This study aimed to assemble a comprehensive collection of articles for bibliometric analysis. A structured query was constructed in the WoS database to capture literature relevant to carotid webs (full search string provided in the Supplementary material). WoS was chosen due to its broadest coverage of academic literature at the journal, article, and reference level [8,17]. The search was conducted on August 20, 2025, and all identified records were exported with associated metadata. Only English-language publications were retained. Articles were manually screened for relevance.

The analysis spanned publications from 1968 through 2025. To assess international collaboration, additional analyses restricted to 2016-2025 were subdivided into three periods (2016-2018, 2019-2021, 2022-2025) to capture recent acceleration.

Bibliometric analyses were performed in R (The R Foundation, Vienna, Austria) using the Bibliometrix R-Tool, as described in our previous work [17,22]. We supplemented this with open-source Python (Python Software Foundation, Wilmington, DE, USA) packages and libraries, as described previously [6-8]. Python regular expressions (re module) were applied for textual parsing of article titles. Data were aggregated with the Pandas library, and visualizations were generated using seaborn, Matplotlib, and networkx libraries, with network analyses further processed in Gephi (v0.10.1). Coauthorship networks were detected using the Clauset-Newman-Moore greedy modularity optimization algorithm [16,17].

Author-submitted keywords and WoS KeyWords Plus, an algorithmically generated set of words or phrases related to an article, were analyzed to identify emerging terms and topics. RPYS, a bibliometrics method analyzing the frequency with which references are cited in a collection of documents, was used to explore the historical roots of the field, as in our prior work [6-8].

Specifically, RPYS is a hypothesis-free bibliometrics method that analyzes a set of publications, counting the number of times cited references from a particular publication year appear in the citation lists of the other publications. A spectrogram plot is generated, with prominent peaks denoting the most frequently cited publication years and marking periods when foundational studies were published. RPYS then identifies the most influential publications, defined as those with substantially more citations than others from the same year. In doing so, RPYS identifies papers that significantly shaped a field’s early development and research directions, even if they are less frequently cited today [8]. Applied to the carotid web literature, RPYS assessed citation patterns to detect such influential works.

When referring to citation metrics in text, “global” refers to the number of citations a publication has received from documents contained in the entire WoS database, whereas “local” refers to the number of citations a publications has received from documents included in the analyzed collection (e.g., all publications concerning carotid webs).

Author productivity was assessed using the h-, m-, and g-indices. The h-index reflects the number of publications (h) with at least h citations; the m-index normalizes this value by years since first publication; and the g-index identifies the largest number (g) of papers that collectively have at least g² citations. Together, these metrics provide complementary measures of productivity, continuity, and impact.

Our WoS query initially returned 305 documents; these were manually screened to include only publications relating to carotid webs. Thus, the final database included 281 documents from 109 sources, authored by 1,129 unique individuals and citing 1,361 references (Fig. 1). The average document age in our collection was 4.79 years. There was a 6.15% mean annual growth rate in articles published over the time span analyzed (1968-2025).

Across the entire database from 1968 to 2025, the proportion of studies with international coauthorship was 9.25%. When separated into three time periods, international coauthorship was 10.53% for the 2016-2018 period, 10.87% for 2019-2021, and 8.82% for 2022-2025 (Fig. 2A-C). Within our Web of Science query, international collaboration was first noted in 2016. Across all time points, the United States had the greatest number of corresponding authors on all articles (n=405), followed by China (n=152), France (n=146), and Canada (n=80) (Fig. 2D-E).

We extracted unique n-grams from keywords, and the percentage of total mentions accumulated was calculated. Eight keywords occurring frequently across the entire database were identified and included “carotid web,” “stroke,” “fibromuscular dysplasia,” “management,” “risk,” “recurrent,” and “Afro-Caribbean patients” (Fig. 3).

The earliest keywords to appear with notable frequency were “stroke” and “fibromuscular dysplasia,” both of which rose sharply after 2015. By 2017-2019, additional terms such as “carotid web,” “ischemic stroke,” “recurrent,” and “Afro-Caribbean patients” began increasing in use. Notably, by 2021-2022, “carotid web” and “ischemic stroke” had surpassed “fibromuscular dysplasia” in frequency [4,31].

The rise in “ischemic stroke” usage likely paralleled the growing understanding of carotid webs’ mechanism of promoting distal embolism. Similarly, the increased frequency of the term “recurrent” corresponded with reports of high recurrence rates in untreated patients. The emergence of “Afro-Caribbean patients” as an increasingly frequent keyword beginning in 2019-2021 likely reflects growing awareness of the disproportionate prevalence of carotid webs among this population. Finally, in the most recent period (2019-2025), keywords such as “risk” and “management” have shown growth, consistent with the field’s current focus on treatment strategies (e.g., carotid endarterectomy or stenting) [13].

