INTRODUCTION
In the year 1961, Petersdorf and Beeson defined pyrexia of unknown origin (PUO) as persistent fever for more than 3 weeks or in spite of investigations in the hospital for more than 1 week. PUO is now defined as persistent fever at the end of three outpatient visits or 3 days in the hospital without finding a cause or in spite of 1 week of relevant investigations. The common causes of PUO include bacterial infections [tuberculosis (TB), subacute bacterial endocarditis, and brucellosis], viral infections [EB virus (Epstein–Barr virus) and CMV (cytomegalovirus)], fungal and parasitic infections (kala-azar), noninfective illnesses (rheumatic diseases and malignancy), and rarely central fever due to hypothalamic disturbances, hyperthyroidism, heat fever, and drug fever. Analysis of detailed history and focused physical examination are the prerequisites of a rational approach to diagnosis which is then confirmed by relevant tests.
CASE 1
A 6-year-old child presented with fever for the last 3 months. The child was well prior to the onset of present illness when he started with high fever. After few days of symptomatic therapy, he was treated with multiple antibiotics such as amoxicillin, gentamycin, and ceftriaxone but fever persisted to varying degree. He had poor appetite and lost 4 kg weight. There were no other symptoms. There was no history of recent travel or contact with animals. There was no significant family history. The physical exami-nation showed chronically sick child, with weight 17 kg, height 109 cm, moderate pallor, enlarged liver with 10-cm span, firm in consistency, not tender, spleen 4 cm, no jaundice, ascites, significant pallor, lymphadenopathy, skin rash, and joint involvement. Other systems were normal.
Q1. What is differential diagnosis?
Ans. This child has progressive inflammatory disease as suggested by prolonged fever with loss of weight over 3 months. Hepatosplenomegaly is the only positive finding in this child. Absence of jaundice rules out 2hepatocyte disease and absence of ascites rules out portal hypertension. So, hepatosplenomegaly in this child is likely due to reticuloepithelial cell involvement secondary to infection or noninfective disorder such as systemic inflammatory disease or malignancy. Systemic inflammatory disorder would have manifested with skin rash, arthritis, mouth ulcers, or other organ involvement over few weeks and so is unlikely. Absence of severe anemia, purpura, or lymphadenopathy rules out hematological malignancy though histiocytosis including hemophagocytic lymphohistiocytosis (HLH) is a possibility as anemia and thrombocytopenia may be subclinical. Slowly progressive chronic infections include tuberculosis, brucellosis, CMV, malaria, and kala-azar besides fungal infection. Malaria and kala-azar would have severe pallor and large splenomegaly while CMV would have localized in some organs causing dysfunction such as jaundice in hepatitis. Fungal infection presents in an immunocompromised host and so unlikely in this child. Primary streptococcal infection may present with prolonged fever due to toxins produced by bacteria but fever may not last so long for 3 months and hence unlikely in this child. So, differential diagnosis in this child would include infections such as tuberculosis and brucellosis, though histiocytosis including HLH secondary to infection is not obvious.
Q2. What investigations would you consider?
Ans. Complete blood count (CBC), peripheral smear, and erythrocyte sedimentation rate (ESR)—neutrophilic leukocytosis may favor brucellosis, anemia, and thrombocytopenia may suggest HLH. High ESR is expected in all three conditions:
- Abdominal ultrasonography (USG)—may reveal liver and spleen echostructure that may offer clue to diagnosis
- Chest X-ray—to look for tuberculosis focus
- Results of these tests should decide further tests.
Laboratory Test Results
- CBC: Showed moderate neutrophilic leukocytosis with thrombocytosis and moderate anemia—normocytic normochromic with ESR 108 mm
- Chest X-ray: Normal
- Abdominal USG: Showed multiple small abscesses in liver and spleen.
Q3. What further tests would you order?
Ans. Multiple small abscesses may suggest brucellosis or abscesses may be mistaken for granulomas. So one must order serological tests for brucellosis; it is a noninvasive test and if the antibody test is positive, diagnosis can be reasonably confirmed. Computed tomography (CT) scan of abdomen is done to confirm abscesses in liver and spleen as granulomas may be mistaken for abscesses on USG.
Blood culture and pus culture from USG-guided drainage are performed to rule out common bacterial infections. As common bacterial infection is 3not likely and it is not easy to culture Brucella organism, this test may not help. PCR may be the most specific test for diagnosis of infections but often it is not available.
Laboratory Test Result
Brucella antibody test was positive though this antibody is known to cross-react with many other bacteria and so is not confirmative.
