Original Article

Prostate Cancer, Comorbidity, and the Risk of Venous Thromboembolism: A Cohort Study of 44,035 Danish Prostate Cancer Patients, 1995-2011 bet Horva th-Puho  , PhD1; Timothy L. Lash, DSc1,2; Vera Ehrenstein, DSc1; Michael Borre, PhD3; Anne G. Ording, PhD1; Erzse Mogens Vyberg, PhD4; and Henrik T. Sørensen, PhD1

BACKGROUND: Venous thromboembolism (VTE) is a serious complication of cancer. It is unknown whether comorbidity interacts clinically with prostate cancer (PC) to increase the VTE rate beyond that explained by PC and comorbidity alone, for example, by delaying diagnosis or precluding treatment. METHODS: A nationwide, registry-based cohort study of all 44,035 Danish patients diagnosed with PC from 1995 to 2011 and 213,810 men from the general population matched 5:1 on age, calendar time, and comorbidities. The authors calculated VTE rate ratios and the interaction contrast as a measure on the additive scale of the excess VTE rate explained by synergy between PC and comorbidity. RESULTS: In total, 849 patients in the PC cohort and 2360 men from the general population had VTE during 5 years of follow-up, and their risk of VTE was 2.2% and 1.3%, respectively. The 1-year VTE standardized rate among PC patients who had high comorbidity levels was 15 per 1000 person-years (PYs) (95% confidence interval, 6.8-24 per 1000 PYs), and 29% of that rate was explained by an interaction between PC and comorbidity. The VTE risk was increased among older patients, those with metastases, those with high Gleason scores, those in the D’Amico high-risk group, and those who underwent surgery. CONCLUSIONS: PC interacted clinically with high comorbidity levels and increased the VTE rate. Because of the large PC burden, reducing VTEs associated with comorbidities may have an impact on VTE risk and the potential to improve prognosis. Clinical interactions between high levels of comorbidity and PC on the risk of VTE were observed. Almost 30% of all episodes of VTE C 2015 American Cancer Society. occurred among patients who had high levels of comorbidity. Cancer 2015;121:3692-9. V KEYWORDS: comorbidity, epidemiology, hemostasis, morbidity, neoplasm grading, prostate cancer, pulmonary embolism, venous thrombosis.

INTRODUCTION Venous thromboembolism (VTE) occurs in about 1 in 1000 individuals, and the incidence of thrombosis and of multimorbidity—the presence of several health conditions—increases with age.1,2 Aging is also the most important predictor of prostate cancer (PC), which represents 23% of all primary cancers among men in Europe.3 Approximately 40% of patients with PC have other chronic conditions at diagnosis.4 PC management is complicated by other chronic conditions; and, as PC survival improves, many patients may die of other causes.4-7 VTE represents a major cause of complications in malignancy, worsening prognosis after cancer and other chronic diseases.8,9 Therefore, preventing complications of PC, its treatment, and comorbidities may improve survival and quality of life. VTE occurs among 1% to 2% of patients with PC, and, during the 10 years after PC diagnosis, their risk of VTE is up to 3-fold greater than that among men in the general population.10,11 The factors implicated in this increased risk include tumor-induced hypercoagulability; vascular injury caused by tumor, treatment, or surgery; and, among bedridden cancer patients, venous stasis because of immobilization.10 Yet it is unknown whether comorbid conditions interact with PC to increase the risk of VTE compared with that in men from the general population without PC but with diagnoses of similar comorbidities. Such interaction may arise if comorbidities delay diagnosis, preclude treatment, or worsen prognosis. Therefore, we examined whether PC interacts with comorbidities to affect the rate of VTE beyond that

Correction added on 9 November 2015, after first online publication: on page 3694 in the section on Results/Characteristics, the 3rd sentence has been changed to read “. . .only 15.5% had a CCI score >1.” Corresponding author: Anne G. Ording, PhD, Department of Clinical Epidemiology, Aarhus University Hospital, Olof Palmes All e 43-45, 8200 Aarhus N, Denmark; Fax: (011) 45 871-67-215; [email protected] 1 Department of Clinical Epidemiology, Aarhus University Hospital, Aarhus, Denmark; 2Department of Epidemiology, Rollins School of Public Health, Emory University, Atlanta, Georgia; 3Department of Urology, Aarhus University Hospital, Aarhus, Denmark; 4Institute of Pathology, Aalborg University Hospital, Aalborg, Denmark

