Clinical characteristics of cervical lymphadenopathy in Kawasaki disease and the generation of related cervical complications: a retrospective cohort study

Abstract

Objectives

Cervical lymphadenopathy (CL) is a significant but underexplored feature of Kawasaki disease (KD), especially during its early stages. Although CL is a recognized diagnostic criterion for KD, it is frequently misdiagnosed as bacterial cervical lymphadenitis (BCL). This study aimed to investigate the clinical and imaging characteristics of CL in KD, particularly in cases, where KD initially presents with fever and cervical lymphadenopathy (LKD). The study also focused on differentiating LKD from BCL and examining its association with systemic and cervical complications.

Methods

In this study, 681 KD patients and 101 BCL patients hospitalized at the Yuying Children's Hospital affiliated with Wenzhou Medical University from January 1, 2021, to October 1, 2023, were enrolled. Among the KD patients, 403 exhibited cervical lymphadenopathy, including 86 who presented only with fever and cervical lymphadenopathy upon admission (LKD group), 317 who displayed typical features of cervical lymphadenopathy (Other KD group), and 243 who did not exhibit cervical lymphadenopathy. Demographic, clinical, laboratory, and imaging data were compared across these groups to identify distinguishing characteristics and potential complications.

Results

The findings indicated that patients with CL in KD were younger than those with BCL, experienced a longer duration of fever prior to treatment, and exhibited more severe inflammatory responses and liver abnormalities. KD patients with enlarged cervical lymph nodes were more likely to exhibit systemic involvement. Imaging studies revealed bilateral lymph node enlargement without liquefaction in KD, which contrasts with the abscess formation commonly observed in BCL. Compared to KD patients without lymphadenopathy, those with CL were older, more likely to present with conjunctival congestion, had higher neutrophil ratios, elevated liver damage markers, an increased rate of intravenous immunoglobulin (IVIG) resistance, and a higher incidence of KD shock syndrome (KDSS). Among these, LKD patients were older, nearly all reported neck pain, had a higher referral rate, elevated neutrophil counts and C-reactive protein (CRP) levels, and a higher prevalence of cervical complications, such as atlantoaxial subluxation and parotitis. Ultrasound examinations of cervical lymph nodes showed that the LKD group had larger diameters of enlarged lymph nodes, with the largest lymph node frequently found in the Level II region.

Conclusions

IVIG resistance and KDSS are more prevalent among patients with cervical lymph node enlargement in Kawasaki disease compared to those without lymph node involvement. When cervical B-ultrasonography reveals bilateral multiple lymph nodes in Level II regions, without liquefaction or significant abscess formation, clinicians should be vigilant for the possibility of Kawasaki disease.

Background

Kawasaki disease (KD) is a systemic vasculitis predominantly affecting children under the age of 5 [1]. The most frequent complications of KD are coronary artery lesions (CALs), which include coronary artery dilatation (CADs) and coronary artery aneurysms (CAAs) [2, 3]. KD is recognized as a significant cause of acquired heart disease in children in developed nations [4, 5]. Approximately 12% of KD cases initially manifest with fever and cervical lymphadenitis, which can affect the soft tissues of the neck, causing symptoms, such as neck pain, neck masses, limited neck movement, and torticollis [6, 7]. Although lymphadenopathy is a less common diagnostic criterion for KD, occurring in 50–75% of patients, the other four major clinical criteria are present in 90% of cases [8]. As a result, KD patients presenting with cervical lymphadenopathy (CL) are often misdiagnosed as having bacterial cervical lymphadenitis (BCL), leading to delayed diagnosis and treatment [9,10,11].

Cervical lymphadenitis, a localized inflammatory condition, not only alters the pathognomonic features of KD but also contributes to the development of cervical complications. Previous studies have reported that approximately 3.6% of KD cases were accompanied by retropharyngeal inflammation [12] or primarily presented with parapharyngeal inflammation [13, 14]. Furthermore, KD complicated by parotitis is considered an uncommon manifestation of the disease [15, 16]. Atlantoaxial subluxation (AARS) in KD was first documented by researchers in 1989 and has been found to be more prevalent in children presenting with fever and CL at onset [17, 18]. Our center observed that KD patients with CL are at risk of developing these cervical complications. This study aimed to analyze the clinical characteristics and risks associated with CL in KD by comparing them with those of BCL patients and KD patients without cervical lymphadenopathy.

