Cervical lymphadenopathies in children: A prospective clinical cohort study

Cervical lymphadenopathies in children: A prospective clinical cohort study

International Journal of Pediatric Otorhinolaryngology 82 (2016) 81–87 Contents lists available at ScienceDirect International Journal of Pediatric ...

395KB Sizes 2 Downloads 42 Views

International Journal of Pediatric Otorhinolaryngology 82 (2016) 81–87

Contents lists available at ScienceDirect

International Journal of Pediatric Otorhinolaryngology journal homepage: www.elsevier.com/locate/ijporl

Cervical lymphadenopathies in children: A prospective clinical cohort study Serdar Bozlak a, Muhammet Ali Varkal b, Ismail Yildiz b, Sadik Toprak c, Serap Karaman d, Oguz Bu¨lent Erol e, Ensar Yekeler e, Ays¸egul Unuvar d, Ayse Kilic b, Fatma Oguz f, Emin Unuvar b,* a

Istanbul University, Istanbul Faculty of Medicine, Department of Pediatrics, Istanbul, Turkey Istanbul University, Istanbul Faculty of Medicine, Department of Pediatrics, Division of Ambulatory Pediatrics, Istanbul, Turkey c Bulent Ecevit University, Faculty of Medicine, Department of Forensic Medicine, Zonguldak, Turkey d Istanbul University, Istanbul Faculty of Medicine, Department of Pediatrics, Division of Pediatric Hematology & Oncology, Istanbul, Turkey e Istanbul University, Istanbul Faculty of Medicine, Department of Radiology, Division of Pediatric Radiology, Istanbul, Turkey f Istanbul University, Institute of Child Health, Istanbul, Turkey b

A R T I C L E I N F O

A B S T R A C T

Article history: Received 3 September 2015 Received in revised form 31 December 2015 Accepted 3 January 2016 Available online 12 January 2016

Aim: Cervical lymphadenopathy (LAP) is a common sign and may raise fears about serious illnesses. The aim of our study was to evaluate the patients with cervical LAPs in a general pediatrics clinic setting, and to evaluate follow-up results for potential causes and risk factors for malignancies. Material and methods: Two hundred-eighteen patients aged between 79.4  46.7 months with LAP were enrolled in this prospective cohort study. The patients were examined in terms of demographics, clinical, radiologic and serologic aspects like Epstein–Barr virus (EBV), cytomegalovirus (CMV), parvovirus B19. A lymph node biopsy was performed in selected patients. The patients were followed-up for 8 weeks and risk factors for malignancy were evaluated. Results: Seventy patients (41.3%) had specific etiology and 6 (2.7%) had malignant causes. The causes were as follows: 27% (n = 59) infections; 2.7% (n = 6) malignancies; 11.4% (n = 25) other causes. EBV was responsible for 27% of infectious causes. The other common infectious etiologies were CMV 4.3%, parvovirus B-19 2.9%, and group-A beta-hemolytic streptococcus (GAS) 10.8%. Four of the six malignancies were lymphomas. Predictive factors for malignancy were having LAP larger than 30 mm, rubbery lymph node, high serum CRP and LDH values, no hilum in ultrasonography, and enlargement of lymph node in follow-up. High uric acid levels and leucopenia were also common in the malignancy group. Conclusion: Etiology of cervical LAPs was diagnosed in 41.3% patients. Infectious causes were the most common cause with 27%. Malignancy was diagnosed in 2.7% and lymphoma was the most common malignancy. ß 2016 Elsevier Ireland Ltd. All rights reserved.

Keywords: Child Lymphadenopathy Malignancy

1. Introduction Lymphadenomegaly is a palpation examination finding of an enlarged lymph node. The term lymphadenopathy (LAP) is generally synonymously used because diseases often cause

* Corresponding author. Tel.: +90 0212 414 20 00x31668. E-mail addresses: [email protected] (S. Bozlak), [email protected] (M.A. Varkal), [email protected] (I. Yildiz), [email protected] (S. Toprak), [email protected] (S. Karaman), [email protected] (O.B. Erol), [email protected] (E. Yekeler), [email protected] (A. Unuvar), [email protected] (A. Kilic), [email protected] (F. Oguz), [email protected] (E. Unuvar). http://dx.doi.org/10.1016/j.ijporl.2016.01.002 0165-5876/ß 2016 Elsevier Ireland Ltd. All rights reserved.

enlarged lymph nodes [1]. LAP of the head and neck is a common finding in pediatric ambulatory and emergency departments [1,2]. Lymph nodes react to a new antigen with hyperplasia, which is more excessive in children [3–5]. LAP is often caused by infectious diseases, but malignancies, autoimmune diseases, and chronic inflammatory processes may also be the reason. Mostly self-limited acute upper respiratory viral infections in children may lead to LAP, and these are regresses spontaneously. Some infections may also cause chronic LAPs. Pediatricians differentiate between malignancy and benign causes with clinical and laboratory findings [3,4]. Malignancies that present with LAP may be primary or metastasis. Malignant lymph nodes may enlarge in a rash or with a slow course. Hodgkin’s lymphoma (HL)

