Case 34 Starting a new treatment

Sanna, a 14-year-old girl with severe Crohn’s disease, who has relapsed after multiple courses of corticosteroids and azathioprine, is evaluated for treatment with the anti-Tumour Necrosis Factor (TNF) agent, infliximab.

What risks of biological agents should be discussed with the girl and her family?

Before commencing treatment Sanna’s vaccination status and the presence of anti-Varicella antibodies are confirmed and she has a tuberculin skin test and a CXR. She had received neonatal BCG (and has a scar over her left deltoid) because her parents are originally from Pakistan. The tuberculin skin test shows no induration at 48 hours, and the ST2 doctor records ‘negative tuberculin test’ in Sanna’s notes. The CXR is normal.

How would you interpret these results?

Sanna receives 3 infusions of infliximab over the next 6 weeks and has a substantial improvement in her condition. Five weeks later she presents to her local paediatric unit complaining of fever and is seen by a ST1. She has felt increasingly unwell over the last week, with intermittent fever, loss of appetite, abdominal pain, poor sleep, vomiting, and her parents report that she has become extremely withdrawn. On examination, Sanna has lost 2 kg over the last week 5 weeks, she has a temperature of 38.4°C and mild central abdominal tenderness without guarding or rebound tenderness. Otherwise cardiovascular, respiratory, abdominal and ear, nose and throat examinations are unremarkable.

What causes should be considered for this illness?

After discussion with the registrar, Sanna is admitted to the paediatric ward, with a tentative diagnosis of a relapse of Crohn’s disease and a differential of sepsis or viral gastroenteritis. She has some routine blood tests including blood cultures. These reveal:

How would you manage this girl?

Transfer to the regional referral centre is not possible due to a bed shortage, but Sanna’s case is discussed with the on-call paediatric gastroenterology registrar, Dr Brennan. Following Dr Brennan’s advice, a CXR is performed to rule out tuberculosis, blood, urine and stool cultures are taken, intravenous ceftriaxone and metronidazole are commenced, and abdominal and pelvic USs are performed to exclude an abscess. Over the next five days Sanna continues to have intermittent fevers but all her cultures are negative and the CXR is normal. She is extremely withdrawn and starts to become confused and drowsy with neck stiffness on examination. At this stage an urgent CT scan of her head is performed which shows basal meningeal enhancement and moderate hydrocephalus.

What is the likely diagnosis and what would you do now?

In view of the radiological features, a diagnosis of tuberculous meningitis is suggested by the radiologist. A LP is not performed due to Sanna’s fluctuating conscious level. She goes on to have a complicated course with several acute neurological deteriorations requiring admission to PICU and her parents claim that she subsequently seems to have a reduced intellectual performance. For this reason they sue the hospitals involved, stating that Sanna was not properly assessed for risk of tuberculosis and that the diagnosis was delayed.

Expert opinion

Although pulmonary disease is the most common manifestation of tuberculosis, extrathoracic and disseminated disease can occur without chest signs, particularly in infants and the immunocompromised. Tuberculous meningitis (TBM) often causes vague symptoms initially such as fever, weight loss and anorexia, and classical signs of meningism only evolve later. For this reason TBM is frequently diagnosed late, when there is already hydrocephalus and an increased risk of neurological sequelae. To prevent this it is particularly important to identify risk factors for tuberculosis and to have a low threshold for investigating early in a child with suggestive symptoms. There is increasing use of anticytokine biological agents, many of which increase the risk of sepsis, and the risk of reactivation of latent tuberculosis is particularly increased by anti-TNF agents. Prior to commencing treatment with Infliximab, evaluation for latent tuberculosis is recommended. Unfortunately immunosuppression, prolonged illness and malnutrition can all contribute to a falsely negative tuberculin skin test. In this case it is highly suspicious that there was no response to tuberculin in a child who had previously received BCG. This suggests that either Sanna was anergic due to immunosuppression or that the tuberculin was administered incorrectly; in either case the test becomes invalid. Here, the significance of the unreactive tuberculin test was not realized, and Sanna was managed as if there was no risk of reactivation of latent tuberculosis. Ideally there should have been an additional risk assessment from a detailed history of possible tuberculosis exposure, an interferon-gamma release assay, and either treatment for latent tuberculosis or very careful observation and education of the family about the risk. When Sanna developed symptoms, the reliance on a CXR to rule out tuberculosis neglected the fact that reactivation in immunosuppressed patients may not manifest as pulmonary disease. Failure to consider this possibility led to a delay in the diagnosis and treatment of TBM, and may have contributed to the outcome. The most common neurological sequelae of TBM include intellectual impairment and motor deficits.

Legal comment

The actions of the ST2 doctor who conducted the tuberculin skin test need to be considered. It seems possible that the false negative result was due to his not administering the test correctly. If so, that would clearly be negligent.

However, if he did administer the test correctly, is there a responsible body of medical opinion which would hold that it was acceptable for him (and the doctors who then administered the infliximab) to rely on that test result? Were the circumstances such that the skin test might responsibly be thought to be a sufficient evaluation for latent tuberculosis?

If not, then there appears to have been a breach of duty at that point. Even if there was no breach of duty, the actions of the registrar, Dr Brennan, still need to be considered. Is there a responsible body of medical opinion which would hold that it was acceptable in the circumstances to rely on a CXR to rule out tuberculosis? If not, then there is liability for the consequence of that failure.

If the Expert Opinion is that there has been a breach of duty, then the question is, what were the consequences of that breach? If there is intellectual impairment, then it seems likely to be linked to the TBM. Would earlier treatment have made a difference? If the answer is ‘probably yes’, then Sanna will need to be assessed by a neurologist some years later, when the full extent of the disability and its likely impact become clear. The case may therefore take some time to conclude. It is, of course, a potentially costly case.