Journal of Cancer Survivorship

, Volume 2, Issue 3, pp 128–137

Predicted cardiovascular mortality and reported cardiovascular morbidity in testicular cancer survivors

  • H. S. Haugnes
  • N. Aass
  • S. D. Fosså
  • O. Dahl
  • O. Klepp
  • E. A. Wist
  • T. Wilsgaard
  • R. M. Bremnes
Article

DOI: 10.1007/s11764-008-0054-1

Cite this article as:
Haugnes, H.S., Aass, N., Fosså, S.D. et al. J Cancer Surviv (2008) 2: 128. doi:10.1007/s11764-008-0054-1

Abstract

Introduction

We examined if testicular cancer (TC) treatment is associated with any risk for cardiovascular morbidity or predicted mortality according to the SCORE model, in which a 10-year future risk of ≥5% for developing a fatal cardiovascular event qualify for high-risk status.

Methods

One thousand one hundred thirty-four TC survivors treated 1980–1994 participated in this study (1998–2002). Patients were categorised in four treatment groups: surgery (n = 225), radiotherapy (n = 445), and two chemotherapy groups: cumulative cisplatin dose ≤850 mg (n = 375) and >850 mg (cis>850, n = 89). Patients with cardiovascular disease, diabetes or SCORE ≥5% constituted a high-risk group, and those with SCORE >1% an intermediate/high risk group.

Results

Age-adjusted mean SCORE was 0.93% for the surgery group. In comparison, chemotherapy treated patients had significantly higher SCORE (1.07%, p = 0.01). Only 15% of patients were scored to be at high-risk, while 53% qualified for the intermediate/high risk group. Patients in the cis>850 group had increased odds for having intermediate/high risk, compared with the surgery group (OR 3.4, 95% CI 1.3–8.7). Only 23 cardiovascular events had occurred since the testicular cancer diagnosis.

Conclusion

The SCORE model indicates that patients treated with cisplatin-based chemotherapy have a significantly increased future risk of a fatal cardiovascular event.

Implications for cancer survivors

TC survivors should be followed regularly with respect to cardiovascular risk profile beyond the routine 10-year clinical follow-up.

Keywords

Testicular cancer Cisplatin Cardiovascular Mortality Morbidity SCORE 

Introduction

Germ-cell testicular cancer (TC) is the most common malignancy among young men, and the incidence is rising [1]. Due to more reliable radiologic staging, disease monitoring by tumor markers, the introduction of cisplatin-based chemotherapy in the late 1970s and a multimodal treatment approach [2], the outcome of TC has improved considerably during the last three decades. The current cure rate exceeds 95% [1] and TC patients have a near normal life expectancy once they achieve durable remission.

The growing number of TC survivors has led to an increased attention towards late serious side effects caused by the treatment. Several studies have reported an association between cisplatin-based therapy and an unfavorable cardiovascular (CV) risk profile and CV morbidity [3, 4, 5, 6, 7, 8, 9, 10, 11]. Moreover, radiotherapy has also been associated with CV morbidity and mortality [10, 11, 12, 13, 14, 15]. The interval between TC treatment and the occurrence of the first therapy-related cardiovascular disease (CVD) event appears to be decades rather than years [11]. Hence, the routine oncological 10-year follow-up period may be too short to detect late CV morbidity. A possible alternative is to calculate the risk for future CVD events using risk models.

The European guidelines on CVD prevention [16, 17] recommend the recently developed SCORE (systematic coronary risk evaluation) risk model to identify individuals at high risk for developing fatal CVD [18]. High-risk individuals are defined as persons with established CVD or diabetes or asymptomatic persons with a 10-year future risk of ≥5% for developing a fatal CVD event [16]. In this study, we apply the SCORE risk model to calculate the 10-year future risk for developing a fatal CVD event in asymptomatic long-term TC survivors.

The aim of our study is to (1) by using the SCORE model, evaluate the future risk for developing fatal CVD in a large cohort of Norwegian long-term TC survivors across the treatment modalities (surgery, radiotherapy or chemotherapy) and (2) examine the prevalence of CV morbidity in the same cohort.

Patients and methods

Patients

All Norwegian TC patients surviving ≥5 years after the diagnosis were invited to participate in a national multi-centre follow-up survey [5, 7, 19]. The patients had been treated in the period 1980–1994, and the survey was conducted during 1998–2002 at five university hospitals. The follow-up survey consisted of a 219-item mailed questionnaire and an outpatient clinical examination including laboratory tests.

