Pediatric Hematology and Oncology, 19:211± 218, 2002Copyright C° 2002 Taylor & Francis0888-0018/02 $12.00 + .00

Review Article

ASSESSMENT OF PERIPHERAL LYMPHADENOPATHIES: Experienceat a Pediatric Hematology–Oncology Department in Turkey

Abdullah Kumral, MD, Nur Olgun, MD, Kamer MutafoÆglu Uysal, MD, FundaCorapcíoÆglu, MD, Hale Oren, MD, and Faik SaríalioÆglu, MD h Departmentof Pediatric Hematology± Oncology, Dokuz Eyl Èul University Faculty of Medicine,ÇIzmir, Turkey

h Since a large variety of disorders may lead to lymph node enlargement , determining the cause ofperipheral lymphadenopath y (LAP) in children can be dif® cult. This retrospective study evaluated200 children who were admitted to an Oncology± Hematology department because of lymphadenopathyand aimed to determine the clinical and laborator y ® ndings that were valuable for differentialdiagnosis. A speci® c cause for lymphadenopathy was documented in 93 (46.5%) cases. One hundredforty (70%) children were classi ® ed as having a benign cause for lymph node enlargements. Fourteen(10%) of these cases underwent an excisional lymph node biopsy, and histopathological examinationshowed a reactive hyperplasia. Sixty (30%) cases were classi ® ed as having a malignant disease-causing lymphadenopathy . In terms of differential diagnosis, some associated systemic symptoms,physical ® ndings, and laborator y investigations showed signi® cant difference between benign andmalignant lymphadenopath y groups. The following ® ndings were determined as being important toalert the physician about the probability of a malignant disorder: location of the lymphadenopathy(supraclavicular and posterior auricular), duration of the lymph node enlargement (>4 weeks), sizeof the lymph node (>3 cm), abnormal complete blood cell ® ndings, abnormalities in chest X-ray, andabdominal ultrasonography.

Keywords. children, etiology, peripheral lymphadenopath y

Lymphadenopathy is a common ® nding in children as a result of localor systemic infections, but may be an important indicator of a malignantdisorder. Determining the cause of peripheral lymphadenopathy in childrencan sometimes be a dif® cult diagnostic challenge [1 ± 3].

In this study, we aimed to de® ne the demographic, clinical, and labora-tory ® ndings in children with peripheral lymphadenopathy (LAP) and alsoto determine which of these ® ndings were suf® ciently alarming to search fora malignant disorder.

Received 24 July 2001; accepted 8 December 2001.Address correspondence to Abdullah Kumral, MD, Dokuz Eyl Èul University Faculty of Medicine,

Department of Pediatrics, ÇInciraltõ - ÇIzmir, Turkey. E-mail: abdullah.kumral@deu.edu.tr

211

Pedi

atr

Hem

atol

Onc

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Cal

gary

on

09/2

5/13

For

pers

onal

use

onl

y.

212 A. Kumral et al.

PATIENTS AND METHODS

This study was performed retrospectively in children who had been re-ferred to our Pediatric Hematology ± Oncology department between January1988 and December 1998 because of enlarged peripheral lymph nodes.Patient charts were evaluated to extract the following information: age; sex;duration of lymph node enlargement; features of lymph nodes, including lo-cation, size, mobility, and extension; associated local and systemic symptoms;preceeding BCG vaccination; diagnostic laboratory investigations; radiologicinvestigations; and histopathologic examination(s) and ® nal diagnosis.

Lymph node enlargements were classi® ed according to the size, exten-sion, and duration of the LAP according to the following criteria:

² Size of the lymph node : <1 cm, 1 ± 3 cm, and >3 cm² Extention of the lymph nodes : localized LAP (a single or multiple but adjacent

lymph node involvement) and generalized LAP (more than two and non-adjacent lymph node involvement)

² Duration of the lymph node enlargement : acute (·4 weeks of duration) andchronic (>4 weeks of duration)

Patients who had a lymph node biopsy were also classi® ed according to thetime of diagnostic biopsy: early biopsy group (within the ® rst 2 weeks of pre-sentation) and late biopsy group (after the ® rst 2 weeks of presentation) [4].Some features of LAPs, including the duration, location, size, extention, mo-bility, ¯ uctuation, and local warmth over the enlarged LAP, were comparedbetween malignant and benign LAPs. The frequency of laboratory and ra-diological pathologic ® ndings in patients with malignant and benign LAPwere also compared.

