Improving lesion volume measurements on digital mammograms.

Lesion volume is an important predictor for prognosis in breast cancer. However, it is currently impossible to compute lesion volumes accurately from digital mammography data, which is the most popular and readily available imaging modality for breast cancer. We make a step towards a more accurate lesion volume measurement on digital mammograms by developing a model that allows to estimate lesion volumes on processed mammogram. Processed mammograms are the images routinely used by radiologists in clinical practice as well as in breast cancer screening and are available in medical centers. Processed mammograms are obtained from raw mammograms, which are the X-ray data coming directly from the scanner, by applying certain vendor-specific non-linear transformations. At the core of our volume estimation method is a physics-based algorithm for measuring lesion volumes on raw mammograms. We subsequently extend this algorithm to processed mammograms via a deep learning image-to-image translation model that produces synthetic raw mammograms from processed mammograms in a multi-vendor setting. We assess the reliability and validity of our method using a dataset of 1778 mammograms with an annotated mass. Firstly, we investigate the correlations between lesion volumes computed from mediolateral oblique and craniocaudal views, with a resulting Pearson correlation of 0.93 [95% confidence interval (CI) 0.92 - 0.93]. Secondly, we compare the resulting lesion volumes from true and synthetic raw data, with a resulting Pearson correlation of 0.998 [95%CI 0.998 - 0.998] . Finally, for a subset of 100 mammograms with a malignant mass and concurrent MRI examination available, we analyze the agreement between lesion volume on mammography and MRI, resulting in an intraclass correlation coefficient of 0.81 [95%CI 0.73 - 0.87] for consistency and 0.78 [95%CI 0.66 - 0.86] for absolute agreement. In conclusion, we developed an algorithm to measure mammographic lesion volume that reached excellent reliability and good validity, when using MRI as ground truth. The algorithm may play a role in lesion characterization and breast cancer prognostication on mammograms.

The prognostic potential of mammographic growth rate of invasive breast cancer in the Nijmegen breast cancer screening cohort.

OBJECTIVES: Insight into the aggressiveness of potential breast cancers found in screening may optimize recall decisions. Specific growth rate (SGR), measured on mammograms, may provide valuable prognostic information. This study addresses the association of SGR with prognostic factors and overall survival in patients with invasive carcinoma of no special type (NST) from a screened population. METHODS: In this historic cohort study, 293 women with NST were identified from all participants in the Nijmegen screening program (2003-2007). Information on clinicopathological factors was retrieved from patient files and follow-up on vital status through municipalities. On consecutive mammograms, tumor volumes were estimated. After comparing five growth functions, SGR was calculated using the best-fitting function. Regression and multivariable survival analyses described associations between SGR and prognostic factors as well as overall survival. RESULTS: Each one standard deviation increase in SGR was associated with an increase in the Nottingham prognostic index by 0.34 [95% confidence interval (CI): 0.21-0.46]. Each one standard deviation increase in SGR increased the odds of a tumor with an unfavorable subtype (based on histologic grade and hormone receptors; odds ratio 2.14 [95% CI: 1.45-3.15]) and increased the odds of diagnosis as an interval cancer (versus screen-detected; odds ratio 1.57 [95% CI: 1.20-2.06]). After a median of 12.4 years of follow-up, 78 deaths occurred. SGR was not associated with overall survival (hazard ratio 1.12 [95% CI: 0.87-1.43]). CONCLUSIONS: SGR may indicate prognostically relevant differences in tumor aggressiveness if serial mammograms are available. A potential association with cause-specific survival could not be determined and is of interest for future research.

Impact of high electromagnetic field levels on childhood leukemia incidence.

The increasing exposure to electromagnetic fields (EMFs) has raised concern, as increased exposure may result in an increased risk of childhood leukemia (CL). Besides a short introduction of CL and EMF, our article gives an evaluation of the evidence of a causal relation between EMF and CL by critically appraising the epidemiological and biological evidence. The potential impact is also estimated by the population attributable risk. The etiology of CL is largely unknown, but is probably multifactorial. EMF may be one of the environmental exposures involved. Three pooled analyses of case-control studies showed a 1.4- to 1.7-fold increased CL risk for extremely low-frequency EMF (ELF-EMF) exposure levels above 0.3 µT. Several biases may have played a role in these studies, but are unlikely to fully explain the increased risk. For effects of radiofrequency ELF evidence is lacking. None of the proposed biological mechanisms by which ELF-EMF might cause CL have been confirmed. The estimated overall population attributable risk was 1.9%, with the highest estimates in Northern America and Brazil (4.2% and 4.1%, respectively). The potential impact of EMF exposure on public health is probably limited, although in some countries exposure might be relatively high and thus might have a more substantial impact. We recommend nationwide surveys to gain more insight into the contemporary exposure levels among children. Reducing exposure from power lines near densely populated areas and schools is advised. Future epidemiological studies should focus on limiting bias.

