Childhood lymphoma: Diagnostic accuracy of chest radiography for severe pulmonary complications

H. T. Winer-Muram, S. A. Rubin, W. M. Kauffman, S. G. Jennings, Kristopher Arheart, J. T. Sandlund, P. M. Bozeman

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

Introduction: We sought to determine whether chest radiography can be reliably used to distinguish persistent or relapsing pulmonary lymphoma from a variety of infectious and noninfectious pulmonary conditions that can occur in children receiving treatment for lymphoma. Methods: We studied chest radiographs of 37 patients (30 with non-Hodgkin's lymphoma, and seven with Hodgkin's disease) who died of paediatric lymphoma or of treatment complications. Pulmonary findings at autopsy comprised lung tumour (n=14), pleural tumour (n=12), pneumonia (n=22), adult respiratory distress syndrome (ARDS; n=16), haemorrhage (n=27), and infarction (n=13). Using a 4-point scale and without knowledge of autopsy findings, three radiologists independently rated antemortem radiographs for the presence of pulmonary tumour, pleural tumour, pneumonia in general, pneumonia caused by viral, bacterial, fungal, and protozoan pathogens, ARDS, pulmonary haemorrhage, and pulmonary infarction. Diagnostic accuracy was defined by the area under the receiveroperating-characteristic curve (Az). Results: Diagnostic accuracy was good for pulmonary tumour (Az, 0.71±0.6), protozoan pneumonia (Az, 0.77±0.06), and ARDS (Az, 0.86±0.07) but poor for all other conditions. The absence of both pleural effusions and mediastinal/right hilar lymphadenopathy was significantly associated (P≤0.04) with the absence of lung tumour. Discussion: The pulmonary processes in these patients can all demonstrate diffuse airspace opacification, and many patients had multiple lung abnormalities at autopsy. The radiologist-readers were unable to identify which pulmonary conditions were responsible for radiographic findings in most patients. The readers were able to identify patients who did not have pulmonary lymphoma. If pulmonary involvement with lymphoma is unlikely, bronchoscopy with bronchoalveolar lavage may be sufficient to establish a diagnosis. When pulmonary lymphoma is a clinical consideration, open lung biopsy is usually required for diagnosis.

Original languageEnglish
Pages (from-to)842-847
Number of pages6
JournalClinical Radiology
Volume50
Issue number12
DOIs
StatePublished - Jan 1 1995
Externally publishedYes

Fingerprint

Radiography
Lymphoma
Thorax
Lung
Autopsy
Neoplasms
Pneumonia
Pulmonary Infarction
Viral Pneumonia
Hemorrhage
Multiple Abnormalities
Adult Respiratory Distress Syndrome
Bronchoscopy
Bronchoalveolar Lavage
Pleural Effusion
Hodgkin Disease
Non-Hodgkin's Lymphoma
Infarction

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging
  • Oncology

Cite this

Winer-Muram, H. T., Rubin, S. A., Kauffman, W. M., Jennings, S. G., Arheart, K., Sandlund, J. T., & Bozeman, P. M. (1995). Childhood lymphoma: Diagnostic accuracy of chest radiography for severe pulmonary complications. Clinical Radiology, 50(12), 842-847. https://doi.org/10.1016/S0009-9260(05)83105-8

Childhood lymphoma : Diagnostic accuracy of chest radiography for severe pulmonary complications. / Winer-Muram, H. T.; Rubin, S. A.; Kauffman, W. M.; Jennings, S. G.; Arheart, Kristopher; Sandlund, J. T.; Bozeman, P. M.

In: Clinical Radiology, Vol. 50, No. 12, 01.01.1995, p. 842-847.