Methods

Medical records of patients hospitalized with KD at the Wenzhou Medical University affiliated Yuying Children’s Hospital were reviewed. After estimating the number of all KD patients who met the diagnostic criteria and screening the specific information for each case, we calculated the effect size by combining previous studies comparing KD complicated by cervical lymphadenitis and BCL. Then, we used the statistical software G.Power to estimate the required sample size, taking into account a 90% confidence interval, 95% response distribution, and a 5% margin of error. Based on the predicted minimum sample size of at least 50–60 cases per group, we included more cases in each group. From January 1, 2021, to October 1, 2023, a total of 681 KD patients and 101 BCL patients were enrolled in the study. Among the KD patients, 18 cases were excluded due to the absence of IVIG treatment, and 17 cases were excluded due to incomplete laboratory data. The remaining 646 patients included 403 cases complicated by cervical lymphadenopathy, of which 86 cases presented with fever and cervical lymphadenopathy as the initial manifestation LKD group (who show only fever and cervical lymphadenopathy before other clinical signs of KD appear [1]), and 317 cases exhibited typical features of cervical lymphadenopathy (Other KD group). In addition, there were 243 cases without cervical lymphadenopathy (Fig. 1).

Fig. 1

Patients flow chart. Flow chart showing the demographic and clinical information of all study participants. From January 1, 2021, to October 1, 2023, 782 children in our KD database were enrolled

The diagnosis of complete KD is based on the criteria defined by the American Heart Association (AHA) [2]: fever lasting at least 5 days, along with four out of the five major clinical manifestations: rash, bilateral conjunctivitis without exudate, inflammation of oral mucosa, cervical lymphadenopathy and extremity changes. The diagnosis of incomplete KD follows AHA guidelines as well, where fever lasting more than 5 days is associated with two or three clinical criteria, in conjunction with or without CAA [2]. All KD cases included in this study were diagnosed by a KD specialist.

BCL is defined as acute cervical lymphadenitis in which antibiotic treatment improves fever symptoms; alternatively, bacterial identification can occur through aspiration or surgical biopsy [1].

Coronary artery abnormalities identified through echocardiography were classified based on the Z-score scheme, with the following definitions: (1) no involvement: Z-score always < 2.0 mm. (2) Dilation: Z-score ≥ 2.0 mm to < 2.5 mm. (3) Small aneurysm: Z-score ≥ 2.5 mm to < 5.0 mm. (4) Medium aneurysm: Z-score ≥ 5.0 mm to < 10.0 mm, and absolute dimension < 8 mm. (5) Large or giant aneurysm: Z-score ≥ 10.0 mm, or absolute dimension ≥ 8 mm [2]. Given that the peak incidence of CALs in KD typically occurs between 2 and 4 week post-onset [3], we at least two echocardiography examinations were required after KD onset, with a minimum interval of 1 week between assessments. To avoid missed diagnoses, any CALs meeting the aforementioned criteria detected within 1 month of KD onset were considered to have occurred.

Treatment regimen: All patients received aspirin at a daily dose of 30–50 mg/kg during the acute phase. The IVIG regimen was determined based on the total immunoglobulin dose and the child’s body weight, and was categorized into two types: the standard regimen (2 g/kg administered intravenously as a single dose) and the non-standard regimen (for instance, 1 g/kg/day for two consecutive days) [4,5,6].

Duration of IVIG treatment: IVIG treatment administered more than 10 days after onset of illness was classified as delayed IVIG treatment. Conversely, treatments administered within 10 days were categorized as non-delayed IVIG treatment.

Intravenous immunoglobulin treatment efficacy: IVIG efficacy was categorized as IVIG-resistant and IVIG-sensitive. The IVIG-resistant was defined as patients who experienced persistent fever above 38.0 °C for 48 h following the completion of IVIG infusion, otherwise defined as the latter [4].

