Determinants of malignancy in paediatric neck lymphadenopathy—emphasis on the neutrophil-to-lymphocyte ratio

Introduction

Neck masses are common in children and most often represent benign, self-resolving conditions. The major diagnostic challenge of persistent neck lump is their wide differential spectrum, which can include congenital, inflammatory and neoplastic entities. Moreover, cervical masses can arise from any anatomical structure, including salivary glands, thyroid gland, blood vessels, and lymph nodes. A thorough history and clinical examination, along with imaging when required, usually allows refinement of the differential diagnosis (1,2). However, the need for further work-up is not always obvious based on clinical presentation (3).

Cervical lymphadenopathy is by far the leading cause of neck swelling in the paediatric population. Palpable lymphadenopathy is very common in children aged 4–8 years. It is usually bilateral and the lymph nodes tend to be smaller than 1 cm (4). Etiologically, infective or inflammatory lymphadenitis is the most common cause. Clinically useful classifications include acute versus chronic lymphadenitis, infectious versus non-infectious, and specific versus nonspecific (when no causal agent is identified) (5-7).

The main priority of the diagnostic work-up in children with cervical lymphadenopathy is identifying those that require specific treatments, and in particular, those with malignancies. While malignancies in childhood are overall rare, their incidence has been slowly increasing over the last two decades in Australia, and this trend is expected to continue. It is estimated that 5% of all paediatric malignancies occur in the head and neck, region, with lymphoma and to a lesser extent rhabdomyosarcoma as leading pathologies (8,9).

The goal of this study is to analyze a surgical paediatric cohort with cervical lymphadenopathy, in particular the diagnoses and risk factors for malignancy, with an emphasis on the neutrophil-to-lymphocyte ratio (NLR).

Methods

Included patients were those managed at the Royal Children’s Hospital (RCH) Melbourne, Australia, a tertiary paediatric hospital, between January 2017 and December 2020. The institution surgical database was queried with relevant surgical codes for lymph node biopsy of the neck. All surgeries at RCH are coded using the “Medicare Benefits Schedule”, encompassing all the medical services subsidised by the Australian government. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). This study was approved by the Research Governance and Ethics Committee, Royal Children’s Hospital Melbourne (2021/27). Due to the retrospective nature of the research, the requirement for informed consent was waived. The study is reported according to the STROBE reporting checklist (available at https://www.theajo.com/article/view/10.21037/ajo-23-64/rc).

The inclusion criteria were as follows: (I) age below 17 years; (II) surgical resection, partial or total, (incisional or excisional) of a cervical lymph node for diagnostic purposes; (III) sufficient sampling (conclusive histopathological diagnosis); (IV) baseline demographics available. Patients were excluded if, in addition to not fulfilling the inclusion criteria: (I) biopsies were performed at an external institution; (II) histopathological diagnosis was performed at an external institution; (III) biopsies performed for restaging purposes with a previously-known diagnosis; (IV) cases of acute or abscessed lymphadenitis; (V) definitive histology showing a non-nodal structure (e.g., branchial derivative).

Patients included were anonymized using a unique identifying number attributed to all patients managed at RCH. The following items were extracted from the computer-based patient records and entered in a standardised Statistical Package for the Social Sciences (SPSS) spreadsheet for further statistical analysis: age, gender, onset mode (acute ≤4 weeks, chronic >4 weeks), imaging modalities performed, blood tests performed and pertinent results [full blood count, differential blood count, lactate dehydrogenase (LDH), serologies], definitive histopathological analysis (specific diagnosis as well as dichotomized benign vs. malignant), treatment type (none, antibiotics, corticosteroids, surgery, oncological management).

Summary statistics were calculated. Ad hoc subgroup comparisons were then performed (i.e., benign vs. malignant), using Student’s t-test to compare averages and Fisher’s exact test to compare proportions. Diagnostic values were presented as specificity, sensitivity, positive and negative predictive values. Cut-off values were chosen by using receiver-operating characteristic (ROC) curves. Multivariable analysis was performed to identify independent risk factors using binomial logistic regression. All P values are two-sided and a P value inferior to 0.05 was deemed to show statistically significant differences. All figures and tables are original and were created specifically for this paper.

Results

One-hundred and two children were included in this study (Figure 1). All children underwent open lymph node biopsies that were subject to histopathological examination. The definitive histopathologic diagnoses are shown in Figure 2.

Figure 1 Workflow with screened, reviewed and included patients, with reasons for exclusion. MBS, medicare benefits schedule; MAC, mycobacterium avium complex.