Research output: Contribution to journalArticle

Winer-Muram, HT, Rubin, SA, Kauffman, WM, Jennings, SG, Arheart, K, Sandlund, JT & Bozeman, PM 1995, 'Childhood lymphoma: Diagnostic accuracy of chest radiography for severe pulmonary complications', Clinical Radiology, vol. 50, no. 12, pp. 842-847. https://doi.org/10.1016/S0009-9260(05)83105-8
Winer-Muram HT, Rubin SA, Kauffman WM, Jennings SG, Arheart K, Sandlund JT et al. Childhood lymphoma: Diagnostic accuracy of chest radiography for severe pulmonary complications. Clinical Radiology. 1995 Jan 1;50(12):842-847. https://doi.org/10.1016/S0009-9260(05)83105-8
Winer-Muram, H. T. ; Rubin, S. A. ; Kauffman, W. M. ; Jennings, S. G. ; Arheart, Kristopher ; Sandlund, J. T. ; Bozeman, P. M. / Childhood lymphoma : Diagnostic accuracy of chest radiography for severe pulmonary complications. In: Clinical Radiology. 1995 ; Vol. 50, No. 12. pp. 842-847.
@article{4ca2962a395743538bb1bd48666d19a0,
title = "Childhood lymphoma: Diagnostic accuracy of chest radiography for severe pulmonary complications",
abstract = "Introduction: We sought to determine whether chest radiography can be reliably used to distinguish persistent or relapsing pulmonary lymphoma from a variety of infectious and noninfectious pulmonary conditions that can occur in children receiving treatment for lymphoma. Methods: We studied chest radiographs of 37 patients (30 with non-Hodgkin's lymphoma, and seven with Hodgkin's disease) who died of paediatric lymphoma or of treatment complications. Pulmonary findings at autopsy comprised lung tumour (n=14), pleural tumour (n=12), pneumonia (n=22), adult respiratory distress syndrome (ARDS; n=16), haemorrhage (n=27), and infarction (n=13). Using a 4-point scale and without knowledge of autopsy findings, three radiologists independently rated antemortem radiographs for the presence of pulmonary tumour, pleural tumour, pneumonia in general, pneumonia caused by viral, bacterial, fungal, and protozoan pathogens, ARDS, pulmonary haemorrhage, and pulmonary infarction. Diagnostic accuracy was defined by the area under the receiveroperating-characteristic curve (Az). Results: Diagnostic accuracy was good for pulmonary tumour (Az, 0.71±0.6), protozoan pneumonia (Az, 0.77±0.06), and ARDS (Az, 0.86±0.07) but poor for all other conditions. The absence of both pleural effusions and mediastinal/right hilar lymphadenopathy was significantly associated (P≤0.04) with the absence of lung tumour. Discussion: The pulmonary processes in these patients can all demonstrate diffuse airspace opacification, and many patients had multiple lung abnormalities at autopsy. The radiologist-readers were unable to identify which pulmonary conditions were responsible for radiographic findings in most patients. The readers were able to identify patients who did not have pulmonary lymphoma. If pulmonary involvement with lymphoma is unlikely, bronchoscopy with bronchoalveolar lavage may be sufficient to establish a diagnosis. When pulmonary lymphoma is a clinical consideration, open lung biopsy is usually required for diagnosis.",
author = "Winer-Muram, {H. T.} and Rubin, {S. A.} and Kauffman, {W. M.} and Jennings, {S. G.} and Kristopher Arheart and Sandlund, {J. T.} and Bozeman, {P. M.}",
year = "1995",
month = "1",
day = "1",
doi = "10.1016/S0009-9260(05)83105-8",
language = "English",
volume = "50",
pages = "842--847",
journal = "Clinical Radiology",
issn = "0009-9260",
publisher = "W.B. Saunders Ltd",
number = "12",

}

TY - JOUR

T1 - Childhood lymphoma

T2 - Diagnostic accuracy of chest radiography for severe pulmonary complications

AU - Winer-Muram, H. T.

AU - Rubin, S. A.

AU - Kauffman, W. M.

AU - Jennings, S. G.

AU - Arheart, Kristopher

AU - Sandlund, J. T.

AU - Bozeman, P. M.

