmet need. Incomplete and temporally uneven author-keyword coverage, together with the coverage and indexing practices of Web of Science, also limited the precision of the temporal estimates. Nevertheless, this analysis provides a longitudinal map of how bullous pemphigoid research has evolved and offers an empirical basis for directing future synthesis and investigation toward emerging clinical and therapeutic questions.
Funding Statement
This research received no external funding.
AI Use Declaration
During corpus construction, ChatGPT (GPT-5.6 Sol; OpenAI, San Francisco, CA, USA) was used for the prespecified first-pass relevance screen described in Methods. ChatGPT (GPT-5.5) was also used during manuscript preparation for language editing, organization of selected text, and troubleshooting of selected R code. All AI-generated outputs were reviewed by the authors against the underlying records, software outputs, and cited sources. The tools did not make final eligibility decisions, independently calculate bibliometric indicators, generate references, or determine the interpretation of the findings. The authors take full responsibility for the accuracy, integrity, and originality of the work.
Data Sharing Statement
All bibliometric records analyzed in this study were retrieved from the Web of Science Core Collection and are available from the publisher under standard licensing.
Author Contributions
All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.
Disclosure
The authors report no conflicts of interest.
References
- 1.Akbarialiabad H, Schmidt E, Patsatsi A, et al. Bullous pemphigoid. Nat Rev Dis Primers. 2025;11(1):12. doi: 10.1038/s41572-025-00595-5 [DOI] [PubMed] [Google Scholar]
- 2.Werth VP, Murrell DF, Joly P, Ardeleanu M, Hultsch V. Bullous pemphigoid burden of disease, management and unmet therapeutic needs. J Eur Acad Dermatol Venereol. 2025;39(2):290–22. doi: 10.1111/jdv.20313 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Persson MSM, Begum N, Grainge MJ, Harman KE, Grindlay D, Gran S. The global incidence of bullous pemphigoid: a systematic review and meta-analysis. Br J Dermatol. 2022;186(3):414–425. doi: 10.1111/bjd.20743 [DOI] [PubMed] [Google Scholar]
- 4.van Beek N, Weidinger A, Schneider SW, et al. Incidence of pemphigoid diseases in Northern Germany in 2016 - first data from the Schleswig-Holstein Registry of Autoimmune Bullous Diseases. J Eur Acad Dermatol Venereol. 2021;35(5):1197–1202. doi: 10.1111/jdv.17107 [DOI] [PubMed] [Google Scholar]
- 5.Bahloul D, Dubucq H, Thomas RB, et al. Burden of disease of bullous pemphigoid: a targeted literature review. Dermatology. 2024;240(5–6):823–832. doi: 10.1159/000540480 [DOI] [PubMed] [Google Scholar]
- 6.Ito T, Kimura A, Muramatsu K, et al. Eosinophil-derived neurotoxin and eosinophil cationic protein are key molecules for blister formation in bullous pemphigoid. J Allergy Clin Immunol. 2025;156(3):705–718. doi: 10.1016/j.jaci.2025.04.028 [DOI] [PubMed] [Google Scholar]
- 7.Gellrich FF, Rakhmetova S, Günther C, Beissert S, Hobelsberger S. Real-time, noninvasive diagnosis of bullous pemphigoid: diagnostic performance of line-field confocal optical coherence tomography. Skin Res Technol. 2025;31(9):e70235. doi: 10.1111/srt.70235 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Kaya Ş, Yüceer RO, Tosun M, Mustafaeva R. The role of C4d expression in the diagnosis of bullous pemphigoid: a clinical and histopathological study. Arch Dermatol Res. 2025;317(1):569. doi: 10.1007/s00403-025-04031-2 [DOI] [PubMed] [Google Scholar]
- 9.Miao H, Yang J, Bai Y, et al. Aberrant glycosylation patterns as potential biomarkers for diagnosis and disease progression in bullous pemphigoid. Front Immunol. 2025;16:1538126. doi: 10.3389/fimmu.2025.1538126 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Williams HC, Wojnarowska F, Kirtschig G, et al. Doxycycline versus prednisolone as an initial treatment strategy for bullous pemphigoid: a pragmatic, non-inferiority, randomised controlled trial. Lancet. 2017;389(10079):1630–1638. doi: 10.1016/S0140-6736(17)30560-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Kasperkiewicz M, van Beek N, Schmidt E. Autoimmune bullous diseases: therapeutic update. Drugs. 