Advice
Clinical and technical evidence
Clinical and technical evidence
A literature search was carried out for this briefing in accordance with the published process and methods statement. This briefing includes the most relevant or best available published evidence relating to the clinical effectiveness of the technology. Further information about how the evidence for this briefing was selected is available on request by contacting mibs@nice.org.uk.
Published evidence
Three prospective observational studies were selected for inclusion (n=2,232 lymph node [LN] specimens from 253 people) and reported that the diagnostic performance of OSNA was better than standard histopathology.
Table 2 summarises the clinical evidence as well as its strengths and limitations.
Table 2 Summary of selected studies
|
Study |
Intervention and comparators |
Outcomes |
Strengths and limitations |
|
313 LNs from 22 patients (stages I, II and III), prospective observational study. Single centre (Switzerland). |
OSNA Standard single-level H&E staining histopathology. Intensive histopathology (5 levels of H&E staining and immuno-histochemistry analyses). |
OSNA had a diagnostic performance similar to intensive histopathology and was better than standard H&E staining analysis. Cancer staged as LN negative by standard H&E staining histopathology in 2 of 13 patients was upstaged as LN positive by OSNA. Compared with intensive histopathology, OSNA had 94.5% sensitivity, 97.6% specificity, and a concordance rate of 97.1%. |
The study received manufacturer funding. The study did not analyse whole LNs. A key innovation of the OSNA system is that whole LNs can be analysed, reducing the risk of tissue allocation bias. To prevent selection bias, each LN was divided, with 1 portion being analysed using OSNA and the remaining portion being analysed using standard histopathology or intensive histopathology. The sample included a mix of patients with early and advanced stage colon cancer. OSNA is designed for use in people with early stage colon cancer. |
|
1,594 LNs from 103 patients (stages I and II), prospective observational study. Multicentre (Germany, Switzerland, Spain). |
OSNA Standard single level H&E staining histopathology. |
Cancer staged as LN negative by standard H&E staining histopathology in 26 of 103 patients was upstaged as LN positive by OSNA. |
The manufacturer sponsored the study. The study did not analyse whole LNs. |
|
325 sentinel LNs from 128 patients (stages I and II), prospective observational study. Multicentre (The Netherlands). |
OSNA Standard single level H&E staining histopathology. Intensive histopathology (4 levels of H&E staining and immunohisto-chemistry analyses). |
Cancer staged as LN negative by standard H&E staining histopathology in 20 of 99 patients was upstaged as LN positive by OSNA. Cancer staged as LN negative by standard H&E staining histopathology in 36 of 99 patients was upstaged as LN positive by intensive histopathology. |
The study did not analyse whole LNs. OSNA is designed to detect levels of a tumour biomarker rather than isolated tumour cells, which may account for the difference in results between OSNA and intensive histopathology. Sampling bias may also have affected the results, because the authors noted that tumour cells were only present in specimens analysed by intensive histopathology. The manufacturer provided technical support and study material. |
|
Abbreviations: H&E, haematoxylin and eosin; LN, lymph node. |
|||
Strengths and limitations of the evidence
None of the studies were done in the UK, which may affect their generalisability to the NHS. Although all the studies were prospective in design, none were randomised controlled trials which would have helped to systematically control for biases. All 3 studies were observational rather than randomised controlled studies.
Two studies investigated OSNA in LNs, and 1 study assessed OSNA in sentinel LNs. LNs have a differing potential to sentinel LNs to harbour metastases, so the study results may not be directly comparable.
Two studies assessed at least 12 LNs for each patient. Croner et al. (2014) examined a median of 14 LNs (range 5–46), and Guller et al. (2012) a median of 13 LNs (range 6–24) using OSNA. A third study examined a maximum of 3 sentinel LNs for each patient (Vogelaar et al. 2014). The standard in the NHS is to analyse 12 or more LNs.
Overall, the population complied with the indication for using the test. Two studies only included people with early stage colon cancer, but the third study also included people with advanced colon cancer. OSNA is designed to be used in people with early stage colon cancer. People with advanced colon cancer (defined as TNM stage III or IV) will by definition already have been diagnosed with metastatic disease in the LNs.
Post-operative histopathology is the usual approach in the NHS to stage and grade colon cancer. The number of levels examined using histopathology can affect accuracy; more levels can reduce the risk of tissue allocation bias. Typically, a single section is cut from each LN, however the current practice for handling LNs is not uniform and so may vary locally. All studies used single level histopathology as the standard comparator. The number of levels used in intensive histopathology varied in Vogelaar et al. (2014) and Guller et al. (2012) so study results may not be directly comparable. The intensive histopathology approach (including immunohistochemistry) in Guller et al. (2012) may have been more accurate because more levels were examined.