Figure 2 Definitive diagnoses in children with neck lymphadenopathy. EBV, Epstein-Barr virus; MAC, mycobacterium avium complex.

Sixty percent (60%) of the children had benign conditions, with non-specific lymphadenitis representing the single most common diagnosis (30.4%). The leading diagnosis in cases of specific infectious lymphadenitis was mycobacterium avium complex (MAC) lymphadenitis (17.6%), which was confirmed by cultures, presumed based on the presence of necrotizing granulomas and acid-fast bacilli, or on histopathological examination. Malignancies were diagnosed in 39.2% of this cohort’s children, with Hodgkin (28.4%) and non-Hodgkin (8.8%) lymphomas as the most common diagnoses. The diagnostic methods used are listed in Table 1 and the main diagnoses per age group are shown in Figure 2. Over two-thirds of children underwent excisional biopsies. In three cases, non-surgical biopsy was undertaken prior to definitive histopathologic diagnosis with open biopsy. Fine-needle aspiration cytology (FNAC) was obtained only in one case, a child with MAC, correctly identifying the benign nature of the lymph node and suggesting an infectious process. Core-needle biopsies (CNBs) were performed in two cases. Of these, a lymphoma was eventually diagnosed in one case (CNB was non-diagnostic and not suggestive of malignancy), and the other case was non-specific lymphadenitis (CNB correctly identified the definitive diagnosis). In all cases, FNAC and CNB were performed under general anesthesia and coupled with an imaging study. Serologies were performed in 60% of the children.

Comparison of baseline demographics and clinical features in children with benign versus malignant lymphadenopathy, revealed that malignancies were more significantly unilateral and supraclavicular (P values 0.04 and <0.001, respectively). Age, gender and clinical presentation did not significantly differ between these groups (Table 2). We then compared the same features in children up to 5 years of age and below 5 children of age, given on one hand that MAC lymphadenitis occurred almost exclusively in children age 5 or less, along with the low incidence of malignancies seen below this age (11.5% vs. 48.7%, P=0.001, respectively; Table 2, Figure 2). Moreover, children with MAC lymphadenitis were significantly younger than those with malignancies (Figure 3A). Constitutional symptoms at presentation were significantly more common in children above 5 years of age, and in line with the significantly higher prevalence of malignancy in children above 5, supraclavicular lymphadenopathies were more common in this group. Conversely, the only feature significantly more common in children below 5 years of age was the presence of the typical skin changes (violaceous coloration) seen in MAC lymphadenitis (Table 2).

Figure 3 Laboratory results comparisons. (A) Age. (B) LDH. (C) Neutrophils. (D) Lymphocytes. (E) NLR. (F) WBC. (G) CRP. (H) ESR. Only significant differences are displayed. Data are presented as average ± SD. MAC, mycobacterium avium complex; LDH, lactate dehydrogenase; NLR, neutrophil-to-lymphocyte ratio; WBC, white blood cell; CRP, C-reactive protein; ESR, erythrocyte sedimentation rate; SD, standard deviation.

Laboratory findings among the three main diagnoses (malignancy, MAC and non-specific lymphadenitis; Figure 3A-3H) was compared; illustrated in Figure 3C-3E. Patients with MAC and non-specific lymphadenitis had significantly higher lymphocyte counts (P<0.001 and P=0.04 respectively) and lower neutrophil counts (P<0.01 for both groups) compared to the patient group with malignancy [mean 2.1, standard deviation (SD) 1.4]. Consequently, the NLR was significantly higher in the malignancies group (mean 4.2, SD 4.1) compared to patients with MAC (mean 1, SD 0.7, P<0.01) or non-specific lymphadenitis (mean 1.7, SD 1.9, P<0.01).

Dedicated preoperative neck imaging was available for 75% of this cohort’s children, and 18.6% of the children underwent two different imaging modalities for the neck (Table 3). The most commonly performed imaging modality was chest X-ray (CXR) (63.7% patients), which was pathological in 21 cases (32.3% of X-rays performed), usually showing a mediastinal enlargement and/or prominent hilar lymph nodes. A pathological CXR was significantly associated with malignancy (6.1% vs. 59.4%, P<0.001). Neck ultrasound was available in 55.9% of the patients. The minimal size in benign versus malignant lymphadenopathy was significantly different (12 vs. 19 mm respectively, P<0.01). No imaging features of computed tomography (CT)-scan and magnetic resonance imaging (MRI) correlated with malignancy (Table 3). In general, all imaging modalities showed lower specificity and positive predictive value than sensitivity and negative predictive values. MRI showed the highest sensitivity and negative predictive values at 100% (Table 3). Imaging and serologies done after nodal biopsies were not recorded for this study, as the focus was on diagnosis.