Final Diagnosis
Diagnosis of brucellosis was made on circumstantial evidence of prolonged fever resistant to commonly used antibiotics and presenting with multiple abscesses in liver and spleen. Child was treated with oral tetracycline and improved.
Q4. Why did this child not respond to multiple antibiotics?
Ans. Antibiotics of choice for brucellosis are tetracycline, rifampicin, and aminoglycosides. As Brucella bacteria multiply inside the cells, antibiotic needs to be continued for 4–6 weeks. Empirical antibiotic use is usually restricted to few days and in case of no response, antibiotic is changed. This explains why gentamycin did not work in this child.
Take Home Message
Prolonged fever may be due to infection or noninfective disorder. If it is due to infection, it may be chronic infection, partially treated infection, or infection-induced immunological or toxin-mediated disorder. Partially treated infection often localizes to kidneys, lungs, or brain and usually provides some clues. Partially treated typhoid fever may present without localization, but fever does not last for 3 months. Toxin produced by bacterial infection such as streptococci may prolong fever but not for 3 months. Noninfective cause must be ruled out in a child with fever for 3 months and include evolving systemic inflammatory disorder or malignancy, often hematological.
CASE 2
A 10-year-old child presented with fever for last 3 months. He was well prior to onset of present illness. It started with low-grade fever that has varied in intensity over time in spite of various drugs used including antibiotics. He has poor appetite and has lost some weight. Over the last one month, he seems to get tired with accustomed exertion and is reluctant to go out to play. There are no other symptoms. Physical examination showed mild fever in a child who looked chronically sick. His pulse rate was 120/min and respiratory rate 25/min. He had moderate hepatomegaly firm in consistency with span of 10 cm, mildly tender, spleen not palpable, engorged neck veins and absent hepatojugular reflux (HJR), no murmur or cardiomegaly, chest was clear, and other systems normal.4
Q1. What is differential diagnosis?
Ans. This child has presented with subacute onset of slowly progressive disease with exertional tiredness, disproportionate tachycardia (heart rate faster than that expected with fever), and mildly tender hepatomegaly without jaundice. It suggests cardiac and liver involvement of long duration. Mild hepatic tenderness indicates either mild inflammation or congestion. Bacterial or amoebic liver abscess is unlikely as there would have been severe tenderness with fast progression. Chronic hepatitis should have presented with jaundice. However, mild jaundice may not be visible clinically and may be picked up by blood test. If not inflammation, then it may be congested liver and engorged neck veins and disproportionate tachycardia support such a possibility. However, it is not cardiac failure as there is no cardiomegaly and HJR is absent. Engorged neck veins without HJR suggest obstruction to superior vena cava and congested liver is due to obstructed inferior vena cava. So diagnosis is constrictive pericarditis. Fever of long duration is in favor of chronic infection, most likely to be tuberculosis. Restrictive cardiomegaly would be another possibility as it also presents with gradually progressive cardiac disease without cardiomegaly. So, differential diagnosis in this child stands to be constrictive pericarditis due to tuberculosis and restrictive cardiomyopathy must be ruled out.
Q2. What investigations would you consider?
Ans.
- CBC: May not be directly helpful except to pick up comorbid deficiency anemia
- ESR: Significantly high ESR may favor infection
- Chest X-ray: Expected to show normal size heart without signs of cardiac failure, to look for focus of tuberculosis
- 2D echocardiogram: Expected to reveal diminished filling of cardiac chambers and may show thickened pericardium
- CT chest: May not add any more information
- Mantoux test: Positive test does not help to diagnose tuberculosis while negative test at the age of 10 years may be taken against diagnosis of tuberculosis
- Gastric aspirate for AFB (acid-fast bacilli): It may be worth it though in absence of obvious lung lesion, test may be negative
- Biopsy: It is too invasive test and better avoided.
Laboratory Test Results
- CBC was within normal limits
- ESR was 85 mm/end of 1 hour
- 2D echocardiogram showed thickened pericardium
- Gastric aspirate did not show AFB.
Final Diagnosis
- Constrictive pericarditis due to TB
- Child was treated with anti-TB treatment and improved.
Take Home Message
Diagnosis in a child presenting with prolonged fever evolves over time as physical signs appear as disease progresses. Periodic physical examination often offers a clue to diagnosis more than randomly ordered laboratory tests. Physical finding of enlarged liver demands observation for engorged neck veins, if any that may suggest not a primary liver disease but evidence of venous congestion or obstruction. Empirical antibiotic therapy should be avoided and minimum relevant laboratory tests must be ordered prior to antibiotic therapy. Partially treated bacterial infections pose a challenge to diagnosis and usually end up with increased morbidity. Noninfective diseases should always be kept in mind.