Of 1,814 eligible patients, 1,463 (81%) signed the informed consent form and participated in the study. In total 129 patients aged above 60 years were excluded, as almost all males aged above 60 in a Norwegian study were classified as high-risk individuals with regard to CVD risk according to the European guidelines [20]. One patient has recently withdrawn from the database, and the remaining 1,134 patients constitute the present study sample (Fig. 1). The Ethical Review Board of the Southern Health region of Norway approved the study.
FIGURE 1

Number of patients forming the study population.

Treatment principles and treatment groups

The study patients were either treated within the Swenoteca collaboration [21, 22, 23] or according to EORTC/MRC protocols [24, 25, 26, 27, 28]. All patients were initially orchiectomized, and staging was performed according to the Royal Marsden Staging System [29].

Seminomas

Most patients with early stages (≤IIA) of seminomas were treated with infra-diaphragmatic radiotherapy. The dog-leg technique involving radiation to the para-aortic and ipsilateral iliac nodes was generally used, but 29 patients received radiation to the para-aortic area only, as this technique was introduced at one institution in 1989. Only five patients had received additional mediastinal irradiation. From early 1980s to mid 1990s the standard radiotherapy dose was gradually reduced from 36–40 to 25.2–30 Gy. The majority of patients with more advanced stages received cisplatin-based chemotherapy followed by retroperitoneal surgery or radiation in selected cases.

Non-seminomas

Patients with early stages (≤IIA) of non-seminomas were until around 1990 treated with primary retroperitoneal lymph node dissection (RPLND), followed by cisplatin-based chemotherapy if lymph node metastases were present. Later, the diagnostic RPLND was replaced by surveillance or adjuvant chemotherapy for clinical stage I patients. Patients with more advanced stages received cisplatin-based combination chemotherapy [22, 27], followed by RPLND and further chemotherapy in case of residual malignant cells.

Treatment groups

Based on these treatment principles, the TC survivors were categorized into four treatment groups according to initial and eventual relapse treatment: (1) surgery only, including RPLND; (2) radiotherapy only; (3) chemotherapy with a cumulative dose of cisplatin ≤850 mg (cis≤850); (4) chemotherapy with a cumulative dose of cisplatin>850 mg (cis>850).

The cutoff point for the two chemotherapy groups was set at 850 mg cisplatin to differentiate between (1) patients who received standard four courses or less and (2) those who received more than four courses or “higher dose” chemotherapy regimens due to serious prognosis, poor response, progression or relapse [19]. Most chemotherapy-treated patients (n = 442, 95%) received cisplatin-based chemotherapy, primarily in combination with etoposide and bleomycin (BEP) or vinblastine and bleomycin (CVB). Twenty-two patients (5%) who received carboplatin instead of cisplatin in research protocols [26, 27] were included in the cis≤850 group.

SCORE risk model

A new European risk scoring system for fatal CVD was presented in 2003, based on the SCORE project [18]. This risk model is derived from a pooled dataset of cohort studies from 12 European countries, including 205,178 persons with baseline examinations in the 1970s and 1980s. The SCORE risk model estimates the individual absolute 10-year probability (in percent) of having a coronary or non-coronary fatal CVD event, with separate equations for geographical low-risk and high-risk regions. Since Norway has been considered a high-risk country with regard to CVD [17], we have applied the equation for high-risk regions.

The calculation of SCORE is based on gender, age, systolic blood pressure, total cholesterol and current smoking status, and is valid only for persons without previous CVD. The risk function in SCORE was calculated using a Weibull proportional hazards model [18]. SCORE is embedded in the current versions of the European guidelines on prevention of cardiovascular disease [16, 17]. According to these guidelines, high-risk individuals are asymptomatic persons with SCORE ≥5%, while persons with SCORE ≤1% are considered low-risk individuals [16, 30]. Furthermore, persons with established CVD or diabetes are also considered to be high-risk individuals.

Assessments

Patient-reported events of angina, myocardial infarction, stroke and other atherosclerotic arterial diseases were classified as CVD. Information regarding the experience of CVD events and diabetes, smoking status and antihypertensive, antidiabetic and/or lipid lowering medication was obtained from the questionnaire. Patient-reported diagnoses of stroke and angina were verified using the medical records at oncological units. With regard to other diagnoses (myocardial infarction, claudication, retinal embolus), the patients had generally given detailed information in the questionnaire. Respondents reporting a diagnosis of diabetes and/or use of antidiabetic medication were classified as having diabetes. Smoking habits were classified according to prevalent cigarette smoking (yes/no).