82

S. Bozlak et al. / International Journal of Pediatric Otorhinolaryngology 82 (2016) 81–87

shows slow progress in lymph node size, but Non-Hodgkin’s lymphoma (NHL) is abrupt [5]. Some clues may be helpful to differentiate benign causes from malignant. LAP located supraclavicular is mostly malign; whereas LAP with concomitant maculopapulary rash is often benign [5–7]. No decrease in size in the follow-up, organomegaly, leucocytosis, leucopenia, high erythrocyte sedimentation rate (ESR), Lactate dehydrogenase (LDH) and uric acid levels may indicate malignancy [8,9]. The aim of our study was to evaluate the patients with cervical LAPs in a general pediatrics clinic setting, and to evaluate followup results for potential causes and risk factors for malignancies. 2. Methods This prospective single-center-cohort clinical study was performed in children who visited Istanbul University, Istanbul Medical Faculty, Ambulatory Clinics of Department of Pediatrics. Informed consent was obtained from all participants’ parents. Approval of the Research Ethics Committee of Istanbul University was obtained. This study was supported by Scientific Research Projects Coordination Unit of Istanbul University (project number: 35906). 2.1. The study group and inclusion criteria Children with LAP of the neck and head who had given informed consent were included in the study group. Lymph nodes that were enlarged more than 10 mm in the cervical, submental, and submandibular region, and 5 mm in the suboccipital, preauriculary and postauriculary region, were accepted as LAP. Any palpable lymph node in the supraclaviculary region was accepted as LAP [3,6,9]. LAP that lasts less than 4 weeks was termed acute LAP and 4 weeks or more as chronic LAP. Patients with any swelling other than lymph node, chronic disease, who were treated or diagnosed in another clinic, or had lymph node enlargement that did not fulfill our LAP criteria were not included in the study. An anamnesis form were used to collect the following information: age, sex, start date of symptoms, history of travel, animal contact, tuberculosis contact, immunization, B symptoms for lymphoma, medication causing LAP (e.g., phenytoin, allopurinol, valproic acid), dental symptoms and history of upper or low respiratory tract infections. LAP was classified depending on location, size, and existence of inflammation findings (redness on the skin, pain on palpation, tenderness and increased heat on the skin), relationship with peripheral tissue, like as fixed or not, and consistency. LAP was grouped depending on its consistency as soft, rubbery or hard. Other signs like arthritis, splenomegaly or rash were noted. B symptoms include: fever greater than 38 8C with no known cause, drenching night sweats, and unintentional weight loss of more than 10% of body weight over a period of 6 months or less. Presence of one was accepted as B symptom positive. Patients left study for any reason before 8 weeks were excluded. 2.2. Laboratory tests Investigation of hemogram (CBC), C-reactive protein (CRP), ESR, LDH, serum uric acid level, and blood smear were routinely performed all patients. Other diagnostic tests such as throat cultures, viral serological analysis, ultrasonography of LAP, tuberculin skin test, bone marrow aspiration, and excisional biopsy were undertaken if necessary. Patients were followed-up for 8 weeks after the diagnosis. If there was no suspicion of malignancy after the first-step evaluation, we treated patients with appropriate antibiotics against probable Gram-positive bacteria with as amoxicillin, amoxicillin–clavulanate, cefuroxime or ampicillin–sulbactam for

