Article Text

An all-consuming case
  1. E Jameson2,
  2. P Rawling1,
  3. D Thiruchunapalli3,
  4. J Stahlschmidt4,
  5. S Frazer1
  1. 1Bradford Teaching Hospitals NHS Foundation Trust, Bradford, Yorkshire, UK
  2. 2Royal Manchester Children's Hospital, Manchester, UK
  3. 3Leeds Radiology Academy, Leeds, UK
  4. 4St James's University Hospital, Leeds, UK
  1. Correspondence to Dr Simon Frazer, Bradford Royal Infirmary, Duckworth Lane, Bradford, Yorkshire BD9 6RJ, UK; simon.frazer{at}


This is a case of a 6-month-old child who presented to the local paediatric department with a relatively short history of developmental regression, hypotonia, weight loss, irritability and hepatosplenomegaly. The child then proceeded to spike temperatures up to 39°C.

The case demonstrates the logical investigative process involved in such a case and discusses the differential diagnoses at each stage. It also highlights the importance of multi-specialty work in complex cases and the potential consequence of premature closure of diagnosis.

Statistics from

A health visitor referred a 6-month-old male Asian to the local paediatric department because of weight loss and significant concerns about the child's development. The patient had weight loss of 400 g over the past 8 weeks, which had initially been attributed to inadequate breast feeding; thus, weaning had commenced. His growth chart showed that his weight had crossed from the 50th to below the 0.4th centile. The parents reported increasingly unsettled behaviour with writhing of the limbs and arching of the back and an occasional cough and temperature. His mother felt that over the past 2–3 weeks, he had “taken a step backwards”. The mother had presented to the general practitioner on numerous occasions with these concerns and high temperatures. There were no other reported symptoms.

The patient was the youngest of three children born to consanguineous parents (first cousins). His mother had one previous termination for fetal abnormality. The siblings and parents were all reported to be well. There was no significant family history.

The patient was born at 36 weeks by a normal vaginal delivery. Antenatal scans had shown extra fluid in one of the ventricles of the brain, but a postnatal scan was normal so follow-up was not indicated. He had no previous admissions to hospital but had received antibiotics at the age of 8 weeks for infected eczema; this had subsequently improved with topical treatments.

On initial examination, he was apyrexial but mildly dehydrated. He was not obviously dysmorphic but did have low-set ears. He was irritable on handling and hypotonic. He had poor head control and was unable to sit even with support, and he was not observed rolling. He did have a bilateral grasp, but it was weak and he could not transfer objects. There were no obvious concerns regarding his vision and hearing, although his interaction was poor. He was noted to have significant hepatosplenomegaly.

In summary, the main features in the case at presentation were developmental regression, hypotonia, weight loss, irritability and hepatosplenomegaly. The patient was, therefore, investigated as per the local guideline (see figure 1 and refer to table 1) that explains the disorders screened for with the tests. In addition, routine blood samples were sent, that is, full blood count, urea and electrolyte levels, bone profile, ferritin level, thyroid function and chromosomes. An abdominal ultrasound in view of organomegaly, a brain magnetic resonance imaging (MRI) and an ophthalmologic review were also requested.

Figure 1

Investigation of a child with developmental delay in two or more areas.

Table 1

Explanation of metabolic tests done in figure 1

Comment 1: developmental delay with organomegaly

The combination of developmental regression and hypotonia in a child where the parents are first cousins suggested the possibility of a genetic neurodegenerative disorder; the presence of organomegaly would lean more towards this being a storage disorder. There are many storage disorders, but this constellation of symptoms with the absence of dysmorphism is most suggestive of Gaucher or Niemann–Pick disease.

Gaucher disease

This is a sphingolipid metabolic disorder with the primary defect being a block in the lysosomal degradation of glucosylceramide and glucosylsphingosine. There are three clinical phenotypes:

  • Type 1: 80%–90% of all cases with a particularly high prevalence in Ashkenazi Jews (carrier frequency 1 in 13). It is defined by the lack of neurological symptoms, and although it can present at any age, it is usually diagnosed in adulthood. The clinical phenotype varies from asymptomatic through to visceral, haematological and skeletal disease.

  • Type 2: Acute neuronopathic Gaucher disease classically presents in early infancy with brainstem dysfunction, organomegaly, pronounced spasticity, failure to thrive and cachexia. There is a rapid clinical deterioration resulting in death by 2 years of age.

  • Type 3: Subacute or chronic Gaucher disease is recognised by severe systemic involvement and supranuclear saccadic horizontal gaze palsy, with or without developmental delay. Some patients will develop a myoclonic encephalopathy. Differentiation between this and type 2 is difficult but vital for prognostic reasons.

