Article Text

Methodology for assessing patterns of interstitial pneumonia in children
  1. ANDREW G NICHOLSON, Consultant Histopathologist
  1. ANDREW BUSH, Reader and Honorary Consultant in Paediatric Respirology, Royal Brompton and Harefield
  1. Royal Brompton and Harefield NHS Trust
  2. Royal Brompton Hospital, London SW3 6NP, UK
  3. a.nicholson{at}
  4. NHS Trust, UK
  1. Dr D Hacking, Paediatric Specialist Registrar, Department of Paediatrics, Wexham Park Hospital, Slough, UK (hackandmack{at}

Statistics from

Editor,—The report of Hackinget al,1 of a series of infants with very early onset interstitial lung disease (ILD) with good prognosis, is of great clinical interest but sadly represents a lost investigative opportunity.

Firstly, their statement that percutaneous open lung biopsy has fewer side effects than open lung biopsy (OLB) is not supported by any direct comparative trial, and cannot be allowed to stand. Indeed, the largest published series using this technique2 was heavily criticised both for the number of complications and the often non-diagnostic samples obtained.3 ,4 By contrast OLB is safe,5 permits direct inspection of the site of biopsy, and allows the acquisition of specimens large enough to determine the lung architecture in order more precisely to classify the different types of paediatric ILD. OLB thus allows appreciation of the distribution of disease involvement within the acinus, allowing more precise identification of different histopathological patterns. There is therefore no reason to favour percutaneous biopsy over OLB in skilled hands; indeed the weight of evidence is in favour of OLB. Secondly, their selected nomenclature of paediatric ILD is open to criticism. It is questionable whether the use of term “idiopathic pulmonary fibrosis (IPF)” is still appropriate in children. IPF is generally used synonymously with lone cryptogenic fibrosing alveolitis, which in adults is most often represented histopathologically by the pattern of “usual interstitial pneumonia (UIP)”. However UIP is rarely if ever seen in children; much more common are lymphocytic interstitial pneumonia (LIP), desquamative interstitial pneumonia (DIP), non-specific interstitial pneumonia (NSIP), and chronic pneumonitis of infancy.6 Identifying these histological patterns may point towards specific investigations with regard to aetiology, and may also provide prognostic and treatment data, and we consider that it is a pity that this or a similar histopathological classification was not used in this report.

We suggest that more will be learned about these rare conditions if diagnostic precision is maximised by comparison of pre-biopsy computed tomography with properly classified histological findings.7 Interdisciplinary collaboration is needed to achieve this, and it is unfortunate that more details of imaging and an up to date classification of histology were not included in an otherwise informative paper.


Dr Hacking et al respond

Editor,—In their response to our article, Nicholson and Bush suggest that an investigative opportunity has been missed. We do not agree. These authors have repeated their previously reported criticism of percutaneous lung biopsy (PLB)1-1 and have suggestedthat this technique is prone both to more complications and to a greater number of non-diagnostic samples. We have shown that PLB in a series of nine patients was adequate for diagnosis in all cases and did not result in pneumothoraces significant enough to require thoracocentesis.1-2 In our present report of 11 patients, PLB was not associated with any major complications and failed to provide a histological diagnosis in only one patient. This compares favourably to Nicholson and Bush's own report of open lung biopsy (OLB) in 27 cases where three patients experienced significant complications—that is, a pneumothorax, a haemothorax, and a pleural space infection.1-3 Moreover, five previously self ventilating patients required ventilation after biopsy, and five patients returned from biopsy with a chest drain which had been inserted in the course of the procedure. We do not agree that OLB is superior to PLB.

Nicholson and Bush go on to question the nomenclature of paediatric idiopathic pulmonary fibrosis and briefly describe the histological classifications of usual pneumonia (UIP) and desquamative interstitial pneumonia (DIP) as we did in the introduction to our article. They suggest the “these histological patterns may . . . provide prognostic and treatment data”. However, the distinction between UIP and DIP is questionable,1-4 as they may represent different stages in the same disease process.1-5 1-6 In common with previous reports,1-7-1-9 we have shown that the severity of histological change did not relate to patient's response to steroids or their eventual outcome.

We agree with Nicholson and Bush when they state that “diagnostic precision is maximised by comparison of pre-biopsy computed tomography with properly classified histological findings” as this was practised throughout our series. We fear that they have missed the most important aspect of our report which is that idiopathic pulmonary fibrosis in children has a diverse natural history and a variable prognosis that can be favourable. The good prognosis seen in our series is different from previous case reports indicating a greater than 50% mortality.1-8


  1. 1-1.
  2. 1-2.
  3. 1-3.
  4. 1-4.
  5. 1-5.
  6. 1-6.
  7. 1-7.
  8. 1-8.
  9. 1-9.

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.