14 days and called them for a follow-up check afterwards. If there was no regression after two weeks, the patients underwent further examinations like tuberculin skin test, ultrasonography, chest radiography, and serology for EBV, CMV, Parvovirus B19, and Mycoplasma pneumonia. If there was still no suspicion for malignancy after reevaluation, we followed-up the patient routinely. In the event that we found a risk for malignancy, we contacted the Division of Pediatric Hematology–Oncology of Istanbul Medical Faculty for additional opinions. All biochemical analyses were undertaken in the Department of Biochemistry of Istanbul Medical Faculty, Istanbul University. Venous blood samples were taken in appropriate tubes and centrifuged. Uric acid, CRP, and LDH enzyme activities were measured using a Cobas Integra 800 analyzer (Roche Diagnostics GmbH, Mannheim, Germany). Hemograms were measured using an LH 780 (Beckman Coulter, USA) and ESR with THERMA NE (Linear Chemicals S.L, Spain). All serologic analyses were studied in specific ELISA kits [10,11]. 2.3. Radiologic investigations All radiologic images were taken in the Department of Radiology, Istanbul Medical Faculty, Istanbul University. Ultrasonography for LAP was performed with the patient in the supine position with their neck slightly extended. A Logiq 9 with 13–5 mHz probe (GE Healthcare, Milwaukee, WI, USA) was used. Bilateral submandibular, submental, postauriculary and preauriculary nodes, posterior cervical triangle, midjugulary chain area and lower jugulary chain area were checked. Long axis/short axis ratio of lymph node (L/S), echogenicity of hilum, vascular pattern were checked. If L/S around 1, and on echogenicity showing deformation, irregular margins, and chaotic vascular patterns were determined, these are defined as compatible with malignancy [12–15]. 2.4. Histopathologic evaluation Lymph node excisional biopsies were performed under general anesthesia [16–18]. Pathologic examination of bone marrow and lymph node tissues were performed in the Department of Pathology in Istanbul Medical Faculty, Istanbul University. All specimens were evaluated by a senior pathologist expert on hematology. Bone marrow aspiration was done in some patients with any clinical symptoms considered acute lymphoblastic leukemia or lymphoma. We had not any complication for excisional lymph node biopsies. Average time referral to biopsy and diagnosis were 2–3 weeks. 2.5. Statistical analysis Number Cruncher Statistical System 2007 (NCSS) and Power Analysis and Sample Size (PASS) 2008 statistical software 2008 were used for statistical analyses. In the evaluation of the qualitative and quantitative parameters, descriptive methods (mean, standard deviation, median, and frequency) were used. Mann–Whitney U test was used for quantitative parameters with non-normal distribution. Fisher’s exact test and Fisher–Freeman– Halton test were used for qualitative evaluation. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were assessed for diagnostic tests. Risk factors for malignancy were estimated with Multivariate logistic regression analysis. Significance were taken as p < 0.01 and p < 0.05. 3. Results Two hundred eighteen patients were enrolled in the study (Fig. 1). Among all groups, 34.4% (n = 75) were girls, and the mean age of all groups was 79.4  46.7 months (range, 13–215 months)

S. Bozlak et al. / International Journal of Pediatric Otorhinolaryngology 82 (2016) 81–87

83

323 patients were found appropriate to our criteria.

46 patients gave no informed consents

277 patients were taken to the study

32 patients left the study due to patients’ families’ decision. 27 patients did not come to controls.

Study course were finished with 218 patients.

128 patients had no defined cause of LAP.

59 patients had specific infectious causes. 25 patients had non-infectious benign causes.

4 cases had lymphoma. 1 case had acute lymphoblastic leukemia. 1 case had hemaphagocytic lymphohistiocytosis. Fig. 1. Flow chart of the study (n = 218).

(Table 1). One hundred twenty-eight patients (58.7%) had had acute symptoms, and 41.3% (n = 90) had chronic symptoms. LAPs were found in the following locations: 63.7% pre-cervical, 19.3% submandibular, 6.9% post-cervical, 9.2% retroauricular, and 0.9%

supraclavicular. Patients were divided into two groups: patients with no malignancy on final diagnosis (group 1; n:212), and patients with malignancy (group 2; n:6). Clinical and laboratory test results were compared between the two groups (Fig. 2).

Table 1 Descriptive analyses of clinical findings of the groups: group-1: patients diagnosed benign lymphadenopathy (LAP) (n = 212), group-2: patients diagnosed malign LAP (n = 6). Variables

Total group (n = 218)

Group-1 (n = 212)

Group-2 (n = 6)

Statistical significance p

Age (month), mean  SD; (median)

79.4  46.7 (70.5)

77.8  45.4 (70)

134.2  60.7 (121)

0.013

Gender, female, n (%) LAP size mm; mean  SD; (median)

75 (34.4%) 24.6  5.3 (25)

73 (34.4%) 24.5  5.2 (24.5)

2 (33.3%) 29.2  5.9 (30.5)

NS 0.06

LAP, onset of symptoms n (%) Acute 4 weeks Chronic >4 weeks

128 (58.7) 90 (41.3)

125 (59.0) 87 (41.0)

3 (50) 3 (50)

NS

LAP, localization (%) Pre-cervical Submandibular Post-cervical Retroauricular Supraclavicular

139 (63.7) 42 (19.3) 15 (6.9) 20 (9.2) 2 (0.9)

136 (64.8) 42 (20.0) 14 (6.7) 20 (9.5) 0

3 0 1 0 2

NS

History of infectious disease, n; (%) LAP, immobile, fixed, n (%)

100 (45.9) 23 (10.6)

98 (46.2) 20 (9.4)

2 (33.3) 3 (50.0)

NS 0.017

LAP, consistency, n; (%) Hard Rubbery Soft

16 (7.3) 34 (15.6) 168 (77.1)

14 (6.6) 31 (14.6) 167 (78.8)

2 (33.3) 3 (50) 1 (16.7)

0.002

5 (83.3)

0.001

B symptoms, n; (%) NS: non-significant statistically.