The data were analyzed using the Statistical Package for the Social Sci-ences Software Package (SPSS) (version 7.5). Comparisons between groupswere made using appropriate statistical methods (Pearson chi-square andFisher’s exact test). Logistic regression analysis was used to evaluate the riskfactors. A p value of less than .05 was considered signi® cant.

RESULTS

A total of 200 children with peripheral LAP were analyzed. Some systemicand local features of lymph nodes associated with benign and malignant dis-eases are given in Table 1. The other systemic symptom was skin eruption(n D 10) (5%). The medical history revealed previous BCG vaccination withinthe last 3 months in only one (0.5%) patient. There were statistically signi® -cant differences between the 2 groups according to the duration, site, size,and extension (p < .05). The duration of the lymph node enlargement wasnot known in 15 (7.5%) cases.

Pedi

atr

Hem

atol

Onc

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Cal

gary

on

09/2

5/13

For

pers

onal

use

onl

y.

Peripheral Lymphadenopathy in Children 213

TABLE 1 Associated Systemic Symptoms and Local Features of LAP in Benign and MalignantLAP Groups

Benign Malignantdisorders (n D 140) disorders (n D 60) Total (n D 200)

n (%) n (%) n (%) p value

Systemic symptomsFever 76 (54.3) 22 (36.7) 98 (49.0) 0.0300Night sweats 14 (10) 4 (6.7) 18 (9.0) 0.5930Weight loss 2 (1.4) 13 (21.7) 15 (7.5) 0.0001

Local ® ndingsDuration

Acute 76 (60.8) 15 (25.0) 91 (45.5) 0.0001Chronic 49 (39.2) 45 (75.0) 94 (47.0) 0.0001

Locationa

Suboccipital 15 (10.7) 3 (5.0) 18 (9.0) 0.2820Postauricular 10 (7.1) 10 (16.7) 20 (10.0) 0.0300Preauricular 1 (0.7) 1 (1.7) 2 (1.0) 0.5110Cervical 134 (95.7) 55 (91.7) 189 (94.5) 0.2050Supraclavicular 6 (4.3) 26 (43.3) 32 (16.0) 0.0001Axillary 23 (16.4) 10 (16.7) 33 (16.5) 0.5590Inguinal 8 (5.7) 9 (15.0) 17 (8.5) 0.0340

Size<1 cm 50 (36.5) 3 (5.0) 53 (26.5) 0.00011± 3 cm 61 (44.5) 22 (36.7) 83 (41.5) 0.2270>3 cm 26 (19.0) 35 (58.3) 61 (30.5) 0.0001

ExtensionGeneralized 45 (32.1) 34 (56.7) 79 (39.5) 0.0200Local 95 (67.9) 26 (43.3) 121 (60.5) 0.0200

MobilityFixed 13 (9.3) 8 (13.3) 21 (10.5) 0.2680Mobile 127 (90.7) 52 (86.7) 179 (89.5) 0.2680

FluctuationPresent 4 (2.9) 0 4 (2.0) 0.3190Not present 136 (97.1) 60 (100) 196 (98.0) 0.2370

Local heatPresent 7 (5.0) 0 7 (3.5) 0.1050

aEach site was evaluated separately even if involved with another site; therefore, the sum is notequal to 100%.

One hundred forty (70%) children were classi® ed as having a benigncause for lymph node enlargements (benign LAP group) and 60 (30%) caseswere classi® ed as having a malignant disease causing lymphadenopathy (ma-lignant LAP group). In the benign LAP group, the most frequently involvedsite was cervical (95.8%) for localized LAPs, and cervical and axillar (33.3%)for generalized LAPs. In 140 children with benign LAP, a speci® c cause forlymph node enlargement could be determined in 33 (23.6%) cases: CMV in2 (1.4%), toxoplasma infection in 5 (3.7%), EMN in 22 (15.7%), tuberculosislymphadenitis in 3 (2.1%), and BCG lymphadenitis in 1 (0.7%) patient.

