Variability among breast radiation oncologists in delineation of the postsurgical lumpectomy cavity

Partial breast irradiation (PBI) is becoming more widely used. Accurate determination of the surgical lumpectomy cavity volume is more critical with PBI than with whole breast radiation therapy. We examined the interobserver variability in delineation of the lumpectomy cavity among four academic radiation oncologists who specialize in the treatment of breast cancer.

Thirty-four lumpectomy cavities in 33 consecutive patients were evaluated. Each physician contoured the cavity and a 1.5-cm margin was added to define the planning target volume (PTV). A cavity visualization score (CVS) was assigned (1-5). To eliminate bias, the physician of record was eliminated from the analysis in all cases. Three measures of variability of the PTV were developed: average shift of the center of mass (COM), average percent overlap between the PTV of two physicians (PVO), and standard deviation of the PTV.

Of variables examined, pathologic resection volume was significantly correlated with CVS, with larger volumes more easily visualized. Shift of the COM decreased and PVO increased significantly as CVS increased. For CVS 4 and 5 cases, the average COM shift was 3 mm and 2 mm, respectively, and PVO was 77% and 87%, respectively. In multiple linear regression, pathologic diameter >4 cm and CVS > or =3 were significantly associated with smaller COM shift. When CVS was omitted from analysis, PVO was significantly larger with pathologic diameter > or =5 cm, days to planning <36, and older age.

Even among radiation oncologists who specialize in breast radiotherapy, there can be substantial differences in delineation of the postsurgical radiotherapy target volume. Large treatment margins may be prudent if the cavity is not clearly defined.