Arthritic presentation of malignancies in children: a retrospective study from two centers in South India

Abstract

Children with malignancies can present with varied symptoms mimicking rheumatological or orthopedic conditions. Symptoms such as fever, myalgia, arthralgia, and arthritis usually suggest an underlying musculoskeletal condition. However, malignancies in children can also present with such symptoms. The objective of this study was to analyze the clinical and laboratory features of children with malignancies presenting with arthritic manifestations to the paediatric rheumatology clinic and to raise awareness of these presentations among practising physicians. A retrospective case review was carried out in 53 patients who presented to 2 paediatric rheumatology units in 2 tertiary care hospitals in South India. These children presented with musculoskeletal symptoms and had a final diagnosis of malignancy. The median age was 6.1 years with a range from 1 to 15 years and male:female ratio of 1.12:1. The most common presentation was bone pain (75%), followed by fever (53%), polyarthralgia (51%), refusal to bear weight in lower limbs (40%), night pain (40%), and joint swelling (15%). Anemia with Hb < 8 g/dl was observed in 26% of the patients, white cell count (WCC) < 4000 cells/mm³ in 17%, WCC > 12,000 cells/mm³ in 15%, platelets < 150,000/ml in 43%, and erythrocyte sedimentation rate > 20 mm/hr in 77%.The peripheral smear was positive for malignancy in only 40% of the patients. Before referral to tertiary units, 34% were already treated with steroids with a suspected diagnosis of juvenile idiopathic arthritis. Treatment with steroids could mask the symptoms of malignancy and could lead to a delay in diagnosis and a poor outcome.

INTRODUCTION

Musculoskeletal complaints constitute a common cause for children being brought to outpatient departments. A majority of these children tend to have benign nonspecific conditions like hypermobility or growing pains [1]. However, a small proportion of these patients may have either inflammatory or infective arthritis or rarely an underlying malignancy [2].

Juvenile idiopathic arthritis (JIA) is a diagnosis of exclusion. Children with JIA may present with typical complaints of pain and swelling of the joints with morning stiffness. In the absence of diagnostic markers, the presence of swollen joints, raised white cell counts with neutrophilia, and elevated inflammatory markers may help in clinching the diagnosis [3]. However, at times arriving at a diagnosis can be challenging, especially in nonverbalizing young children who may present with loss of function, refusal to bear weight, irritability, and altered behavior [4]. In children with systemic juvenile idiopathic arthritis (SJIA), fever maybe the only initial manifestation of SJIA without the typical evanescent rash and arthritis [5,6]. Children with malignancies can present in a similar fashion with fever and musculoskeletal symptoms with laboratory parameters being within the near normal range [7].

Careful and detailed assessment may be required in such children to rule out malignancies. Moreover, in developing countries, where there is no structured referral system, there is a risk of these patients with malignancy being wrongly diagnosed as JIA and being started on steroid treatment. This can result in the masking of symptoms, further complicating the clinical picture and resulting in the delayed treatment of underlying conditions. The purpose of this study is to highlight the importance of differentiating malignancies from JIA in children presenting with musculoskeletal manifestations.

MATERIALS AND METHODS

Retrospective study of hospital records of 53 paediatric patients who presented with musculoskeletal manifestations and had a final diagnosis of malignancy was carried out. These patients were seen by the paediatric rheumatologists in two tertiary hospitals in South India over a period of 8 years from 2013 to 2020. The following details of patients were collected: age, sex, clinical symptoms, laboratory parameters, history of prior treatment with steroids, relevant imaging studies, referral, and final diagnosis.

RESULTS

A total of 53 patients were seen in both hospitals over a period of 8 years (Table 1). Patients who presented to the paediatric rheumatology clinic with musculoskeletal symptoms and later on diagnosed with malignancy and patients already started on therapy for JIA elsewhere but later found to have an underlying malignancy were included in the study.

Table 1.

Clinical details and investigations.