See editorial on pages 3574-6 this issue. Additional supporting information may be found in the online version of this article. DOI: 10.1002/cncr.29535, Received: February 24, 2015; Revised: April 16, 2015; Accepted: April 22, 2015, Published online July 6, 2015 in Wiley Online Library (wileyonlinelibrary.com)

3692

Cancer

October 15, 2015

Prostate Cancer and Venous Thrombosis/Ording et al

expected from individual conditions acting alone. This information could contribute important knowledge for preventive and treatment strategies. MATERIALS AND METHODS Design and Setting

We conducted a cohort study of all patients with PC in Denmark between 1995 and 2011 and a comparison cohort of men from the general population. Denmark has a national tax-supported health care system. Access to health care is free of charge for the entire population.12 Starting in 1968, all Danish inhabitants have been assigned a unique 10-digit civil personal registration (CPR) number at birth or upon immigration. The CPR number is recorded in all registries, ensuring accurate linkage of individuals.13 We used the CPR number to establish a cohort of patients with PC registered in the Danish Cancer Registry (DCR) from 1995 through 2011. The DCR includes data on all cancers diagnosed in Denmark starting in 1943,14 coded according to the International Classification of Diseases, 10th revision (ICD-10).14 PC stage was categorized as localized, regional, distant, and unknown. Among PC patients, the positive predictive value (PPV) and sensitivity of VTE has been shown to be 86% and 98%, respectively, in the Danish National Patient Registry (DNPR).15 The general population (GP) cohort was established based on the Civil Registration System (CRS), which holds information on the civil and vital status of all Danish inhabitants.13 We matched up to 5 men from the general population without a PC diagnosis to each patient in the PC cohort based on age and specific comorbidities. The index date was defined as the date of PC diagnosis and the corresponding date of matching for the GP cohort. Men from the GP cohort who had incident PC diagnosed after study inclusion were censored in that cohort and were included in the PC cohort. All men in the PC cohort (n 5 1305) and in the GP comparison cohort (n 5 5165) who had a previous diagnosis of VTE were excluded from the study. Comorbid Conditions and Treatment

All available ICD, 8th Revision (ICD-8) and ICD-10 discharge diagnoses of conditions included in the Charlson Comorbidity Index (CCI) were obtained from the DNPR.16,17 We also used this registry to collect information on prostatectomy and pharmacologic and surgical androgen-deprivation therapy (ADT) (Supporting Table 1). The DNPR includes all nonpsychiatric discharge diagnoses from inpatient hospitalizations since 1977 and all Cancer

October 15, 2015

outpatient and specialist visits since 1995.16 ICD codes are recorded using ICD-8 up to 1994 and using ICD-10 thereafter. We excluded PC and nonmelanoma skin cancers from the CCI. Comorbidities were included separately and were summarized by CCI scores (0, 1, 2-3, or 4). It has been demonstrated that the PPVs for comorbidities are consistently high, ranging from 82% to 100%.18 The PPVs for pharmacologic ADT and surgical ADT are 93% and 100%, respectively.19 Follow-Up and VTE

The study outcome was the time to VTE, defined as the first inpatient or outpatient discharge diagnosis of pulmonary embolism (PE) or deep venous thrombosis (DVT) recorded in the DNPR after the diagnosis/index date. We did not include VTE diagnoses that were recorded solely at emergency departments, because these are working diagnoses and are not confirmed.15 Patients who had first occurrences of both PE and DVT recorded on the same date were classified as PE patients. We defined postsurgery VTE as VTE associated with a record of any surgery in the DNPR within 3 months before VTE diagnosis. We used the DNPR and the CRS to follow the PC cohort and the GP cohort from the index date to the first occurrence of VTE, death, emigration, 5 years of follow-up, or December 31, 2012. We stratified the follow-up period by the first year and by years 2 through 5 after the index date. Definition of Analytic Variables

We characterized men from the 2 cohorts within age categories (ages 0-69, 70-74, 75-79, 80-84, and 85 years), index year (1995-1999, 2000-2004, and 2005-2011), baseline comorbidity level (CCI score: 0, none; 1, low; 23, moderate; and high, 4), and individual comorbidities; and men in the PC cohort were characterized according cancer stage at PC diagnosis and treatment with prostatectomy, pharmacologic ADT, and bilateral orchiectomy. For treatment data, we restricted the analysis to patients who were diagnosed from 1996, when the Nordic Medico-Statistical Committee (NOMESCO) classification of surgical procedures was introduced in Denmark. Data on TNM classification have been included in the DCR since 2004.14 By linkage to the National Pathology Registry, which has recorded all histologic examinations in Denmark since 1997,20 we also included data on Gleason score for patients who were diagnosed in 2004 and later. Gleason scores were categorized as from 2 to 6, 7, and from 8 to 10. By using laboratory data for patients in the Central Denmark and North Denmark Regions, we 3693