Delineation of the neck node levels for ultrasound imaging: according to the nomenclature proposed by the American Head and Neck Society and the American Academy of Otolaryngology-Head and Neck Surgery, and in alignment with the TNM atlas for lymph nodes in the neck, 7 node groups (some being divided into several levels) were defined [7]. Level I includes submental group (Ia) and submandibular group (Ib). Level Ia is a median region located between the anterior belly of the digastric muscles. Level Ib contains the submandibular nodes located in the space between the inner sides of the mandible laterally and the digastric muscle medially. Level II contains the upper jugular nodes located around the upper one-third of the internal jugular vein (IJV) and the upper spinal accessory nerve (SAN). Level III contains the middle jugular nodes located around the middle third of the IJV. Level IVa contains the lower jugular lymph nodes located around the inferior third of the IJV. Level IVb contains the medial supraclavicular lymph nodes located in the continuation of level IVa down to the cranial edge of the sternal manubrium. Level V are the posterior cervical triangle and supraclavicular lymphoid group, which are divided into two parts bounded by the inferior border of the cricoid cartilage, with Level Va above and Level Vb below. Level VI is the central lymphatic group, subdivided into Level VIa, located between the anterior edges of the sternocleidomastoid muscles, and Level VIb situated between the 2 common carotid arteries. Level VII is the upper mediastinal zone extending from the upper sternal border to the upper border of the aortic arch (Fig. 2).

Fig. 2

There are seven neck node groups (some being divided into several levels) were defined

Clinical findings, laboratory parameters, and radiological data were systematically documented. This included laboratory tests results upon admission and before treatment initiation, such as white blood cell count (WBC), neutrophil percentage, C-reactive protein (CRP), albumin (ALB), alanine aminotransferase (ALT), platelet count (PLT), erythrocyte sedimentation rate (ESR), N-terminal pro-brain natriuretic peptide (NT-proBNP), as well as imaging studies of cervical lymph nodes, atlantoaxial joint, parotid gland, chest X-ray. Detailed records of treatment regimens, including IVIG administration and management of cervical complications, were also maintained. Patient outcomes, including resolution of primary disease and cervical symptoms, were meticulously recorded.

Statistical methods

First, demographic, clinical characteristics and imaging differences were compared between patients groups. Continuous variables were analyzed using two independent samples t test or rank-sum test as appropriate; categorical variables were evaluated using the chi-square test or Fisher’s exact tests. The Cohen’s d effect size (0.07), chi-square test results (P = 0.03), and odds ratio with 95% CI (2.233, 1.047–4.763).

Multiple logistic regression models were employed to assess the impact of cervical lymph nodes enlargement and other factors on IVIG resistance. To evaluate potential multicollinearity in the regression models, we conducted Variance Inflation Factor (VIF) analyses. A VIF threshold of > 5 was used to identify significant multicollinearity among variables. In model 1, adjustments were made for age (in months) and gender only. In model 2, additional adjustments included KD type, presence of coronary artery damage, IVIG therapeutic effect, treatment regimen, and time of IVIG treatment. In model 3, further adjustments were made for laboratory indices.

The statistical analyses were performed using IBM SPSS (version 26.0). All tests were two-tailed. P < 0.05 was considered statistically significant.

Results

Inter-group comparison of clinical features

In this study, we analyzed a cohort 403 KD patients with CL, including 101 patients with BCL, and 243 KD patients without CL was analyzed. Compared to BCL patients, KD patients with CL were significantly younger, with a median age of 36.0 months vs. 68.5 months. In addition, they experienced a prolonged duration of fever prior to treatment initiation (P = 0.005) and exhibited significantly higher levels of C-reactive protein (CRP) (P = 0.002), alanine aminotransferase (ALT) (P < 0.001), N-terminal pro b-type natriuretic peptide (NT-proBNP) (P = 0.001), and D-Dimer (P < 0.001), as well as lower albumin (ALB) levels (P < 0.001). Patients with BCL exhibited a higher likelihood of concurrent parotitis, suppurative tonsillitis, and submandibular gland enlargement. No significant difference was observed in the incidence of atlantoaxial subluxation between the two groups (see Table 1). According to Table 2, KD patients with CL were older than those without lymphadenopathy (median age: 36.0 months vs. 17.2 months) and demonstrated a higher prevalence of lip and oral mucosal abnormalities. Although there was no significant difference in white blood cell count, patients with CL had a higher percentage of neutrophils (P = 0.015) and a lower percentage of lymphocytes (P < 0.001), as well as a higher incidence of liver damage (P = 0.038), and IVIG resistance (P = 0.033), which was confirmed in subsequent logistic regression analysis (Table S2). Furthermore, patients with CL had a higher likelihood of progressing to KDSS (P = 0.009).

