Diagnosis of Kikuchi-Fujimoto disease in an 11-year-old girl with fever and sickle cell disease

Abstract

Kikuchi-Fujimoto disease (KFD) is a rare lymphohistiocytic disorder which can cause prolonged fever and other systemic B symptoms including diffuse lymphadenopathy. Given its clinical presentation, there is often initial concern for lymphoma and diagnosis requires lymph node biopsy. It most frequently affects young women of Asian descent; it is less commonly encountered in paediatric patients. KFD is typically a benign, self-limited process, however, there is an association with development of systemic lupus erythematosus. Given its rarity, it remains unclear if KFD is associated with other chronic conditions. Here we present the third case of KFD occurring in a paediatric patient with sickle cell disease.

Background

Fever is a very common presenting symptom in children. In many instances, fever is secondary to a self-limited viral process or a bacterial source can be identified on history and examination. However, there are occasions where no source can be identified, and fever persists. In these circumstances, the differential is quite broad including oncologic, rheumatologic and infectious aetiologies. The broad number of underlying processes poses a diagnostic challenge given that causes can range from benign to potentially fatal if left untreated. Underlying medical conditions, such as sickle cell disease (SCD), must be considered in diagnostic evaluation, as particular medical conditions can influence workup and management of fever of unknown origin.

Kikuchi-Fujimoto disease (KFD) is one cause of fever of unknown origin in paediatric patients; it is a lymphohistiocytic disorder with clinical symptoms that include fever and lymphadenopathy. Diagnosis is made by histopathology, which typically shows disruption of the normal architecture of the lymph node with necrosis and apoptotic karyorrhectic debris. There can also be proliferation of histiocytes and CD8⁺ T cells with an absence of neutrophils. Histiocytes often express CD68 and myeloperoxidase.¹ The pathogenesis of the disease remains unknown, and it is rare in the paediatric population, with literature suggesting that 20% of patients affected by KFD are children.² The incidence remains unknown but a large review identified around 700 cases worldwide since 1970s.^{1 2} In younger patients, there is a male predominance,³ and available literature suggests that presenting symptoms in children are similar to adults, although these may differ in frequency; for example, children are more likely to have tender lymphadenopathy in addition to fever and rash than adults.²

We present here a case of an 11-year-old girl with SCD, who presented with prolonged fever ultimately diagnosed with KFD, which to our knowledge is the third reported case of KFD in a child with SCD.^{4 5} Given the rarity of KFD, it is important to characterise the presentation in paediatrics, especially when comorbid conditions exist to complicate the diagnosis.

Case presentation

An 11-year-old girl with a previous medical history significant for SCD, partial splenectomy and prematurity presented as a transfer from an outside institution to the paediatric floor with fever of unknown origin.

She had been in her usual state of health until 3 weeks prior to presentation when she developed fever as high as 102°F. She continued to have daily fevers for the following 5 days; fevers were associated with abdominal pain and headache at time of fever but otherwise she was asymptomatic. She was evaluated by her paediatrician and was empirically treated with levofloxacin without resolution of symptoms.

On day 6 of fever, she was admitted to an outside institution for fever and a presumed sickle cell pain crisis. She had chest, spine and abdominal radiographs which were all reported as normal. Abdominal ultrasound was unrevealing without mentioning of lymphadenopathy. She had a robust negative infectious workup including blood and urine cultures, Epstein-Barr virus (EBV) titres, cytomegalovirus titres, Bartonella titres, Rickettsia titres and Lyme testing. She was treated with broad-spectrum antibiotics including piperacillin/tazobactam and doxycycline but daily fevers persisted. She was also given intravenous pain medications for abdominal pain and back pain for presumed vaso-occlusive pain crisis. Given the persistent prolonged fever without a source identified, she was transferred to a tertiary care centre for multidisciplinary evaluation.

Her review of systems was significant for one recent painful oral ulcer on the right side of her mouth. Specifically, she denied any rashes. She denied joint pain or swelling. She reported that the fevers changed in quality since they started; they began by occurring at any time of day but evolved to occur daily in the late afternoon. She described a rapid rise in temperature with shaking and rigours, but otherwise felt well outside of the febrile episodes.

The patient had no family history of autoimmune or autoinflammatory disorders. She had no history of recent travel, no contact with international travellers, no exposure to animals, no known exposure to ticks or wooded areas and no other notable exposures.

On examination at admission, she was afebrile and haemodynamically stable. She was well appearing without significant examination findings. She did not have any notable lymphadenopathy or rash. Her abdominal examination was normal without organomegaly.

