Proportional incidence and radiological review of large (T2+) breast cancers as surrogate indicators of screening programme performance.

OBJECTIVES: Surrogate measures of screening performance [e.g. interval cancer (IC) proportional incidence] allow timely monitoring of sensitivity and quality. This study explored measures using large (T2+) breast cancers as potential indicators of screening performance. METHODS: The proportional incidence of T2+ cancers (observed/expected cases) in a population-based screening programme (Trento, 2001-2009) was estimated. A parallel review of 'negative' preceding mammograms for screen-detected T2+ and for all ICs, using 'blinded' independent readings and case-mixes (54 T2+, 50 ICs, 170 controls) was also performed. RESULTS: T2+ cancers were observed in 168 screening participants: 48 at first screen, 67 at repeat screening and 53 ICs. The T2+ estimated proportional incidence was 68% (observed/expected = 168/247), corresponding to an estimated 32% reduction in the rate of T2+ cancers in screening participants relative to that expected without screening. Majority review classified 27.8% (15/54) of T2+ and 28% (14/50) of ICs as screening error (P = 0.84), with variable recall rates amongst radiologists (8.8-15.2%). CONCLUSIONS: T2+ review could be integrated as part of quality monitoring and potentially prove more feasible than IC review for some screening services. KEY POINTS: • Interval breast cancers, assumed as screening failures, are monitored to estimate screening performance • Large (T2+) cancers at screening may also represent failed prior screening detection • Analysis of T2+ lesions may be more feasible than assessing interval cancers • Analysis of T2+ cancers is a potential further measure of screening performance.

Analysis of proportional incidence and review of interval cancer cases observed within the mammography screening programme in Trento province, Italy.

PURPOSE: The aim of this study was to assess the performance of the mammography screening programme in Trento Province, Italy, by analysing the interval cancers (IC) observed from 2001 to 2008. MATERIAL AND METHODS: IC were identified by linking screening archives with local cancer registry and pathology archives as well as with hospital discharge records. Proportional incidence was determined according to breast cancers expected in the absence of screening, estimated on the basis of patients/year at risk and age-specific incidence. The review of screening mammograms preceding ICs was performed by an external (three radiologists) and an internal (five radiologists) panel and aimed at assessing the proportion of IC reviewed as screening errors. Results were compared with European Community (EC) recommended standards. RESULTS: IC proportional incidence was 15.90% for the first year (EC standard <30%) and 25.77% for the second year (EC standard <50%) of the interval. At external review, 18.4% of cases were reviewed as screening errors (identified by at least two of three reviewers), whereas at internal review (identified by at least three of five reviewers) it was 17.4% (EC standard <20%). No significant difference was observed between external and internal review (mean recall rate 8.3% vs. 9.0%; mean identification rate 19.7% vs 19.6%). CONCLUSIONS: The study confirms good performance of the mammography screening programme in Trento Province, Italy. Comparable results of external and internal reviews suggest that the latter, no doubt easier to be implemented, might be adopted as a routine procedure to assess this early efficacy indicator.

Benefits of double reading of screening mammograms: retrospective study on a consecutive series.

PURPOSE: The purpose of this study was to assess the performance of delayed second reading of screening mammograms when added to real-time reading plus immediate assessment. MATERIAL AND METHODS: The study setting was the mammography screening programme of an Italian Local Health Unit. Recall rate and cancer detection rate at first reading or informed second reading only were assessed in a cohort of 23,629 women aged 50-69 years screened during 2007-2008. Incremental recall rate, incremental cancer detection rate and incremental cost of second reading were determined. RESULTS: Recall rate was 13.0% at first and 2.7% at second reading (incremental recall rate +21.1%). Overall, recalls were more frequent in the younger decade and in the presence of denser breasts. Cancer detection rate was 7.06‰ (n=167) at first and 0.93‰ (n=22) at second reading (incremental cancer detection rate +13.1%). Compared with first reading, second reading detected more cancers depicted as isolated microcalcifications and distortions (40.9% vs. 16.2%, p=0.02) and at a lower stage (stage 0-I 81.8% vs. 69.5%, p=0.34). The cost of adding delayed second reading was + 3.65 per screened individual or 3,926.61 per incremental cancer detected. CONCLUSIONS: The study confirms the efficacy of second reading, even as an adjunct to real-time single reading plus immediate assessment. Incremental recall rate is acceptable in view of the incremental cancer detection rate, and both figures are within the range of literature reports on double-reading performance.