Resting blood pressure was measured manually or with an automatic device. Body mass index (BMI) was calculated as weight in kilograms divided by the square of height in meters (kg/m2). Blood samples were drawn non-fasting by venipuncture at each hospital laboratory between 0800 and 1200. Total serum cholesterol was measured enzymatically, while levels of serum total testosterone were determined using a commercial immunoassay, with similar reference ranges at each hospital.

Statistical analysis

Mean doses of cytotoxic drugs in the two chemotherapy groups were compared using Students t test. Continuous variables were analyzed using multiple linear regression, while dichotomous variables were analyzed using multiple logistic regression. Only CVD events occurring after the TC diagnosis were included in the analysis of CVD. All regression analyses were adjusted for age. When comparing the impact of the treatment modalities, the surgery group was used as reference. Additional adjustments for antihypertensive treatment, lipid-lowering medication, BMI and total testosterone were performed without influencing the results. All p values are two-tailed with statistical significance set at p < 0.05. The data were analyzed using SPSS 14.0 (SPSS Inc., Chicago, IL).

SCORE was calculated for all TC survivors without CVD prior to the follow-up. As the SCORE values in our study population had a skewed distribution, they were transformed to approximate a normal distribution. SAS software (SAS Institute, Cary, NC) was used to calculate the most optimal transformation (transformed SCORE = SCORE0.1). Predicted transformed SCORE values were back transformed and reported accordingly. SCORE values were also categorized and analyzed using ordinal logit regression, with the SCORE values, divided into quartiles, as the dependent variable. However, the proportional odds assumption was not met, as the test of parallel lines was significant (p < 0.001). We instead present the results of logistic regression analyses with dichotomization of the quartiles as follows: The three highest versus the lowest quartile, the two highest versus the two lowest quartiles, and the highest versus the three lowest quartiles.

Since only a small minority of the TC survivors belonged to the high-risk group, we performed additional analyses defining an intermediate/high risk group with individuals displaying SCORE >1% [16, 30]. Persons with established CVD and/or diabetes were also included in this group, as they already have an increased future fatal CVD risk [17].

Results

Study patients

There were no differences with respect to baseline clinical characteristics between the responders and the non-responders [5]. Table 1 presents clinical characteristics of the study patients, according to treatment group. The median observation time for all patients was 11.1 years. Compared with the surgery group, the radiotherapy group was older at diagnosis and follow-up (p < 0.001, both), while the cis>850 group was younger at diagnosis (p = 0.007) and follow-up (p < 0.001) and had a shorter observation time (p = 0.001). Patients in the cis>850 group received more etoposide (p < 0.001) and bleomycin (p = 0.02) than the cis≤850 group. The mean vinblastine dose (p = 0.59) did not differ between the groups. Only five patients had received additional mediastinal irradiation.
Table 1

Characteristics of 1,134 study patients

Characteristic

Surgery

Radiotherapy

Cis≤850 mg

Cis>850 mg

All patients

N = 225

N = 445

N = 374

N = 90

N = 1134

Age, years, median (range)

 At diagnosis

29 (16–53)

34 (18–51)

29 (15–52)

26 (15–48)

31 (15–53)

 At follow-up

41 (24–60)

45 (28–60)

42 (23–60)

36 (25–59)

43 (23–60)

Follow-up, years, median (range)

11.8 (5–21)

10.8 (5–21)

11.8 (5–22)a

9.4 (5–20)

11.1 (5–22)a

Initial RMH stageb

 I

220 (98)

421 (95)

132 (35)

11 (12)

784 (69)

 IM/II

5 (2)

24 (5)

178 (48)

25 (28)

232 (21)

 III

  

15 (4)

9 (10)

24 (2)

 IV

  

49 (13)

45 (50)

94 (8)

Histology

 Non-seminoma

219 (97)

2 (0.5)

305 (82)

80 (89)

606 (53)

 Seminoma

6 (3)

443 (99.5)

69 (18)

10 (11)

528 (47)

Chemotherapy doses, median (range)c

 Cisplatin (mg)

  

740 (185–850)