Antibiotics were administered preoperatively in 33/102 (32.4%) of the patients, leading to clinical improvement with reduction in lymph node size in one child whose final diagnosis was Hodgkin lymphoma. After histopathological diagnosis, all patients with malignancies (39%) received chemotherapy and 10.8% underwent radiation therapy. Antibiotics were administered after nodal biopsies in 23.5% of the patients and another 7% received other medications [corticosteroids =3, naproxen =2, cyclophosphamide =1 and granulocyte colony-stimulating factor (G-CSF) =1]. No further treatment was required in 25.5% of the children.

A multivariable logistic regression analysis was performed to assess the effect of gender, supraclavicular location, nodal laterality, age and nodal localization and laterality, CXR findings and neutrophil to lymphocyte ratio (NLR) on the likelihood of malignancy. Multicollinearity assessment showed a significant interaction between pathological CXR and presence of supraclavicular lymphadenopathy and thus CXR was removed from the final model. An NLR threshold of 1.5 was determined by ROC analysis. Results showed that both supraclavicular location [odds ratio (OR) 9.2, 95% confidence interval (CI): 2.7–31.5, P≤0.001] and an NLR >1.5 (OR 7.4, 95% CI: 2.2–25.2, P=0.001) were significant independent risk factors for malignancy.

Discussion

In this study, the clinical, radiological, and laboratory findings in a cohort of children presenting with cervical lymphadenopathy who underwent diagnostic surgical biopsies was reviewed. The main findings were as follows: (I) although reactive or non-specific lymphadenitis is the most common diagnosis, there was a very substantial rate of malignancies (40%); (II) MAC lymphadenitis is the most common diagnosis in children younger than five years of age, while malignancies (lymphoma) are most common after this age, particularly in teenagers; (III) imaging studies have limited diagnostic performance to distinguish between benign and malignant lymphadenopathies; (IV) the NLR seems to be the most useful marker to distinguish between malignant and benign entities, while classical markers such as LDH showed little diagnostic performance for this indication; and (V) a supraclavicular location along with an NLR superior to 1.5, are significant independent risk factors for malignancy.

Cervical lymphadenopathy in children is a common problem that tends to cause diagnostic uncertainty due to the broad range of differentials. A recent systematic review by Deosthali et al. (10) compiled data of seven studies from five countries (none of which were Australian cohorts) and found two thirds of paediatric cervical lymphadenopathy cases did not have an identifiable etiology. Epstein-Barr virus (8.89%), malignancy (4.69%) and granulomatous disease (4.06%) were the next most common etiologies. Of malignant conditions, non-Hodgkin lymphoma was the most commonly identified, followed by Hodgkin lymphoma. Our study’s findings were consistent with these cohorts as a large proportion of patients were found to have non-specific lymphadenitis (30.4%). However, malignancy of all causes comprised a greater proportion of our cohort (39.2%) compared with Deosthali et al. (10). This is certainly due to this study inclusion criteria requiring resection or biopsy for confirmation of diagnosis and therefore inherent selection bias, as encountered in other studies, increasing the proportion of malignancies (7,11,12).

While the criteria guiding the need for a histological diagnosis for a paediatric neck lymphadenopathy are not clearly standardized, a number of studies have attempted to define the profile of high-risk patients. In particular, a systematic review by Locke et al. (13) revealed that neither clinical presentation (i.e., fever, weight loss and lethargy) nor supplementary laboratory or radiological studies help predicting the need for open biopsy. In the RCH cohort, supraclavicular location was the only clinical feature that was a significant independent risk factor for malignancy.

In line with this, an interesting observation from the RCH cohort was that the stereotypical presentation with B-type symptoms (fever, weight loss, night sweats) and raised LDH levels was not more common in patients diagnosed with lymphomas than in young children with MAC. With this in mind, a single lymphadenopathy or nodal conglomerate in an infant with constitutional symptoms should be considered as warranting surgical resection if possible. As indicated by previous studies, the preferred management for MAC is surgical resection, given the reduced efficacy of antibiotics for this indication (14,15). Even in the case of lymphomas, surgical resection of singular, or otherwise non-disseminated lesions, can potentially reduce the need for adjuvant therapy.

It is important to underline that previous studies performed in different geographic settings do not necessarily reflect the regional epidemiological features of Australia. This is particularly relevant when it comes to MAC lymphadenitis. Lymphadenitis is the most common manifestation of non-tuberculous mycobacterial disease in healthy, immunocompetent children, particularly those aged 1–5 years old (16,17). The diagnosis of MAC lymphadenitis tends to be tedious given the need for prolonged culture and the variable diagnostic accuracy of polymerase chain reaction (PCR). As such, in case of persistent unilateral lymphadenitis, particularly submandibular, with typical discoloration or discharge, a presumptive diagnosis can usually be made in regions with relatively high incidences (18).