On the role of arbitration of discordant double readings of screening mammography: experience from two Italian programmes.

PURPOSE: The authors sought to assess the role of arbitration by a third reader of discordant double readings to reduce the rate of recalls to diagnostic assessment. MATERIALS AND METHODS: A consecutive series of 7,660 double readings of screening examinations were considered. Discordant recalls were arbitrated by an expert reader (negative/positive). Diagnostic assessment was performed irrespective of arbitration results, and its outcome was used as reference standard for the study purpose. Assuming that negative arbitration would deny recall, its impact was assessed in terms of reduced recall rate and reduced cancer detection rate. Cost analysis of introducing arbitration was performed according to these results. RESULTS: Recalls at double reading were 528 (6.8%), of which 230 (43.5%) were concordant and 298 (56.5%) were discordant. The latter underwent arbitration, which was negative in 216 (72.4%) and positive in 82 (27.6%) cases, respectively. Overall, 49 cancers were detected (6.39 ‰ screened, 9.2% recalled): 43 cancers were detected among concordant (5.6 ‰ screened, 18.6% concordant) and six among discordant recalls (0.7 ‰ screened, 2.0% discordant). Six cancers were observed among arbitrated cases: five (6%) in positive and one (4.6 ‰) in negative arbitrations. Negative arbitration would have spared 216 assessment procedures (2.8% absolute, 40.9% relative reduction of recall rate) while missing one cancer case (0.13 ‰ absolute, 2.0% relative reduction of cancer detection rate). Arbitration cost was 74 euro, whereas 216 spared assessment procedures would have cost 14,558.4-23,346 euro. CONCLUSIONS: Arbitration is a cost-effective procedure that could be employed as a first measure to counterbalance excess recall rate observed in a double-reading scenario.

Analysis of interval cancers observed in an Italian mammography screening programme (2000-2006).

PURPOSE: The purpose of this study was to assess the interval cancer (IC) proportional incidence and review IC cases observed in an Italian mammography screening programme during 2000-2006. MATERIALS AND METHODS: ICs were identified through linkage of a screening database with the local cancer registry and hospital discharge records to calculate proportional (observed/expected) incidence. Negatively reported mammograms preceding ICs underwent blind review (randomly mixed with negative controls in a 2:1 ratio) by three expert radiologists and classified according to European guidelines criteria (OC=occult, MS=minimal sign, SE=screening error) according to majority report. Proportional IC incidence and rate of reviewed IC classified as SE were compared with European guideline standards. RESULTS: Proportional IC incidence was 10.8% in the first and 40.0% in the second year of the interval (European standard=30% or 50%, respectively). Sensitivity estimate for the 2-year interval was 74.6%. ICs were reviewed as SE, MS or OC in 15.0%, 14.0% or 71.0% of cases, respectively. Corresponding review results for negative controls were 7.0%, 25.0% or 68.0%, respectively. Positive predictive value for IC was 51.7% for SE and 21.8% for MS reporting category, respectively (p=0.008). European standard (<20% reviewed as SE) was reached. CONCLUSIONS: The study shows that the sensitivity of the mammography programme was good, complying with European guideline recommendations. Assessment of IC-based early indicators of screening efficacy is feasible in a current screening programme and should become a routine procedure.

Autologous chondrocyte implantation: prospective MRI evaluation with clinical correlation.

PURPOSE: This study was done to assess the progression of cartilage repair after autologous chondrocyte implantation (ACI) with magnetic resonance imaging (MRI) and to correlate the findings with the clinical outcome. MATERIALS AND METHODS: Forty-one patients (mean age 30 years) affected by chondral defects of the knee (27 patients) and ankle joint (14 patients) who underwent arthroscopic autologous osteochondral grafting were studied 6 months and 1 year postoperatively with MRI. Cartilage repair after chondrocyte implantation was studied by assessing the degree of defect filling, graft integration, graft signal intensity, integrity of the subchondral lamina and trabecular oedema underneath the graft. MR findings were correlated with clinical data. RESULTS: Postoperative MRI evaluation at 6 months demonstrated complete filling of the osteochondral defect in 12/41 cases, complete integration in 18/41, mild hyperintensity in 28/41, intact subchondral lamina in 38/41 and trabecular oedema in 11/41. Postoperative MRI evaluation at 1 year demonstrated complete filling of the osteochondral defect in 9/41 patients, complete integration in 22/41, mild hyperintensity in 23/41, intact subchondral lamina in 36/41 and trabecular oedema in 8/41. Filling of the osteochondral defect and incomplete integration, nonintact subchondral lamina, high signal intensity and absence of oedema were found to correlate with worse clinical-functional outcomes. CONCLUSIONS: MRI shows direct prognostic signs of the clinical outcome of ACI.

