Diagnostic Yield of Biopsies in Eosinophilic Esophagitis
Diagnostic Yield of Biopsies in Eosinophilic Esophagitis
This data set represents perhaps the most comprehensive in vivo mapping of EoE within individual subjects to date. We have shown that endoscopic manifestations of EoE predict eosinophil densities. We previously characterised the variability of eosinophilic density in an EoE esophagectomy specimen obtained from a patient with distal longitudinal furrows. Clearly, routine ~2 mm oesophageal biopsy samples may easily miss heterogeneously dispersed eosinophil infiltrates. Indeed, by calculating peak eosinophil counts per biopsy rather than obtaining the peak of five biopsies (as recommended in current guidelines) the variability of the disease becomes quite apparent with several biopsies demonstrating no eosinophils at all, resulting in a lower average eosinophilia than is commonly reported in the disease due to averaging of the counts. Indeed, perhaps due to our protocol and extensive proximal biopsy selection, only 37.8% of our total biopsies contained at least 15 eos/HPF. Thus, the importance of biopsy site selection should not be underemphasised as both diagnosis and treatment success is interpreted via the peak eosinophils/HPF.
Typical endoscopic features of EoE include concentric mucosal rings, linear furrowing, white plaques or exudates and narrow calibre oesophagi termed strictures, though endoscopic features appear normal in some cases. Here, we evaluate the relationship between endoscopic appearances and eosinophil density by comprehensively biopsying multiple oesophagi. Our study is the first prospective study to determine appropriate location and quantity of biopsies based on phenotypes in suspected EoE patients. Validated endoscopic scoring systems at the time of the study to specifically document oedema or severity of strictures were unavailable when the endoscopies were performed. Hirano has subsequently published recommendations on endoscopic scoring which recommend grading the extent of plaques, oedema and furrows. Rather than grade the furrows as present or not as recommended in Hirano's paper, we defi-ned the furrows as lines or furrows (Figure 2). Thus, as endoscopists move more towards this endoscopic scoring system, information regarding associations of mucosal changes to eosinophil density may become more definitive.
The 2011 consensus diagnostic guidelines by Liacouras et al., recognise the need for better understanding of the biopsy location (proximal/middle/distal) as well as the optimal number of tissue samples to survey patients suspected of EoE. Our study generated 'histological maps' to better understand the eosinophilic distribution and density within each varying area of endoscopic appearances. We show that exudates and more advanced furrows typically reflect higher eosinophil densities. Rings may represent a fixed feature and were seen in some of the resolved EoE patients as well as in inflamed EoE subjects and do not indicate active eosinophilic inflammation unless accompanied by furrows or exudates. Thus, the presence of rings is not an indicator of active disease unless they are seen in the presence of other features of inflammation such as furrows or exudates. Indeed, our data suggest that the presence of rings with a diminished luminal circumference, may warrant dilation rather than medical therapy as the eosinophilia is mild in this setting.
We determined that the histologic mucosal changes correlated only weakly with the degree of local eosinophilia. This suggests that the eosinophilia may vary over time – with histological changes lagging behind the eosinophilic stimuli. Hence, it is likely that eosinophilic peaks may vary over time, suggesting that the only way to define this disease in its entirety is to either repeat mapping studies at various times or to find a way to 'define' the eosinophilic involvement in the entire oesophagus during one specific moment in time either by imaging or other modalities.
Our study has a few limitations. One is that we performed biopsies on 10 EoE subjects (12 sets of biopsies) which limits the representation of disease presentation due to the variability in the disease itself. However, the majority of our patients demonstrated variable presentation of disease within their oesophagus creating a more equally representative environment. We feel that our population represents a significant proportion of patients suffering from EoE and our results should be applicable to the diseased population. Additionally, unlike prior studies, we have been able to extensively biopsy and assess eosinophilia between unique individuals.
While a spectrum of pathological changes have been reported, the sole histological criterion required for a diagnosis of EoE remains high eosinophil density (≥15 eos/HPF) without data to warrant consideration of the distribution of eosinophils. Many of the EoE patients displayed multiple areas of low or no eosinophil density, particularly in the proximal oesophagus. Thus, the importance of biopsies 'targeted' to endoscopic areas of more active disease cannot be ignored. Although some patients present with marked proximal eosinophilia, our histological mapping of patient oesophagi shows eosinophil density to be more dense distally in many patients. Combined, these results provide a new opportunity to accurately identify and quantify burden of disease based upon endoscopic findings within individuals.