Regional recurrence in breast cancer patients with sentinel node micrometastases and isolated tumor cells.

OBJECTIVE: The impact of axillary treatment in daily practice on 5-year regional recurrence rate in breast cancer patients with isolated tumor cells or micrometastases in the sentinel node (SLN). BACKGROUND: Axillary dissection is recommended in patients with tumor-positive SLNs. But, in recent studies, regional recurrence rates seemed low if dissection was omitted. METHODS: We identified all patients in The Netherlands with invasive breast cancer who had an SLN biopsy before 2006, favorable primary tumor characteristics, and node-negative disease, isolated tumor cells or micrometastases as final nodal status. The primary endpoint was regional recurrence rate. To investigate differences in recurrence rates between patients with and without axillary treatment, a proportional hazard regression was carried out correcting for potential confounders. RESULTS: In total, 857 patients with node-negative disease, 795 patients with isolated tumor cells, and 1028 patients with micrometastases in the SLN were included. Without axillary treatment, the 5-year regional recurrence rates were 2.3%, 2.0%, and 5.6%, respectively. Compared with patients who underwent axillary treatment, the adjusted hazard ratio for regional recurrence in patients who underwent an SLN procedure only was 1.08 (95% CI, 0.23-4.98) for node-negative disease, 2.39 (95% CI, 0.67-8.48) for isolated tumor cells, and 4.39 (95% CI, 1.46-13.24) for micrometastases. Doubling of tumor size, grade 3 and negative hormone receptor status were also significantly associated with recurrence. CONCLUSIONS: Not performing axillary treatment in patients with SLN micrometastases is associated with an increased 5-year regional recurrence rate. Axillary treatment is recommended in patients with SLN micrometastases and unfavorable tumor characteristics.

Micrometastases or isolated tumor cells and the outcome of breast cancer.

BACKGROUND: The association of isolated tumor cells and micrometastases in regional lymph nodes with the clinical outcome of breast cancer is unclear. METHODS: We identified all patients in The Netherlands who underwent a sentinel-node biopsy for breast cancer before 2006 and had breast cancer with favorable primary-tumor characteristics and isolated tumor cells or micrometastases in the regional lymph nodes. Patients with node-negative disease were randomly selected from the years 2000 and 2001. The primary end point was disease-free survival. RESULTS: We identified 856 patients with node-negative disease who had not received systemic adjuvant therapy (the node-negative, no-adjuvant-therapy cohort), 856 patients with isolated tumor cells or micrometastases who had not received systemic adjuvant therapy (the node-positive, no-adjuvant-therapy cohort), and 995 patients with isolated tumor cells or micrometastases who had received such treatment (the node-positive, adjuvant-therapy cohort). The median follow-up was 5.1 years. The adjusted hazard ratio for disease events among patients with isolated tumor cells who did not receive systemic therapy, as compared with women with node-negative disease, was 1.50 (95% confidence interval [CI], 1.15 to 1.94); among patients with micrometastases, the adjusted hazard ratio was 1.56 (95% CI, 1.15 to 2.13). Among patients with isolated tumor cells or micrometastases, the adjusted hazard ratio was 0.57 (95% CI, 0.45 to 0.73) in the node-positive, adjuvant-therapy cohort, as compared with the node-positive, no-adjuvant-therapy cohort. CONCLUSIONS: Isolated tumor cells or micrometastases in regional lymph nodes were associated with a reduced 5-year rate of disease-free survival among women with favorable early-stage breast cancer who did not receive adjuvant therapy. In patients with isolated tumor cells or micrometastases who received adjuvant therapy, disease-free survival was improved.

Prognostic impact of isolated tumor cells in breast cancer axillary nodes: single tumor cell(s) versus tumor cell cluster(s) and microanatomic location.