Variables	Normal values	Frequency
Total patients		53
Age range		1-15 years
Mean		6.1 years
M:F ratio		28:25 (1.12:1)
Bone pain		40 (75%)
Fever		28 (53%)
Polyarthralgia		27 (51%)
Nonweight bearing		21 (40%)
Night pain		21 (40%)
Joint swelling		8 (15%)
Hb (gm/dl)	11-17
<8		14 (26%)
8-11		27 (51%)
>11		12 (23%)
WCC (cells/cumm)	4,000-12,000
<4,000		9 (17%)
4,000-12,000		36 (68%)
>12,000		8 (15%)
Neutropenia <1,500		21 (40%)
Normal neutrophil count		32 (60%)
Platelet count per ml	150,000-450,000
<150,000		23 (43%)
>150,000		30 (57%)
ESR (at 1 hour)	<20
Raised		41 (77%)
Normal		11 (21%)
Not available		1 (2%)
Peripheral smear for malignancy
Positive		21 (40%)
Negative		32 (60%)
Prior treatment with steroids
Yes		18 (34%)
No		34 (64%)
Not known		1 (2%)

M:F ratio, Male:female ratio; Hb, Hemoglobin; WCC, White cell count; ESR, Erythrocyte sedimentation rate.

Although the age range was 1-15 years, the mean age was 6.1 years and 31/53 (58.5%) patients were either 5 years or younger. The male:female ratio was 1.12:1. The most commonly presenting symptom was bone pain in 40/53 (75%) patients, followed by fever in 28/53 (53%), polyarthralgia in 27/53 (51%), refusal to bear weight in 21/53 (40%), night pain in 21/53 (40%), and joint swelling in 8/53 (15%).

Hemoglobin of <8 g/dl was seen in 14 patients, between 8 and 11 g/dl in 27 patients, and >11 g/dl in 12 patients. Leukopenia of <4,000 cells/mm³ was seen in 9 patients and leukocytosis of >12,000 cells/mm³ in 8 patients. Surprisingly, majority of the patients (36/53) had total white cell counts within the normal range. Normal neutrophil count was seen in 32 patients and neutropenia less than 1,500 was seen in 21 patients. 30 out of 53 patients had a normal platelet count > 150,000 per ml and 23/53 patients had a platelet count < 150,000 per ml. Significant thrombocytopenia (<50,000 per ml) was seen in only five patients. Peripheral smear was positive for malignancy in 21/53 patients. 32 /53 patients did not show evidence of malignancy in the peripheral smear. A raised erythrocyte sedimentation rate (ESR) was seen in 41/52 patients, in 1 patient ESR was not known, and the rest had normal ESR.

18 out of 53 patients had received steroids elsewhere with a presumed diagnosis of inflammatory arthritis. Status of pretreatment with steroid was not known in one patient. 34 patients had not received steroids prior to presentation to paediatric rheumatology. 16 patients had abnormal radiological findings which helped in the diagnostic evaluation, of which 9 children had periosteal reaction, 4 had lucent zones (Figure 1), 1 patient showed a sclerotic area (Figure 2), and 2 patients had lytic lesions (Figure 3).

Figure 1.

Lucent zones seen in the distal end of tibia and fibula (arrows).

Figure 2.

Sclerosis seen in distal part of tibia (arrow).

Figure 3.

Lytic lesion seen in proximal part of humerus (arrow).

The most common diagnosis was acute lymphocytic leukemia (ALL) in 45 (84.9%) patients; Ewing’s sarcoma in 3 (5.6%); neuroblastoma in 2 (3.7%); and Hodgkin’s lymphoma, Burkitt lymphoma, and acute myeloid leukemia in 1 each. The average duration to diagnosis was 2.5 months with a range from 2 weeks to 5 months.

DISCUSSION

In this retrospective study carried out in two tertiary centers in South India, we analyzed the clinical features and lab parameters in children presenting with musculoskeletal manifestations and were later on diagnosed with malignancy.

Compared to two other major studies of this type, the age range and the median age of the patients were similar in our study [8,9]. However, our study showed an M:F of 1.12:1, whereas in the other studies the M:F ratio was 2:1. Musculoskeletal problems like arthralgia, bone pain, refusal to bear weight, night pain, and joint swelling were predominant complaints in this group of patients and fever was noticed in 53% of the patients. Cabral and Tucker [8] also noted similar findings in their study.