Investigations

On admission to the tertiary care institution, laboratory evaluation revealed erythrocyte sedimentation rate of 59 mm/hour (reference: 0–13 mm/hour), C-reactive protein of 14.7 mg/dL (reference: <0.6 mg/dL), aspartate aminotransferase 698 U/L (reference: 15–41 U/L) and alanine aminotransferase 637 U/L (14–54 U/L). Otherwise, labs were only significant for haemoglobin of 94 g/L (reference: 114–155 g/L) which was her baseline value given her SCD. She was not coagulopathic and had no evidence of haemolysis. On arrival, paediatric infectious disease and rheumatology were consulted, and she was admitted to the haematology/oncology service.

Rheumatology recommended additional laboratory evaluation including autoantibody testing which revealed a negative antinuclear antibody, negative antidouble-stranded DNA antibody, negative anti-Smith antibody, negative anti-ribonucleoprotein antibody, negative anti-Ro antibody, negative anti-La antibody and negative antinuclear cytoplasmic antibody. Her antibody testing was significant for a positive low-titre antismooth muscle antibody (1:80). Her complements (C3 and C4) were normal. She had normal IgG subclasses.

Infectious disease recommended additional infectious testings including herpes simplex virus PCR, hepatitis E testing, adenovirus testing, enterovirus testing, EBV antibodies, parvovirus testing, human herpes virus 6, Ehrlichia testing, histoplasma testing and a purified protein derivative skin test was placed. All of these studies resulted in negative.

Given no clear source of fever on multidisciplinary evaluation, the decision was made to pursue imaging to evaluate for occult malignancy, abscess or inflammatory process. After discussion with radiology, optimal imaging modality was determined to be whole-body MRI. The study revealed significant axillary and retroperitoneal lymphadenopathy, not previously appreciated on physical examination. There were no skeletal abnormalities appreciated and no evidence of osteomyelitis. In addition, the abdominal organs were normal in appearance aside from evidence of prior partial splenectomy. There were no abdominal neoplasms visualised.

Given the concern for a malignant process, she subsequently underwent excisional biopsy of a 2 cm retroperitoneal lymph node. Pathology revealed altered nodal architecture and on high magnification demonstrated patchy areas of necrosis with surrounding histiocytic cell proliferation and increased apoptosis. There were many histiocytic cells that were CD68 positive as well as some histiocytic cells that were positive for myeloperoxidase (figure 1). Given these findings in the absence of any other clinical symptoms or serologically consistent signs of systemic lupus erythematosus (SLE), she was diagnosed with KFD.

Figure 1.

Histopathologic evaluation of excisional biopsy of a retroperitoneal lymph node. (A) H&E stained section shows an altered nodal architecture with paracortical lymphoid hyperplasia (right field of the image) and geographic necrosis (left of the image), ×100. (B) A high magnification demonstrates patchy areas of necrosis with surrounding histiocytic cell proliferation and increased apoptosis. Note a few histiocytic cells with nucleus compressed to the periphery (crescentic histiocytes). H&E stain, ×400. (C) In the necrotic area, many cells are positive for CD68, a histiocytic antigen marker. Immunohistochemistry for CD68, ×100. (D) Some histiocytic cells around the necrosis/apoptotic area are positive for myeloperoxidase. Immunohistochemistry for myeloperoxidase, ×400.

Differential diagnosis

The differential diagnosis of fever of unknown origin is quite broad in the paediatric population. It includes infectious, rheumatic and oncologic aetiologies with infectious aetiologies being most common when a source is identified.⁶ Often with prolonged fever without a source, a multidisciplinary approach is required and the differential evolves over time as testing results become available.

In patients with SCD, serious bacterial infections must be considered and fully evaluated given these patients have functional asplenia and are at risk for invasive bacterial infections. These patients can have fever in the setting of SCD-related illnesses such as acute chest syndrome as well. Thus, fever in this population requires prompt evaluation and empiric treatment for bacterial infections. This patient received empiric antibiotics at the beginning of her illness without resolution of fever.

Other aetiologies considered included viral processes like EBV or cytomegalovirus. Bartonella was considered given it is a common cause of fever of unknown origin. She also lived in an area where tick-borne illnesses such as Rocky Mountain spotted fever and ehrlichiosis were common, so testing was performed for these infections and she was empirically treated with doxycycline which was later discontinued.

Fevers persisted despite antimicrobial treatment, thus oncological processes such as lymphoma were considered, particularly after the whole body MRI demonstrated significant retroperitoneal lymphadenopathy. However, she had not had night sweats or unintentional weight loss, less consistent with an oncologic process like lymphoma.

Rheumatologic processes considered initially were inflammatory conditions associated with fever such as systemic-onset juvenile arthritis but seemed less likely in the absence of joint findings, rash and characteristic quotidian fever. SLE was considered as well, but she lacked clinical features of the disease and her serologic studies (including autoantibody testing) did not support this diagnosis. IgG4-related disease was considered, IgG subclasses were normal, and it was requested that tissue sample be stained for IgG4 to further evaluate for this condition, which was negative.