Treatment where indicated is with either enzyme replacement therapy or substrate reduction therapy. Enzyme replacement therapy aims at restoring the degradation of the substrate, whereas substrate reduction therapy aims to reduce disease burden by slowing the rate of substrate synthesis.

Niemann-Pick disease

Niemann–Pick disease is a heterogenous group of disorders that lie on a clinical spectrum. Historically, it has been classified into types A and B (due to sphinomyelinase deficiencies) and type C (due to lipid trafficking defects). Niemann–Pick disease due to sphingomyelinase deficiencies results in accumulation of sphingomyelin in systemic organs and the brain and has early- and late-onset forms. The early form classically presents with diarrhoea, vomiting and failure to thrive in the first weeks of life. Hepatosplenomegaly and lymphadenopathy occur by 3–4 months of age, with psychomotor retardation being evident by 6 months of age. Patients eventually become spastic and rigid, with death usually occurring between 18 months and 3 years of age. A cherry-red spot is seen in 50% of patients. The late-onset form has a more chronic onset presenting in late infancy or early childhood with splenomegaly or hepatosplenomegaly, commonly associated with faltering growth. Life expectancy may be reduced owing to associated lung disease and massive hepatosplenomegaly.

Niemann–Pick disease due to lipid trafficking defects also has a heterogenous clinical course. All patients will develop neurological symptoms, and classification is based on the age of onset of these. There are thus infantile, late infantile/juvenile and adult onset pictures. The juvenile form is the most common, and hepatosplenomegaly is virtually always present. The child will typically present with clumsiness due to ataxia at 3–5 years of age. The ataxia progresses to dysphagia, dysarthria and, ultimately, dementia and death. The classic clinical finding is of vertical supranuclear gaze palsy, and 20% of patients will have gelastic epilepsy.

Within a couple of days of admission, spikes in temperature up to 39°C were recorded. A septic screen was performed (full blood count, C-reactive protein (CRP) clotting, blood cultures, urine for microscopy and culture). A lumbar puncture was considered but not performed at this stage because of deranged clotting (international normalised ratio 1.4). The results of note were haemoglobin level of 9.1g/dL, white blood cell count of 22 10*9/L (neutrophil count 11), CRP level of 46 mg/L (peaked at 93), albumin level of 27 g/L and ferritin level of 134 ug/L. He was given cefotaxime, and acyclovir was added in to cover for herpes encephalitis. A dose of vitamin K was given to correct the clotting.

Despite a 2-week course of antibiotics, the patient continued to spike temperatures. The results of several blood cultures taken during these 2 weeks were negative, as were those of the urine and cerebrospinal fluid (CSF) cultures, and a chest x ray result was normal. A “drug holiday” was tried to rule out the possibility of antibiotic-related fever, but this was unsuccessful. This development of pyrexia of unknown origin (PUO) could not be readily linked to a possible metabolic disorder; thus, second-line PUO investigations were done in the form of Quantiferon and human immunodeficiency virus testings. CSF was examined for acid–alkali fast bacilli. Blood cultures taken while off antibiotics remained negative. The Quantiferon was equivocal.

While the antibiotics were discontinued, he developed a swollen, tender left upper leg. This was initially investigated by Doppler ultrasound (to exclude thrombosis) and x rays. Antibiotics were recommenced in case of osteomyelitis. In addition, during this time, the blood film was found to show the presence of vacuolated lymphocytes, and specialised metabolic testing showed elevated chitotriosidase on white cell enzyme testing. These results made the differential diagnosis strongly suggestive of either a storage or a haematological disorder; however, there were further possibilities, as highlighted in table 2.

Table 2

Differential diagnosis: factors for and against each while radiological investigations were awaited

During the admission, the patient underwent radiological investigation as follows:

  • Abdominal ultrasound in view of hepatosplenomegaly (figure 2): the latter was confirmed. The liver showed fatty enlargement with small hypoechoic areas throughout.

  • Doppler ultrasound of the legs to exclude thrombosis: normal.

  • x Ray of the left leg and arm (figures 3 and 4): lytic lesion in bones.

  • Skeletal survey triggered by the finding of the lytic lesion: numerous lytic lesions in the long bones and the skull.

  • MRI of the legs done to visualise the bony lesions further: bone marrow changes but not suggestive of osteomyelitis.

  • MRI head in view of regression: cerebral atrophy.

Figure 2

Abdominal ultrasound showing liver lesion.

Figure 3

x Ray of the leg showing lytic lesions.

Figure 4

x Ray of the arm showing lytic lesions.

