European Journal of Cancer (2015) 51, 1630–1637

Available at www.sciencedirect.com

ScienceDirect journal homepage: www.ejcancer.com

Cancer surveillance using registry data: Results and recommendations for the Lithuanian national prostate cancer early detection programme Adam Gondos a,⇑, Agne Krilaviciute b, Giedre Smailyte b, Albertas Ulys c, Hermann Brenner a,d a

Division of Clinical Epidemiology and Aging Research, German Cancer Research Center, Heidelberg, Germany Lithuanian Cancer Registry, National Cancer Institute, Lithuania c National Cancer Institute, Oncourology Department, Lithuania d German Cancer Consortium (DKTK), Heidelberg, Germany b

Received 21 March 2015; received in revised form 12 April 2015; accepted 14 April 2015 Available online 1 June 2015

KEYWORDS Prostate cancer National early detection programme PSA-test Epidemiology

Abstract Introduction: We describe long term trends in prostate cancer epidemiology in Lithuania, where a national prostate specific antigen (PSA) test based early detection programme has been running since 2006. Methods: We used population-based cancer registry data, supplemented by information on PSA testing, life expectancy and mortality from Lithuania to examine age-specific prostate cancer incidence, mortality and survival trends among men aged 40+ between 1978 and 2009, as well as life expectancy of screening-eligible men, and the proportion of men with a first PSA test per year since the programme started. Results: The number of prostate cancer patients rose from 2.237 in 1990–1994 to 15.294 in 2005–2009. By 2010, around 70% of the eligible population was tested, on average around two times. The early detection programme brought about the highest prostate cancer incidence peaks ever seen in a country to date. Recent incidence and survival rises in the age

⇑ Corresponding author: Division of Clinical Epidemiology and Aging Research, German Cancer Research Center, INF 581, 69120 Heidelberg, Germany. Tel.: +49 6221 421348; fax: +49 6221 421302. E-mail address: [email protected] (A. Gondos).

http://dx.doi.org/10.1016/j.ejca.2015.04.009 0959-8049/Ó 2015 Elsevier Ltd. All rights reserved.

A. Gondos et al. / European Journal of Cancer 51 (2015) 1630–1637

1631

groups 75–84 suggest PSA testing in the elderly non-eligible population. Life expectancy of men aged 70–74 indicates that less than 30% of patients will live for 15 years and may have a chance to benefit from early detection. Conclusions: Early detection among men aged 70–74, and particularly among the elderly (75+) may have to be reconsidered. Life expectancy assessment before testing, avoiding a second test among men with low PSA values and increasing the threshold for further evaluation and the screening interval may help reducing harm. Publishing information on treatment modalities, side-effects and patient reported quality of life is recommended. Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction To our knowledge, Lithuania is the only country in the world with a prostate specific antigen (PSA) test based official national prostate cancer early detection programme [1]. The programme was launched in 2006, when men aged 50–74, as well as men aged 45–49 with family history of prostate cancer became eligible to undergo PSA testing once a year, while since 2008, testing is offered every two years. The programme works on an invitation by opportunity basis currently, and tests are offered when men visit their general practitioner for some reason. Further urological evaluation is offered when PSA levels exceed 3 ng/ml. Although the European Randomized study of screening for prostate cancer (ERSPC) demonstrated a potential for mortality decline [2,3], the very long lead time of cancer detection [4], the potentially very serious side-effects of the treatment and large variation in the life expectancy of different populations [5], as well as concern about the overall validity of the ERSPC results [6] keep PSA-based early detection highly controversial [7–9]. In this paper, we provide a comprehensive analysis of population-based trends of prostate cancer incidence, survival and mortality using data from the Lithuanian Cancer Registry [10] and other relevant accompanying data sources, and compare recent trends in prostate cancer incidence and mortality in Lithuania to those seen in the United States (US), the country which probably experienced the highest intensity of PSA testing in the recent two decades. Additionally, we include an analysis of recent life expectancy in Lithuanian men to consider the age-specific variation in chances for a potential benefit from testing. 2. Materials and methods 2.1. Data sources and calculation of basic epidemiologic measures Incidence data for this analysis came from the population-based Lithuanian Cancer Registry, which covers the entire population (3 million people in 2011) of the country, and has been in operation since 1984 (yet incidence data are available since 1978) [10]. The

registry’s principal sources of information are from primary, secondary and tertiary health care centres and pathology laboratories, with additional notifications from all physicians, hospitals and other health institutions. These notifications are supplemented by death certificate information, and population registry information to verify vital status – these data were available since 1990. National mortality data were available from Statistics Lithuania [11] for the years 1978–2009. Here, cancer information of patients aged 40 years and older and diagnosed with prostate cancer (ICD-10 C61) between 1978 and 2009 was considered. We calculated age-specific incidence and mortality trends in detail for 5-year age groups between 45–49 and 85+. Additional data on the number of first tested men in the eligible population in each year between 2006 and 2010, and the total number of tests administered in those years were obtained by request from the National Health Insurance Fund Database to assess testing coverage and intensity. Prostate cancer incidence and mortality data in the US among Caucasians for matched years and age groups were obtained from the SEER’s online database [12]. 2.2. Survival analysis Only patients who were not registered by death certificates only (DCO) or autopsy were included in the survival analysis. Five-year relative survival estimates were calculated for the periods 1995–1999, 2000–2004 and 2005–2009. Relative survival estimates were derived as a ratio of the absolute survival of the cancer patients divided by the expected survival of the age-matched Lithuanian male population. Expected survival estimates were derived using the Ederer II method [13]. Life expectancy was calculated on the basis of national life tables for Lithuania [14]. Survival calculations were done with the STATA statistical package, using the freely available ‘srtr’ command [15], which was set up to allow for model based period analysis [16]. Age standardisation of the period estimates and standard errors was done using the age groups 40–49, 50–59, 60–69, 70–79 and 80+, using weights from International Cancer Survival Standards (ICSS) proposed by Corazziari et al. [17].