CASE 3
A 5-year-old child presented with fever off and on for last 3 months. He was well prior to present illness. It started as mild-to-moderate fever that increased over next few days to higher degree. Fever would spike 3–4 times a day at the interval of 6–8 hours and reduce in severity after antipyretic drugs. Fever pattern hardly changed over last 3 months though there were periods of low-grade fever interspersed with high fever. There were no other significant accompanying symptoms. Few antibiotics were tried in succession for a period varying from 4–7 days without any sustained improvement. Several tests were done that included CBC (repeated several times), blood and urine cultures (often done after antibiotic therapy), cerebrospinal fluid (CSF) examination and culture, serology for various infections, imaging studies, rheumatological tests, and bone marrow examination. Except persistent neutrophilic leukocytosis with thrombocytosis, there were no significant abnormalities in other test results. Considering systemic inflammatory disease, steroids were tried but stopped after few days as the child seemed to deteriorate. This child was referred for further evaluation. Physical examination at the end of 3 months revealed the following:
- Weight 14 kg (had lost 5 kg)
- Height 104 cm
- Temperature 100°F
- Pulse 140/min
- Respiration 25/min
- Blood pressure 90/55 mm
- Mild pallor
- No other significant findings on general examination
Q1. What is differential diagnosis?
Ans. This child is progressively deteriorating as evident by loss of 5 kg over last 3 months in a previously healthy child. Mild-to-moderate fever gradually increasing in severity over next few days may suggest initial bacteremia that settled in some organ resulting in higher degree of fever. However, organ in which it may have settled has not manifested with specific symptoms. This is what happens typically in a typhoid fever. However, it would have usually responded to several antibiotics, unless this bacterial strain was partially resistant. Thus, nonlocalizing partially treated bacterial infection is probable in this child. Other infections including tuberculosis, chronic viral, fungal, or parasitic infections are unlikely as they would have been localized by now. Fever of long duration may also be systemic onset of inflammatory disorder but would have by now developed joint involvement, skin rash, or organ affection. Common malignant disorders in children are hematological and by now would have manifested with symptoms such as pallor, or bleeding. Thus, this child is probably suffering from bacterial infection that is not properly treated. The only significant physical finding in this child is disproportionate tachycardia for age, degree of fever, and anemia while respiratory rate and blood pressure are within normal limits. This suggests cardiac involvement without any evidence of structural defect or it is an acquired disease and fever denotes inflammation, mostly infective as there are no other signs of systemic inflammatory disorder. As this is a chronic progressive infection, it favors diagnosis of subacute bacterial endocarditis.
Q2. What investigations would you consider?
Ans. Specific investigation would be echocardiogram for evidence of endocarditis and any structural defect.
- Chest X-ray for cardiomegaly and status of pulmonary circulation
- Electrocardiogram (ECG) for any rhythm disturbance
- Blood culture is done for etiological diagnosis. Multiple blood samples would enhance bacterial yield
- CBC is done for evidence of acute bacterial infection
- Abdominal USG is done to assess spread of infection in other organs such as liver, spleen, or kidneys even in absence of clinical findings.
Laboratory Test Results
- 2D echocardiogram showed vegetation on mitral valve. There was no structural defect in the heart
- Chest X-ray revealed mild cardiomegaly
- ECG was within normal limits
- Blood culture grew Streptococcus viridans sensitive to penicillin
- WBC 18,000/mm3, P 72, L 26, M 2, E 0, Hb 10 g%, Pl 3.2 lakhs
- Abdominal USG normal.
Final Diagnosis
Subacute bacterial endocarditis without any valvular defect is diagnosed. Diagnosis is based on vegetation seen on echocardiogram and blood culture 7showing growth of S. viridans. This type of infection is commonly seen in damaged valve while infection in normal valve presents more acutely. This child presented with subacute illness due to partial treatment with antibiotics. As antibiotics in such a case need to be continued for at least 4 weeks, early discontinuation or change of antibiotics must have caused acute illness behave like subacute disease.
The child was treated initially with intravenous penicillin and gentamycin but after blood culture report suggested sensitivity to penicillin, gentamy-cin was omitted. Therapy has to be continued for 4–6 weeks till vegetation disappear and so also reversal of all other laboratory and clinical parameters. Supportive treatment includes symptom relief and nutritional rehabilitation.