In the malignant LAP group, the most frequently involved site was cer-vical (80.9%) for localized LAPs, and supraclavicular together with cervi-cal (63.6%) for generalized LAPs. In this latter group ® nal diagnoses are

Pedi

atr

Hem

atol

Onc

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Cal

gary

on

09/2

5/13

For

pers

onal

use

onl

y.

214 A. Kumral et al.

TABLE 2 Final Diagnosis in the Malignant Group

Diagnosis n %

Hodgkin disease 27 45ALL 15 25NHL 10 16.8Nasopharynx carcinoma 4 6.6Hypopharynx carcinoma 1 1.7Thyroid carcinoma 1 1.7Rhabdomyosarcoma 1 1.7ANLL 1 1.7

Note. ALL, acute lymphocytic leukemia; ANLL, acute non-lymphocytic leukemia; NHL, non-Hodgkin lymphoma.

shown in Table 2. All the patients who had solid tumors had histopathologicdiagnosis. All leukemias and patients with stage IV NHL were diagnosed withbone marrow examinations.

The results of the diagnostic tests are given in Table 3, comparing the databetween malignant and benign LAPs. All the results of available laboratoryand radiological examinations showed signi® cant differences between thebenign and malignant LAP groups except for leukopenia and ESR. Atypicallymphocytes were seen in 38 (19%) children; in 22 cases these cells were

TABLE 3 Laboratory and Radiological Investigation Results in Benign and Malignant LAPs

Benign MalignantLaboratory/ disorders (n D 140) disorders (n D 60) Total (n D 200)

radiological ® nding n (%) n (%) n (%) p value

Anemia 54 (38.6) 36 (60.0) 90 (45.0) 0.0050Thrombocytopenia 1 (0.7) 16 (26.7) 17 (8.4) 0.0001WBC

Leukocytosis 16 (11.4) 23 (38.3) 39 (19.5) 0.0020Leukopenia 6 (4.3) 4 (6.7) 10 (5.0) 0.4910

Peripheral blood smearAtypical lymphocytesa 38 (27.1) Ð 38 (19.0) 0.0001Shift to left 38 (27.1) Ð 38 (19.0) 0.0001Blast Ð 19 (31.7) 19 (9.5) 0.0001

ESR<20 mm/h 49 (35) 24 (40) 73 (36.5) 0.3030>20 mm/h 33 (23.6) 14 (23.3) 47 (23.5) 0.5630

Chest X-ray 136 (68)Normal 93 (97.9) 27 (65.9) 120 (60.0) 0.0001Hilar LAP 2 (2.1) 4 (9.8) 6 (3.0) 0.0300Mediastinal LAP Ð 8 (19.5) 8 (4.0) 0.0001Mediastinal C Hiler Ð 2 (4.8) 2 (1.0) 0.0001

LAPAbdominal USG 62 (31.0)

Normal 23 (88.5) 20 (55.6) 43 (21.5) 0.0300Lymphadenopathy Ð 8 (22.2) 8 (4.0) 0.0001Hepatosplenomegaly 3 (11.5) 8 (22.2) 11 (5.5) 0.0300

a Consistent with infectious mononucleosis or virocytes.

Pedi

atr

Hem

atol

Onc

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Cal

gary

on

09/2

5/13

For

pers

onal

use

onl

y.

Peripheral Lymphadenopathy in Children 215

consistent with infectious mononucleosis (EMN), and in 16 children theywere thought to represent virocytes. In differential count, a shift to left wasdetermined in 38 (19%) children to represent infectious diseases. Serologictests were performed in 44(22%) cases for toxoplasma infection, and IgM waspositive in 5 (11.4%) children. Cytomegalovirus (CMV) IgM was investigatedfor 42 (21%) cases, with a positivity rate of 4.8%. In 52 (26%) children,Monospot test was performed for EBV infection and 7 (13.1%) cases werefound positive.

Excisional lymph node biopsy was performed in a total of 57 (28.5%)cases. Forty-three (21.5%) cases had early and 14 (7%) cases had late biopsy.Early biopsy was done in 43 patients for the following indications: an enlarg-ing node or nodes that remain unchanged after 2 weeks, especially if asso-ciated with unexplained fever, weight loss, and hepatosplenomegaly, supra-clavicular LAP and enlarged nodes associated with mediastinal and/or hilarLAP in chest ® lm. All the cases who underwent early biopsy had been diag-nosed as having a malignant disorder. In the late biopsy group, there were14 cases in whom the size of the LAP showed no regression after 4 weeks offollow-up. Thirteen were found to have reactive lymphoid hyperplasia andone case had tuberculosis lymphadenitis.

