Hossein Jadvar MD, PhD, Presenter: Nothing to Disclose

Milton C. Weinstein PhD, Abstract Co-Author: Nothing to Disclose

We conducted a decision tree sensitivity analysis of the comparative role of [F-18]fluorodeoxyglucose positron emission tomography (FDG PET) in staging of the axilla in women with newly-diagnosed breast cancer.

We used the TreeAge Pro Suite software package (TreeAge Software, Inc. Williamstown, MA) to construct a decision tree to compare decisions in relation to staging of the axilla in women with newly-diagnosed breast cancer (T1/T2) and clinically negative (N0) axilla. The root decision for axillary staging was analyzed through three options of the standard of reference axillary lymph nodal dissection (ALND), sentinel lymph node (SLN) biopsy, and FDG PET. SLN findings were considered to be TP, TN, or FN (i.e. skip metastasis). The probabilities for the relevant tree branches and average procedural complication rate were calculated based on literature. The utilities were assigned according to comparative face validity criteria. Sensitivity analyses were performed on prevalence of axillary nodal disease (v), sensitivities of PET (p) and SLN (q), the complication rates of SLN (m), and ALND (n), and the utility of a true positive nodal diagnosis (c).

SLN is the best choice for staging of the axilla, which is compatible with the current clinical practice with v=0.3. As v increases, the PET advantage increases gradually such that at v≥0.9, PET with any sensitivity (at fixed specificity of 0.8) will win over SLN as the procedure of choice. At all values of c, SLN is still the best choice since it is more accurate than PET and has lower complication rate than ALND. For most combinations of m and n, SLN continues to be the best choice. Only when n<0.1, ALND would be preferred. At 0.1≤n≤0.33, the choice between SLN and ALND depends on m. At n≥0.33, SLN is always preferred over ALND. Only for m≥0.9 and n≥0.6 (not plausible based on current clinical literature), PET has an advantage due to its lack of morbidity.

SLN is the best choice for staging of the axilla in T1/T2 clinical N0 breast cancer. FDG PET may be preferred only when the expected prevalence of axillary disease is very high.

We conducted a decision tree sensitivity analysis of the comparative role of [F-18]fluorodeoxyglucose positron emission tomography (FDG PET) in staging of the axilla in women with newly-diagnosed breast cancer.

We used the TreeAge Pro Suite software package (TreeAge Software, Inc. Williamstown, MA) to construct a decision tree to compare decisions in relation to staging of the axilla in women with newly-diagnosed breast cancer (T1/T2) and clinically negative (N0) axilla. The root decision for axillary staging was analyzed through three options of the standard of reference axillary lymph nodal dissection (ALND), sentinel lymph node (SLN) biopsy, and FDG PET. SLN findings were considered to be TP, TN, or FN (i.e. skip metastasis). The probabilities for the relevant tree branches and average procedural complication rate were calculated based on literature. The utilities were assigned according to comparative face validity criteria. Sensitivity analyses were performed on prevalence of axillary nodal disease (v), sensitivities of PET (p) and SLN (q), the complication rates of SLN (m), and ALND (n), and the utility of a true positive nodal diagnosis (c).

SLN is the best choice for staging of the axilla, which is compatible with the current clinical practice with v=0.3. As v increases, the PET advantage increases gradually such that at v≥0.9, PET with any sensitivity (at fixed specificity of 0.8) will win over SLN as the procedure of choice. At all values of c, SLN is still the best choice since it is more accurate than PET and has lower complication rate than ALND. For most combinations of m and n, SLN continues to be the best choice. Only when n<0.1, ALND would be preferred. At 0.1≤n≤0.33, the choice between SLN and ALND depends on m. At n≥0.33, SLN is always preferred over ALND. Only for m≥0.9 and n≥0.6 (not plausible based on current clinical literature), PET has an advantage due to its lack of morbidity.

SLN is the best choice for staging of the axilla in T1/T2 clinical N0 breast cancer. FDG PET may be preferred only when the expected prevalence of axillary disease is very high.

Jadvar, H, Weinstein, M, Decision Tree Sensitivity Analysis of FDG PET in Axillary Staging of T1/T2 Clinical N0 Breast Cancer.  Radiological Society of North America 2005 Scientific Assembly and Annual Meeting, November 27 - December 2, 2005 ,Chicago IL. http://archive.rsna.org/2005/4407678.html