False-Positive Result and Reattendance in the Ontario Breast Screening Program

Objective:

To determine the association between initial screen result and returning for a second screen in an organised breast screening programme for women with a biennial screening recommendation.

Setting:

Women who attended the Ontario Breast Screening Program (OBSP).

Methods:

A retrospective cohort study was conducted of 140,723 Ontario women aged 50 years ond older who had an initial screen at the OBSP between 1 July 1990 and 31 December 1995 and were followed until 30 June 1998. Rescreening rates at 36 months and risk ratio estimates were calculated using survival methods. Age of women, year of screen, region (within Ontario) and initial screen result were compared. For initial screen results, returning for a second screen was examined by integration of screening centre with an assessment programme and by modality of referral.

Results:

Women with a false-positive result were less likely to return for a second screen as were women aged 70 and older and those living in regions of Ontario with fewer OBSP screening centres. However, there were minimal differences in reattendance behaviour by initial screen result for women screened at the OBSP centre with an assessment programme.

Conclusions:

Integration of breast screening and assessment services improved reattendance of women with false-positive screen results within an organised screening programme.

Predictors of survival in patients with non-curative stage IV cancer and malignant bowel obstruction

PURPOSE

Malignant bowel obstruction (MBO) occurs in up to 15% of patients admitted to palliative care wards and management can be clinically challenging. Survival is generally poor with a reported median survival of 1-3 months; however, there are no studies describing predictors of survival for patients with MBO.

PATIENTS AND METHODS

All patients admitted to a tertiary care hospital with a MBO were approached between March 1, 2006 and March 31, 2008 to enter the study. Demographic, clinical, laboratory, and radiographic information were prospectively collected from patient charts and the patient's functional status (Eastern Cooperative Oncology Group score, ECOG) at admission was recorded. Follow-up was until death or the end of the study (August 2008). Survival was estimated using Kaplan-Meier plots and Cox regression models were used to evaluate prognostic factors for survival.

RESULTS

Thirty-five patients were recruited. Median patient age was 61% and 46% were female. Median survival of the cohort was 80 days (range 7-873). Median survival for patients with an ECOG performance status of 0-1 (n = 15) was 222 days, for ECOG 2 patients (n = 9), 63 days and for patients with an ECOG 3/4 score (n = 11) it was 27 days. ECOG status was the strongest predictor of survival on the multivariate analysis. In addition, a low blood urea nitrogen level or a high albumin on admission was also associated with prolonged survival.

CONCLUSION

An ECOG score of 0/1 for patients with MBO in the setting of Stage IV non-curative cancer is the strongest predictor of overall survival.

Room for improvement? A quality-of-life assessment in patients with malignant bowel obstruction

This prospective study followed 35 patients admitted to hospital with malignant bowel obstruction (MBO) to evaluate quality of life (QOL). Subjects completed the Edmonton Symptom Assessment Scale (ESAS) and Rotterdam Symptom Checklist (RSCL) at recruitment, and at one week, one month and three months.The highest ranked ESAS scores at recruitment (which was generally 18-36 hours post admission to hospital) included loss of appetite (median=7.5), fatigue (6.5) and overall well-being (6.0). The total ESAS score improved by 7.5, 11.5 and 11.0 points respectively at one week, one month and three months (p<0.05, p<0.01, NS).RSCL median scores for physical and psychological subscales were high at baseline (36.2, 42.9) and improved significantly at one week and one month (p<0.05). Psychological functioning appeared to be worsening by three months and at no time did activity level improve significantly. The overall QOL score was extremely poor at baseline (6.0 median) improving to 3.3 at one month (median fall=1.0, p<0.05) and 3.4 at three months.Further work should address the lack of improvement in activity and apparent deterioration in psychological functioning after one month.

Verification of imatiNib cost-effectiveness in advanced gastrointestinal stromal tumor in British Columbia (VINCE)