Take Home Message
Bacterial endocarditis is often subacute or chronic infection because heart valves have no dedicated blood supply and so when bacteria get attached to the valve to form vegetation, antibiotics do not reach the site of disease. It is also the reason why antibiotics need to be administered intravenously for 4–6 weeks to ensure control of infection. Thus this is one of the bacterial infections that may continue to manifest for many weeks, and hence referred to as subacute bacterial endocarditis. Thus, this is one example of chronic active bacterial infection.
CASE 4
An 8-year-old child presented with fever for last 3 months. He was apparently well prior to onset of present illness that started with an erratic fever pattern. At times, he would get severe rigors with high fever ending with sweating and normal temperature while at other times, fever would be continuous, low grade for several hours. He had no other significant accompanying symptoms. However over the last 3 months, he had poor appetite and had lost considerable amount of weight and became severely cachectic. Several tests were done without any clue to diagnosis though tests revealed severe anemia for which he received packed red blood cell transfusion and so also few courses of antibiotics and antimalarial drugs were tried without benefit.
On physical examination, the child looked wasted and chronically sick with severe pallor.
Weight 16 kg, height 120 cm, pulse rate 134/min, RR 36/min, no lymphadenopathy, bone and joints normal, no edema.
Abdomen distended, more in upper part, liver 5 cm below costal margin, firm, not tender, liver span 12 cm, spleen 8 cm firm, no ascites, soft systolic murmur over precordium and other systems were normal.
Q1. What is differential diagnosis?
Ans. This child has severe progressive disease as suggested by development of cachexia and had also resulted in severe anemia requiring transfusion.8 It indicates a probable hematological disorder. Long-duration fever may favor diagnosis of malaria; however, it would have been easy to prove by simple tests and moreover this child did not respond to trial with antimalarial drugs. Severe anemia requiring transfusion at this age is not likely to be deficiency anemia. There has been no history of blood loss or evidence of hemolysis in the form of jaundice. Hence, anemia in this child must have resulted from bone marrow disorder. Bone marrow aplasia is ruled out as hepatosplenomegaly is not a feature and it would have presented with bleeding manifestations in the form of purpura or ecchymosis. Acute lymphatic leukemia is less likely as disease often manifests over short time and cachexia is not a feature. Chronic myeloid leukemia mostly presents without significant anemia. Bone marrow infiltration is often a slow progression and those with prolonged fever may be due to myelofibrosis or HLH; both are commonly secondary to infection that may go unnoticed. Another possibility is a chronic infection itself such as kala-azar (leishmaniasis). In fact cachexia due to disease, as often seen in malignant disorders in adults, is rare in children but untreated kala-azar does present in similar way.
Q2. What investigations would you consider?
Ans.
- CBC, peripheral smear, and ESR
- Reticulocyte count
- Bone marrow examination
- Depending upon bone marrow examination result, further tests need to be planned to rule out either HLH or kala-azar.
Laboratory Test Results
- Hb 4 g% microcytic hypochromic anemia
- WBC 16500, P 63, L 30, M 4, E 3, Platelet 0.35 lakh, ESR 120 mm
- Corrected reticulocyte count 3%
- Bone marrow examination showed Leishmania donovani (LD) bodies
- No further tests were carried out as diagnosis is confirmed. Serological tests are not dependable.
Final Diagnosis
Bone marrow showing LD bodies is the gold standard of diagnosis of kala-azar. He was treated with amphotericin B. Liposomal preparation is most preferred but costly. Antimony compounds are also used in the treatment of kala-azar such as sodium stibogluconate. Child recovered completely.
Take Home Message
Severe anemia as a significant feature often suggests hematological disease, either primary or secondary. Anemia with hepatosplenomegaly is a feature of either hemolytic anemia or bone marrow infiltration. A chronically sick and febrile child almost favors bone marrow infiltrative disorders. Diagnosis is 9confirmed only on bone marrow examination and at times marrow aspiration fails and marrow biopsy may be necessary. Similarly in case of strong clinical suspicion of bone marrow involvement, single bone marrow examination may not pick up the diagnosis and repeat examination may be necessary. This is because many diseases evolve over time to offer classical laboratory results.
SUGGESTED READING
- Bennett JE, Dolin R, Blaser MJ. Mandell, Douglas, and Bennett's Principles and Practice of Infectious Diseases. Elsevier Inc., 2014. 3697 p.
- Longo DL, Fauci AS, Kasper DL, Hauser SL, Jameson J, Loscalzo J (Eds). Harrison's Principles of Internal Medicine, 18th edn. New York, NY: McGraw-Hill; 2012.
- Parthasarthy A. IAP Textbook of Pediatrics, 7th edn. India: Jaypee Brothers Medical Publishers; 2019.