1632

A. Gondos et al. / European Journal of Cancer 51 (2015) 1630–1637

Overall, extremely large increases in 5-year relative survival were seen, with estimates rising from below 50% units in 1995–1999 to up to and above 90% in all age groups except the oldest, in which patients had a 5-year relative survival of 85% by 2005–2009 (Table 2). Among those with localised disease, 5-year relative survival rose by 40% units between 1995–1999 and 2005– 2009, and exceeded 100% in the latest period, indicating no prostate cancer related excess mortality compared to the general population in the first 5 years after diagnosis. Among patients with locally advanced disease, survival also rose by 40% units, and reached 86% by 2005– 2009. Less dramatic, but still very large rise was seen among patients with regional disease, while the smallest rise, 9% units, was seen among patients diagnosed with metastasis, who continued to have low 5-year survival (22%) in 2005–2009. In 2006, the first year of the programme, testing occurred most frequently in the age groups 65–69 and 70–74, in which 30% of the eligible population received a PSA test, compared to ca. 25% in the age groups 55–59 and 60–64, and 17% in the age group 50–54 (Table 3). In the youngest eligible age group (50–54), into which yet untested men enter each year, proportions with first tests declined only slightly in the subsequent years, while in older age groups, the percent of men newly tested declined strongly. Around 72–78% of the total eligible male population has received at least one PSA test in

3. Results Table 1.a presents the number of diagnosed prostate cancer cases by age group and calendar period between 1990 and 1994 and 2005–2009. Compared to 1990–1994, when ca. 2.200 cancers were registered, the total number of diagnoses rose by roughly 50% by 1995–1999, and approximately doubled in each consecutive period thereafter, resulting in more than 15.000 diagnoses by 2005–2009. Between 1990–1994 and 2005–2009, absolute rises in the number of cases was largest in the age groups 60–69 and 70–79, which contributed 75% of the rise in case numbers, while the proportionately largest rises (10- to 12-fold) occurred in the two youngest age groups. In 1990–1994, around 75% of patients were diagnosed with localised or locally advanced stage, while around 20% were found to have metastasis at diagnosis (Table 1.b). In the following two periods, the proportion of patients diagnosed with locally advanced disease increased substantially, while the proportion of patients with metastasis at diagnosis declined considerably. In the latest period, after a further radical shift in the stage distribution, localised or locally advanced disease was found in 63% and 33% of the all patients, respectively, and only a small fraction of all patients were diagnosed with regional or metastatic disease (around 1% and 4%, respectively).

Table 1.a Number of cases of prostate cancer in Lithuania, by age group and calendar period. Age group

40–49 50–59 60–69 70–79 80+ Total * **

1990–1994

1995–1999

2000–2004

2005–2009

Rise in case numbers**

Total

N*

%

N

%

N

%

N

%

N

%

N

%

15 178 697 781 566 2,237

0.7 8.0 31.2 34.9 25.3 100.0

22 267 1,097 1,404 654 3,444

0.6 7.8 31.9 40.8 19.0 100.0

40 518 2,140 3,137 1,108 6,943

0.6 7.5 30.8 45.2 16.0 100.0

164 2,467 5,882 5,479 1,302 15,294

1.1 16.1 38.5 35.8 8.5 100.0

241 3,430 9,816 10,801 3,630 27,918

0.9 12.3 35.2 38.7 13.0 100.0

149 2,289 5,185 4,698 736 13,057

993 1,286 744 602 130 584

N: number of cases. Numerical and percent rise between 1990–1994 and 2005–2009.

Table 1.b Number of cases of prostate cancer in Lithuania, by stage and calendar period. Stage

Unknown Of those known: Localised Locally advanced Regional Metastatic Total

1990–1994

1995–1999

2000–2004

2005–2009

Total

N

%

N

%

N

%

N

%

N

%

318

14.2

200

5.8

451

6.5

1,360

8.9

2,329

8.3

727 724 86 382 1,919

37.9 37.7 4.5 19.9

1,070 1,460 132 582 3,244

33.0 45.0 4.1 17.9

2,215 3,358 82 837 6,492

34.1 51.7 1.3 12.9

8,749 4,543 102 540 13,934

62.8 32.6 0.7 3.9

12,761 10,085 402 2,341 25,589

49.9 39.4 1.6 9.1

A. Gondos et al. / European Journal of Cancer 51 (2015) 1630–1637

1633

Table 2 Period estimates of 5–year relative survival of patients with prostate cancer in Lithuania, by age group and calendar period, and by stage and calendar period. 1995–1999

2000–2004

2005–2009

Change

P-value

PE

SE

PE

SE

PE

SE

Age group 40–49 50–59 60–69 70–79 80+ Total 

20.3 42.3 42.6 48.3 46.4 42.8

8.7 3.9 2.2 2.3 3.9 1.5

41.9 61.8 67.0 65.3 58.7 62.0

12.1 3.2 1.7 1.8 3.9 1.6

90.7 91.8 94.2 90.0 85.1 90.8

4.6 1.4 1.0 1.1 3.3 0.8

+70.4 +49.5 +51.6 +41.7 +38.7 +48.0

Cancer surveillance using registry data: Results and recommendations for the Lithuanian national prostate cancer early detection programme.

We describe long term trends in prostate cancer epidemiology in Lithuania, where a national prostate specific antigen (PSA) test based early detection...
406KB Sizes 0 Downloads 7 Views