9 (4.1)

4 (1.9)

S. Bozlak et al. / International Journal of Pediatric Otorhinolaryngology 82 (2016) 81–87

84

Simple main test (n: 212) (CBC, CRP), ESR, LDH, uric acid)

Posive serologic test (n:62 )

EBV (n:37), CMV (n:5), Parvovirus B-19 (n:3), M. pneumoniae (n:6), GAS (n:11)

Chest X-ray (n: 67)

Pneumonic infiltration (n:9) Widened mediastinum (n:1)

USG evaluaon (n:111)

Caseous necrosis (n: 3) Malignant image (n:2)

Group 1 (n: 212)

Lymph node biopsies (n:18 )

Bone marrow aspiraon (n:4 )

Study group (n: 218)

Simple main test (n: 6) (CBC, CRP), ESR, LDH, uric acid)

Posive serology (n: 1 )

EBV (n:1)

Chest X-ray (n: 6)

Widened mediastinum (n:2)

USG evaluaon (n:5)

Malignant image (n:4)

Group 2 (n: 6)

Lymph node biopsies (n:5 )

Bone marrow aspiraon (n:6 )

Hodgkin lymphoma (n:2) Non-Hodgkin lymphoma (n:2) Hemaphagocitic lymphohistiositosis (n:1) Acute leukemia (n:1) Lymphoma (n:2) Malignant histiocytosis (n:1)

Fig. 2. Patients progression on study groups (n:218).

The mean age of the patients of the malignant group was older than that of the benign group (p < 0.05). The male/female ratio was similar between the two groups. Mean lymph node size was 24.5  5.2 mm in group 1 and 29.2  5.9 mm in group 2. Lymph node size in group 2 was larger, but there was no statistical significance (p > 0.05). A cut-off size of 30 mm showed a sensitivity of 66% and specificity of 80% for malignancy. One hundred (45.9%) patients had a history of infection; there was no difference between the groups. Twenty-three (10.6%) patients had fixed- and 195 (89.4%) had mobile lymph nodes on palpation. Inmobile lymph nodes were seen more often in group 1 than group 2; the difference was statistically significant (p < 0.05). The consistency of lymph nodes were 7.3% (n = 16) hard, 15.6% (n = 34) rubbery, and 77.1% (n = 168) generally soft. When the two groups were compared, we found that the lymph nodes in group 1 were hard in 6.6% (n = 14), rubbery in 14.6% (n = 31), and soft in 78.8% (n = 167). In group 2, they were hard in 33.3% (n = 2), rubbery in 50% (n = 3), and soft in 16.7% (n = 1). Lymph nodes in group 1 were more mostly soft, whereas they were more frequently rubbery in group 2. There was a statistically significant difference

in the consistency of lymph nodes between the groups (p < 0.01). Palpation of a rubbery lymph node has a sensitivity of 50% and specificity of 85.3% for malignancy (odds ratio [OR]: 5.83; 95% CI: 1.13–30.25). B symptoms were found in 9 patients (4.1%). Existence of B symptoms were more common in group 2, which was statistically significant (p < 0.01). Sensitivity and specificity of B symptoms for malignancy were 83.3% and 98.1%, respectively (OR: 260; 95% CI: 24.45–2764.7). 4. Biochemical evaluation The percentage of the patients with a hemoglobin (Hb) level under 11.5 g/dL was 18% (n = 33) and 50% (n = 3) in groups 1 and 2, respectively. The mean Hb in group 1 was 12.3  1.0 g/dL and 11.0  1.1 g/dL in group 2. The ratio of patients with anemia higher in group 2 than group 1 (p < 0.05). ESR was similar in both groups (Table 2). Some 3.6% (n = 7) of patients had a white blood cell (WBC) count under 4500/mm3, and 9.4% (n = 18) above were 15,000/mm3. Of all

S. Bozlak et al. / International Journal of Pediatric Otorhinolaryngology 82 (2016) 81–87

85

Table 2 Analyses of laboratory tests of the groups: group-1: patients diagnosed benign LAP (n = 212), group-2: patients diagnosed malignancy (n = 6). Variables (mean  SD; (median)),

Total group (n = 218)

Group-1 (n = 212)

Group-2 (n = 6)

Statistical significance p

Leucocyte (/mm3)

10,520  7179 (9590)

10,140  3848 (9590)

22,296  35,338 (8360)

0.56

Hemoglobin (g/dL)

12.3  1.1 (12.2)

12.3  1.0 (12.30)

11.0  1.1 (11.35)

0.01

Thrombocytes (/mm3)

293,943  107,380 (300,500)

295,475  106,400 (302,000)

239,833  138,096 (202,000)

0.18

ESR (mm/h)

23.75  21.43 (15)

23.26  21.03 (15)

38.17  29.84 (34)

0.21

CRP (mg/L)

11.7  26.1 (0.84)

10.7  25.7 (0.77)

39.4  25.7 (33)

0.001

LDH (IU/L)

536.5  433.7 (460)

488.1  140.7 (458.5)

1861.0  1879.1 (1289.5)

0.034

Uric acid (mg/dL)

3.7  1.3 (3.5)

3.6  1.1 (3.5)

6.4  4.1 (5.1)