10049

Background: To evaluate the cost-effectiveness of imatinib in British Columbia Cancer Agency (BCCA) patients with advanced gastrointestinal stromal tumors (GIST) in terms of median overall survival (OS) and median progression-free survival (PFS). Results were also compared to imatinib literature reports. Methods: A pragmatic, retrospective medical record review identified BCCA patients with advanced GIST who received imatinib or historical treatment during successive, pre-specified time periods. Data was collected on survival, duration of therapy, progression free status, and cost of drug, labour and supplies. Primary outcome was the cost-effectiveness of imatinib based on median-OS (Kaplan Meier method). Secondary outcomes were cost-effectiveness based on median overall-PFS, cost of therapy per patient, 1-year OS, and 1-year PFS. This study was conducted from the perspective of the BCCA. Sensitivity analyses varying effectiveness over the 95% CI, cost over the range of treatment duration, discounting level at 0, 3 and 5%, and comparing effectiveness to literature controls were performed. Life expectancy (mean survival), was also calculated and used in the CE sensitivity analysis. Results: Forty-six and 47 patients in the imatinib and historical groups respectively were analyzed. Median-OS in the imatinib group was 66.7 months (95% CI 61.7, infinity) compared to 7.7 months (95% CI 6.0, 12.6) in the historical group. Median-PFS was 45.3 months (95% CI 24.4, infinity) and 5.6 months (95% CI 3.5, 8.5) respectively. The 1-year OS and 1-year PFS for the imatinib group was 95.4% (95% CI 82.9, 99.2) and 81.4% (95% CI 66.1, 91.1) compared to 32.6% (95% CI 20.0, 48.1) and 17.4% (95% CI 8.3 - 32.0) in the historical group. No difference was seen between the imatinib study results and literature reports. The annual incremental cost-effectiveness ratio was $15,882 per median life year gained and $23,603 per median year of PFS for imatinib. Conclusions: Imatinib for advanced GIST is a cost- effective treatment in BC. Results were robust across the range of sensitivity analyses performed.

No significant financial relationships to disclose.

Cost-effectiveness (CE) analysis of CHOP and rituximab for diffuse large B-cell lymphoma (DLBCL) in British Columbia (BC)

6623

Background: The BC Cancer Agency (BCCA) provides province-wide, population-based care. Outcomes are monitored to verify therapeutic effectiveness and justify funding for systemic treatment policies. The CE of rituximab with CHOP (cyclophosphamide, doxorubicin, vincristine, and prednisone) (CHOP-R) for DLBCL was compared to its predecessor, CHOP alone. Methods: This was a pragmatic population-based CE analysis based on the original cohort of advanced DLBCL patients described by Sehn et al (JCO 2005) who received either CHOP or CHOP-R between Sept 1999 and Aug 2002 (18-months pre and post availability of rituximab in BC) according to standard BCCA treatment policy at the time. The primary endpoint was CE in terms of life-expectancy (LE) at a median follow-up of 4 years (cost-per-life-year-gained). Costs were incorporated into a decision analysis including primary systemic therapy and downstream chemotherapy, radiotherapy, and stem-cell transplant (SCT). Actual incidence of each downstream therapy was converted to a probability for each group. Downstream therapy costs were then multiplied by these probabilities and added to the respective primary treatment costs. The CE analysis took the BCCA perspective which includes all direct costs for active cancer treatment, and hospitalization for SCT, but not ambulatory supportive care. Sensitivity analyses varying LE to the extremes of its 95% CI, modeling out to 15 years and discounting at 0, 3, and 5% were performed. Results: 292 patients were included and categorized to treatment received: 148 CHOP and 144 CHOP-R (median follow- up 5.4 and 4 years respectively). LE to 4 years was 30.18 months for CHOP and 39.44 for CHOP-R. OS at 4 years was 48.8% and 70.1% for CHOP and CHOP-R respectively (p<0.0001) Respective costs of primary and downstream therapy were $4,682 and $7,198 for CHOP versus $26,366 and $6,228 for CHOP-R. The incremental CE ratio at 4 years median follow-up was $26,844 CDN per life year gained. Results were robust across univariate sensitivity analyses conducted. Conclusions: At 4 years median follow-up, CHOP-R improves LE and appears to be economically attractive at conventional thresholds. CE is an increasingly useful tool for the BCCA in making decisions about new cancer therapies.

No significant financial relationships to disclose.

A Computed Tomography-based Protocol vs Conventional Clinical Mark-up for Breast Electron Boost

AIMS

Computed tomography planning of whole breast radiotherapy (WBRT) improves breast coverage and reduces the normal tissue dose. Computed tomography planning may increase tumour bed boost treatment accuracy. The aims of this investigation were (1) to compare the breast boost volume treated with clinical mark-up with the volume delineated with computed tomography planning and (2) to study tumour bed volume changes between the initial planning computed tomography scan and a second computed tomography scan at the time of breast boost mark-up.