Discussion
This data set represents perhaps the most comprehensive in vivo mapping of EoE within individual subjects to date. We have shown that endoscopic manifestations of EoE predict eosinophil densities. We previously characterised the variability of eosinophilic density in an EoE esophagectomy specimen obtained from a patient with distal longitudinal furrows. Clearly, routine ~2 mm oesophageal biopsy samples may easily miss heterogeneously dispersed eosinophil infiltrates. Indeed, by calculating peak eosinophil counts per biopsy rather than obtaining the peak of five biopsies (as recommended in current guidelines) the variability of the disease becomes quite apparent with several biopsies demonstrating no eosinophils at all, resulting in a lower average eosinophilia than is commonly reported in the disease due to averaging of the counts. Indeed, perhaps due to our protocol and extensive proximal biopsy selection, only 37.8% of our total biopsies contained at least 15 eos/HPF. Thus, the importance of biopsy site selection should not be underemphasised as both diagnosis and treatment success is interpreted via the peak eosinophils/HPF.
Typical endoscopic features of EoE include concentric mucosal rings, linear furrowing, white plaques or exudates and narrow calibre oesophagi termed strictures, though endoscopic features appear normal in some cases. Here, we evaluate the relationship between endoscopic appearances and eosinophil density by comprehensively biopsying multiple oesophagi. Our study is the first prospective study to determine appropriate location and quantity of biopsies based on phenotypes in suspected EoE patients. Validated endoscopic scoring systems at the time of the study to specifically document oedema or severity of strictures were unavailable when the endoscopies were performed. Hirano has subsequently published recommendations on endoscopic scoring which recommend grading the extent of plaques, oedema and furrows. Rather than grade the furrows as present or not as recommended in Hirano's paper, we defi-ned the furrows as lines or furrows (Figure 2). Thus, as endoscopists move more towards this endoscopic scoring system, information regarding associations of mucosal changes to eosinophil density may become more definitive.
The 2011 consensus diagnostic guidelines by Liacouras et al., recognise the need for better understanding of the biopsy location (proximal/middle/distal) as well as the optimal number of tissue samples to survey patients suspected of EoE. Our study generated 'histological maps' to better understand the eosinophilic distribution and density within each varying area of endoscopic appearances. We show that exudates and more advanced furrows typically reflect higher eosinophil densities. Rings may represent a fixed feature and were seen in some of the resolved EoE patients as well as in inflamed EoE subjects and do not indicate active eosinophilic inflammation unless accompanied by furrows or exudates. Thus, the presence of rings is not an indicator of active disease unless they are seen in the presence of other features of inflammation such as furrows or exudates. Indeed, our data suggest that the presence of rings with a diminished luminal circumference, may warrant dilation rather than medical therapy as the eosinophilia is mild in this setting.
We determined that the histologic mucosal changes correlated only weakly with the degree of local eosinophilia. This suggests that the eosinophilia may vary over time – with histological changes lagging behind the eosinophilic stimuli. Hence, it is likely that eosinophilic peaks may vary over time, suggesting that the only way to define this disease in its entirety is to either repeat mapping studies at various times or to find a way to 'define' the eosinophilic involvement in the entire oesophagus during one specific moment in time either by imaging or other modalities.
Our study has a few limitations. One is that we performed biopsies on 10 EoE subjects (12 sets of biopsies) which limits the representation of disease presentation due to the variability in the disease itself. However, the majority of our patients demonstrated variable presentation of disease within their oesophagus creating a more equally representative environment. We feel that our population represents a significant proportion of patients suffering from EoE and our results should be applicable to the diseased population. Additionally, unlike prior studies, we have been able to extensively biopsy and assess eosinophilia between unique individuals.
While a spectrum of pathological changes have been reported, the sole histological criterion required for a diagnosis of EoE remains high eosinophil density (≥15 eos/HPF) without data to warrant consideration of the distribution of eosinophils. Many of the EoE patients displayed multiple areas of low or no eosinophil density, particularly in the proximal oesophagus. Thus, the importance of biopsies 'targeted' to endoscopic areas of more active disease cannot be ignored. Although some patients present with marked proximal eosinophilia, our histological mapping of patient oesophagi shows eosinophil density to be more dense distally in many patients. Combined, these results provide a new opportunity to accurately identify and quantify burden of disease based upon endoscopic findings within individuals.
Source...