Positive Predictive Value Of Imaging Research Articles

Local extension staging of sinonasal tumours: retrospective comparison between CT/MRI assessment and pathological findings

To assess the reliability of clinical staging with CT and MRI in sinonasal cancers. We conducted a retrospective review of patients who underwent surgery for a sinonasal cancer. The 7th edition of the TNM classification was used to establish tumour staging. Standardised preoperative CT/MRI staging was compared with the pathological staging based on specimens obtained during surgery from each subsite within and around the tumour. We analysed data from 68 patients between January 2010 and December 2014. A comparison of cT and pT stages was established for 49 naso-ethmoidal and 16 maxillary tumours. Clinical staging for naso-ethmoidal cT1 and maxillary cT2 was consistent with pathological results. Clinical staging for naso-ethmoidal cT2, cT3 and cT4b was overstated in comparison with pT findings. The positive predictive value of imaging was <65% for the lamina papyracea, the cribriform plate, the dura, and the frontal and sphenoid sinuses. Sensitivity was over 75% for each anatomical site except for the cribriform plate (73.3%) and the sphenoid sinus (57.2%). Systematic pathologic analysis of the anatomical areas around the sinonasal cancer has to be applied in further studies to improve our therapeutic management. CT/MRI mapping cannot replace accurate assessment of tumour extension during surgery.

The impact of combined interpretation of localization studies on image-guided surgical approaches for primary hyperparathyroidism.

The impact of single and combined interpretations of ultrasonography and sestamibi scintigraphy to select the appropriate surgical approach in patients with primary hyperparathyroidism were evaluated retrospectively. A total of 183 patients with primary hyperparathyroidism who were evaluated preoperatively using both ultrasonography and sestamibi scintigraphy were included in the study. The results of preoperative localization studies were correlated with intraoperative findings and postoperative histopathological results. The localization rates of individual and combined interpretations of ultrasonography and sestamibi scintigraphy were evaluated. The overall sensitivity and the positive predictive value of ultrasonography and sestamibi scintigraphy were 76% and 90%, and 81% and 91%, respectively. Both imaging studies were concordant for the same localization(s) in 121 (66%) of 183 patients. The prevalence rates of single-gland and multiglandular disease were 90% (N.=109) and 10% (N.=12), respectively, in patients with concordant results (P=0.0001). The overall rate of localization was 91% (N.=110) in these patients. In these patients with concordant results, the sensitivity and the positive predictive value of imaging were 91% and 100%, respectively. The localization rates were 96% and 42% for single-gland and multiglandular disease, respectively (P=0.0001). Of the remaining 62 patients, 50 had negative imaging with either ultrasonography (N.=29) or MIBI (N.=21). Of the 29 patients with negative ultrasonography results, sestamibi scintigraphy was positive in 23. Of the 21 patients with negative sestamibi scintigraphy imaging, ultrasonography was positive in 15. Thus, 38 patients had a single positive imaging result. The majority (95%) of these 38 patients had single-gland disease, and the rate of multiglandular disease was 5% (P=0.0001). The rate of localization was 95% (36/38) in patients with a single positive imaging study. Eighteen patients had discordant imaging results. The overall rate of localization in primary hyperparathyroidism is greater than 90% if ultrasonography and sestamibi scintigraphy are concordantly positive for the same localization, and the prevalence of multiglandular disease is low in patients with concordant imaging. An image-guided surgical approach and selective parathyroidectomy exhibit high cure rates in the setting of concordantly positive sestamibi and ultrasonography results, even if intraoperative parathormone monitoring is not used.

Core Biopsy Following Neoadjuvant Chemotherapy Predicts Tumour Response and Aids in Breast Preservation.

Abstract Back groundAdvancement in chemotherapy regimens has improved pathological tumour response following neoadjuvant chemotherapy (NAC). This has subsequently helped in increasing incidence of breast preservation. Here we report how core biopsy of the tumour bed following NAC is useful in predicting tumour response and aids in breast preservation.Materials and MethodsData were collected retrospectively from consecutive newly diagnosed breast cancer patients at West Middlesex University Hospital between January 2002 and December 2008 who underwent NAC. Patients were assessed with clinical examination, mammogram, ultrasound (USS) and Magnetic Resonance Imaging (MRI). NAC consisted of 6-8 cycles of FEC or had FEC followed by Taxanes. The tumour response was assessed by USS and MRI. In addition, multiple cores were performed both in the central and periphery of the tumour bed. Patients subsequently underwent either mastectomy or wide local Excision (WLE) with axillary clearance. The extent of resection was determined by the presence of tumour in the peripheries.Results43 patients received neo-adjuvant chemotherapy and the median age was 49 years (range 24-72). The median size of tumour by MRI scan was 4.5 cm (range 2.5-10). One patient declined surgery and 8 patients did not have core biopsy following neoadjuvant chemotherapy so all the nine cases were excluded from analysis.The sensitivity of USS, MRI and core biopsy in predicting residual disease in our series was 84%, 84% and 68% respectively. The specificity and the positive predictive value (PPV) of imaging was 25% and 75% whereas the specificity and PPV of core biopsy was both 100%.Our overall initial breast conservation rate was 88% (30 of 34) and 4 of them had subsequent mastectomy following breast conservation surgery (BCS) due to positive margins following BCS reducing it to 76% (26/34).The mean size of the T2 tumour was 2.9 cm and 90% had BCS (9/10) including re excision of margins in one case. The average size of the T3 tumour was 6.0 cm and 76% (13/17) had BCS. Re excision of positive margins was done twice in one case. The mean size of T4 tumour in our series was 6.3 cm and 57% had BCS (4/7).DiscussionThe tumour response following NAC is traditionally assessed by imaging and further surgery is determined by the extent of response. The low specificity and PPV of imaging is due to overestimation of the presence of residual tumour. Absence of enhancement in MRI means either there is necrosis or fragmentation. This is especially true when the tumour mass has completely resolved. The presence of architectural distortion in these cases can complicate MRI interpretation. It is interesting to note that MRI has a high negative predictive value in diagnosing primary tumour as well as recurrence. However this ability is lost following chemotherapy for above mentioned reasons. So core biopsy is an extremely useful tool here when it is positive irrespective of what the imaging indicates. The rate of breast conservation is also increased by assessing tumour response to decide the extent of resection. Extensive sampling of the central and periphery of the tumour following chemotherapy increases the negative surgical margins when breast conservation is considered. Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 4121.

Assessment of a scoring system for breast imaging

Triple assessment is the standard method of assessing symptomatic breast lumps. There is an accepted format for reporting cytology but not for the reporting of mammograms or breast ultrasonographic images. This study describes a scoring system for reporting breast imaging methods. Patients with symptomatic breast lumps seen during 1 year were included. All patients underwent triple assessment. Imaging studies were reported using a grading system from 1 to 5. The results of triple assessment were compared with the final histology. Some 127 women had both mammography and ultrasonographic imaging; the final grade was identical in 60 per cent. The positive predictive value of imaging reported as grade 5 or 4 combined with cytology of C5 or C4 was 100 per cent. All lesions with C3 cytology and benign imaging were benign on histological examination whereas if a C3 grade was combined with imaging grade 5 or 4, all the lesions were malignant. Combining the imaging grade with the results of aspiration cytology and clinical examination often predicts the final histology in patients with breast disease.