MATERIALS AND METHODS

Women receiving adjuvant WBRT and an electron boost after breast-conserving surgery were eligible. As per standard practice, WBRT was computed tomography planned while the boost electron portal was clinically defined. Electron field borders were then traced with wire and a second computed tomography scan was carried out in the boost treatment position. Post-surgical radiological abnormalities were contoured to create a tumour bed clinical target volume (CTV) on both scans (CTV1 and CTV2). A 1cm margin to CTV2 defined the planning target volume (PTV). The proportions of the CTV2 and PTV receiving 90% (V90) and 80% (V80) of the dose were calculated. Changes in volume between CTV1 and CTV2 were analysed.

RESULTS

Data from 47 eligible patients were analysed. The mean V90 for the PTV was 61%. Lower electron energy (P<0.001) and small field sizes (P=0.004) were associated with a low V90. The mean CTV decreased by 4.3 cm3 (P=0.014) and was smaller in those with a long surgery to computed tomography interval (P=0.008). On average, the 90% isodose covered 61 cm3 of normal tissue.

CONCLUSIONS

Conventional clinical breast boost planning is inaccurate. Electron boost computed tomography planning together with appropriate surgical clip placement and the use of mammograms and pathological information should provide optimal coverage of the tumour site. The boost could usually be planned from the initial computed tomography scan.

Setup variations in locoregional radiotherapy for breast cancer: an electronic portal imaging study

Recent trials demonstrating a survival benefit with locoregional radiotherapy (LRRT) to the chest wall and regional nodes in women with node-positive breast cancer have led to increased use of complex techniques to match three or more radiation fields, but information on setup reproducibility with LRRT for breast cancer is scarce. This study reports the magnitude and directions of random and systematic deviations in LRRT for breast cancer using an offline electronic portal imaging verification protocol. Electronic portal images (EPIs) of 46 consecutive women treated with LRRT for breast cancer from March 2001 to February 2002 with LRRT were analysed. Comparisons of EPIs to the corresponding digitally reconstructed radiographs were performed offline with anatomy matching. Displacements in mm were recorded in the superior-inferior (SI), medial-lateral (ML), and anterior-posterior (AP) directions. Random errors ranged from 2.0 mm to 2.5 mm for the breast/chest wall tangential treatments and 2.3 mm to 3.9 mm for the supraclavicular nodal treatments. Systematic errors occurred to a greater degree in the AP direction for the tangential fields and in the ML direction for the supraclavicular field. Displacements of > or =10 mm were found in 1.2% of breast/chest wall tangential treatments and in 6.2% of supraclavicular nodal treatments. These data demonstrate that EPI is a useful tool to verify setup reproducibility in LRRT for breast cancer.

CT slice index and thickness: impact on organ contouring in radiation treatment planning for prostate cancer

Objective: To assess the impact of CT slice index and thickness (3 mm versus 5 mm) on (i) prostate volume, dimensions, and isocenter coordinates, (ii) bladder and rectal volumes, and (iii) DRR quality, in the treatment of prostate cancer. Methods: 16 patients with prostate cancer underwent two planning CT‐scans using 3 and 5 mm slice index/thickness. Prostate, bladder, and rectum were outlined on all scans. Prostate isocenter coordinates, maximum dimensions, and volumes were compared along with bladder and rectal volumes. Bladder volumes and maximum diameters were further investigated using a second observer. A comparative analysis of DRR quality was conducted as well as a dosimetric analysis using DVH. Results: The differences in measurements of prostate volume, isocenter coordinates and maximum dimensions between the 3 and 5 mm scans, were small and not statistically significant. Similar finding was seen for rectal volume. However, bladder volume was always larger on the 3 mm scan (mean difference=27.9 cc;  SE=4.8 cc;  95% CI:  17.7−38.2 cc;  p<0.001) and the findings were reproduced with the second observer (mean difference=31.9 cc;  SE=4.7 cc;  95% CI:  21.9−41.9 cc;  p<0.001). The differences in volume are caused by a slight increase in (1) the measurement of the longitudinal dimensions on the 3 mm scans, and (2) the slice by slice measured bladder area on the 3 mm scans. The latter is due to partial volume effect. The 3 mm DRR were slightly better than the 5 mm DRR. The bladder DVH differed significantly in some patients. Conclusion: Bladder volume is significantly larger on the 3 mm scans. Differences in contoured areas may be accounted for, in part, by the partial volume effect.

PACS number(s): 87.57.–s, 87.53.–j

Trends in the incidence of testicular germ cell cancer in Ontario by histologic subgroup, 1964-1996

BACKGROUND

Testicular cancer is rare but is notable because it affects mainly young men. The incidence of this disease has been increasing in developed countries throughout the world for several decades. The authors examined trends in the incidence of testicular germ cell cancer in Ontario for the period 1964-1996 according to the 2 main histologic groups, seminoma and non-seminoma.