Original Article

retrieved information on prostate-specific antigen measurements obtained within 90 days before or after the date of PC diagnosis and closest to the date of diagnosis to generate the D’Amico score, which was categorized as low risk, intermediate risk, and high risk.21 The initial PC cohort consisted of 44,341 patients. We were unable to identify matches in the general population for 306 men (0.7%). The unmatched men in the PC cohort were older (40% of unmatched men vs 20% of matched men were aged 80 years and older) and had a higher prevalence of CCI scores 3 than the matched men in the PC cohort (98% vs 5.7%, respectively). Statistical Analysis

Cumulative incidence estimates and curves were computed to estimate the risk of VTE in the PC and GP cohorts according to baseline characteristics, considering death as a competing risk. The matching was dissolved in adjusted analyses and when stratifying by follow-up period. In those analyses, we estimated age-standardized incidence rates of VTE using age weights from the PC cohort as of the index date. These standardized rates allowed for a comparison of VTE rates in various populations with the PC population at the index date. We then created Cox regression models to compute hazard ratios as a measure of the VTE rate ratio based on the standardized rates. We adjusted for the CCI score in overall analyses comparing men in the PC cohort with men in the GP cohort, and we also adjusted for age (continuous) and index year categories in analyses within strata of comorbidity levels. We stratified analyses on PE and DVT, VTE provoked and not provoked by surgery, nonmetastatic versus metastatic PC, prostatectomy, and pharmacologic and surgical ADT. For analyses stratified by treatment, we started follow-up on the date of the first treatment course, excluding all patients who were censored between the date of PC diagnosis and treatment. In analyses restricted to PC patients, we evaluated the effect on the VTE rate of increasing Gleason score and increasing D’Amico risk group. The interaction contrast (IC) was calculated to examine the impact of interaction between PC and comorbidity on the VTE rate. The IC is a measure of the synergistic or antagonistic effect between 2 factors that cannot be explained by the additivity of their individual effects22: IC 5 VTE ratePC & CCI  1 2 (VTE rateGP & CCI  1 2 VTE rateGP & CCI 5 0) 2 (VTE ratePC & CCI 5 0 2 VTE rateGP & CCI 5 0) 2 VTE rateGP & CCI 5 0, where VTEratePC & CCI  1 is the total VTE rate among patients with PC and comorbidity, VTE rateGP & CCI  1 2 VTE rateGP & CCI 5 0 is the effect of comorbidity among men 3694

in the GP cohort, VTE ratePC & CCI 5 0 2 VTE rateGP & CCI 5 0 is the effect of PC, and VTE rateGP & CCI 5 0 is the background rate among men from the GP cohort without PC or comorbidity. The IC is a measure of the overall VTE rate and the expected VTE rate given the baseline VTE rate, the effect of PC on the VTE rate, and the effect of comorbidity on the VTE rate. If the rate among patients who have PC with comorbidity equals the sum of the baseline VTE rate, the VTE rate in patients with PC but no comorbidity, and the VTE rate of men without PC but with comorbidity, then there will be no evidence of a clinical interaction between PC and comorbidity. When the VTE rate among patients who have PC with comorbidity is different from the sum of the baseline VTE rate, the VTE rate in patients who have PC but no comorbidity, and the VTE rate of men without PC but with comorbidity, then PC and comorbidity interact synergistically or antagonistically to increase or reduce the VTE rate compared with the sum of the individual effects. All analyses were conducted using SAS version 9.2 (SAS Institute Inc, Cary, NC). The study was approved by the Danish Data Protection Agency (record 2011-416174). In Denmark, no further permissions are needed to conduct observational studies that do not involve participant contact. RESULTS Characteristics