1180 (855–3095)

 

 Bleomycin (mg)

  

300 (30–390)

300 (90–540)

 

 Etoposide (mg)

  

3005 (300–8550)

3858 (67–10580)

 

 Vinblastine (mg)

  

72 (18–108)

64 (19–100)

 

Additional treatment in chemotherapy groups

 RPLND

  

233 (62)

71 (80)

 

 Radiotherapy

  

37 (10)

10 (11)

 

Blood pressure, mean (SD)

 Systolic blood pressure

129 (17)

133 (17)

133 (20)

132 (17)

132 (18)

 Diastolic blood pressure

81 (11)

84 (11)

83 (12)

84 (12)

83 (12)

Total cholesterol (mmol/l) mean (SD)

5.70 (1.13)

5.71 (1.05)

5.64 (1.16)

5.75 (1.11)

5.69 (1.11)

Body mass index (kg/m2) mean (SD)

26.6 (3.5)

26.4 (3.5)

26.4 (4.4)

27.2 (4.6)

26.5 (3.9)

Total testosterone (nmol/l) mean (SD)

16.2 (4.9)

15.5 (5.3)

15.6 (5.8)

14.7 (5.7)

15.6 (5.4)

Antihypertensive treatment

10 (4.5)

39 (8.8)

24 (6.4)

8 (9.0)

81 (7.2)

Lipid-lowering treatment

8 (3.6)

10 (2.2)

4 (1.1)

2 (2.2)

24 (2.1)

Cigarette smokers

65 (30)

157 (36)

145 (39)

30 (34)

397 (36)

Prevalence of CVD eventsd

2 (0.9)

21 (4.9)

7 (2.0)

2 (2.4)

32 (2.9)

Prevalence of diabetes

4 (1.8)

13 (2.9)

11 (2.9)

2 (2.2)

30 (2.6)

High-risk groupe

29 (13)

86 (20)

46 (13)

6 (7)

167 (15)

Intermediate/high risk groupf

95 (43)

274 (65)

174 (50)

28 (34)

571 (53)

There are some missing data for the following variables: blood pressure (n = 12), total cholesterol (n = 13), body mass index (n = 9), total testosterone (n = 6), antihypertensive treatment (n = 6), smoking status (n = 22), CVD events (n = 45), high-risk group (n = 66) and intermediate risk group (n = 56). Data are numbers (percentage) unless otherwise specified.

aOne patient had an observation time <5 years (4.3 years).

bRoyal Marsden Staging System [29]

cChemotherapy doses are listed for those who received the actual chemotherapy agent.

dCardiovascular disease (CVD) events include myocardial infarction, angina, stroke and other atherosclerotic disease.

eSystematic coronary risk evaluation (SCORE) ≥5% and/or CVD event and/or diabetes

fSCORE >1% and/or CVD event and/or diabetes

SCORE

Mean SCORE percentages for treatment groups, according to age, are presented in Table 2. Age-adjusted means of SCORE according to treatment groups and the result of linear regression are presented in Table 3. Only the cis≤850 group was significantly different from the surgery group (Δ SCORE 0.12%, p = 0.02), while the cis>850 group had the highest age-adjusted mean SCORE (Δ SCORE 0.15%, p = 0.07). All chemotherapy treated patients had significantly higher age-adjusted mean SCORE (Δ SCORE 0.14%, p = 0.01), compared with the surgery group. The SCORE percentages for patients treated with combined radiotherapy and chemotherapy did not differ from those treated with chemotherapy alone.
Table 2

SCORE percentages in study patients (according to treatment group), categorized by age groups

 

Surgery N = 218

Radiotherapy N = 408

Cisplatin≤850 N = 350

Cisplatin>850 N = 84

All patients N = 1060

Age <30 years, n (%)

10 (5)

5 (1)

27 (8)

19 (22)

61 (6)

 Age (years) median

28

29

28

28

28

 Score, mean (SD)

0.05 (0.03)

0.09 (0.02)

0.07 (0.05)

0.08 (0.07)

0.07 (0.05)

Age 30–39 years, n (%)

91 (42)

105 (26)

113 (32)

41 (49)

350 (33)

 Age (years) median

36

38

36

35

36

 Score, mean (SD)

0.36 (0.35)

0.49 (0.37)

0.44 (0.36)

0.44 (0.44)

0.43 (0.37)