0.72

the patients, 2.1% (n = 4) had a platelet count under 150,000/mm3 and 15.1% (n = 29) were above 400,000/mm3. The number of patients with thrombocytopenia was statistically significantly higher in group 2 (p < 0.01). There was no difference in total WBC count between the groups. Groups 1 and 2 had mean lymphocyte counts 4053.9  2014.5/mm3 and 1854.0  1258.3, respectively. Group 1 had a significantly higher lymphocyte count (p < 0.01). Normal blood smears were observed in 77.3% (n = 150), 20.1% (n = 39) had Downey cells, 2.6% (n = 4) had atypical cells, and 0.5% (n = 1) eosinophilia. Normal blood smears were seen in 78.7% (n = 148) of group 1 and 33.3% (n = 2) of group 2. The higher frequency of normal blood smear results found in group 1 compared with group 2 was statistically significant (p < 0.01). CRP above 5 mg/L was 27.2% (n = 49) in group 1; the remainder of group 1 had normal CRP (72.8%; n = 131). CRP was high in all patients in group 2 (n = 6). Group 2 had a statistically significant higher mean CRP value than group 1, 39.4  25.7 mg/L and 10.7  25.7 mg/L, respectively (p < 0.05). The CRP value showed 100% sensitivity and 72.7% specificity for malignancy. Patients with high LDH level (normal range: 135–435 U/L) above 435 U/L were 56.1% (n = 92) in group 1 and 83.3% (n = 5) in group 2. The mean LDH value (p = 0.034) was significantly higher in group 2 (1861.0  1879.1) than in group 1 (488.1  140.7). Patients with high uric acid level (normal range: 1.1–6.4 mg/dL) above 6.4 mg/dL were 0.6% (n = 1) in group 1 and 20% (n = 1) in group 2. There was no statistical significance between groups (p = 0.072; p > 0.05). 5. Serologic evaluation EBV VCA IgM positivity of all patients was 27.1% (n = 38) and there was no statistical difference between the groups. Serologic positivity of all patients were as follows; 4.3% (n = 5) CMV IgM, 2.9% (n = 3) Parvovirus IgM, 5.8% (n = 6) Mycoplasma pneumonia IgM, and 10.8% (n = 11) throat culture positivity for group A streptococci, (p > 0.05) (Table 3). 6. Radiologic results Some 33.5% of patients had a chest X-ray (n = 73). Of the X-rays, 83.6% (n = 61) showed no pathology, 12.3% (n = 9) had pneumonic infiltration, and 4.1% (n = 3) had a widened mediastinum. There were more patients with a widened mediastinum in Group 2 (p < 0.05). One hundred sixteen patients (53.2%) underwent

ultrasonography. Ultrasonography showed reactive small nodes in 107 of patients (92.2%), malignant LAP in 6 patients (5.2%), and caseous necrosis in 3 patients (2.6%). Malignant lymph node images were found in 4 patients of group 2. Group 2 had a higher ratio of ultrasonography requirement and hilum pathology was more common in group 2 (p < 0.01). 6.1. Histopathologic examination Twenty-three lymph node biopsies (10.5% of all patients) were performed. Eighteen (8.4%) of the 212 patients in group 1 underwent a biopsy, whereas this was 83% (n = 5) in group 2. Ten (4.6%) patients underwent bone marrow aspiration. The results of malignancy were as follows: 1 case of acute leukemia, 4 cases of lymphoma, and 1 case of malignant histiocytosis (Table 4). 6.2. Follow-up period We reassessed patients after 8 weeks. Palpable lymph nodes had disappeared in 92.2% (n = 201) of patients in follow-up. The size of 5% (n = 11) of patients’ LAP had not changed after two weeks and 2.7% (n = 6) of patients’ LAP had enlarged. The number of patients with a larger LAP in follow-up was higher in group 2, which was statistically significant (p < 0.01).

Table 3 Serology tests results of the groups: group-1: patients diagnosed benign LAP (n = 212), group-2: patients diagnosed malignancy (n = 6). Test, positive/total (%)

Total group

Group-1

Group-2

Statistical significance p

EBV VCA IgM CMV IgM Parvovirus B-19 IgM M. pneumoniae IgM Throat culture—GAS

38/140 (27.1%) 5/116 (4.3%) 3/103 (2.9%)

37/139 (26.6) 5/116 (4.3%) 3/103 (2.9%)

1/1 0 0

NS* NS NS

6/103 (5.8%)

6/103 (5.8%)

0

NS

11/98 (10.8%)

11/98 (10.8%)

0

NS

p > 0.05. All serologic analyses were studied in the following ELISA kits; dia pro diagnostic kits, Italy (for CMV IgM, CMV IgG, EBV VCA IgM and EBV VCA IgG), Novatec immune diagnostic GmBh, Germany (for Mycoplasma pneumonia IgM, mycoplasma pneumonia IgG, parvovirus B19 IgM, parvovirus B19 IgG) (10–11). * NS: Statistical insignificant.