In breast cancer, it has been shown that pN0(i+) and pN1mi have a comparable negative impact on disease-free survival, compared with pN0. However, pN0(i+) is considered to be a heterogeneous group. We determined the effect of metastatic size and microanatomic location within the pN0(i+) group on breast cancer recurrence. We included all Dutch breast cancer patients diagnosed in 1998-2005 with favorable primary tumor characteristics and a final nodal status of pN0(i+). For this analysis, only patients without adjuvant systemic therapy were eligible (n = 513). Presence of single tumor cells versus cell clusters, metastatic size and microanatomic location were recorded. Primary endpoint was disease-free survival. Analyses were adjusted for age at diagnosis, tumor size, tumor grade, axillary treatment and hormone receptor status. The 5-year disease-free survival of patients with single tumor cell(s) (n = 93) was 78.6% and with tumor cell cluster(s) (n = 404) 77.1%. The hazard ratio for disease events was 1.05 (95% CI 0.63-1.76) for cell cluster(s) compared with single cell(s). In a Cox regression model, doubling of metastatic tumor size corresponded to a hazard ratio of 1.21 (95% CI 1.02-1.43). The adjusted hazard ratio was 0.90 (95% CI 0.54-1.50) for parenchymal (n = 112) versus sinusoidal location (n = 395). Single tumor cells bear similar prognostic information as small tumor cell clusters, even though results do suggest that within the pN0(i+) group, increasing size of nodal involvement is associated with reduced survival. Microanatomic location does not seem to have prognostic relevance.

Pattern of follow-up care and early relapse detection in breast cancer patients.

Routine breast cancer follow-up aims at detecting second primary breast cancers and loco regional recurrences preclinically. We studied breast cancer follow-up practice and mode of relapse detection during the first 5 years of follow-up to determine the efficiency of the follow-up schedule. The Netherlands Cancer Registry provided data of 6,509 women, operated for invasive non-metastatic breast cancer in 2003-2004. In a random sample including 144 patients, adherence to follow-up guideline recommendations was studied. Mode of relapse detection was studied in 124 patients with a second primary breast cancer and 160 patients with a loco regional recurrence. On average 13 visits were performed during the first 5 years of the follow-up, whereas nine were recommended. With one, two and three medical disciplines involved, the number of visits was 9, 14 and 18, respectively. Seventy-five percent (93/124) of patients with a second primary breast cancer, 42 % (31/74) of patients with a loco regional recurrence after breast conserving surgery and 28 % (24/86) of patients with a loco regional recurrence after mastectomy had no symptoms at detection. To detect one loco regional recurrence or second primary breast cancer preclinically, 1,349 physical examinations versus 262 mammography and/or MRI tests were performed. Follow-up provided by only one discipline may decrease the number of unnecessary follow-up visits. Breast imaging plays a major and physical examination a minor role in the early detection of second primary breast cancers and loco regional recurrences. The yield of physical examination to detect relapses early is low and should therefore be minimised.

Impact of routine follow-up examinations on life expectancy in ovarian cancer patients: a simulation study.

OBJECTIVE: The clinical benefit of routine follow-up in patients treated for ovarian cancer is subject to debate. In this study, the magnitude of the potential survival benefit of routine examinations was evaluated by Markov modeling. METHODS: The clinical course of ovarian cancer was simulated using a 4-state nonstationary Markov model. Risk of recurrence and mortality probabilities were derived from individual patient data and Statistics Netherlands. The life expectancy was simulated for 3 follow-up scenarios: a current, withholding (all recurrences detected symptomatically), and perfect follow-up program (all recurrences detected asymptomatically). The impact of effective recurrence treatment in the future was modeled by varying the mortality ratio between patients with asymptomatically versus symptomatically detected recurrences. The model was validated using empirical data. RESULTS: The mean life expectancy of patients, aged 58 years and in complete clinical remission after primary treatment, was 10.8 years. Varying the transition probabilities with ±25% changed the life expectancy by up to 1.1 years. The modeled life expectancy for the withholding and perfect follow-up scenarios was also 10.8 years and insensitive to model assumptions. In patients with stages IIB to IV, the life expectancy was 7.0 years, irrespective of follow-up strategy. A mortality ratio of 0.8 for patients with asymptomatically versus symptomatically detected recurrences resulted in a gain in life expectancy of 5 months for withholding versus perfect follow-up. CONCLUSIONS: Routine follow-up in ovarian cancer patients is not expected to improve the life expectancy. The timing of detection of recurrent ovarian cancer is immaterial until markedly improved treatment options become available.