This study included 44,035 patients who were diagnosed with PC from 1995 to 2011 and 213,810 men from the general population (Table 1). Because of matching, the median age in both cohorts was 72 years (interquartile range [IQR], 65-78 years). The vast majority of men in the PC cohort had no comorbidities (67%) at diagnosis, and only 15.5% had a CCI score >1. The PC cohort was followed for a median of 3.2 years (IQR, 1.6-5.0 years), and the matched cohort was followed for a median of 4.5 years (IQR, 2.4-5.0 years). The difference in mean follow-up was because of the excess mortality hazard for men in the PC cohort. VTE 5 Years After Prostate Cancer Diagnosis

In total, 849 men in the PC cohort and 2360 men in the GP cohort were diagnosed with VTE within 5 years after the PC diagnosis/index date. The cumulative incidence of VTE was 2.2% (95% confidence interval [CI], 2.1%2.4%) in the PC cohort and 1.3% (95% CI, 1.26%1.37%) in the GP cohort. The VTE risks did not change Cancer

October 15, 2015

Prostate Cancer and Venous Thrombosis/Ording et al

TABLE 1. Descriptive Characteristics of the Prostate Cancer Cohort and the General Population Comparison Cohort: Denmark, 1995 to 2011 PC Cohort (N 5 44,035) Characteristic Age, y 69 70–74 75–79 80–84 85 Index year 1995–1999 2000–2004 2005–2011 Individual Charlson Comorbidity diseases Myocardial infarction Congestive heart failure Peripheral vascular disease Cerebrovascular disease Dementia Chronic pulmonary disease Connective tissue disease Ulcer disease Mild liver disease Diabetes type I and II Hemiplegia Moderate to severe renal disease Diabetes with end-organ damage Any tumor except prostate cancer Leukemia Lymphoma Moderate to severe liver disease Metastatic solid tumor AIDS Charlson Comorbidity Index score 0 1 2–3 4 Prostate cancer stage Localized Regional Distant Unknown Gleason scorea 2–6 7 8–10 Missing D’Amico scorea Low risk Intermediate risk High risk Missing Prostatectomyb Pharmacologic androgen-deprivation therapyc Surgical androgen-deprivation therapyb

GP Cohort (N 5 213,810)

No.

%

No.

%

18,988 8720 7680 5364 3283

43 20 17 12 7.5

93,940 41,839 36,938 25,360 15,733

44 20 17 12 7.4

8041 11,092 24,902

18 25 57

39,164 53,896 120,750

18 25 57

2302 1916 1749 3477 281 2688 742 1399 193 1999 45 733 905 2101 71 166 36 276 8

5.2 4.4 4.0 7.9 0.6 6.1 1.7 3.2 0.4 4.5 0.1 1.7 2.1 4.8 0.2 0.4 0.1 0.6 0.0

10,756 8590 7934 16,415 1223 12,426 3346 6375 847 9177 176 3152 4054 9818 295 729 143 1189 26

5.0 4.0 3.7 7.7 0.6 5.8 1.6 3.0 0.4 4.3 0.1 1.5 1.9 4.6 0.1 0.3 0.1 0.6 0.0

29,694 7625 5625 1091

67 17 13 2.5

146,045 36,897 26,715 4153

68 17 13 1.9

18,527 1846 8105 15,557

42 4.2 18 35

NA NA NA NA

27 27 21 25

NA NA NA NA

38 4.4 56 1.8 35 35 13

NA NA NA NA NA NA NA

2303 2377 1826 2176 3286 379 4865 152 15,110 12,498 5659

AIDS indicates acquired immunodeficiency syndrome; GP cohort, general population cohort; NA, not available; PC cohort, prostate cancer cohort. a Analyses were restricted to patients who were diagnosed from 2004. b Analyses were restricted to patients who were diagnosed from 1996. c Analyses were restricted to patients who were diagnosed from 2000.

Cancer

October 15, 2015

3695

Original Article

action between PC and comorbidity was only evident among patients who had high comorbidity levels, in whom an interaction accounted for 22% of all incidents of VTE. VTE rate ratios decreased with increasing comorbidity levels from 2.1 (95% CI, 1.9-2.4) among patients without comorbidity to 1.5 (95% CI, 0.80-2.8) among patients who had high comorbidity levels compared with men from the general population. These findings are presented in Table 2. Stratified Analyses

Figure 1. The cumulative risk of venous thromboembolism (VTE) during 5 years of follow-up is illustrated for the current cohort of patients with prostate cancer and the cohort of men from the general population who were matched on age and comorbidities (Denmark, 1995-2011).