PY - 1995/1/1

Y1 - 1995/1/1

N2 - Introduction: We sought to determine whether chest radiography can be reliably used to distinguish persistent or relapsing pulmonary lymphoma from a variety of infectious and noninfectious pulmonary conditions that can occur in children receiving treatment for lymphoma. Methods: We studied chest radiographs of 37 patients (30 with non-Hodgkin's lymphoma, and seven with Hodgkin's disease) who died of paediatric lymphoma or of treatment complications. Pulmonary findings at autopsy comprised lung tumour (n=14), pleural tumour (n=12), pneumonia (n=22), adult respiratory distress syndrome (ARDS; n=16), haemorrhage (n=27), and infarction (n=13). Using a 4-point scale and without knowledge of autopsy findings, three radiologists independently rated antemortem radiographs for the presence of pulmonary tumour, pleural tumour, pneumonia in general, pneumonia caused by viral, bacterial, fungal, and protozoan pathogens, ARDS, pulmonary haemorrhage, and pulmonary infarction. Diagnostic accuracy was defined by the area under the receiveroperating-characteristic curve (Az). Results: Diagnostic accuracy was good for pulmonary tumour (Az, 0.71±0.6), protozoan pneumonia (Az, 0.77±0.06), and ARDS (Az, 0.86±0.07) but poor for all other conditions. The absence of both pleural effusions and mediastinal/right hilar lymphadenopathy was significantly associated (P≤0.04) with the absence of lung tumour. Discussion: The pulmonary processes in these patients can all demonstrate diffuse airspace opacification, and many patients had multiple lung abnormalities at autopsy. The radiologist-readers were unable to identify which pulmonary conditions were responsible for radiographic findings in most patients. The readers were able to identify patients who did not have pulmonary lymphoma. If pulmonary involvement with lymphoma is unlikely, bronchoscopy with bronchoalveolar lavage may be sufficient to establish a diagnosis. When pulmonary lymphoma is a clinical consideration, open lung biopsy is usually required for diagnosis.

AB - Introduction: We sought to determine whether chest radiography can be reliably used to distinguish persistent or relapsing pulmonary lymphoma from a variety of infectious and noninfectious pulmonary conditions that can occur in children receiving treatment for lymphoma. Methods: We studied chest radiographs of 37 patients (30 with non-Hodgkin's lymphoma, and seven with Hodgkin's disease) who died of paediatric lymphoma or of treatment complications. Pulmonary findings at autopsy comprised lung tumour (n=14), pleural tumour (n=12), pneumonia (n=22), adult respiratory distress syndrome (ARDS; n=16), haemorrhage (n=27), and infarction (n=13). Using a 4-point scale and without knowledge of autopsy findings, three radiologists independently rated antemortem radiographs for the presence of pulmonary tumour, pleural tumour, pneumonia in general, pneumonia caused by viral, bacterial, fungal, and protozoan pathogens, ARDS, pulmonary haemorrhage, and pulmonary infarction. Diagnostic accuracy was defined by the area under the receiveroperating-characteristic curve (Az). Results: Diagnostic accuracy was good for pulmonary tumour (Az, 0.71±0.6), protozoan pneumonia (Az, 0.77±0.06), and ARDS (Az, 0.86±0.07) but poor for all other conditions. The absence of both pleural effusions and mediastinal/right hilar lymphadenopathy was significantly associated (P≤0.04) with the absence of lung tumour. Discussion: The pulmonary processes in these patients can all demonstrate diffuse airspace opacification, and many patients had multiple lung abnormalities at autopsy. The radiologist-readers were unable to identify which pulmonary conditions were responsible for radiographic findings in most patients. The readers were able to identify patients who did not have pulmonary lymphoma. If pulmonary involvement with lymphoma is unlikely, bronchoscopy with bronchoalveolar lavage may be sufficient to establish a diagnosis. When pulmonary lymphoma is a clinical consideration, open lung biopsy is usually required for diagnosis.

UR - http://www.scopus.com/inward/record.url?scp=0028785987&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0028785987&partnerID=8YFLogxK

U2 - 10.1016/S0009-9260(05)83105-8

DO - 10.1016/S0009-9260(05)83105-8

M3 - Article

C2 - 8536395

AN - SCOPUS:0028785987

VL - 50

SP - 842

EP - 847

JO - Clinical Radiology

JF - Clinical Radiology

SN - 0009-9260

IS - 12

ER -