Across the entire database from 1968 to 2025, the mean growth in the number of new articles per year relating to carotid webs was 6.15%; the largest period of year-over-year growth occurred from 2020 to 2025, in which the total number of published articles tripled. In fact, despite the first report of carotid web being published in 1968, 90% of articles in the database were published after 2016 (Fig. 4A-D).

Stroke had the greatest number of published documents (23 articles) followed by Journal of Vascular Surgery (19). Stroke was also the most locally cited source with 635 citations, followed by American Journal of Neuroradiology at 360 citations and Neurology at 164. The top five sources by production were Stroke, Journal of Vascular Surgery, International Journal of Stroke, Journal of Stroke & Cerebrovascular Diseases, and Journal of NeuroInterventional Surgery.

Stroke and Journal of Vascular Surgery have experienced significant and earlier growth relative to other sources. Additionally, Stroke had the highest h-index of 7 in the context of carotid web literature, followed by Journal of NeuroInterventional Surgery at 6.

We identified the most prolific authors within the carotid web research field. To accurately measure research impact, we evaluated the 25 most productive authors through their h-, m-, and g-indices to give a multifaceted assessment (Fig. 5A-C).

A coauthor network graph was created for all authors with 10 or more publications within the body of literature (Fig. 5D). Three coauthorship networks were detected algorithmically, which remained relatively small due to the field’s niche subject and recent emergence as a more prominent area of research.

RPYS was used to identify citation patterns across the carotid web literature (Fig. 5E). Peaks in publication years correspond to the appearance of influential works that shaped the trajectory of the field. This analysis revealed 19 seminal publications spanning 1968-2021, representing the foundation of current carotid web research and delineating how the field moved from early descriptive pathology to recognition of carotid web as a distinct, clinically relevant, and potentially treatable cause of ischemic stroke (Table 1).

International collaboration in carotid web research remains limited. Our analysis revealed that the first international collaboration in carotid web research occurred in 2016 and has remained at modest levels since, involving only 9-11% of studies. The United States has produced the largest number of publications, followed by China, France, and Canada. Unlike other cerebrovascular or neurosurgical fields where international coauthorship has markedly grown, carotid web research remains largely concentrated within a few countries [15,25,27,38].

This reflects both the field’s relative nascency and the predominance of retrospective single-center case series [10]. Given the disproportionate prevalence of carotid webs among patients of Afro-Caribbean descent, broader global collaboration will be essential to ensure studied populations are representative of those most at risk [18]. Broader international cooperation will also be critical to address unresolved questions regarding incidence, natural history, recurrence risk, and optimal management strategies.

Although our keyword analysis was limited to a relatively small corpus, it revealed a fundamental shift in the conceptualization of carotid web over time. Early publications frequently referred to “fibromuscular dysplasia” and “stroke,” reflecting a period of diagnostic uncertainty and the absence of a defined nosology. In this period, webs were conceptualized as a histological variant rather than a unique pathology [9,26]. By the late 2010s, “carotid web” emerged as a distinct keyword, paralleled by terms such as “ischemic stroke,” “recurrent,” and “Afro-Caribbean patients.” This period mapped the transition from case-based descriptions to a broader epidemiological understanding of who is at risk and the high likelihood of treatment failure with antiplatelet therapy alone. Most recently, the rise of “risk” and “management” reflects the field’s turn toward quantitatively evaluating stroke risk and investigating optimal intervention strategies [2,3,24]. This complete trajectory mirrors the clinical evolution of carotid web research: from descriptive pathology, to recognition of its embolic mechanism, to contemporary concerns with recurrence prevention and intervention.

Publication sources have shifted over time, reflecting changes in clinical interest. Stroke and Journal of Vascular Surgery have been consistent contributors to the field, producing early and high-impact publications. Stroke achieved the highest h-index and local citation count, illustrating its central role in publishing carotid web research. However, more recent years have seen diversification across neurosurgical, neurointerventional, neurological, and radiological journals as carotid web research grows increasingly multidisciplinary in nature.

Our analysis of author productivity demonstrates that a relatively small number of investigators currently drive much of the literature. Established contributors such as Haussen, Nogueira, and Frankel have among the highest h- and g-indices, reflecting consistent productivity and citation impact. Rising investigators with high m-indices, such as Benali, Bana, Alhabli, and Oshinski, illustrate how newer authors have rapidly influenced the field within a shorter time span.

Coauthorship analysis identified three relatively small clusters with limited cross-collaboration, likely reflecting the field’s nascency and the rarity of the vascular lesion itself. These findings illustrate both a challenge and an opportunity: as carotid web research matures, stronger inter-institutional and interdisciplinary collaborations will be essential to accelerate progress [1,27].