The most interesting finding was the diagnostic value of NLR and in particular its ability to distinguish between benign and malignant neck lymphadenopathy in children. NLR is a biomarker which is calculated through differential blood counts and therefore does not require additional clinical resources in a paediatric population. NLR is a surrogate marker of systemic inflammation, and its significance, primarily as a marker of poor prognosis, has been comprehensively assessed in patients with head and neck squamous cell carcinomas (19). In contrast however, its diagnostic significance, particularly in the setting of lymphomas, remains largely unexplored (20). Reference normal values for NLR increase with age due to a physiological increase in neutrophils, with NLR 50^th centile values increasing from 0.99 to 1.76 from the ages 3–18 years of age (21). Through our multivariate analysis, we found that an NLR greater than 1.5 was a significant independent risk factor for malignancy (OR 7.4). The age range in the RCH cohort was not significantly different between benign and malignant groups and when comparing the mean age of each group to the Moosmann et al. reference normal values for age groups, the 50^th centile for NLR is below 1.5 for both males and females (21). Kaplama et al. (22) recently reported a significantly higher median NLR in patients with lymphoma than in those with reactive lymphadenopathy (3.5 vs. 1.7, respectively). A multivariate analysis was not performed in that paper. Comparatively, this study found that a value slightly above the median for benign conditions was a good threshold to distinguish between benign and malignant conditions. The authors suggest that in case of persistent, unilateral and/or supraclavicular lymphadenopathies, especially in children above 5 years of age, the NLR be taken into account to establish the indication for surgical biopsies in children with cervical lymphadenopathies. Further observational, multicentric studies are warranted to establish the role of NLR as a diagnostic biomarker in children with cervical lymphadenopathy.

The main limitations of this study are its retrospective design and the selection bias inherent to the surgical nature of the cohort. The retrospective design had an impact on inconsistent documentation of clinical histories and sometimes clinical findings. This was not an issue regarding complementary laboratory tests and imaging findings, which were systematically documented and kept in the electronic medical records.

Although most children with a neck lymphadenopathy do not undergo tissue sampling, the nature of the cohort was dictated by the need for a histopathological gold-standard. While the findings are consistent with previously reported studies, it is important underlining that this cohort certainly underestimates the diagnostic value of serologies, given that when a serological diagnosis is made, most patients do not require confirmatory nodal biopsies. It is estimated that definitive serological diagnosis can be made in approximately 10% of children with cervical lymphadenopathy (13). Therefore, the authors reiterate the importance of further studies to determine the value of NLR in all comers.

Conclusions

The likelihood of malignancy in paediatric lymphadenopathy increases with a supraclavicular location and a NLR greater than 1.5. There are other supportive factors including unilaterality, constitutional symptoms and mediastinal abnormality on CXR. The most common malignancy in the cohort is lymphoma and therefore any tissue investigation is to be directed towards ruling in or confirming the diagnosis of lymphoma and their subtype. The younger child under 5 is more likely in this population to have an infectious cause such as atypical mycobacterial lymphadenitis.

Acknowledgments

The abstract was presented as a poster at the summer congress of the Swiss Society of Otorhinolaryngology-Head and Neck Cancer (Lausanne, June 2022).

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://www.theajo.com/article/view/10.21037/ajo-23-64/rc

Data Sharing Statement: Available at https://www.theajo.com/article/view/10.21037/ajo-23-64/dss