Treatment

After undergoing lymph node resection, she clinically improved in the days following the resection without any additional treatment. Four days after the lymph node resection, she defervesced and her inflammatory markers started to normalise.

She was evaluated by hepatology during admission, given her transaminitis and positive antismooth antibody testing. The team determined that these abnormal labs were difficult to interpret in the clinical setting of KFD and this diagnosis might interfere with the ability to definitively diagnose autoimmune hepatitis. After lymph node resection, her liver enzymes began to normalise prior to her discharge from the hospital.

Outcome and follow-up

The patient was discharged home after being afebrile for several days, and returning to her baseline appetite, energy level and pain control. She was seen in follow-up 1 month after her hospitalisation and had not had return of fever. She had normalisation of her inflammatory markers and liver enzymes. Imaging done 2 months after hospitalisation showed significant decrease in retroperitoneal lymphadenopathy.

Discussion

Fever of unknown origin is not uncommon in the paediatrics population. It is a challenging presentation, and often there is no identified aetiology despite robust multidisciplinary evaluation. Infectious causes are the most commonly reported identified aetiology. However, oncologic and rheumatologic aetiologies must be considered on the differential diagnosis as well.⁶ KFD is a very rare cause of prolonged fever and lymphadenopathy in paediatric patients and can often be a difficult diagnosis to make as symptoms mimic oncologic processes like lymphoma and tissue pathology is required for definitive diagnosis.

We reviewed the medical literature for reports of paediatric patients with SCD who developed KFD and found two other case reports: one of a 13-year-old boy⁴ and the other of a 17-year-old boy.⁵ To our knowledge, this case is the third case reported in the literature. We did find two additional case reports of young adults with SCD who developed KFD. Both of these patients developed autoimmune thyroid disease after their diagnosis of KFD.⁷ Given the rarity of KFD, it is unclear whether there is an association between SCD and KFD, or if therapies for treatment of SCD could be a causative factor.

Though there are few reported case reports, KFD should be considered in patients with SCD who present with significant lymphadenopathy and fever without improvement on empiric antibiotics, and no alternative infectious or oncologic source of fever can be identified. Greater recognition of KFD as a cause of fever of unknown origin in this patient population may limit unnecessary long-term broad antimicrobial exposure.

KFD is typically self-limited and benign. Given that the aetiology and pathogenesis remain unknown, standardisation of treatment is challenging. At this time, there are no known clinical predictors of disease duration which further complicates the decision to initiate treatment. There are varying reported rates of recurrence of KFD, with reports of up to 15% and recurrence can occur years after the initial presentation.^{8 9} Medications for treatment are largely based on case reports and experts’ opinion and include nonsteroidal anti-inflammatory medications, short courses of steroids and/or hydroxychloroquine.^{1 10 11} There are special considerations when managing patients who have SCD and KFD; steroids are typically avoided in patients with SCD for multiple reasons given concern for infection, risk of avascular necrosis and studies have demonstrated increased readmission rates for pain in vaso-occlusive pain crises with short courses of intravenous steroids as compared with placebo.¹²

It is also important to recognise the association of KFD with SLE. In several case reports, patients developed SLE after diagnosis of KFD¹³; however, patients can have both of these entities concurrently.⁹ Though SLE is often on the differential for fever of unknown origin, it is important to do a thorough investigation for SLE once the diagnosis of KFD is made. There are other reports (as mentioned above) that document development of other autoimmune disorders after KFD, such as autoimmune thyroid disease and Sjogren’s syndrome¹⁴, which prompts the question: is the development of KFD autoimmune in nature, or does it simply trigger an autoimmune process manifested later? The reported ranges of the association between KFD and SLE are up to 25% of patients with higher rates in the Asian population.¹⁵

Here, we describe the third paediatric patient with SCD who developed KFD, which was self-limited and resolved without treatment. Further studies and reviews are needed to both determine the aetiology and pathogenesis of KFD as well as characterise any clinical features which prognosticate risk of complicated disease and development of autoimmune disorders such as SLE.

Learning points.

• The differential of fever of unknown origin is quite broad including infectious, rheumatologic and oncologic aetiologies with infections being most common.

• Kikuchi-Fujimoto disease (KFD) is a rare lymphohistiocytic disorder with typical clinical symptoms of fever and lymphadenopathy.

• Though rarely reported, KFD should be considered in patients with sickle cell disease presenting with typical clinic symptoms such as fever and lymphadenopathy where no alternative aetiology is identified, and there are special treatment considerations in this patient population.

• KFD is associated with the development of systemic lupus erythematosus.