Age 40–49 years, n (%)

73 (33)

186 (46)

154 (44)

18 (22)

431 (41)

 Age (years) median

44

45

44

45

44

 Score, mean (SD)

1.7 (1.5)

1.9 (1.6)

1.7 (1.3)

1.9 (1.1)

1.8 (1.4)

Age 50–59 years, n (%)

44 (20)

112 (27)

56 (16)

6 (7)

218 (20)

 Age (years) median

54

53

53

52

53

 Score, mean (SD)

5.5 (3.4)

5.4 (3.9)

6.4 (5.7)

4.6 (3.4)

5.7 (4.3)

Thirty-two persons with cardiovascular disease are excluded from the analysis. Missing data for 42 patients. Age is categorized according to age at follow-up.

Table 3

Linear regression and age-adjusted means of SCORE, with 95% CI

Treatment group

Age-adjusted mean (%; 95% CI)

Δ SCORE (%)

P valuea

Surgery

0.93 (0.88–1.02)

Reference

Radiotherapy

1.00 (0.95–1.07)

0.07

0.20

Cisplatin≤850

1.05 (1.00–1.12)

0.12

0.02

Cisplatin>850

1.08 (0.95–1.21)

0.15

0.07

Age-adjusted means with 95% CI of SCORE and Δ SCORE (difference versus surgery group) are evaluated at age 42.7 years (mean age for all patients).

aLinear regression with the transformed SCORE as the dependent variable

In age-adjusted analyses, the transformed SCORE was positively associated with cumulative cisplatin (p = 0.02), bleomycin (p = 0.01) and etoposide doses (p = 0.04). Cumulative vinblastine was not associated with the transformed SCORE (p = 0.20).

Table 4 gives the results of logistic regression analyses with the dichotomized quartiles of SCORE. Compared with the surgery group, the other treatment groups had significantly increased odds for belonging to the three highest quartiles of SCORE. Only the cis>850 group had increased odds for belonging to the two highest quartiles of SCORE (OR 3.4, 95% CI 1.2–9.9).
Table 4

Logistic regression with odds ratios (OR) for being in higher SCORE categories

Treatment group

1st versus 2nd, 3rd and 4th quartile of SCORE

1st and 2nd versus 3rd and 4th quartile of SCORE

1st, 2nd and 3rd versus 4th quartile of SCORE

OR

95% CI

OR

95% CI

OR

95% CI

Surgery

1.00

Reference

1.00

Reference

1.00

Reference

Radiotherapy

2.33

1.18–4.61

1.38

0.75–2.53

0.97

0.53–1.81

Cisplatin≤850

3.06

1.53–6.13

1.37

0.74–2.52

0.90

0.46–1.76

Cisplatin>850

2.69

1.04–6.93

3.41

1.18–9.85

1.52

0.49–4.76

The SCORE variable is divided into quartiles. All analyses are age-adjusted.

High-risk and intermediate/high risk groups

In total, 167 (15%) patients belonged to the high-risk group, with the highest percentage in the radiotherapy group (Table 1). Age-adjusted odds for belonging to the high-risk group did not differ across the treatment groups when using surgery group as reference (data not shown).

Overall, 571 (53%) patients belonged to the intermediate/high risk group (Table 1). The cis>850 group had significantly higher age-adjusted odds for belonging to this group compared with the surgery group (OR 3.4, 95% CI 1.3–8.7; Fig. 2). Figure 3 presents the percentages of study patients (according to treatment group) with intermediate/high risk within 10-year age groups. The percentage classified as intermediate/high risk increased substantially with age, and in the 50–59 years group, all patients were classified as intermediate/high risk. In the 40–49 years group, a significantly larger fraction of the cis>850 group were classified as intermediate/high risk compared with the surgery group (OR 6.6, 95% CI 1.3–35.5).
FIGURE 2

Age-adjusted odds ratios (OR) for belonging to the intermediate/high risk group (SCORE>1% and/or cardiovascular disease and/or diabetes) in different treatment groups using the surgery group as reference. Bars indicate 95% confidence intervals for OR.

FIGURE 3

Percentage of individuals classified as intermediate/high risk (SCORE >1% and/or cardiovascular disease and/or diabetes), according to treatment group and age at follow-up within 10-year age intervals.

