Boy with fever, occasional vomiting and diarrhea, enlarged lymph node

What’s your diagnosis?

ByJames H. Brien, DO

Fact checked byCarol L. DiBerardino, MLA, ELS

Add topic to email alerts

Please provide your email address to receive an email when new articles are posted on Practice Management in Rheumatology.

A previously healthy, 8-year-old Korean boy is admitted with a 14-day history of daily low-grade but rising fever and an 8-day history of occasional vomiting and nonbloody diarrhea.

Despite these complaints, between episodes of fever, he seems to feel well, with normal appetite and play activity and no weight loss. No one else in the home is sick, and there are no known sick contacts. There has been no recent travel or animal exposure. However, it is summer, and he is exposed to a variety of biting insects because he spends much of the day playing outside.

The patient’s past medical history is remarkable only for an episode of Kikuchi-Fujimoto disease 1 year earlier, diagnosed by excisional biopsy of a painfully swollen cervical lymph node. He has been healthy since then prior to the current complaints.

Examination revealed an alert, well-appearing 8-year-old male, in no distress, with a temperature of 102.4°F and otherwise normal vital signs and no weight loss. The significant findings included an enlarged, mildly tender, nonerythematous, nonfluctuant right-sided cervical lymph node measuring 3 cm by 1.5 cm (Figures 1 and 2). He also had some mild, painless hepatosplenomegaly. The rest of his examination was normal.

Figure 1. Cervical lymph node. Image: James H. Brien, DO

Figure 2. Cervical lymph node. Image: James H. Brien, DO

An array of blood tests revealed elevated inflammatory markers; a normal complete blood count (with no atypical lymphocytes); a normal metabolic profile, including liver enzymes; and a negative respiratory viral panel. Other negative tests included serologies for cat scratch disease, toxoplasmosis and a heterophile test. However, a rectal swab for enterovirus by PCR was positive.

A biopsy of the swollen node seen in the figures was performed. Gram and acid-fast stains were negative; however, since I’m not good at reading hematoxylin and eosin (H&E) cellular stains, the pathologist’s reading was “polymorphous reactive lymphoid cells with numerous phagocytic histiocytes and minimal lymphocytes” (Figure 3). There were no malignant cells seen, and flow cytometry revealed “a nonspecific immunophenotype, with a mixture of B and T cells, and T-lymphocyte predominance, supporting a reactive process.”

Figure 3. Lymph node biopsy, 400x hematoxylin and eosin stain revealing polymorphous reactive lymphoid cells with numerous phagocytic histiocytes and minimal lymphocytes. Image: James H. Brien, DO

What’s your diagnosis?

A. Infectious mononucleosis

B. Kikuchi-Fujimoto disease

C. Kimura disease

D. Scrofula

Answer and discussion

The answer is B, Kikuchi-Fujimoto disease (K-F disease; often referred to as Kikuchi disease), which was independently described by Kikuchi and Fujimoto in Japan in 1972. This is a very rare disease. According to the NIH, less than 5,000 cases have been reported in the United States so far. The diagnosis is made by recognizing the characteristic lymph node biopsy results (polymorphous reactive lymphoid cells with numerous phagocytic histiocytes and minimal to no lymphocytes). At the same time, it is also obviously important to rule out cancer, which may share similar symptoms early on. The clues in this case were the history of being previously diagnosed with this condition by biopsy, in a Korean child, raising the likelihood of a recurrence. One Korean paper found as many as 40% of pediatric patients had a recurrence.

Infectious disease triggers have been proposed by many but without proof. In this case, the finding of enterovirus in the stool raises the question of association — but again, it is not proof of cause, especially since enteroviruses tend to peak in the summer months, making coincidence possible. Recommendations for treatment also suffer from a lack of evidence-based studies due to the rarity of the disease. There is a suggestion that hydroxychloroquine may be helpful. Steroids and nonsteroidal anti-inflammatory agents are also used. For more information, an excellent resource can be found by searching the NIH Genetic and Rare Diseases Information Center (https://rarediseases.info.nih.gov/diseases/6834/kikuchi-fujimoto-disease). The patient recovered without further concerns.

Infectious mononucleosis, usually caused by Epstein-Barr virus (EBV), may result in similar prolonged symptoms, including splenomegaly, but severe pharyngitis/tonsillitis is commonly seen, which was not seen in this patient. Additionally, mononucleosis would be more likely in an older child. The negative heterophile test, as well as there being no atypical lymphocytes in peripheral blood, leads away from this diagnosis. However, if not for the previous history of K-F disease, EBV serologies might have been done to confirm the negative heterophile test, which has a 25% false-negative rate.

Kimura disease, another rare benign inflammatory condition of unknown etiology primarily affecting males of Asian descent, is characterized by subcutaneous nodular lesions, mostly in the head and neck area, and can also result in cervical lymph node swelling. On biopsy, eosinophilic inflammation is found, which has led experts to believe that the condition is likely triggered by an allergic response. This has guided therapeutic recommendations, including steroids and other immunosuppressive agents. Depending on the location, radiation therapy may be used, but many resort to surgical removal. A quick and free review of this rare disorder can be found at https://dermnetnz.org/topics/kimura-disease.

Lastly, scrofula is the old term for tuberculous or nontuberculous mycobacteria (NTM) infection of the cervical lymph node(s). The term comes from the Latin word scrofa, later becoming scrofula, meaning brood or breeding sow, as a comparative description of the jowls or neck of the sow with the appearance of the cervical lesions of the patient, which may be multiple (Figure 4, NTM cervical adenitis; courtesy of J.W. Bass collection), and prolonged in the days prior to effective therapy. Most recommend surgical removal of NTM-infected lymph nodes, with or without medical treatment. Medical treatment can be with azithromycin or clarithromycin for 6 to 12 weeks, with or without ethambutol or rifampin, depending on sensitivities. Thorough surgical debridement, as seen in Figures 5 to 8, it should be curable. Tuberculous adenitis should be treated with prolonged multidrug therapy (three to four drugs) depending on sensitivities. Surgical removal is usually not necessary, but consultation with an infectious disease specialist with either diagnosis is also recommended. Guidance can be found in Nelson’s Pediatric Antimicrobial Therapy, 2024 edition, page 4.

Figure 4. Nontuberculous mycobacteria infection of cervical nodes, one with postsurgical removal. Image: Courtesy of J.W. Bass Collection

Figure 5. Nontuberculous mycobacteria infection of cervical node (scrofula). Image: James H. Brien, DO

Figure 6. Nontuberculous mycobacteria infection of cervical node (scrofula) surgical debridement. Image: James H. Brien, DO

Figure 7. Nontuberculous mycobacteria infection of cervical node (scrofula) postsurgical debridement. Image: James H. Brien, DO

Figure 8. Granulation tissue found in nontuberculous mycobacteria infection of cervical node (scrofula) seen in Figures 4 to 6. Image: James H. Brien, DO

Columnist comments

I hesitate to show a disease you will likely never see, but for those readers who may come across this unusual condition, this case may awaken a neuron from a long sleep. As usual, if you have a case with a publishable picture that you would like to see appear in this column, please send it to me for review. If suitable, you can be a coveted “guest columnist.”

Please keep in touch: jhbrien@aol.com.