2026;86(4):485–506. doi: 10.1007/s40265-026-02292-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Cao P, Xu W, Zhang L. Rituximab, omalizumab, and dupilumab treatment outcomes in bullous pemphigoid: a systematic review. Front Immunol. 2022;13:928621. doi: 10.3389/fimmu.2022.928621 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Lin X, Li X, Zhai Z, Zhang M. JAK-STAT pathway, type I/II cytokines, and new potential therapeutic strategy for autoimmune bullous diseases: update on pemphigus vulgaris and bullous pemphigoid. Front Immunol. 2025;16:1563286. doi: 10.3389/fimmu.2025.1563286 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Olbrich H, Sadik CD. Emerging drugs for the treatment of bullous pemphigoid: what’s new on the horizon? Expert Opin Emerg Drugs. 2025;30(3):169–178. doi: 10.1080/14728214.2025.2555080 [DOI] [PubMed] [Google Scholar]
- 15.Lai YC, Yew YW, Lambert WC. Bullous pemphigoid and its association with neurological diseases: a systematic review and meta-analysis. J Eur Acad Dermatol Venereol. 2016;30(12):2007–2015. doi: 10.1111/jdv.13660 [DOI] [PubMed] [Google Scholar]
- 16.Siranart N, Chumpangern Y, Phutinart S, et al. Chronic disease associated with bullous pemphigoid risk: a systematic review and meta-analysis. JAAD Int. 2024;17:141–152. doi: 10.1016/j.jdin.2024.08.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Liu SD, Chen WT, Chi CC. Association between medication use and bullous pemphigoid: a systematic review and meta-analysis. JAMA Dermatol. 2020;156(8):891–900. doi: 10.1001/jamadermatol.2020.1587 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Atzmony L, Mimouni I, Reiter O, et al. Association of bullous pemphigoid with malignancy: a systematic review and meta-analysis. J Am Acad Dermatol. 2017;77(4):691–699. doi: 10.1016/j.jaad.2017.05.006 [DOI] [PubMed] [Google Scholar]
- 19.Asdourian MS, Shah N, Jacoby TV, Reynolds KL, Chen ST. Association of bullous pemphigoid with immune checkpoint inhibitor therapy in patients with cancer: a systematic review. JAMA Dermatol. 2022;158(8):933–941. doi: 10.1001/jamadermatol.2022.1624 [DOI] [PubMed] [Google Scholar]
- 20.Myers EL, Culton DA. A narrative review of pemphigoid diseases: bridging associations, comorbidities, and management. Dermatol Ther. 2025;15(7):1755–1770. doi: 10.1007/s13555-025-01444-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Powers CM, Thakker S, Gulati N, et al. Bullous pemphigoid: a practical approach to diagnosis and management in the modern era. J Am Acad Dermatol. 2025;92(6):1337–1350. doi: 10.1016/j.jaad.2025.01.086 [DOI] [PubMed] [Google Scholar]
- 22.Huang SC, Chiu TM, Lee CY, Chang HC, Wu WJ, Gau SY. Researching trends in pemphigoid diseases: a bibliometric study of the top 100 most cited publications. Front Med. 2023;9:1088083. doi: 10.3389/fmed.2022.1088083 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Öztekin A, Öztekin C. The evolution of Pemphigus publications: a bibliometric analysis with research trends and global productivity. Medicine. 2024;103(18):e38047. doi: 10.1097/MD.0000000000038047 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Montazeri A, Mohammadi SM, Hesari P, Ghaemi M, Riazi H, Sheikhi-Mobarakeh Z. Preliminary guideline for reporting bibliometric reviews of the biomedical literature (BIBLIO): a minimum requirements. Syst Rev. 2023;12(1):239. doi: 10.1186/s13643-023-02410-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Aria M, Cuccurullo C. bibliometrix: an R-tool for comprehensive science mapping analysis. J Informetr. 2017;11(4):959–975. doi: 10.1016/j.joi.2017.08.007 [DOI] [Google Scholar]
- 26.van Eck NJ, Waltman L. Software survey: VOSviewer, a computer program for bibliometric mapping. Scientometrics. 2010;84(2):523–538. doi: 10.1007/s11192-009-0146-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Schmidt E, Zillikens D. Pemphigoid diseases. Lancet. 2013;381(9863):320–332. doi: 10.1016/S0140-6736(12)61140-4 [DOI] [PubMed] [Google Scholar]
- 28.Liu Z, Diaz LA, Troy JL, et al. A passive transfer model of the organ-specific autoimmune disease, bullous pemphigoid, using antibodies generated against the hemidesmosomal antigen, BP180. J Clin Invest. 1993;92(5):2480–2488. doi: 10.1172/JCI116856 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Joly P, Roujeau JC, Benichou J, et al. A comparison of oral and topical corticosteroids in patients with bullous pemphigoid. N Engl J Med. 2002;346(5):321–327. doi: 10.1056/NEJMoa011592 [DOI] [PubMed] [Google Scholar]