Table 1 Comparison of clinical characteristics between bacterial cervical lymphadenitis and lymphadenopathy KD patients

Table 2 Comparison of clinical characteristics between lymphadenopathy KD and the without lymphadenopathy patients

Subsequently, we divided the 403 KD patients with CL into the LKD group and the Other KD group. As illustrated in Table 3, the LKD group exhibited a significantly higher median age (53.5 months) compared to the Other KD group (31.0 months). In addition, nearly all patients in the LKD group initially presented with neck pain symptoms and had a notably higher rate of inter-hospital or inter-departmental transfers (P = 0.028). However, the occurrence of the other four typical signs of KD was lower in the LKD group compared to typical cervical lymphadenopathy KD patients. The LKD group also demonstrated a higher percentage of neutrophils (P = 0.006), a lower percentage of lymphocytes (P < 0.001), and significantly elevated CRP levels (P = 0.039). Importantly, our findings indicated that the LKD group was more prone to cervical complications, including 8 cases (9.3%) of spontaneous atlantoaxial subluxation, 6 cases (7.0%) of acute parotitis, and 1 case (1.2%) of pharyngeal infection.

Table 3 Comparison of characteristics between LKD and the other KD patients

In Table S1, patients were stratified into two groups based on the size of the cervical lymph nodes: those with lymph nodes ≥ 1.5 cm and those with lymph nodes < 1.5 cm group. In the ≥ 1.5 cm group, the age of onset was significantly older (P < 0.001), and 52.6% of the patients reported symptoms of neck pain. Comparable to the < 1.5 cm group, this cohort exhibited a higher percentage of neutrophils (P < 0.001) and a greater incidence of liver damage (P = 0.016).These patients were also more susceptible to the aforementioned complications in the neck region.

Inter-group comparison of imaging

We observed that, compared to KD patients with CL, BCL patients exhibited a significantly larger maximum lymph node diameter (P < 0.001), a higher incidence of abscess formation (P = 0.001), and a greater likelihood of liquefaction (P < 0.001) as assessed by ultrasound. However, KD patients with CL showed bilateral cervical lymph node involvement (P = 0.001), with no significant difference in cervical lymph node segmentation (Tables 1 and 4). When comparing the location of enlarged lymph nodes using ultrasound, a notable difference was found in the location of the largest lymph node (P < 0.001) between the LKD and Other KD groups (Table 4).Specifically, in the LKD group, 32 cases (61.5%) were located in Level II, whereas in the other KD group, 18 cases (23.7%) were identified in this level. Occurrences in Levels I, III, and V were less common, with no occurrences noted at Level IV.

Table 4 Comparison of the location of the largest lymph node in groups

Discussion

Cervical lymphadenopathy, as a primary clinical manifestation, accounts for approximately 9–23% of acute KD admissions [8, 9]. Although CL is less frequently observed among the five principal diagnostic criteria for KD, there have been instances, where patients initially present solely with fever and cervical lymphadenopathy [10]. Such presentations can lead to delayed diagnosis and treatment, thereby increasing the risk of KD-associated cardiac complications and IVIG resistance [11]. Furthermore, these cases may be misdiagnosed as bacterial lymphadenitis or other lymphadenopathy-related conditions. Therefore, early identification of KD patients with CL can be facilitated by evaluating clinical and imaging features.