markedly within index-year categories. Cumulative incidence curves are presented in Figure 1. First Year of Follow-Up

During the first year of follow-up, the VTE risk was 0.7% (95% CI, 0.6%-0.8%) in the PC cohort and 0.27% (95% CI, 0.25%-0.29%) in the GP cohort. The overall VTE rate was almost 3-fold higher in the PC cohort compared with that in the GP cohort (adjusted VTE rate ratio, 2.8; 95% CI, 2.4-3.2). Among men who had high comorbidity levels (CCI score, 4), the VTE rate per 1000 personyears (PYs) was 15.2 (95% CI, 6.8-24 per 1000 PYs) in the PC cohort compared with 5.8 (95% CI, 3.2-8.3 per 1000 PYs) in the GP cohort, corresponding to a rate difference of 9.4 VTE incidents. Similarly, among men without comorbidity, the standardized rate difference between the PC and GP cohorts was 5.0 VTEs per 1000 PYs. The IC—the difference between these 2 rates (95% CI, 9.45.0)—was 4.4 VTE incidents per 1000 PYs, corresponding to 29% of all incidents of VTE among patients with high comorbidity levels attributable to an interaction between comorbidity and PC. A negative interaction was observed among the men who had low and moderate comorbidity levels. Subsequent Follow-Up

During years 2 through 5 of follow-up, men in the PC cohort had an almost 2-fold increased risk of VTE (adjusted VTE rate ratio, 1.9; 95% CI, 1.7-2.1). An inter3696

The rate of postsurgery VTE (ie, surgery within 3 months before VTE) was 2.2 per 1000 PYs (95% CI, 1.7-2.7 per 1000 PYs) among patients with PC and 0.1 per 1000 PYs (95% CI, 0.0-0.1 per 1000 PYs) among patients without PC in the first year after diagnosis, translating into a 40fold increased rate ratio of postsurgery VTE during the first year of follow-up and a 6-fold increased rate ratio during 2 to 5 years of follow-up relative to the GP cohort. However, men in the PC cohort underwent more surgery than men in the GP cohort. VTE rate ratios that were not provoked by surgery were more constant: they were about 2-fold higher in the PC cohort compared with the GP cohort. PC patients aged 70 years or older, those with metastatic PC, and those who underwent either prostatectomy or bilateral orchiectomy had a particularly marked interaction with high comorbidity levels in the first year after diagnosis. Stratified analyses are presented in Table 3, and the ICs associated with individual comorbidities in are listed in Supporting Figure 1 (see online supporting information). In analyses restricted to the PC cohort, the VTE rate increased with increasing Gleason score and increasing D’Amico risk group to a modest 1.34 (95% CI, 0.65-2.8) increased rate among patients who had Gleason scores from 8 to 10 in the first year after diagnosis compared with those who had scores from 2 to 6. Patients in the D’Amico high-risk group had a 1.6-fold increased rate (95% CI, 0.91-fold to 2.8-fold increased rate) of VTE compared with patients in the D’Amico low-risk group. DISCUSSION In this large cohort study with long-term follow-up of a population-based PC cohort and a matched GP cohort, interaction was observed among patients who had high comorbidity levels, accounting for almost 30% of all incidents of VTE among these patients during the 5 years after diagnosis. This finding is particularly interesting, because the majority of men in the PC cohort (67%) were free of comorbidities at diagnosis. However, older patients Cancer

October 15, 2015

Prostate Cancer and Venous Thrombosis/Ording et al

TABLE 2. Charlson Comorbidity Index Scores, Standardized Venous Thromboembolism Incidence Rates, Interaction Contrasts, and Rate Ratios of Venous Thromboembolism in the Prostate Cancer Cohort and the Matched General Population Comparison Cohort: Denmark, 1995 to 2011

CCI Score 0–1 Year of follow-up Overall PC cohort GP cohort 0 PC cohort GP cohort 1 PC cohort GP cohort 2–3 PC cohort GP cohort 4 PC cohort GP cohort 2–5 Years of follow-up Overall PC cohort GP cohort 0 PC cohort GP cohort 1 PC cohort GP cohort 2–3 PC cohort GP cohort 4 PC cohort GP cohort

Interaction Contrast (95% CI)

VTE Rate Ratio (95% CI)a

Proportion of the Rate Explained by Interaction, %

No. of Individuals

No. of VTEs

Person-Years of Follow-Up

Standardized Rate (95% CI)