Cumulative cisplatin dose (p = 0.018) and etoposide dose (p = 0.006) was positively associated with the intermediate/high risk group. Cumulative vinblastine dose (p = 0.75) and bleomycin dose (p = 0.13) was not associated with the intermediate/high risk group.

Cardiovascular morbidity

In total, 32 (3.0%) of the study patients reported CVD events. In nine of these, the CVD events occurred prior to the TC diagnosis, and the majority was treated with radiotherapy and had two or more CVD risk factors (hypertension, hypercholesterolemia, obesity, diabetes and smoking) at follow-up (data not shown).

Of interest are the remaining 23 patients who experienced a CVD event after the cancer diagnosis (Table 5). The median age at the event was 47 years (range 29–55), and the event occurred median 7 years (range 1–14) after the TC diagnosis. The majority of those patients had two or more CV risk factors at follow-up. Only one of 47 patients treated with both chemotherapy and radiotherapy reported a CVD event after treatment for TC (Table 5, patient no. 18). All irradiated patients who experienced CVD had received a dog-leg field [below the diaphragm, median dose 36 Gy (range 25–40)]. For the six chemotherapy-treated patients with CVD, the median cisplatin dose was 780 mg (range 410–1,300). All groups administered cytotoxic therapy seemed to have an increased risk of CVD compared with the surgery group (radiotherapy: OR 6.5, 95% CI 0.8–49.5; cis≤850: OR 3.2, 95% CI 0.4–28.2; cis>850: OR 4.1, 95% CI 0.3–67.8), but due to the limited number of events the differences were not statistically significant.
Table 5

Cardiovascular disease (CVD) events occurring after the testicular cancer (TC) diagnosis

Patient

Age at TC diagnosis, years

Treatment

Radiation/cisplatindose

Cardiovascular disease

Age at event, years

Years since diagnosis

Cardiovascular risk factors

1

39

RPLNDa

 

MI

40

1

Hypertension, diabetes, hypercholesterolemia

2

35

Radiotherapyb

25.2 Gy

MI

40

5

Hypertension, obesity, hypercholesterolemia

3

38

Radiotherapy

27.0 Gy

MI

42

4

Hypertension, obesity, hypercholesterolemia

4

43

Radiotherapy

30.0 Gy

MI

49

6

Hypercholesterolemia

5

36

Radiotherapy

36.0 Gy

MI

48

12

Hypertension, hypercholesterolemia

6

41

Radiotherapy

36.0 Gy

MI

55

14

Hypertension, hypercholesterolemia, smoker

7

44

Radiotherapy

36.0 Gy

MI

51

7

Hypercholesterolemia, smoker

8

41

Radiotherapy

38.0 Gy

MI

47

6

Hypertension, obesity, hypercholesterolemia

9

41

Radiotherapy

39.6 Gy

MI

47

6

Hypertension, hypercholesterolemia

10

30

Radiotherapy

40.0 Gy

MI

43

13

Hypertension

11

35

Radiotherapy

40.0 Gy

MI

44

9

Hypertension, hypercholesterolemia

12

30

Chemotherapyc

800 mg

MI

30

<1 year

Hypertension, obesity, smoker

13

46

Chemotherapy

1,300 mg

MI

50

4

Hypertension, hypercholesterolemia, smoker

14

42

Radiotherapy

30.0 Gy

Angina, by-pass

50

8

Obesity, hypercholesterolemia, smoker

15

44

Radiotherapy

30.0 Gy

Angina, by-pass

51

7

Hypertension

16

41

Chemotherapy

410 mg

Angina, by-pass

47

6

Hypertension, hypercholesterolemia, smoker

17

36

Chemotherapy

600 mg

Angina

49

13

Hypertension

18

29

Radiotherapy/ chemotherapy

40.0 Gy/760 mg

Angina

41

12

Hypercholesterolemia, smoker

19

27

Radiotherapy

36.0 Gy

Stroke

29

2

None

20

31

Radiotherapy

36.0 Gy

Stroke

44

13

Hypercholesterolemia, smoker

21

40

Chemotherapy

800 mg

Stroke

47

7

Hypertension, obesity, hypercholesterolemia

22

45

Radiotherapy

27.0 Gy

Retinal embolus

48

4

Hypertension, hypercholesterolemia

23

24

Radiotherapy

36.0 Gy

Claudication

38

14

Hypercholesterolemia, smoker

CVD events include myocardial infarction (MI), angina, stroke and other atherosclerotic arterial disease.

aRetroperitoneal lymph node dissection

bAll irradiated patients were treated by the dog-leg technique, none received mediastinal irradiation.

cAmong the chemotherapy treated patients, three patients received the CVB regimen, two the BEP regimen and one was treated with a combination of ifosfamide, vincristine and cisplatin.