Increase in the incidence of gestational trophoblastic disease in The Netherlands.

OBJECTIVE: The objective of this study is to determine the incidence and time trends of gestational trophoblastic disease (GTD) in The Netherlands using population-based data. METHODS: Data on patients with a pathologically confirmed diagnosis of GTD from 1995 to 2008 were obtained from PALGA, a national archive containing all histopathology reports in The Netherlands. Data on number of deliveries were obtained from the Database of Statistics Netherlands. RESULTS: During the study period, 4249 GTD patients were registered. Overall incidence rates of hydatidiform mole (HM), choriocarcinoma and placental site trophoblastic tumor (PSTT) were 1.34 per 1000 deliveries, 3.1 per 100,000 deliveries, and 1.0 per 100,000 deliveries, respectively. Incidence rates of HM increased from 1.02 per 1000 deliveries in 1995 to 1.56 per 1000 in 2001, an increase of 0.091 per year (95% CI 0.081-0.101). After 2001 incidence rates remained constant (increase per year -0.010, 95% CI -0.045-0.024). Maternal age and ethnicity are known to influence the risk of HM. Highest incidences were observed in women under 20 and over 40years of age. The proportion of deliveries accounted for by women over 40years of age increased from 1.5% to 2.9%, whereas women under 20 accounted for 1.5% of deliveries. The proportion of live births of Asian descent increased from 2.6% to 3.7%. CONCLUSION: The incidence of GTD in The Netherlands increased significantly from 1995 to 2008. This can partially be explained by increased maternal age and increased proportion of live births of Asian descent. Part of the increase might result from improved diagnostic techniques. However, these factors do not seem to account for the total observed increase and part of the increase therefore remains unexplained.

No supportive evidence for clinical benefit of routine follow-up in ovarian cancer: a Dutch multicenter study.

INTRODUCTION: Routine follow-up is standard medical practice in ovarian cancer patients treated with curative intent. However, no strong evidence exists indicating that prognosis is improved. The objective of this study was to evaluate the routine follow-up schedule for ovarian cancer patients regarding the adherence to the Dutch protocol, the detection of recurrences, and the follow-up's impact on overall survival. METHODS: All 579 consecutive patients diagnosed with epithelial ovarian, primary peritoneal, or fallopian tube cancer in 4 Dutch hospitals between 1996 and 2006 were selected. Only patients in complete clinical remission after primary treatment were studied. Compliance to the Dutch follow-up guideline was assessed in a random sample of 68 patients. Of the 127 patients with recurrence, the mode of recurrence detection was addressed. Survival time since primary treatment was calculated using the Kaplan-Meier method. RESULTS: The patients received more follow-up visits than was recommended according to the guideline. The cumulative 5-year risk of recurrence was 55% (95% confidence interval [CI], 43%-67%). The survival of patients with recurrent ovarian cancer detected asymptomatically at a routine visit (n = 51) tended to be better compared with patients with symptomatic detection at a routine (n = 31) or diagnosed after an interval visit (n = 31). The median survival times were 44 (95% CI, 38-64), 29 (95% CI, 21-38), and 33 months (95% CI, 19-61), respectively (P = 0.08). The median time from primary treatment to recurrence was similar for the 3 groups: 14, 10, and 11 months, respectively (P = 0.26). CONCLUSIONS: Follow-up in line with (inter)national guidelines yields a seemingly longer life expectancy if the recurrence was detected asymptomatically. However, this result is expected to be explained by differences in tumor biology and length-time bias.

Considering early detection of relapsed ovarian cancer: a review of the literature.