S. Bozlak et al. / International Journal of Pediatric Otorhinolaryngology 82 (2016) 81–87

86

Table 4 List of confirmed etiologies of lymphadenopathy (n:218). Disease n,%

n

%

Infectious diseases (total) Epstein–Barr virus (EBV) Cytomegalovirus (CMV) Parvovirus B-19 Mycoplasma pneumoniae Group A streptococcus (GAS) Tuberculosis Streptococcus pneumoniae

59 38 2 2 3 11 3 1

27 17.4 0.9 0.9 1.3 5 1.3 0.45

Non-infectious diseases (total) Kawasaki disease Familial Mediterranean fever Rosai Dorfman disease Kikuchi Fujimato disease PFAPA syndrome Lymphamoid papillomatosis Reactive nodular hyperplasia

25 5 2 1 1 1 1 14

11.4 2.2 0.9 0.45 0.45 0.45 0.45 6.8

6 2 2 1 1

2.7 0.9 0.9 0.45 0.45

Malignancy (total) Hodgkin lymphoma Non-Hodgkin lymphoma Acute leukemia Hemaphagocytic lymphohistiocytosis

6.3. Risk factors We evaluated with logistic regression analysis (Backwards Stepwise method) following risk factors for malignancy: Leukocytosis, size of LAP and CRP, ESR, LDH, and uric acid levels. The model was found significant with a percentage of 96.2%. Uric acid and leucopenia were significant risk factors for malignancy, with OD 1.68 (95% CI: 1.14–2.48) and 35.15 (95% CI: 2.11–584.76), respectively (Table 5). 7. Discussion We had more male patients than females in the study group (34.3% female, 65.6% male). Oguz et al. conducted a similar study with 457 patients aged between 2 months and 19 years, with 63% male dominance [19]. Yaris et al. performed a study with 98 patients aged between 86.3  5.7 months and with a male percentage of 58% [20]. The age and sex variations of our study group are similar to previous studies [19,20]. Inflammatory processes and malignant diseases are more likely to be seen in boys; boys are more likely to attend hospital than girls. Sex was an independent factor in our study because our study was a prospective cohort study. In our study, the most common cause of malignant LAP was lymphoma, which generally appears in older children. This may be the reason why the mean age of children with malignant LAP was more than the comparison group. Oguz et al. found that chronic symptoms lasting more than 4 weeks were critical for diagnosis. In Oguz et al.’s study, 96.4% of patients with malignant LAP showed symptoms after 4 weeks, whereas 61.8% of patients with benign LAP had acute symptoms Table 5 Logistic regression analysis results of risk factors for malignancy. p High serum uric acid level

0.008

Leucocyte count >15,000/mm3 <4500/mm3

0.8 0.013*

* **

95% CI: 95% confidential interval. Odds: Odds ratio.

Odds**

%95CI* limits

1.68

1.14

2.48

1.47 35.15

0.06 2.11

31.68 584.76

[19]. Kumral et al. demonstrated enhanced risk for malignancy with symptoms that last more than 4 weeks. Whereas 75% of malignant LAP presented symptoms after 4 weeks, only 39.2% of benign LAP had chronic symptoms [20]. Onset time of symptoms was similar between our study groups, which is in disaccord with previous studies. The reason for this may be that our study was a prospective study with close follow-up. Yaris found that the most common sites for cervicofacial LAPs were submandibular (36.7%) and anterior cervical (33.6%) [20]. Kumral et al. [21] and Oguz et al. [19] also found cervical LAPs were the main sites for cervicofacial LAPs. Cervical lymph nodes make up a large portion of lymphatic drainage of the upper respiratory track, which is a common location for infection in the pediatric population. This may be the reason why cervical lymph nodes are most frequently enlarged. We had two patients with supraclaviculary lymph node, both of them had malignancy. We think that, any LAP in this area is carefully evaluated for malignancy. Malignant LAPs were significantly larger than those that were the benign. In Kumral et al.’s study, the size of lymph nodes were as follows; 28% under 1 cm, 41.5% between 1 and 3 cm, and 30.5% were over 3 cm. Some 81.4% of benign LAPs were smaller than 3 cm; only 34.9% of malignant LAPs were smaller than 3 cm. Moreover, 58.4% of malignant LAPs were larger than 3 cm [21]. Oguz et al. also grouped patients according to LAP size; 21.9% were smaller than 1 cm, 32.3% were between 1 and 3 cm, and 46% had LAPs larger than 3 cm. None of the patients with LAPs smaller than 1 cm were malignant. For malignancy, 85.6% of malignant LAPs were over 3 cm [19]. In our study, we showed that the percentage of patients with large LAPs (over 3 cm) was higher in the malignancy group, group-2. Patients who had malignant LAPs in our study mostly had hard or rubbery lymph nodes, and B symptoms, which is in agreement with similar studies [19]. After examining189 patients for hemoglobin concentration, we determined anemia in 18% (n = 33) in group 1 and 50% (n = 3) in group 2. There was a statistically significant difference in the number of patients with anemia, there was no between groups difference in the leucocyte count or ESR levels. Thrombocytopenia, elevated CRP, LDH, and uric acid levels were common findings patients with malignancies, which was statistically significant. CRP, ESR, and LDH were also high in patients with malignant LAPs in Oguz et al.’s study [19]. Kumral et al. found thrombocytopenia, anemia, and leukocytosis ratio significantly higher in patients with malignancies [21]. Yaris reported LDH significantly relevant for evaluation of malignant lymph nodes but there were no differences with anemia, thrombocytopenia, ESR, leukocytosis, and leucopenia [20]. Ultrasonography is a non-invasive technique used for evaluation of lymph nodes by experienced clinicians [22]. Although ultrasonography in itself is harmless, if examining a child with tender lymphadenopathy it can be painful. Doppler ultrasonography is especially useful in the differentiation of malignancies [22]. We evaluated 152 patients (69.7%) with ultrasonography. Failure of echogenic hilar fat and disappearance of hilum were found in 75% of the patients with malignant LAPs in Ahuja’s study [22]. In our study, benign LAPs had normal hilum with typical vascularization; five patients from 6 in group 2 had ultrasonographic findings consistent with malignancy. Lymph node biopsy was performed in 10.5% patients, of which 78.2% (n = 18) were benign and 21.7% (n = 5) were malignant. Oguz et al. reported that 79.8% of biopsies found malignant lymph nodes [19]. We evaluated patients for biopsy criteria with an oncology team beforehand. Having strict criteria for lymph node excision and closely following-up patients may be the reason why many patients underwent a biopsy with benign LAP. Every patient who was diagnosed as having a malignancy underwent bone marrow