Discussion

In this study we analyzed CV morbidity and assessed the future risk for a fatal CVD event in long-term TC survivors. By using the SCORE model, patients treated with cisplatin-based chemotherapy appeared to have an increased risk for a fatal CVD event in the next 10 years, compared with surgery treated TC patients.

The major strength of this study is the large patient population, which enables us to make comparisons across treatment groups to study the impact of the specific treatments. The relatively short follow-up period (median 11 years) limits the number of CVD events in our relatively young study population. Thus, associations between TC treatment and actual CVD events can not be assessed. It may, however, be useful to apply the SCORE model as a surrogate endpoint for CVD.

To our knowledge, this is the first study that applies a CVD risk score in cancer survivors. As the CVD prevalence in young TC survivors is relatively low, calculating the future CVD risk may be a valuable method to detect a possible association between TC treatment and CVD morbidity. The most frequently applied risk models have been derived from the American Framingham Heart study, based on data from 5,573 white Americans aged 30–74 years with baseline examinations in the period 1968–1975 [31, 32]. Multifactorial risk models which estimate the risks of having different CVD events over the next ten years have been produced, including age, gender, blood pressure, total cholesterol, high-density lipoprotein cholesterol, smoking, glucose intolerance and left ventricular hypertrophy. Most commonly used is the “classical” model, estimating the risk of having a fatal or non-fatal coronary heart disease event in the next 10 years [32]. While the classical Framingham risk model defines high risk as a 10-year risk ≥20% of having a fatal or non-fatal coronary heart disease event, the SCORE model uses a 10-year risk ≥5% of having a fatal CVD event as cutoff for high risk.

The Framingham risk models seem to overestimate CVD risk in European countries [33, 34]. Thus, the SCORE project developed a European risk model based on large cohort studies [18], aiming to give more well-adjusted risk estimations of fatal CVD in European populations. Using the SCORE model, a major limitation may be the inclusion of fatal endpoints only, as fatal events represent only part of the CVD incidents [35]. Since obesity is an important risk factor for CVD [36], the absence of obesity in the model may be another limitation. Using the SCORE risk model in our relatively young patient population may have some particular weaknesses. SCORE was evaluated for 45–64 year old persons [18], and it is not validated for the youngest men in our study population. SCORE and other risk functions estimate the absolute risk, and may not intercept young persons at an increased relative risk, but at a low absolute risk [17]. Nevertheless, the major aim of this study was to compare and report possible differences between treatment groups, using the same risk model, irrespective of possible limitations.

A more aggressive treatment of CVD risk factors during the last decade [17] may potentially affect the application of risk models. This is, however, not a relevant issue in our patient population, since the study follow-up period was relatively short (1998–2002). Additionally, our TC survivors are still relatively young and have to a limited extent developed hypertension or hypercholesterolemia.

Only 32 (3.0%) of our TC survivors had ever experienced a CVD event. This rate is lower than the 5% rate found by Huddart et al. [10] among 992 TC survivors at a median follow-up of 10.2 years (excluding CVD mortality). A recently published Dutch study [11] on 2339 TC survivors reported a 20-year actuarial risk of 18.1% for any CVD event, both fatal and non-fatal. The higher CVD risk rate in this study may at least in part be explained by their longer follow-up period (median 18.4 years). The Dutch group [11] included patients treated in the 1960s and 1970s, when CVD incidence rates were significantly higher in most industrialized countries [37, 38]. During this period, TC treatment included higher radiotherapy doses and larger fields, more frequent mediastinal irradiation and other chemotherapy regimens, which all may have contributed to increased CVD incidence rates.

A few previous reports have shown an increased risk for CV morbidity after cisplatin-based chemotherapy in TC survivors [9, 10, 11]. Huddart et al. reported a 2.6-fold increased risk for CVD after chemotherapy compared with surveillance.[10] The larger Dutch study reported a non-significantly increased standardized incidence ratio for myocardial infarction at 1.46 after chemotherapy alone [11]. Their CVB and BEP regimens, respectively, were associated with 1.9 and 1.5-fold increased risk for CVD in comparison to surgery only. Due to a limited number of events, we could not identify any association between chemotherapy treatment and CV morbidity. In fact, only six of the chemotherapy treated patients experienced a CVD event during the follow-up. The lack of such an association may, at least in part, be related to a shorter follow-up time in our cohort, and the relatively low attained age.