In this study, enlargement of cervical lymph nodes was observed in 62.4% of KD patients. We found that patients with CL were generally younger but exhibited elevated levels of CRP, PLT, ESR, NT-proBNP, D-Dimer, and AST, alongside relative anemia. A study proposed that natriuretic peptide levels may serve as a valuable biomarker to differentiate between KD and CL patients [12]. Furthermore, a 4-item scoring system has been proposed to distinguish node-first presentation of Kawasaki disease (NFKD) from BCL patients [13]. However, in our center, KD patients with CL were younger, with no differences found in white blood cell and neutrophil counts. Notably the incidence of pyuria, hepatosplenomegaly, and pulmonary imaging abnormalities was higher among these patients. Given that KD is a vasculitis, it may affect systemic organs, including the respiratory, digestive, and genitourinary systems [14, 15]. In some atypical cases of KD, initial presentations may mimic other diseases but eventually align with KD as the condition progresses. Ultrasound imaging of the neck in KD patients with CL revealed bilateral lymphadenopathy without liquefaction (Fig. 3). Abscesses were more frequently observed in BCL cases, and lymph node diameters were larger; however, imaging findings alone are not definitive for distinguishing between KD and BCL.

Fig. 3

a, b Bilateral lymph node enlargement in Kawasaki disease. c, d Inadequate blood flow in swollen lymph nodes in the neck in Kawasaki disease

In recent years, there has been a notable increase in scholarly interest regarding KD patients who present with fever and cervical lymphadenopathy. Due to the relatively small number of such cases, the clinical features of these patients have not been fully elucidated [8]. This unusual presentation represents one of the common diagnostic challenges in KD, often leading to delayed diagnosis. Among typical KD patients, 8–21% have been categorized as NFKD and initially presented with only cervical lymphadenopathy and fever [10]. We observed that LKD patients accounted for 21.3% of KD patients with cervical lymph nodes swelling. These patients exhibited characteristics including an older age demographic, elevated absolute band counts, and increased CRP levels, alongside decreased platelet counts. However, no significant differences were observed in white blood cell counts, ALT, ALB, or D-Dimer levels. In addition, there were no observed differences in the incidence of CALs or IVIG resistance between patients with LKD and those with typical KD. Previous studies have similarly identified older age, elevated absolute band counts, and increased CRP levels as closely associated with LKD [13], and there remains controversy regarding differences in white blood cell counts [16, 17]. A decision-tree model developed by Kim et al. for LKD patients and BCL patients, based on data from the training cohort, included three determinants: abscess defined neck computed tomography or ultrasonography, percentage change in CRP level, and percentage change in neutrophil count [18]. Kubota et al. reported that LKD patients were older and exhibited higher CRP levels than typical KD patients, although no significant difference was observed in WBC count [8]. Jun et al. also reported that LKD patients showed advanced age, higher neutrophil count, and elevated CRP levels, but no significant difference in WBC count [16].

Older children possess a more developed mucosal immune system and exhibit a stronger inflammatory response [19]. In patients with LKD, there is a notable elevation in systemic inflammatory markers and an increased incidence of localized cervical lymphadenopathy. In our cohort, almost all LKD patients reported symptoms of neck swelling or pain, which may be partly due to the limited verbal communication abilities in younger children. Furthermore, the prevalence of the other four clinical signs in LKD patients was lower compared to those observed in typical KD patients with cervical lymphadenopathy, potentially contributing to increased rates of misdiagnosis and delayed diagnosis.