- 30.Langan SM, Smeeth L, Hubbard R, Fleming KM, Smith CJ, West J. Bullous pemphigoid and pemphigus vulgaris--incidence and mortality in the UK: population based cohort study. BMJ. 2008;337(7662):a180. doi: 10.1136/bmj.a180 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Labib RS, Anhalt GJ, Patel HP, Mutasim DF, Diaz LA. Molecular heterogeneity of the bullous pemphigoid antigens as detected by immunoblotting. J Immunol. 1986;136(4):1231–1235. doi: 10.4049/jimmunol.136.4.1231 [DOI] [PubMed] [Google Scholar]
- 32.Jordon RE, Beutner EH, Witebsky E, Blumental G, Hale WL, Lever WF. Basement zone antibodies in bullous pemphigoid. JAMA. 1967;200(9):751–756. doi: 10.1001/jama.1967.03120220053008 [DOI] [PubMed] [Google Scholar]
- 33.Beutner EH, Jordon RE, Chorzelski TP. The immunopathology of pemphigus and bullous pemphigoid. J Invest Dermatol. 1968;51(2):63–80. doi: 10.1038/jid.1968.94 [DOI] [PubMed] [Google Scholar]
- 34.Bernard P, Vaillant L, Labeille B, et al. Incidence and distribution of subepidermal autoimmune bullous skin diseases in three French regions. Bullous Diseases French Study Group. Arch Dermatol. 1995;131(1):48–52. doi: 10.1001/archderm.1995.01690130050009 [DOI] [PubMed] [Google Scholar]
- 35.Giudice GJ, Emery DJ, Zelickson BD, Anhalt GJ, Liu Z, Diaz LA. Bullous pemphigoid and herpes gestationis autoantibodies recognize a common non-collagenous site on the BP180 ectodomain. J Immunol. 1993;151(10):5742–5750. doi: 10.4049/jimmunol.151.10.5742 [DOI] [PubMed] [Google Scholar]
- 36.Stanley JR, Tanaka T, Mueller S, Klaus-Kovtun V, Roop D. Isolation of complementary DNA for bullous pemphigoid antigen by use of patients’ autoantibodies. J Clin Invest. 1988;82(6):1864–1870. doi: 10.1172/JCI113803 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Chalmers JR, Wojnarowska F, Kirtschig G, et al. A randomised controlled trial to compare the safety, effectiveness and cost-effectiveness of doxycycline (200 mg/day) with that of oral prednisolone (0.5 mg/kg/day) for initial treatment of bullous pemphigoid: the Bullous Pemphigoid Steroids and Tetracyclines (BLISTER) trial. Health Technol Assess. 2017;21(10):1–90. doi: 10.3310/hta21100 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Chen Y, Sun K, Chang J. Dupilumab treatment outcomes in bullous pemphigoid: a systematic review and single-arm meta-analysis. Front Immunol. 2026;17:1651543. doi: 10.3389/fimmu.2026.1651543 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Nieto-Benito LM, Suárez-Fernández R. Unraveling the link between bullous pemphigoid and neurological disease: a single-center study of 257 patients. Actas Dermosifiliogr. 2025;116(8):807–814. doi: 10.1016/j.ad.2024.10.071 [DOI] [PubMed] [Google Scholar]
- 40.Chen X, Zhang Y, Luo Z, et al. Prognostic factors for mortality in bullous pemphigoid: a systematic review and meta-analysis. PLoS One. 2022;17(4):e0264705. doi: 10.1371/journal.pone.0264705 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Kawsar A, Edwards C, Patel P, et al. Checkpoint inhibitor-associated bullous cutaneous immune-related adverse events: a multicentre observational study. Br J Dermatol. 2022;187(6):981–987. doi: 10.1111/bjd.21836 [DOI] [PubMed] [Google Scholar]
- 42.Thompson JA, Schneider BJ, Brahmer J, et al. NCCN Guidelines® Insights: management of immunotherapy-related toxicities, Version 2.2024. J Natl Compr Canc Netw. 2024;22(9):582–592. doi: 10.6004/jnccn.2024.005742 [DOI] [PubMed] [Google Scholar]
- 43.Singh S, Kirtschig G, Anchan VN, et al. Interventions for bullous pemphigoid. Cochrane Database Syst Rev. 2023;8(8):CD002292. doi: 10.1002/14651858.CD002292.pub4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Thompson DR, Clark AM. Measuring research success via bibliometrics: where they fit and how they help and hinder. J Adv Nurs. 2015;71(12):2699–2701. doi: 10.1111/jan.12610 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
All bibliometric records analyzed in this study were retrieved from the Web of Science Core Collection and are available from the publisher under standard licensing.