Whereas mediastinal irradiation has been associated with increased CVD risk in TC patients [11, 12, 13, 14], there are conflicting data regarding the association between infradiaphragmatic irradiation and CVD risk. Zagars et al. [12] reported a 1.8-fold increased risk for cardiac death in seminoma patients treated with infradiaphragmatic radiotherapy alone with follow-up beyond 15 years. A British study [10] reported a 2.4-fold increased risk for CVD after radiotherapy alone compared with surveillance. The authors did not, however, specify whether the CVD events occurred in their 8% of patients who had received mediastinal irradiation. Finally, Fosså et al. [15] noted a significantly increased mortality from circulatory diseases among patients who were younger than 35 years at diagnosis and treated with radiotherapy only (including mediastinal irradiation), with a standardized mortality ratio of 1.70.

Hitherto, the two largest studies addressing CVD risk in patients treated with infradiaphragmatic irradiation only are the Dutch study [11] and the present one. Neither showed any association between radiotherapy and CV morbidity. Combined radiotherapy and chemotherapy treatment was associated with increased CVD risks in both the British [10] and the Dutch study [11]. In the present series, only one of the 47 patients treated with the combined modality experienced a CVD event.

In our study, patients treated with cisplatin-based chemotherapy had an increased risk for a 10-year fatal CVD event compared with the surgery group. They also had increased odds for being in the intermediate/high risk group, particularly after large cumulative cisplatin doses. We have recently reported a high prevalence of hypertension, obesity and metabolic syndrome in heavily cisplatin-treated patients [5, 7]. Cisplatin-based chemotherapy may lead to Leydig cell insufficiency [39]. Low endogenous testosterone is associated with increased levels of cardiovascular risk factors [40, 41, 42], and an increased risk of CVD mortality [43, 44]. However, cisplatin-based treatment was associated with a significant future CVD risk even after adjusting for testosterone, implying additional causative mechanisms. These may, among other unknown causes, be chemotherapy-induced endothelial dysfunction and a possible induction of atherosclerosis [45, 46].

In conclusion, this report provides evidence for an increased future CVD risk in TC survivors treated with cisplatin-based chemotherapy. The limited number of CVD events and the relatively short follow-up signals the great need for a longer follow-up study. Furthermore, our results suggest that TC survivors treated with chemotherapy should be followed regularly beyond the standard 10-year follow-up period. Risk reduction strategies with regard to CVD are important for these young men.

Acknowledgements

Supported in part by grants from Erna and Olav Aakres Legacy and the Norwegian Foundation for Health and Rehabilitation (grant no. 1998/27). Thanks to project secretary Vigdis Opperud for assistance with the database. The study is a national clinical study as part of the Norwegian Urological Cancer Group (NUCG) III project.

Copyright information

© Springer Science+Business Media, LLC 2008

Authors and Affiliations

  • H. S. Haugnes
    • 1
  • N. Aass
    • 2
  • S. D. Fosså
    • 3
    • 4
  • O. Dahl
    • 5
    • 6
  • O. Klepp
    • 7
  • E. A. Wist
    • 8
  • T. Wilsgaard
    • 9
  • R. M. Bremnes
    • 1
    • 10
  1. 1.Department of Oncology, Institute of Clinical MedicineUniversity of TromsøTromsøNorway
  2. 2.Department of OncologyRikshospitalet Medical CenterOsloNorway
  3. 3.Department of Clinical Cancer ResearchRikshospitalet Medical CenterOsloNorway
  4. 4.Medical FacultyUniversity of OsloOsloNorway
  5. 5.Section of Oncology, Institute of MedicineUniversity of BergenBergenNorway
  6. 6.Department of OncologyHaukeland University HospitalBergenNorway
  7. 7.Department of OncologySt. Olav University HospitalTrondheimNorway
  8. 8.Department of OncologyUllevål University HospitalOsloNorway
  9. 9.Institute of Community MedicineUniversity of TromsøTromsøNorway
  10. 10.Department of OncologyUniversity Hospital of North NorwayTromsøNorway

Personalised recommendations