A unique strength of our analysis lies in the use of RPYS to identify 19 critical articles that helped to map the field’s trajectory: from Rainer et al.’s 1968 description of “fibromuscular hyperplasia,” through decades of nosologic uncertainty and overlap with fibromuscular dysplasia, to the pivotal mid-2010s studies that defined carotid web as a distinct, clinically significant vascular lesion with a high risk for recurrent ischemic stroke due to a thromboembolic mechanism [5,24,28]. Recent research has shifted from recognition and diagnosis to assessing stroke recurrence risk and evaluating optimal management strategies, with interventional approaches increasingly favored [1,10,14,24]. Notably, the webs’ natural history and potential primary prevention strategies remain unclear. Identifying these seminal articles is essential for understanding the field’s historical origins and the key turning points. Collectively, these seminal publications form the evidentiary backbone upon which contemporary diagnostic criteria, recurrence risk estimates, and interventional strategies for carotid web-associated stroke are based.

While early studies are foundational, several assumptions from the historical literature require careful reevaluation in light of more recent evidence:

- Nosology and relation to fibromuscular dysplasia (FMD): Early histologic descriptions of intimal fibromyxoid proliferation led many reports to label webs as “atypical FMD,” as FMD was the only known entity with a similar histology, and carotid webs were not yet formally defined [26,32,36]. Contemporary series, however, support carotid web as a distinct lesion characterized by focal intimal proliferation rather than the medial fibroplasia of classic FMD [31,34]. Future work should avoid conflating the two entities and adopt clear, reproducible definitions.

- Posterior ICA Location: Early case reports exclusively described webs on the posterior ICA wall, a finding reiterated across decades. As a result, webs are frequently defined as lesions of the posterior ICA wall. Larger modern series, however, confirm that webs may arise from the medial, lateral, or even anterior ICA wall around the carotid bulb, though the posterior wall remains the most common carotid web location [10,14,35]. Restricting diagnostic attention to the posterior wall risks missed diagnoses; careful multiplanar review of all walls on imaging is warranted [10].

- Pathogenesis: Early reports hypothesized carotid webs to be developmental in origin, including McNamara’s 1987 paper postulating them as an embryologic anomaly of brachiocephalic development [20,23]. However, no definitive evidence supports this. One pediatric stroke cohort, albeit small, was unrevealing for webs, and one case report suggested possible de novo formation in an adult [12,33]. Whether webs are congenital, acquired, or represent a spectrum remains unknown and merits further study.

Our findings illustrate the value of bibliometric methods in mapping rapidly expanding but historically diffuse research areas. In the case of carotid webs, where decades of sporadic case reports have given way to a recent surge in more systematic studies, bibliometric analysis provides both a structured overview and an educational resource. The RPYS-derived set of seminal publications in Table 1 offers a concise reading list for trainees and clinicians seeking to deepen their understanding of the field’s historical development and the current trajectory of carotid web research.

The bibliometric trends identified in this study have direct implications for current practice and future study design. The recent surge in terms related to “management” and “risk” illustrates a shift from diagnostic discovery to therapeutic decision-making, a transition of particular relevance to cerebrovascular surgeons.

The move away from the “FMD” label carries critical management implications. While FMD is often managed conservatively, the bibliometric rise of the term “recurrent” demonstrates that carotid webs are increasingly viewed as high-risk lesions requiring definitive intervention. This shift is supported by emerging evidence of recurrence rates as high as 30-50% in medically managed patients, compared to significantly lower rates following endarterectomy or stenting [1,11,13]. Furthermore, the identification of Afro-Caribbean heritage as a rising focal point in the literature suggests that clinicians should maintain a high index of suspicion for carotid webs in young patients of this demographic presenting with cryptogenic stroke.

The current fragmentation of coauthorship networks and predominance of single-center retrospective series highlight critical gaps in the surgical literature. Multicenter prospective registries are needed to address fundamental questions regarding optimal patient selection criteria for endarterectomy versus stenting, long-term durability of each approach, comparative recurrence rates over prolonged follow-up, and peri-procedural complication profiles. Furthermore, primary prevention for incidentally discovered carotid webs remains largely unexplored, such as whether asymptomatic webs discovered incidentally should be treated prophylactically. This question is directly relevant to neurosurgical practice and warrants systematic investigation.

This study is primarily limited by its reliance on the Web of Science database. While WoS is considered the gold standard for bibliometric metadata and reference-level analysis, it may omit relevant articles from regional or niche journals not indexed therein. This potential for missing research suggests that while our study provides a comprehensive “roadmap” of the field, it should not be viewed as an exhaustive inventory of every published case. Further, the inclusion of only English-language publications introduces a linguistic bias. Additionally, RPYS relies on citation frequency and thus may underrepresent recent influential publications that have not yet accumulated sufficient citations, creating a natural time lag effect. Finally, although bibliometric analysis provides robust quantitative insights into research trends, it cannot capture the full clinical or scientific nuance of individual studies and must be interpreted within that context. As the carotid web literature continues its rapid expansion, future supplementary analyses will be necessary to capture emerging data and fill remaining gaps in the literature.