Angiosonography in suspicious breast lesions with non-diagnostic FNAC: comparison with power Doppler US.

PURPOSE: The purpose of this study was to evaluate the accuracy of angiosonography in comparison with colour Doppler ultrasound (US) in the discrimination of suspicious breast lesions with nondiagnostic fine-needle aspiration cytology (FNAC). MATERIALS AND METHODS: Pre-operative Power Doppler US and angiosonography were prospectively performed in 20 suspicious breast lesions with non-diagnostic FNAC. A second-generation US contrast agent was utilised with a high-frequency transducer and a contrast-specific algorithm (low acoustic pressure CnTI). The enhancement characteristics of all lesions were analysed using qualitative and quantitative parameters obtained from time-intensity curves with the different imaging modalities. The final diagnosis was confirmed at pathology in all cases. Microvessel density (MVD) was assessed in the surgical specimen using CD34. RESULTS: The correct assessment of biological behaviour was achieved in all cases by angiosonography (sensitivity of 100%; specificity of 91%) and colour Doppler US (45% sensitivity; 78% specificity). MVD correlated with the biological behaviour. CONCLUSIONS: Angiosonography is more accurate than colour Doppler US in the correct assessment of biological behaviour of suspicious breast lesions.

[Digital radiography of the thorax with selenium detector. Clinical results and optimization potential of the technique]. / Radiografia digitale del torace con detettore al selenio. Risultati clinici e possibilità di ottimizzazione della tecnica.

INTRODUCTION: We compared the image quality of the chest radiograph obtained with a digital selenium detector and with a conventional system and investigated the possible improvements in the digital technique resulting from kilovoltage (kV) lowering, antiscattering grid addition and image format reduction. MATERIAL AND METHODS: 150 subjects in the first series were submitted to posteroanterior chest radiography with both the selenium and the conventional systems. Image quality was compared by giving a score to the depiction of anatominal and pathologic findings. Thirty-two and 31 subjects from two other series were submitted to two digital chest radiographs each: the former at high (150) and low (90) kV, and the latter at 150 kV with the antiscattering grid and at 90 kV without the grid. Comparisons were made by choosing the better of the two images of each subjects. A score was given to the depiction quality of several difficult-to-detect findings in full-size and small-size format digital images obtained in another series of 27 subjects. RESULTS: As for anatomical detailing, digital selenium images were of much better quality than conventional images: the mean scores given by 3 observers to digital images (5.32; 5.55; 6.68) are higher than those given to the corresponding conventional images (4.49; 5.02; 5.81) and the difference is statistically significant (p < 0.001 in all cases). The advantage of digital over conventional images is also significant with reference to diagnostic confidence in the identification of pathologic findings (p < 0.001; p < 0.005; p < 0.01), but to a lessere extent (mean scores: 3.98; 4.22; 3.60 for the digital system, versus 3.43; 3.69; 3.38 for the conventional system). The digital images acquired at lower kV (90 kV) were much more frequently chosen by the two observes (87.5% and 96.8% of cases) than the images acquired at 150 kV; the entry dose at lower energies (91 muGy using an anthropomorphic phantom) is not significantly higher than the dose given at 150 kV (85 muGy). No significant difference was found in the two observers' choice between the digital images taken at 90 kV without antiscattering grid and those taken at 150 kV with the grid, the former being preferred in 38.7% and 58% of cases. The level of diagnostic confidence in the detailing of difficult-to-detect findings was slightly higher in full-size digital images (mean scores: 5.33 and 6.77) than in small-size ones (4.88 and 5.96). DISCUSSION AND CONCLUSIONS: Digital selenium images always exibit better quality than conventional images: the difference is very marked relative to anatomical detailing and not so striking, though still significant, in showing pathologic findings. Digital selenium image quality can be improved relative to the manufacturer's guidelines (150 kV exposure with no grid), by lowering the kV and adding the antiscattering grid, without increasing patient exposure too much. Digital image format reduction allows cost containment without affecting diagnostic reliability.