Ultrasonography (US) is the most appropriate imaging modality for initial assessment of pediatric cervical lymphadenopathy due to its widespread availability and non-invasive nature [20]. CT or MRI were not utilized in this study, as all patients were children, and ultrasound avoids radiation exposure and does not necessitate sedation. Ultrasonographic examination revealed that both the LKD group and the Other KD group exhibited multiple enlarged lymph nodes, which did not serve as a distinguishing factor between the two groups. Nevertheless, notable differences were observed in the location and maximum lymph node diameter. Specifically, LKD patients typically presented with bilateral lymphadenopathy and larger maximum nodal diameters, correlating with more prominent clinical manifestations. The cervical region contains over 300 lymph nodes, which are classified into Level I-Level VII based on the TNM classification of AJCC and UICC, 2002 [21]. The study revealed distinct patterns of CL in LKD and KD patients with CL. Specifically, LKD patients exhibited a higher prevalence of enlarged lymph nodes in the Level II region. Conversely, KD patients with CL demonstrated a notable absence of lymph node enlargement in the Level IV region. Level II lymph nodes are subdivided into Level IIa and IIb based on the on the anatomical position of accessory nerve, draining lymph from the oral cavity, nasal cavity, nasopharynx, oropharynx, hypopharynx, larynx, and parotid gland. Level IV lymph nodes are responsible for lymphatic drainage from the nasopharynx, thyroid, cervical esophagus, and larynx [7, 22]. The specific reasons for these differences remain unclear but may be related to the route of infection and regional lymphatic drainage patterns.

Although cervical lymphadenitis is one of the prominent clinical manifestations in KD, lymph node biopsy or histopathological examination is seldom conducted. Notably, in older children, lymphadenopathy often presents as an initial symptom of KD, leading to occasional biopsies to differentiate it from malignant lymphoma. A case report described a cervical lymph node biopsy performed on an 18-month-old girl who was initially suspected of having lymphoma; however, she was subsequently diagnosed with KD following the biopsy. In multivirus real-time PCR and comprehensive direct sequencing of the cervical lymph node biopsy specimen, the genomes of torque teno virus and Streptococcus spp. Genome were identified [23]. Current epidemiological and pathogenic studies indicate that multiple infectious agents may be implicated in the development of KD [24]. For example, it reported that Streptococcus pneumoniae, Mycoplasma pneumoniae, Chlamydia, adenovirus, enterovirus, parainfluenza virus, coronavirus, and Epstein–Barr virus may be associated with KD [25]. Therefore, further studies is warranted to elucidate any potential association between cervical lymphadenitis and the pathogenesis of KD.

Notably, compared to KD patients without CL, those with CL had significantly higher rates of IVIG resistance and KDSS, as well as increased percentages of neutrophils. April et al. [19] found that KD patients with CL had a stronger inflammatory response compared to those without lymphadenopathy. These findings support the hypothesis that presence of CL represents a more severe form of KD with a more intense inflammatory process. A case of KDSS initially presenting as cervical lymphadenitis has been reported, highlighting the potential for NFKD to progress to KDSS [26]. The precise mechanism underlying this progression is not yet clear. However, patients with KDSS are at an elevated risk of IVIG resistance, CAA, and aneurysm formation [27]. In our center, patients with KD complicated by CL exhibited a higher susceptibility to KDSS, which subsequently exacerbated systemic inflammatory responses and may be associated with increased incidences of IVIG resistance.

Currently, there is no consensus regarding whether cervical lymphadenopathy in KD patients constitutes a risk factor for CAL and IVIG resistance [16]. The study by Byung et al. demonstrated no significant difference in the incidence of CAA or IVIG resistance between LKD and typical KD [1], which aligns with our findings. However, after adjusting for confounding factors. This study revealed a higher rate of IVIG resistance among KD patients with CL. This observation is consistent with the findings of Choi et al. [22], who identified cervical lymphadenopathy as a risk factor for IVIG resistance. This phenomenon may not be attributable to delayed or non-standard treatment, but rather could be associated with the severity of the disease. Furthermore, this study observed a relatively high incidence of cervical lymphadenopathy, affecting 62.4% of patients. Hu et al. [28] found that plasma cytokines such as IL- 6 and TNF-α are involved in the acute-phase inflammation of KD. The levels of these inflammatory factors are notably higher in non-responders compared to IVIG responders. This is consistent with our conclusion, as this study found that the levels of ANC and CRP were significantly higher in children with KD complicated by cervical lymphadenitis compared to those with KD without cervical lymphadenitis. Cervical lymphadenopathy in KD reflects the local inflammatory response elicited by these factors and vasculitis [29, 30]. In the study, the incidence rate of IVIG resistance in KD patients with CL was 7.9%, which was higher than KD patients without CL. The IVIG resistance rates in LKD group was 5.8%, and there was no difference in the incidence rate of CALs among the all groups. This indicated the prognosis of LKD patients was relatively better, possibly due to more prominent clinical symptoms, leading to more accurate diagnoses and timely treatment, thereby reducing the incidences of CALs and the need for additional IVIG.