DISCUSSION

Lymphadenopathy is a frequently encountered physical ® nding in chil-dren and may arise from a wide variety of benign or malignant disorders. Clin-ical history, physical ® ndings, and laboratory investigations may give someimportant clues for differential diagnosis, but determining the speci® c causeof the LAP can be quite dif® cult in some patients.

In this study, a speci® c cause could be determined in 46.5% of cases. Inone large series, a speci® c cause could be found in 41% of the children un-dergoing lymph node biopsy [2]. We determined a malignant disorder as thecause of peripheral lymph node enlargement in 30% of the pediatric cases.This high rate of malignant causes for LAP mostly re¯ ects the characteristicsof patients who had been referred to a hematology ± oncology unit. A recentstudy from another hematology ± oncology unit showed 27% of malignantcauses in 382 children with peripheral LAPs [5].

The duration of LAP was signi® cantly longer for malignant disorders.In our study, most of the chronic LAPs were related to a malignant causeand more than half of the benign LAPs had a short duration by history. Ininfectious or reactive LAPs, the clinical history is usually short [6].

In this study, generalized LAP was more frequent encountered in themalignant LAP group. This result mostly re¯ ects the characteristics of themalignant LAP group, which includes patients with lymphoproliferative ma-lignant disorders in 86.6% of cases. A variety of infectious and in¯ ammatory

Pedi

atr

Hem

atol

Onc

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Cal

gary

on

09/2

5/13

For

pers

onal

use

onl

y.

216 A. Kumral et al.

diseases as well as malignant disorders may give rise to generalized LAP.However, all of these disorders usually have other distinguishing signs orsymptoms to aid in diagnosis [7, 8].

Cervical nodes were the most frequently involved localization for boththe benign and malignant LAP groups. The differential diagnosis of local-ized LAP in the head and neck area is more problematic because both acuteand chronic infectious and malignant disorders cause nodal enlargement inthis area [9].

The site of involvement was of value for differential diagnosis only for pos-terior auricular and supraclavicular LAPs; these two sites were signi® cantlyrelated to malignant disorders. Most children have palpable, small cervi-cal, axillary, and inguinal nodes. Although no site of adenopathy is com-pletely free from serious disorders, adenopathy in the posterior auricular,epitrochlear, or supraclavicular area is de® nitely abnormal [2, 7].

The size of the node was not of diagnostic value when it was between 1 and3 cm. Nodes that were smaller than 1 cm were suggestive of benign disordersand nodes that were larger than 3 cm were of diagnostic value in differen-tial diagnosis for malignant disorders. Although the relevant studies mostlyavoid giving a de® nite size for differential diagnosis, a maximum diameter of>2 cm was considered to be the limit to distinguish malignant or granulo-matous disorder from other causes [5, 10].

Fever was more frequently encountered in patients with benign LAP.This ® nding re¯ ects the high incidence of upper respiratory infections inpatients with cervical node enlargements. Weight loss was a more frequentlyassociated symptom in the malignant LAP group, which can be explained bythe predominance of Hodgkin disease in the malignant group.

The incidence of anemia, thrombocytopenia, and leukocytosis was sig-ni® cantly higher in children with malignant disorders. Lymphoblasts in pe-ripheral blood smear were encountered in either leukemias or stage IV NHL.Atypical lymphocytes and shift to left were signi® cantly higher in benign dis-orders; the former was consistent with EMN, and the latter was consistentwith infectious diseases. ESR did not show any signi® cant difference betweengroups. Abnormalities in chest X-ray and abdominal ultrasonography (USG)were suggestive of a malignant disorder. Hilar and mediastinal LAPs were al-ways encountured in malignant disease except for 2 cases who had hilarLAP associated with Mycobacterium tuberculosis infections. Abdominal USG® ndings showed a signi® cant difference between groups; the presence of en-larged abdominal lymph nodes and hepatosplenomegaly was higher in themalignant LAP group.