METHODS

Data on incident cases of testicular germ cell cancer diagnosed in Ontario residents aged 15-59 years between 1964 and 1996 were extracted from the population-based Ontario Cancer Registry. Annual rates of testicular cancer for the 2 histologic groups were analysed by means of log-linear regression to estimate average annual percent change.

RESULTS

Between 1964 and 1996 the incidence of testicular germ cell cancer increased by 59.4%, from 4.01 to 6.39 per 100,000. This corresponded to an average annual increase of about 2% for both nonseminoma and seminoma. The relative increase in incidence was greatest in the lowest age group (15-29 years) for both histologic groups, although the data suggest that the incidence of nonseminoma cancer in this age group began to decline in the early 1990s. The increase in incidence appears to be due to a birth cohort effect, with more recent cohorts of men at increased risk.

INTERPRETATION

The rise in the incidence of testicular germ cell cancer, not only in Ontario but also in many developed countries, requires investigation. The search for explanatory factors should focus on exposures whose prevalence may have increased over the past few decades and that are common enough to affect population incidence. The similarity of trends for seminoma and nonseminoma cancer suggests that the underlying risk factors are likely the same.

Results from the Ontario breast screening program, 1990-1995

OBJECTIVE

The Ontario breast screening program (OBSP) is a provincial breast screening programme offering two view mammography, clinical breast examination, instruction in breast self examination, and systematic two year recall to Ontario women 50 years and older. This paper presents the results of the programme's intermediate outcomes from 1990 to 1995 and compares them with recommended standards and other published programmes.

METHODS

Programme data from a provincial screening programme in a large Canadian province were collated from 18 sites in operation between July 1990 and December 1995.

RESULTS

In its first five years of operation, 215,738 screens were performed on 142,173 women. The referral rates for initial and rescreens are 13.8% and 8.6% respectively. A total of 1718 women were diagnosed with cancer, 1325 at initial screens and 393 at rescreens, resulting in cancer detection rates of 9.3 and 5.3/1000. The cancer detection rates for invasive cancers were 8.3/1000 at initial screens and 4.5/1000 at rescreens. The benign to malignant biopsy ratio was 1.5:1 at initial screens and 1.3:1 at rescreens. Of the 1358 cancers diagnosed at initial screens, 11.6% were in situ and 50.3% of invasive cancers of known size were < 15 mm. For women with invasive cancer where nodal status was known, 71.3% were node negative. The proportions at rescreens were 15.7%, 60.0%, and 76.0% respectively.

CONCLUSIONS

While the OBSP has achieved the standards suggested by other studies and programmes during its first five years of operation, there is work to be done to increase participation and obtain more complete data on tumour size and nodal status.

An assessment of the influence of clinical breast examination reports on the interpretation of mammograms in a breast screening program. Ontario Breast Screening Program Radiologists Research Group

The population-based Ontario Breast Screening Program (OBSP) provides two-yearly screening by both nurse examiner clinical breast examination (CBE) and two-view mammography to women aged 50 to 69. CBE alone accounts for about 5% of cancer detection. The purpose of this study was to determine whether CBE information affects radiologists' interpretation of mammography. Interpretation was defined by 1) radiologists' referral rates for diagnostic evaluation and 2) radiologists' accuracy in distinguishing cancer from non-cancer on mammograms. Mammograms were obtained from women randomly selected from those screened in the OBSP between 1990 and 1992. Selection was stratified by whether or not the nurse examiner had independently referred for diagnostic evaluation. Additional women diagnosed with breast cancer were selected to increase the precision of the receiver-operating characteristic (ROC) curve. Each participating OBSP radiologist read a set of mammograms twice, approximately three weeks apart. The first reading was based on mammograms alone. At the second reading, the CBE report was included on the reporting form. Among 620 women referred by the nurse, radiologist referral rate increased from 37.7% to 40.8% (p = 0.079) when CBE information was available. Among the 637 not referred by the nurse, radiologist referral rate decreased from 29.8% to 25.6% (p = 0.005). There was little effect on the sensitivity and specificity of radiologist referral and the areas under the two ROC curves (with and without CBE information) were not significantly different (p = 0.571). Although there was some effect of CBE information on radiologists' pattern of referral, there was no effect on accuracy of cancer detection.