Peer Review File: Available at https://www.theajo.com/article/view/10.21037/ajo-23-64/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://www.theajo.com/article/view/10.21037/ajo-23-64/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). This study was approved by the Research Governance and Ethics Committee, Royal Children’s Hospital Melbourne (2021/27). Because of the retrospective nature of the research, the requirement for informed consent was waived.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Lee J, Fernandes R. Neck masses: evaluation and diagnostic approach. Oral Maxillofac Surg Clin North Am 2008;20:321-37. [Crossref] [PubMed]
2. Riva G, Sensini M, Peradotto F, et al. Pediatric neck masses: how clinical and radiological features can drive diagnosis. Eur J Pediatr 2019;178:463-71. [Crossref] [PubMed]
3. Jackson DL. Evaluation and Management of Pediatric Neck Masses: An Otolaryngology Perspective. Physician Assist Clin 2018;3:245-69. [Crossref] [PubMed]
4. Park YW. Evaluation of neck masses in children. Am Fam Physician 1995;51:1904-12. [PubMed]
5. Chang SSY, Xiong M, How CH, et al. An approach to cervical lymphadenopathy in children. Singapore Med J 2020;61:569-77. [Crossref] [PubMed]
6. Friedmann AM. Evaluation and management of lymphadenopathy in children. Pediatr Rev 2008;29:53-60. [Crossref] [PubMed]
7. Rosenberg TL, Nolder AR. Pediatric cervical lymphadenopathy. Otolaryngol Clin North Am 2014;47:721-31. [Crossref] [PubMed]
8. Dickson PV, Davidoff AM. Malignant neoplasms of the head and neck. Semin Pediatr Surg 2006;15:92-8. [Crossref] [PubMed]
9. Youlden DR, Baade PD, Green AC, et al. The incidence of childhood cancer in Australia, 1983-2015, and projections to 2035. Med J Aust 2020;212:113-20. [Crossref] [PubMed]
10. Deosthali A, Donches K, DelVecchio M, et al. Etiologies of Pediatric Cervical Lymphadenopathy: A Systematic Review of 2687 Subjects. Glob Pediatr Health 2019;6:2333794X19865440.
11. Celenk F, Baysal E, Aytac I, et al. Incidence and predictors of malignancy in children with persistent cervical lymphadenopathy. Int J Pediatr Otorhinolaryngol 2013;77:2004-7. [Crossref] [PubMed]
12. Vural S, Genc DB, Celikboya E. Clinical Characteristics of and Cancer Incidence in Children Evaluated for Lymphadenoapthy Referred to Pediatric Oncology Clinics. Sisli Etfal Hastan Tip Bul 2020;54:222-6. [PubMed]
13. Locke R, Comfort R, Kubba H. When does an enlarged cervical lymph node in a child need excision? A systematic review. Int J Pediatr Otorhinolaryngol 2014;78:393-401. [Crossref] [PubMed]
14. Flint D, Mahadevan M, Barber C, et al. Cervical lymphadenitis due to non-tuberculous mycobacteria: surgical treatment and review. Int J Pediatr Otorhinolaryngol 2000;53:187-94. [Crossref] [PubMed]
15. Lindeboom JA. Surgical treatment for nontuberculous mycobacterial (NTM) cervicofacial lymphadenitis in children. J Oral Maxillofac Surg 2012;70:345-8. [Crossref] [PubMed]
16. Griffith DE, Aksamit T, Brown-Elliott BA, et al. An official ATS/IDSA statement: diagnosis, treatment, and prevention of nontuberculous mycobacterial diseases. Am J Respir Crit Care Med 2007;175:367-416. [Crossref] [PubMed]
17. Pham-Huy A, Robinson JL, Tapiéro B, et al. Current trends in nontuberculous mycobacteria infections in Canadian children: A pediatric investigators collaborative network on infections in Canada (PICNIC) study. Paediatr Child Health 2010;15:276-82. [Crossref] [PubMed]
18. Haimi-Cohen Y, Markus-Eidlitz T, Amir J, et al. Long-term Follow-up of Observation-Only Management of Nontuberculous Mycobacterial Lymphadenitis. Clin Pediatr (Phila) 2016;55:1160-4. [Crossref] [PubMed]
19. Takenaka Y, Oya R, Kitamiura T, et al. Prognostic role of neutrophil-to-lymphocyte ratio in head and neck cancer: A meta-analysis. Head Neck 2018;40:647-55. [Crossref] [PubMed]
20. Tezol O, Bozlu G, Sagcan F, et al. Value of neutrophil-to-lymphocyte ratio, monocyte-to-lymphocyte ratio, platelet-to-lymphocyte ratio and red blood cell distribution width in distinguishing between reactive lymphadenopathy and lymphoma in children. Bratisl Lek Listy 2020;121:287-92. [Crossref] [PubMed]
21. Moosmann J, Krusemark A, Dittrich S, et al. Age- and sex-specific pediatric reference intervals for neutrophil-to-lymphocyte ratio, lymphocyte-to-monocyte ratio, and platelet-to-lymphocyte ratio. Int J Lab Hematol 2022;44:296-301. [Crossref] [PubMed]
22. Kaplama ME, Güneş AK, Erden B. Evaluation of the predictive role of neutrophil/lymphocyte ratio in the diagnosis of lymphoma in patients with asymptomatic and ısolated cervical lymphadenopathy. Braz J Otorhinolaryngol 2021;87:210-6. [Crossref] [PubMed]