Comment 2: imaging findings and radiological differential diagnosis

The images demonstrate multiple lytic lesions that posed initial diagnostic dilemmas, as the differential diagnosis is wide:

  • Infantile myofibromatosis, which is a rare mesenchymal disorder of childhood often occurs at an earlier age, usually birth, which can affect the skin, viscera, bone and subcutaneous tissues. Affected bones are usually the skull, the orbit and long bones where there are sharply defined lucent lesions. This has spontaneous resolution with good prognosis, unless there is visceral involvement.

  • Langerhans cell histiocytosis (LCH) can affect any age group, but most cases occur in children <12 years old; although this child was very young to have this in the differential, this cannot be completely excluded. This is a multisystem disorder with a well-known propensity to involve predominantly craniofacial bones as solitary or multiple sharply circumscribed punched out lytic lesions.

  • Metastatic neuroblastoma and leukaemia may have similar radiographic appearances. Two-thirds of cases of neuroblastoma are seen in children aged <5 years. The most common symptom is bone pain with common sites of neuroblastoma metastases being the cortical bone (56% of cases) and the bone marrow (70% of cases).

  • Multifocal pyogenic osteomyelitis may also have similar presentation; however, the MR findings were not suggestive of this condition, and it usually involves the metaphysis.

It was still felt that the most likely diagnosis was Gaucher or Niemann–Pick disease; however, LCH was also a strong possibility in view of the fever. The key to diagnosis was deemed to be biopsy of the lesions. The patient was transferred to the local haematology–oncology centre for the surgical procedure.

While there, the patient became more unwell and developed hypoalbuminaemia with marked oedema. A bone marrow biopsy, an excision biopsy of one of the lesions in his leg (that had pus in it) and a liver biopsy were performed.

The patient was subsequently given methylprednisolone for a presumptive diagnosis of LCH, with dramatic improvement, while the pathological result was awaited.

Comment 3: narrowing it down

Langerhans cell histiocytosis

LCH is a clonal proliferation of pathological Langerhans cells that invade a range of organs. The characteristic findings include lytic bone lesions with 5%–10% of cases showing multisystem disease with organ failure (Letterer–Siwe disease).1 If multi-system disease is present, treatment is with steroids and cytotoxic drugs, although prognosis remains poor with 10%–20% succumbing, and of the survivors, at least 50% will have sequelae.1

Storage disorder

Vacuolated lymphocytes are seen in many storage disorders such as I-cell disease, mucopolysaccharidoses, Niemann–Pick type Ia disease, Pompe disease and GM1 gangliosidosis. The white cell enzyme studies had shown moderately elevated chitotriosidase. Chitotriosidase is a chitinase that is expressed by activated macrophages. It is detected on the white cell enzyme screen. Chitotriosidase levels are massively elevated in Gaucher disease (when it is used as a biomarker) but can also be elevated in other lysosomal storage disorders and non-metabolic conditions such as sarcoidosis and thalassaemia. At the level in this case, it would be more in keeping with Niemann–Pick disease than Gaucher disease, which tends to be associated with higher levels. A bone marrow aspirate can differentiate between the two by identifying either Gaucher cells or, in the case of Niemann–Pick disease, foamy histiocytes.

The pathologic result returned and showed that the soft tissue lesion from the left thigh was characterised by mature fatty tissue with focal fat necrosis and a central abscess cavity lined predominantly by histiocytes. Plasma cells and neutrophilic inflammation was conspicuously sparse. Interestingly, one poorly formed granuloma with no caseous necrosis was identified. The Ziehl–Neelsen stain revealed numerous acid-fast bacilli.

The two liver core biopsies showed numerous epithelioid granulomata in portal tracts and within lobules. The portal tracts were otherwise normal, and there were no biliary features or fibrosis. There were also macrovesicular fatty changes and features of mild chronic cholestasis. Storage cells were not noted. The Ziehl–Neelson stain revealed numerous acid-fast bacilli within granulomas and tissue-specific macrophages, that is, Kupffer cells (see figure 5).

Figure 5

Numerous acid-fast bacilli within granulomata and Kupffer cells (Ziehl–Neelsen stain, original magnification ×20).

In both samples, a diagnosis of granulomatous inflammation with plentiful acid-fast bacilli was made. There was no evidence of myofibromatosis. This was, therefore, consistent with a diagnosis of BCGosis.

Comment 4: pathologic diagnosis

The findings of non-caseating granulomata in a biopsy warrant the performance of a test for acid-fast bacilli. The standard procedure for the detection of mycobacteria in paraffin-embedded tissue is the Ziehl–Neelsen stain that uses the ability of mycobacterium to resist the decolourisation by mineral acids after staining with an arylmethane dye. Compared with special microbiology techniques including polymerase chain reaction for mycobacterium DNA species, histopathological staining methods are unreliable.