A number of algorithms for the workup of the child with adenopa-thy have been proposed. For the child with generalized LAP, a completeblood cell count; blood culture; serologic testing for EBV, toxoplasmosis,cytomegalovirus (and, perhaps, HIV), and fungal disease; a tuberculin skin

Pedi

atr

Hem

atol

Onc

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Cal

gary

on

09/2

5/13

For

pers

onal

use

onl

y.

Peripheral Lymphadenopathy in Children 217

test; and chest radiography will usually uncover a lead to a speci® c diagnosis.If no clues emerge, lymph node biopsy is probably indicated [11].

In this study, 57 out of 200 children (28.5%) underwent excisional lymphnode biopsy. It was performed within the ® rst 2 weeks of presentation in43 (21.5%) cases and after this period in 14 (7%). Most of the children hadan early biopsy and all of these cases had been diagnosed as having a malig-nant disorder. None of the 14 children who underwent a late biopsy was diag-nosed as having a neoplastic disease to cause node enlargement. Histopatho-logic examination showed tuberculosis lymphadenitis in one case and it wasconsistent with reactive lymphoid hyperplasia in 13 nodes. In some cases, aspeci® c cause of LAP cannot be determined, even after excisional biopsy, andnondiagnostic reactive hyperplasia is encountered in the majority of thesecases [1, 2].

In conclusion, we determined the following ® ndings as being importantclues for differential diagnosis of peripheral LAPs:

² The time of the LAP is important. If LAP is present for more than 4 weeks,interventions to reach the ® nal diagnosis (especially lymph node biopsy)should be performed immediately.

² LAPs larger than 3 cm represent a greater risk for malignancy.² Supraclavicular and posterior auricular LAP have a higher risk for repre-

senting a neoplastic node enlargement.² Abnormal CBC ® ndings should alert the physician for further diagnostic

evaluation.² Pathologies identi® ed in chest X-ray and abdominal USG have a great

diagnostic value. On the other hand, cervical USG contributes nothingto our ® ndings if performed on areas where LAPs can be easily palpa-ted. Therefore, its use to evaluate cervical masses should be restrictedto certain cases in which the features of the mass represent a diagnosticchallenge.

REFERENCES

1. Lake AM, Oski FA. Peripheral lymphadenopathy in childhood. AJDC. 1978;132:357 ± 359.2. Knight PJ, Mulne AF, Vassy LE. When is lymph node biopsy indicated in children with enlarged

peripheral nodes? Pediatrics. 1982;69:391 ± 396.3. Barton L. Childhood cervical adenitis. Aust Fam Pediatr. 1984;29:163 ± 166.4. Lee Y, Terry R, Lukes RJ. Lymph node biopsy for diagnosis; a statistical study. J Surg Oncol. 1980;14:53±

60.5. Karadeniz C, Oguz A, Ezer U, Ozturk G, Dursun A. The etiology of peripheral lymphadenopathy in

children. Pediatr Hematol Oncol. 1999;16:525 ± 531.6. Seibel NL, Cossman J, Magrath IT. Lymphoproliferative disorders. In: Pizzo PA, Poplack DG, eds.

Principles and Practice of Pediatric Oncology, 3rd ed. Philadelphia: Lippincott ± Raven; 1997:589± 614.7. Steuber CP, Nesbit ME. Clinical assesment and differential diagnosis of the child with suspected

cancer. In: Pizzo PA, Poplack DG, eds. Principles and Practice of Pediatric Oncology, 3rd ed. Philadelphia:Lippincott± Raven; 1997.

8. Zitelli B. Neck mass in children: adenopathy and malignant disease. Pediatr Clin North Am. 1981;28:813± 884.

Pedi

atr

Hem

atol

Onc

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Cal

gary

on

09/2

5/13

For

pers

onal

use

onl

y.

218 A. Kumral et al.

9. Chesney J. Cervical adenopathy. Pediatr Rev. 1994;15:276 ± 284.10. Slap GB, Brooks JSJ, Schwartz JS. When to perform biopsies of enlarged peripheral lymph nodes in

young patients. JAMA. 1984; 252:1321± 1326.11. Link MP, Donaldson SS. The lymphomas and lymphadenopathy. In: Nathan DG, Orkin SH, eds.

Hematology of Infancy and Childhood, 5th ed. Philadelphia: WB Saunders; 1998.

Pedi

atr

Hem

atol

Onc

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Cal

gary

on

09/2

5/13

For

pers

onal

use

onl

y.