INTRODUCTION: Patients treated for ovarian cancer with curative intent receive an intensive follow-up program in the years after treatment. However, the aimed improved survival through early detection of recurrence is subject to debate. Theoretically, the survival benefit depends on the lead time and the preclinical detection rate and on the effectiveness of recurrence treatment. This systematic review aimed at determining the effectiveness of early detection of recurrent ovarian cancer. METHODS: A systematic literature search in PubMed, EMBASE, MEDLINE, and the Cochrane Library was performed for articles published in 1985 to 2009 in English, German, or Dutch, excluding editorials, letters, and case reports. RESULTS: In total, 67 articles were included. Of 4 observational studies and 1 randomized controlled trial, only 1 observational study reported a better survival for patients who attended routine follow-up compared with those who did not. The sensitivity of cancer antigen 125 for a preclinical recurrence, based on 38 articles using 35 U/mL as a cutoff level, was 65%, with a median lead time of 3 months (range, 1-7 months). Seven studies showed that, on average, 67% (ranging from 20% to 80%) of the 798 relapsed patients had no clinical symptoms when recurrent ovarian cancer was diagnosed. CONCLUSIONS: Routine follow-up may detect 2 of 3 recurrences asymptomatically with a lead time of 3 months. Recurrence treatment may extend survival by several months, but published studies did not show a survival advantage of early detection by routine follow-up examinations. Therefore, the content and aims of routine follow-up should be reconsidered, whereas routine cancer antigen 125 testing with the aim to improve life expectancy should be omitted.

Undertreatment of tracheal carcinoma: multidisciplinary audit of epidemiologic data.

National epidemiologic data were examined to determine the eligibility for curative therapy in tracheal carcinoma. An expert audit of primary tracheal carcinomas registered from 2000 to 2005 with the Netherlands Cancer Registry (NCR) included blinded patient data and radiographic review to assess diagnosis and resectability. Actual treatment was compared with the opinions of a multidisciplinary panel (Radboud panel) and a second reviewer. Of 101 NCR-registered primary tracheal carcinomas, the Radboud panel diagnosis was metastatic disease or local extension of adjacent tumors in 34. Seventeen cases were excluded for missing data. In 50 cases confirmed by panel and a second reviewer, actual treatment consisted of surgery in 12 (24%), radiotherapy in 29 (58%), endobronchial treatment in 6 (12%), and observation in 3 (6%). Both panel and second reviewer identified 16 additional surgical candidates, a total of 28 (56%) of 50. Treatment recommendations of panel and second reviewer disagreed in four cases (8%). One-third of NCR-registered primary tracheal carcinomas were misclassified nontracheal primary tumors involving the trachea. A majority of cases meeting audit criteria for diagnosis and surgical resection was treated with other modalities. Interreviewer disagreement was small. The audit of a national cancer registry suggests that incorrect diagnosis and undertreatment are common in rare airway tumors.

Implementation of sentinel node biopsy in breast cancer patients in the Netherlands.

BACKGROUND: This population-based study describes the implementation of the sentinel node biopsy (SNB) in breast cancer patients in the Netherlands. We examined the extent of use over time of SNB in women who were considered eligible for SNB on the basis of their clinical status. METHODS: The study included a total of 35,465 breast cancer patients who were diagnosed with T1-2 tumours (5.0 cm), negative axillary lymph node status and no distant metastases upon clinical examination between 1st January 1998 and 31st December 2003 in six Comprehensive Cancer Centre regions in the Netherlands. Information on axillary surgery was classified as SNB alone, SNB+axillary lymph node dissection (ALND), ALND alone or none. Patterns of use of axillary surgery were summarised as the proportion of patients receiving each surgery type. RESULTS: Overall, 25.7% of patients underwent SNB alone, 19.1% underwent SNB+ALND, 50.0% had ALND alone and 5.2% did not have axillary surgery. SNB was more common in women who had breast-conserving surgery: 50.5% of patients who received breast-conserving surgery underwent SNB compared to 40.7% of patients who had mastectomy (p<0.0001). Amongst patients receiving breast-conserving treatment, 31.7% had SNB as final axillary surgery, whilst 20.5% of patients who had mastectomy had SNB alone (p<0.0001). The proportion of women who underwent a SNB alone or in combination with ALND increased over the period 1998-2003, from 2.1% to 45.8% and from 6.7% to 24.8%, respectively. There were marked differences in the patterns of dissemination of the use of SNB between regions: by 2003, the difference between the regions with the highest and lowest proportion of use was 25%. CONCLUSIONS: SNB has become the standard-of-care for the treatment of breast cancer patients clinically diagnosed with T1-2 tumours, clinically negative lymph nodes and without distant metastases. In 2003, 70.6% of patients with early breast cancer in the Netherlands received SNB, and within this group, 64.9% of patients had SNB as the final axillary treatment. Implementation of SNB may depend on factors associated with regional organisation of care.