Finally, our findings indicate that cervical lymphadenopathy not only affected the clinical features of KD to a certain extent, but also associated effects on other tissues or areas within the neck. Grisel’s syndrome (GS), characterized by a rare, non-traumatic subluxation of the atlantoaxial joint, has been infrequently reported as a possible complication of KD [31, 32]. Current literature suggest that any neck infection or otolaryngologic surgery may contribute to the development of Grisel’s syndrome [33, 34]. Previous studies have documented cases of KD with GS [35, 36]. We identified eight cases of LKD patients with spontaneous atlantoaxial subluxation, all of which were diagnosed via atlantoaxial plain film (X-ray) or cervical CT scan. Scattered cases have also been observed in BCL patients. In pediatric patients, GS involving KD patients manifests as cervical lymphadenopathy. Approximately 50–70% of KD patients exhibit CL associated with deep neck inflammation, which may result in parapharyngeal and retropharyngeal edema and non-suppurative thrombophlebitis [37], thereby increasing the susceptibility to GS. At least one of the following abnormalities was identified on cervical X-ray or CT scan in the eight patients: alteration in physiological curvature, widening of the atlantodental interval (ADI), and asymmetry in the vertebral body-lateral atlantodental space difference (VBLADS) [38, 39] (Fig. 4). In addition to managing the primary condition, all patients underwent conservative treatment for associated complications, including the application of cervical orthosis or skull traction. Within 1 month of follow-up, all cervical symptoms resolved without any neurological dysfunction or sequelae.

Fig. 4

a Asymmetrical spacing of atlantoaxial teeth on both sides. b Differential value of atlanto-lateral clearance 32.4 mm

To date, previous studies reported cases of KD presenting with symptoms of parotitis and retropharyngeal infection [40, 41]. Consistent with these findings, we observed similar presentations in both LKD and BCL patients. It is currently hypothesized that the occurrence of parotitis in KD is associated with a cascade of inflammatory reactions triggered by the infectious source in susceptible children [42]. KD can lead to retropharyngeal hypodensity, manifesting as retropharyngeal abscess, cellulitis, oral edema, which may extend into deeper cervical spaces and cause changes in adjacent soft tissues [43, 44]. The specific pathophysiological mechanisms remain unclear, microvascular vasculitis associated with tissue edema and inflammation has been identified as a primary mechanism. Therefore, it is hypothesized that CL in KD and BCL patients intensifies local inflammatory responses in the neck, resulting in cascading effects on adjacent tissues or organs. Notably, this phenomenon appears to be more prevalent among LKD patients within the KD cohort.

This study has several limitations. First, this is a retrospective study, and we only reviewed past records, which led to the absence of some data. Potential false-positive results may arise from data processing issues, including confidence interval estimation, reliance on a single model, and subgroup heterogeneity. Therefore, the relatively small sample size may affect the results. Second, confounding factors were not fully controlled, and the diagnosis of KD and BCL in children lacked specific laboratory evidence. The methods for examining cervical lymph nodes were also relatively limited. This single-center study’s geographic scope and unadjusted confounders (e.g., SES and genetics) may limit generalizability. Future work will include multicenter follow-up and mechanistic studies on cytokines and genetics.

Conclusion

Patients with lymphadenitis in KD have simultaneous increases in systemic and liver injury indicators, accompanied by clinical manifestations such as pyuria, hepatosplenomegaly, abnormal lung imaging, as well as increased incidence of IVIG resistance and KDSS. Notably, the identification of bilateral and multifocal lymphadenopathy on ultrasound, particularly in the Level II region, characterized by the absence of liquefaction and a reduced occurrence of abscess formation, further corroborates this association. In KD cases where cervical complications and lymphadenitis are observed, clinicians should carefully consider the possibility of LKD.