Hepatic granulomas may be found in up to 10% of liver biopsies. The non-portal distribution of granulomas in a liver biopsy is usually characteristic for infection with tuberculosis. Caseous necrosis in mycobacterium infection is rare, and acid-fast bacilli are demonstrated in <10% of proven cases.2 Hepatic tuberculosis is rarely described, and in patients with extrapulmonary tuberculosis, the liver appears to be the least infected organ.3 The main differential diagnoses of hepatic granulomas include brucellosis, coccidioidomycosis and Hodgkin lymphoma.4

The lack of any neutrophils in the soft tissue biopsy together with the massive amount of acid-fast bacilli is compatible with an anergic/immunocompromised mycobacterium infection.

At this point, it had become obvious that the patient was infected by mycobacterium. Given the systemic nature of the infection, the history of BCG vaccination and the evidence of infection at more than two sites, a diagnosis of BCGosis (disseminated BCG infection) was made. The imaging findings complemented the pathological findings: the hypoechoic areas on the liver ultrasound were in keeping with focal granulomas, and the multiple lytic lesions on plain x ray film were due to mycobacterial infection affecting the bone marrow.

The patient was given quadruple tuberculosis treatment (see table 3). In all cases of BCGosis, an underlying immunodeficiency must be sought so the patient was transferred to a specialist centre. Here, the underlying diagnosis of severe combined immunodeficiency (SCID) was made.

Table 3

Quadruple therapy5

The patient subsequently had a splenectomy and bone marrow transplant and is now making good progress in all areas including his development.

In summary, the diagnosis in this case was SCID with disseminated BCGosis causing PUO, hepatosplenomegaly, lytic bone lesions and apparent regression. It was a histological diagnosis obtained from biopsy of a bone and liver lesions.


BCG is a vaccine against tuberculosis. It is prepared from a strain of the attenuated live bovine tuberculosis bacillus, Mycobacterium bovis. In the UK, BCG is given at birth to babies born into high-risk groups including those with a family history of tuberculosis, ethnic minorities at increased risk of TB exposure and those born in areas with a high prevalence.

BCG is one of the most widely used vaccines in the world, with an unparalleled safety record, but if given to an immunocompromised patient, such as a patient with SCID, it can cause disseminated or life-threatening infection such as in this case. The documented incidence of this happening is less than one per million immunisations given.6 If, however, it does occur, BCGosis is treated as for tuberculosis using quadruple therapy. Liver toxicity is common in three of four of the drugs; thus, liver function must be tested before initiation of treatment and subsequently monitored. It is also advisable to check visual acuity before administration of ethambutol, although clearly, this is not easy in children <5 years old and so should be used with caution.

SCID usually presents in infancy with severe infections and failure to thrive and is characterised clinically and immunologically by profound abnormalities of cell-mediated and humoral immunities. It typically results in death before the age of 2 years if not treated with a bone marrow transplant.

This case history illustrates the importance of a multi-specialty approach to complex problems. It is easy to make a presumptive diagnosis based on collections of signs and symptoms, especially when there is supporting laboratory evidence (raised chitriosidase, vacuolated lymphocytes and bony lytic lesions). This can lead to a premature diagnosis and for the clinicians to close off to alternative causes. Premature closure is well described in the literature on clinical reasoning, and this case beautifully highlights the importance of considering all the differentials when aspects do not fit comfortably within a pattern. The constant need to question a diagnosis is an art that must be emphasised in medical training.

Key points

  • BCGosis is a difficult diagnosis to make because of its non-specific presentation.

  • In all cases of BCGosis, an underlying immunodeficiency state must be sought.

  • Severe illness in itself can be a cause for developmental regression.

  • A multi-specialty approach is essential in complex cases.

  • Premature closure during diagnosis making is potentially dangerous and is best avoided by acknowledgment of its existence and constant questioning when the clinical picture does not quite fit.


The authors thank Dr Frazer for his support and encouragement in writing this article. The authors also thank Drs Hattingh and Elliot (Consultant Radiologists, Bradford Royal Infirmary) for allowing us to use the images and for providing advice on the radiology section, Drs Elliott and Richards (Consultant Paediatric Oncologist and Haematologist, respectively, St James University Hospital, Leeds) and the bone marrow transplant unit in Newcastle for their input into this case.


View Abstract


  • Competing interests None.

  • Patient consent Obtained.

  • Provenance and peer review Commissioned; externally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.