The impact of histological subtype in developing both ovarian and endometrial cancer: A longstanding nationwide incidence study.

OBJECTIVES: To estimate the incidence of ovarian cancer (OC) and endometrial cancer (EC) separately, as well as double cancers diagnosed in the same calendar year, and to relate the occurrences to histological subtype. STUDY DESIGN: All cases of epithelial OC and EC diagnosed in the Netherlands in 1989-2009 were related to population data. Histologically specific associations were made using the ratio of observed and expected incidence numbers, calculated with age-specific incidence rates. RESULTS: 25,489 OC and 32,729 EC were analyzed, and 649 OC/EC. Life-time risks for OC and EC were 1.8% and 2.4%. Among OC, adenocarcinoma (18%) and serous cancers (33%) were the most prevalent subtypes. In EC, adenocarcinoma (39%) and endometrioid cancer (37%) were highest, with hardly any serous cancers. The observed incidence of OC/EC was 50-fold higher than expected (95% CI, 46-54). For patients aged <55years, the O/E ratio was 274, for the elderly 32, both findings are significant. Of the 2345 OC endometrioid subtype, 294 had EC (12.5%), whereas 1.1 was expected. In EC patients, no particular histological subtype was distinguished with a highly elevated occurrence of OC. The 680 serous EC patients had 11 double cancers (1.6%), of which 8 with the ovarian serous subtype. CONCLUSION: Strong relationships exist between malignancies in the ovary and a second primary malignancy in the endometrium, especially for the endometrioid subtype of ovarian cancer. Viewed from the endometrial site, no special subtype was noted, and the influence of endometrial serous adenocarcinoma in developing serous OC is not plausible.

Local recurrences after different treatment strategies for ductal carcinoma in situ of the breast: a population-based study in the East Netherlands.

PURPOSE: Outcomes after different treatment strategies for ductal carcinoma in situ (DCIS) of the breast were analyzed for a geographically defined population in the East Netherlands. METHODS AND MATERIALS: A total of 798 patients with a first diagnosis of DCIS between January 1989 and December 2003 were included and their medical records were reviewed. Survival rates for ipsilateral recurrences were calculated by the Kaplan-Meier method and a multivariate Cox proportional hazards regression model was used to evaluate the prognostic significance of different variables. RESULTS: The 5-year recurrence-free survival was 75% for breast conserving surgery (BCS) alone (237 patients) compared with 91% for BCS followed by radiation therapy (RT; 153 patients) and 99% for mastectomy (408 patients, p < 0.01). Independent risk factors for local recurrences were treatment strategy, symptomatically detected DCIS, and presence of comedo necrosis. Margin status reached statistical significance only for patients treated by BCS (hazard ratio, 2.0; 95% confidence interval, 1.1-4.0) whereas significance of other prognostic variables did not change. CONCLUSIONS: In a defined population outside a trial setting, RT after BCS for DCIS lowered recurrence rates. Besides the use of RT, a microscopically complete excision of DCIS is essential. This is especially true for patients with symptomatically detected DCIS and with tumors that contain comedo necrosis, as these groups are at particular high risk for recurrent disease.

[The effect of cancer screening on mortality. The case-control study as evaluation method]. / Het effect van kankerscreening op sterfte. Het patiënt-controleonderzoek als evaluatiemethode.

The case-control study is an efficient study design for evaluating the effect of cancer screening. This study method enables calculation of the percentage by which the risk of mortality from cancer decreases in participants in a screening programme if the carcinoma is detected early. Reductions in mortality of 15-65% have been published for participants in a population screening programme for breast, cervical and colorectal cancer. The case-control study is an observational study, in other words it does not have an experimental design. There is, therefore, a risk that confounding and self-selection bias may cause over- or underestimation of the mortality reduction. It is, therefore, important that at publication investigators indicate the extent to which they have succeeded in minimalising the occurrence of these forms of bias in their study design and data analysis.

[Predicting chance of disease: calculation using prediction rules]. / Ziektekansen voorspellen: rekenen met predictieregels.