44,035 213,810

313 577

40,643 207,642

7.7 (6.9–8.6) 2.8 (2.6–3.1)

NA

2.8 (2.4–3.2) Ref

29,694 146,045

202 314

28,061 143,527

7.3 (6.3–8.4) 2.3 (2.1–2.6)

Ref

3.3 (2.8–3.9) Ref

7625 36,897

48 109

6901 35,538

7.2 (5.1–9.3) 2.9 (2.3–3.4)

20.6 (23.1, 1.8)

2.3 (1.6–3.2) Ref



5625 26,715

49 132

4817 24,990

9.3 (6.6–12) 4.9 (4.1–5.8)

20.7 (23.7, 2.3)

1.9 (1.4–2.7) Ref



1091 4153

14 22

864 3587

15 (6.8–24) 5.8 (3.2–8.3)

4.4 (24.5, 13)

2.6 (1.4–5.2)

29

38,098 201,610

536 1,783

96,001 582,260

5.7 (5.2–6.2) 3.2 (3.0–3.3)

NA

1.9 (1.7–2.1) Ref

26,710 140,862

369 1,096

69,127 419,324

5.5 (5.0–6.1) 2.8 (2.6–3.0)

Ref

2.1 (1.9–2.4) Ref

6359 34,179

86 368

15,577 95,055

5.3 (4.2–6.5) 3.7 (3.3–8.2)

21.1 (22.4, 0.3)

1.5 (1.2–1.8) Ref



4286 23,407

67 274

9843 60,672

6.6 (5.0–8.2) 4.3 (3.8–4.8)

20.5 (22.3, 1.3)

1.5 (1.2–2.0) Ref



743 3162

14 45

1454 7209

11 (4.6–17) 5.8 (4.0–7.5)

2.4 (24.2, 9.1)

1.5 (0.8–2.8) Ref

22

NA

CCI indicates Charlson Comorbidity Index; CI, confidence interval; GP cohort, general population cohort; NA, not applicable; PC cohort, prostate cancer cohort; Ref, reference category; VTEs, venous thromboembolisms. a Overall analyses are adjusted for age (continuous), comorbidity levels, and index year categories. Comorbidity stratified analyses are adjusted for age (continuous) and index year categories.

who had severe comorbidity may have been slightly under represented in our study, because we excluded 306 PC patients who had no available GP match available. We observed consistently higher rates of VTE among patients with PC. This rate was increased almost 3-fold during the first year after PC diagnosis and doubled during the second to fifth years of follow-up. Our results are also in concordance with previous research11,23-25 suggesting that surgery procedures for PC are associated with an at least 3-fold increased risk of hospitalization for thromboembolic events and that ADT is associated with an at least 10% increased risk in various settings. Cancer of the prostate is a weaker risk factor for VTE compared with other cancer types,10 but patients with PC may be at higher risk of VTE because of metastatic cancer; multiple comorbidities and medications, Cancer

October 15, 2015

such as glucocorticoids or nonsteroidal anti-inflammatory drugs, affecting cancer treatment choices; and potentially through shared etiology with such conditions as heart diseases.8,26-28 Patients with PC also have a high rate of postsurgery VTE relative to the general population. This result accords with previous findings, likely because of the immobilization and hypercoagulability associated with PC and surgery, and it has been demonstrated that cancer is a risk factor for surgery-related VTE even in procedures for benign conditions.8,29 In particular, men with metastatic PC had a strong interaction with all levels of comorbidity on the VTE rate. Our results suggest that proper anticoagulation prophylaxis of patients with PC and comorbidity may reduce the number of patients with VTE attributable to interaction as well as to comorbidity. 3697

Original Article TABLE 3. Standardized Rates of Venous Thromboembolism in the Prostate Cancer Cohort and Interaction Contrasts Stratified by Postsurgery Venous Thromboembolism, Age Groups, Stage at Diagnosis, and Treatment: Denmark, 1995 to 2011 0–1 Year of Follow-Up

CCI Score Postsurgery VTE 1 2–3 4 Not surgery-related VTE 1 2–3 4 Group aged

Prostate cancer, comorbidity, and the risk of venous thromboembolism: A cohort study of 44,035 Danish prostate cancer patients, 1995-2011.

Venous thromboembolism (VTE) is a serious complication of cancer. It is unknown whether comorbidity interacts clinically with prostate cancer (PC) to ...
132KB Sizes 0 Downloads 10 Views