S. Bozlak et al. / International Journal of Pediatric Otorhinolaryngology 82 (2016) 81–87

examination. We found acute lymphoblastic leukemia in one of these patients. If a patient had symptoms compatible with malignancy like as deep anemia, petechial rash, hepatosplenomegaly, etc. bone marrow aspiration was done for diagnosis of leukemia. Some lymph nodes, being common under malignant LAPs, enlarged in the course of study. As shown in previous studies, we also observed that enlargement of LAPs during follow-up may be a risk factor for malignancy. Infections, especially viral pathogens, are a common cause of LAP in the pediatric population [7]. In our study, 41.3% of patients had a specific etiology. LAPs were 97% (2 1 8) benign and 2.7% malignant. In other studies, idiopathic causes of LAP were found between 45 and 60% [19]. In the present study, 58.7% of patients had non-specific benign LAP. LAPs that become smaller during follow-up are most likely benign. The reason why we have found a high percentage of idiopathic causes may be the result of the high number of patient with acute symptoms in the study group. Kumral et al. found a specific etiology in 16% of their benign patient group. EBV, CMV, toxoplasmosis, and tuberculosis were the most common causes [21]. We define some of the causes of non-infectious benign LAPs; Kawasaki disease (n = 5), familial Mediterranean fever (FMF) (n = 2), Rosai Dorfman disease (n = 1), Kikuchi Fujimato disease (n = 1) and PFAPA syndrome (n = 1). Different ratios (3.2–20%) have been reported for malignant causes of LAP. Studies designed in departments of hematology and oncology clinics have reported higher ratios. Tekgul et al. found a ratio of 22% [23], whereas in our study 2.7% of patients had malignancy. This may be due to our study plan. We prospectively followed up the patients who presented to the ambulatory clinic of general pediatrics, to which a large variety of patients visit. Lymphoma was the most common of the malignant causes, which was consistent with previous studies [22,23]. The strengths of our study were that this was prospective cohort study with a high number of patients, mainly with cervicofacial LAP for which every effort was made to discover the specific etiologies and define the malignancy types. The shortcoming was the low number of patients in the malignancy group, which caused a challenge in the statistical analyses. A further multicenter study may obtain enough patients for clearer results. In conclusion, we studied 218 patients with cervical LAPs. Some features of lymph nodes and clinical conditions of patients must be observed carefully. Lymph nodes that are greater than 30 mm, rubbery on palpation, or show no hilum in ultrasonography may need further evaluation for malignancy. Concomitant signs, as B symptoms, and high CRP and LDH levels must not be overlooked. Some serological testing for microbial agents was beneficial in a number of patients to help reach a diagnosis. Lastly, following-up a patient is important to notice if lymph nodes enlarge, which may be the only sign for malignancy.