Our study showed that 68% of the patients had a normal total white blood cell count almost the same as Cabral and Tucker’s [8] study where 70% of the patients were found to have normal white blood cell counts. A raised ESR was noticed in 41/53 (77%) patients. Sometimes physicians attribute elevated inflammatory markers to an underlying inflammatory condition and even use this as a tool to differentiate from a malignancy. This can be misleading as evidenced in this study, by the number of patients being misdiagnosed with arthritis and being started on steroid therapy. Exposure to corticosteroids prior to the diagnosis of ALL could adversely affect prognosis by masking the clinical features, elimination of blasts on diagnostic investigations like peripheral smear and bone marrow, or also by resulting in a partial treatment response and promoting disease resistance. Prior studies demonstrated poor outcomes among patients who received corticosteroid therapy, most commonly prolonged courses for presumed rheumatologic diseases or other conditions, prior to their diagnosis of ALL. This was shown in our study where 11 children pretreated with steroids had a normal peripheral smear. In such cases where peripheral smear is negative, bone marrow aspiration may be helpful in ruling out leukemia. This is particularly concerning, as the majority of children with aches and pains are seen by primary physicians at their private practice. They should be aware of the red flag signs such as night pain, pain away from joints, low or high white cell count with lymphocyte predominance in the differential count, pain out of proportion to joint involvement, the presence of systemic features, ESR, and platelet count disparity [10]. These findings in a child with musculoskeletal symptoms should prompt a referral to a specialist to rule out an underlying malignancy. Steroid treatment in patients with underlying malignancy makes them a high-risk group with unfavorable diagnosis from the oncology treatment point of view [11].

Elevated ESR and platelet count < 150,000 per ml were seen in 11/53 patients. Of these, two patients had been initiated on steroid therapy. In a child with a polyarticular JIA, a much higher platelet response is expected along with elevated inflammatory markers. If there is a mismatch between these two markers, an alternate diagnosis is more likely. It is important to emphasize that patients with seemingly normal platelet counts above 150,000 per ml or borderline normal platelet counts may harbor an underlying malignancy.

Apart from the patients who had been misdiagnosed with JIA, one of the patients in the study had a history of oligoarticular JIA for 2.5 years prior and had received intraarticular injections and oral methotrexate therapy. A recent onset of backache, which was present throughout the day and night, and severe pain even on slight movement alerted us to the possibility of a malignancy. A bone marrow aspiration confirmed the presence of ALL. This was an exceptional case of a child with a genuine diagnosis of JIA who later developed leukemia [12,13].

Contrary to the myth that significantly elevated white cell counts are present in childhood ALL, a low/low normal total WBC count with lymphocytic predominance and absolute neutropenia with relative lymphocytosis is the standard feature [14]. A normal peripheral smear should not dissuade the physician from requesting a bone marrow study when suspicious clinical features and counts as detailed above are accompanying. In addition, the use of alternative medicine and steroids due to misdiagnosis can also affect peripheral smear appearances.

CONCLUSION

About 40% of the children with malignancies present with musculoskeletal symptoms and about 0.25% of the paediatric patients presenting with musculoskeletal symptoms have an underlying malignancy [15,16]. While assessing children with musculoskeletal complaints, physicians should be able to differentiate between bone and joint pain and look out for red flag signs and symptoms (Table 2).

They have to be aware that the entire complete blood count has to be perused in detail; that a normal peripheral smear does not rule out acute leukemia; and to not start steroids in patients without careful consideration for possible malignancy. Physicians should, therefore, have a low threshold for referral to a specialist if any of the red flag signs are present or if the diagnosis is in doubt. When children present with atypical manifestations of rheumatologic diseases or if there is disparity between clinical manifestations and laboratory parameters, a thorough evaluation is necessary. Pretreatment with steroids in a child with leukemia can be a marker of unfavorable prognosis [17].

Funding

None.

Conflict of interest

The authors declare that there is no conflict of interest regarding the publication of this article.

Ethical approval

The study protocol and other relevant documents for this study were reviewed by the Institutional Ethic Committee (IEC) and ethical clearance was given by the IEC in both institutions for compiling this review (Amrita Institute of Medical Sciences IEC approval number ECASM-AIMS-2021-278 and Sri Ramachandra Institute of Higher Education and Research IEC approval number IEC-N1/21/APR/78/76). Informed consent of caregivers/patients was obtained.