A prediction rule is a statistical model that can be used to predict the presence or absence of a disease based on a limited number of tests or predictive factors. One of the mathematical methods used to formulate prediction rules is a logistic regression analysis of patient data. The discriminatory power of a model is visualizable using box-whisker plots and ROC curves; calibration plots show the match between the predicted chance and the observed frequency of a disease. These graphs are used to assess whether a model adequately reproduces reality. On publication of prediction rules it is important that the regression function is written out and that the chances of a disease on the basis of diagnostic scores are displayed in a histogram. For the practical significance of the model, it is also important to know how often the predicted low, medium or high probabilities of a disease do actually occur in comparison with the advance chance of occurrence.

Age at diagnosis and breast cancer survival in iran.

Background. Tumour characteristics are the most important prognostic factors in breast cancer. Patient-related factors such as young age at diagnosis, obesity, and smoking behaviour may also modify disease outcome. Due to the absence of a unique definition for "young age breast cancer" and the resulting variation in disease management, findings on the association between young age and prognosis of breast cancer are controversial. Methods. This study included 1500 patients with a primary diagnosis of breast cancer in six Iranian hospitals from 5 provinces. We modelled the relative excess risk (RER) of breast cancer death to age at diagnosis and tumour characteristics. Results. Excess risks of death were observed for stage IV disease and poorly differentiated tumours: RER of 4.3 (95% CI: 1.05-17.65) and 3.4 (95% CI: 1.17-9.87), respectively. "Older" patients, particularly those aged 50 and over, presented more often with advanced and poorly differentiated tumours (P = 0.001). After adjustment for stage, histological grade, Her-2 expression, estrogen and progesterone receptors, and place of residency, breast cancer mortality was not significantly different across age groups. Conclusion. We conclude that there is no prognostic effect of age at diagnosis of breast cancer among breast cancer patients treated at cancer centres in different parts of Iran; young and relatively old women have similar risks of dying from breast cancer.

[Responsible cancer screening]. / Verantwoord screenen op kanker.

In the Netherlands, national screening programs for breast and cervical cancer are operating, whilst that for colorectal cancer is in preparation. In the meantime, experimental studies have been conducted into the effectiveness of prostate and lung cancer screening. Death from these five types of cancer is reduced by these screening investigations. However, these screening programmes also have disadvantages, such as unnecessary referral for definitive diagnosis in the hospital. The average hospital would receive on a yearly basis via screening 156 referrals of women with breast cancer, 79 for cervical cancer and nearly 1100 persons for colorectal cancer. n average general practice encounters annually 3 positive screening results for breast cancer, almost 1 referral for cervical carcinoma or an early stage thereof, and every two years a patient with CIN III. For colorectal cancer around 22 referrals can be expected yearly, of which 8 will have adenoma or cancer.

[Vaginal and cervical cancer due to diethylstilbestrol (DES); end epidemic]. / Vagina- en cervixcarcinoom door diëthylstilbestrol (des). Einde epidemie.

OBJECTIVE: To determine the current situation regarding the epidemic of clear cell adenocarcinoma of the vagina and the uterine cervix (CCAC) in relation to the exposure in utero to diethylstilbestrol (DES). DESIGN: Descriptive. METHODS: Patients with CCAC of the uterine cervix or vagina born after 1946 and diagnosed in the period 1969-2005, were identified through the Nationwide network and registry of histo- and cytopathology in the Netherlands and from 2003 onwards through the Netherlands Cancer Registry. Exposure data and clinical data were obtained by means of questionnaires and medical records. The histology slides of tumours were reviewed. For the patients who did not provide consent, only the date of diagnosis and age at diagnosis were known (n = 10). RESULTS: Up to 2005, 144 CCAC patients were registered. Age at diagnosis varied from 8-54 years (mean: 28 years). In the years 1981-2000, the number of new diagnoses in 4 successive 5-year periods was fairly stable (n=26-30) but it was considerably lower in 2001-2005 (n=13). Of the patients whose history of intrauterine exposure to DES was known, 62% had been exposed (76/122). The mean age at diagnosis was 24 years for exposed patients compared to 32 years for non-exposed patients. The 10-year survival rate was 78% (95% CI: 68-87) for exposed and 69% (95% CI: 56-82) for non-exposed patients. CONCLUSION: Since 2000, the incidence of CCAC of the vagina and cervix has decreased markedly compared to the situation in the 1980s and 1990 s. In particular, the number of patients with CCAC exposed in utero to DES has decreased. Whether this decrease shall continue over the coming years remains to be seen.