87

Conflict of interest statement Authors declare no conflict of interests. References [1] K. Rajasekaran, P. Krakovitz, Enlarged neck lymph nodes in children, Pediatr. Clin. North Am. 60 (4) (2013) 923–936. [2] P. Lanzkowsky, Lymphadenopathy and splenomegaly, in: Philip Lanzkowsky (Ed.), Manual of Pediatric Hematology and Oncology, fifth ed., Elsevier Academic Press, New York, NY, 2011, pp. 463–468. [3] L.S. Nield, D. Kamat, Lymphadenopathy in children: when and how to evaluate, Clin. Pediatr. 43 (2004) 25–33. [4] J.R. Gosche, L. Vick, Acute, subacute, and chronic cervical lymphadenitis in children, Semin. Pediatr. Surg. 15 (2006) 99–106. [5] A.K.C. Leung, W.L.M. Robson, Childhood cervical lymphadenopathy, J. Pediatr. Health Care 18 (2004) 3–7. [6] C.J. Stutchfield, J. Tyrrell, Evaluation of lymphadenopathy in children, Paediatr. Child Health 22 (3) (2011) 98–102. [7] C.J. Twist, M.P. Link, Assessment of lymphadenopathy in children, Clin. Pediatr. 49 (2002) 1009–1025. [8] R.N. Abby, Paediatric cervical lymphadenopathy: when to biopsy, Curr. Opin. Otolaryngol. Head Neck Surg. 21 (6) (2013) 567–570. [9] M.F. Alison, Evaluation and management of lymphadenopathy in children, Pediatr. Rev. 29 (2008) 53–60. [10] M. Almasri, A. Papa, E. Souliou, K. Haidopoulou, M. Eboriadou, Respiratory syncytial virus infection in hospitalized children older than 2 years with community-acquired pneumonia, Hippokratia 17 (2) (2013) 146–149. [11] J. Chen, L. Hu, M. Wu, T. Zhonq, Y.H. Zhou, Y. Hu, Kinetics of IgG antibody to cytomegalovirus (CMV) after birth and seroprevalence of anti-CMV IgG in Chinese children, Virol. J. 304 (9) (2012) 1–7. [12] L. Rubaltelli, E. Proto, R. Salmaso, P. Bortoletto, F. Candiani, P. Cagol, Sonography of abnormal lymph nodes in vitro: correlation of sonographic and histolojik findings, AJR 155 (1990) 1241–1244. [13] A. Tschammler, U. Gunzer, E. Reinhart, D. Ho¨hmann, A.C. Feller, W. Mu¨ller, et al., The diagnostic assessment of enlarged lymph nodes by qualitative and semiquantative evaluation of lymph node perfusion with color coded duplex sonography, Rofo 154 (1991) 414–418. [14] H.J. Steinkamp, M. Rausch, J. Maurer, N. Hosten, H. Schedel, R. Langer, et al., Color-coded duplex sonography in the differential diagnosis of cervical lymph node enlargements, Rofo 161 (1994) 226–232. [15] H.J. Steinkamp, J. Maurer, M. Cornehl, D. Knobber, H. Hettwer, R. Felix, Recurrent cervical lymphadenopathy: differential diagnosis with color-duplex sonography, Eur. Arch. Otorhinolaryngol. 251 (1994) 404–409. [16] K. Newman, A.A. Hayes-Jordan, Lymph node disorders, in: J.L. Grosfeld, J.A. O’Neil, E.W. Fonkalsrud, A.G. Coran (Eds.), Pediatric Surgery, sixth ed., Mosby Elsevier, Philadelphia, PA, 2006, pp. 844–849. [17] P.M. Banks, Technical factors in the preparation and evaluation of lymph node biopsies, in: Daniel M. Knowles (Ed.), Neoplastic Hematopathology, second ed., Lippincott Williams & Wilkins, Philadelphia, PA, 2001, pp. 467–483. [18] M.W. Lavrence, W. Pitts, The role of fine needle aspiration biopsy in the diagnosis and management of hemotopetic neolasma, in: Daniel M. Knowles (Ed.), Neoplastic Hematopathology, second ed., Lippincott Williams & Wilkins, Philadelphia, PA, 2001, pp. 483–507. [19] A. Og˘uz, C. Karadeniz, E.A. Temel, E.C. Citak, F.V. Okur, Evaluation of peripheral lymphadenopathy in children, Pediatr. Hematol. Oncol. 23 (2006) 549–561. [20] N. Yaris, M. C¸akır, E. So¨zen, U. Cobanoglu, Analysis of children with peripheral lymphadenopathy, Clin. Pediatr. 45 (2006) 544–549. [21] A. Kumral, N. Olgun, K.M. Uysal, F. Corapcioglu, H. Oren, F. Sarialioglu, Assessment of pheripheral lymphadenopathies experience at a pediatric hematology oncology department in Turkey, Pediatr. Hematol. Oncol. 19 (2002) 211–218. [22] A.T. Ahuja, M. Ying, Grey scale sonography in assessment of cervical lymphadenopathy: review of sonographic appearances and features that may help a beginner, Br. J. Oral Maxillofac. Surg. 38 (2000) 451–459. [23] H. Tekul, S. Oztop, N. Cetingul, S. Soydan, G. Nisli, A prospective study of pheripheral lymphadenopathy in childhood, J. Trop. Pediatr. 43 (1997) 117–118.