Cancer Causes & Control

, Volume 28, Issue 9, pp 971–979 | Cite as

Therapeutic radiation for lymphoma and risk of second primary malignant mesothelioma

  • Ellen T. Chang
  • Edmund C. Lau
  • Fionna S. Mowat
  • M. Jane Teta
Brief report

Abstract

Purpose

This large, population-based U.S. study of lymphoma patients followed for up to four decades enables detailed analysis of second primary mesothelioma risk after radiotherapy.

Methods

U.S. Surveillance, Epidemiology, and End Results data were used to identify second primary mesothelioma among patients diagnosed with Hodgkin lymphoma (HL) or non-Hodgkin lymphoma (NHL) between 1973 and 2014. Standardized incidence ratios (SIRs) were calculated by radiotherapy. Multivariate adjusted associations were examined using competing risks survival analysis.

Results

Among 47,219 HL patients (19,538 irradiated) and 252,090 NHL patients (52,454 irradiated), second primary mesothelioma developed among 28 lymphoma patients who received radiotherapy and 59 who did not. Mesothelioma risk was increased among HL and NHL patients treated with radiotherapy [SIR = 1.78, 95% confidence interval (CI) 1.18–2.58], but not without radiotherapy. After multivariate adjustment, radiotherapy was associated with increased mesothelioma risk (relative risk = 1.64, 95% CI 1.05–2.57), especially in lymphoma patients diagnosed before 1995 and after a latency of at least 10 years, and apparently with younger age at diagnosis.

Conclusions

The increase in second primary mesothelioma risk following radiotherapy for lymphoma is independent of several patient and disease characteristics, and is higher with earlier treatment era and longer latency.

Keywords

Hodgkin lymphoma Non-Hodgkin lymphoma Mesothelioma Radiotherapy Epidemiology SEER Program 

Introduction

Ionizing radiation, a known human carcinogen, has been used since the 1950s to treat a variety of cancers. Consequently, cancer radiotherapy is associated with the development of various second primary cancers. In Hodgkin lymphoma (HL) patients specifically, second primary cancers (along with radiation-induced cardiovascular toxicity) are a leading cause of death in long-term survivors [1]. Second primary cancers also are increased in non-Hodgkin lymphoma (NHL) patients following treatment with radiation (e.g., [2, 3]).

The development of second primary mesothelioma following radiotherapy for cancer has been evaluated in several publications, all of which reported a consistently increased risk of mesothelioma in irradiated cancer survivors, with relative risks ranging from 1.3 to 30, contrasted with a lack of increased risk among cancer survivors who did not undergo irradiation. These included studies of survivors of several primary cancers, including lymphoma [3, 4, 5, 6], breast cancer [7, 8, 9], testicular cancer [10], prostate cancer [11], and other solid tumors [12, 13]. In addition, the risk of cancer—including both first and second primary tumors—following radiotherapy is reported to be increased due to scatter dose to sites not in the field of radiation (e.g., [11, 13, 14, 15, 16, 17]). Only five prior studies have assessed the risk of mesothelioma after radiotherapy specifically for lymphomas, based on a total of 0 [18], 18 [3], 6 [5], 13 [4], and 26 mesothelioma cases [6], respectively. Four of the five studies, excluding the study with no cases and short follow-up [18], reported a positive association with radiotherapy. Other studies (e.g., [19]) have reported an increased risk of mesothelioma after HL in general, but analyses were not restricted to patients who underwent radiotherapy. Additional evidence supporting a causal role of ionizing radiation in mesothelioma development comes from epidemiologic studies documenting excesses of mesothelioma among patients treated with Thorotrast (e.g., [20, 21]).

Few published studies have evaluated whether risk of second primary mesothelioma after cancer radiotherapy varies by patient or disease characteristics, such as sex and latency. De Bruin et al. [4] noted a higher risk in women than men who underwent radiotherapy for HL, whereas Hodgson et al. [5] and Tward et al. [3] found no such sex-based difference in HL and NHL survivors, respectively. Studies of second primary mesothelioma after other cancers either were sex-specific or did not stratify by sex. Brown et al. [8] and Berrington de Gonzalez et al. [7] reported that second primary mesothelioma risk decreased with more recent year at diagnosis of primary breast cancer treated with radiation, whereas Travis et al. [10] and Hodgson et al. [5] found no differences by period of diagnosis after irradiation for testicular cancer and HL, respectively. All studies that stratified by age at diagnosis and latency found that risk of second primary mesothelioma was higher in survivors of various cancers treated at younger ages and those with longer latency [3, 4, 5, 7, 8, 10]. Farioli et al. [11, 13], who did not stratify their analyses by age at diagnosis, also reported higher risks of developing mesothelioma with longer latency after irradiation for solid cancers. Farioli et al. [13] additionally found no significant difference in second primary mesothelioma risk between scattered or direct radiotherapy for prostate cancer.

Given the very low baseline incidence rate of mesothelioma, this rare but deadly malignancy is best studied in large populations with long follow-up. Therefore, we have built upon our prior study of HL and NHL survivors in the U.S. National Cancer Institute Surveillance, Epidemiology, and End Results (SEER) program of cancer registries [6] by extending case ascertainment and follow-up by more than a decade, thereby enabling examination of the relationship between radiotherapy and mesothelioma in greater detail. In this study, by far the largest study to date of mesothelioma risk after radiotherapy for lymphoma, we took advantage of the relatively large number of cases to conduct a multivariate adjusted survival analysis of the independent impact of radiotherapy on second primary mesothelioma risk, accounting for the competing risk of death, and to evaluate differences in risk by sex, age, time period of diagnosis, latency, and nodal status.

Materials and methods

This study used the most recent release of cancer incidence data from the SEER 18 cancer registries in April 2017, including patients diagnosed between 1973 and 2014 (inclusive) and submitted to SEER through November 2016. All patients diagnosed with first, primary, incident HL (SEER cancer site codes 33011 and 33012) or NHL (SEER cancer site codes 33041 and 33042, not including chronic lymphocytic leukemia) between 1973 (nine registries), 1992 (four additional registries), or 2000 (five additional registries) and 2014 were identified from SEER. HL and NHL cases who were treated with non-beam radiotherapy or had unknown age, sex, race, follow-up duration, or treatment (4.1% of HL and 3.7% of NHL cases) were omitted.

Duration of follow-up in months, reported in the SEER data, was calculated from the time of the initial lymphoma diagnosis until the date of mesothelioma diagnosis (latency), the date of death, the date a patient was last known to be alive, or the follow-up cut-off date of 31 December 2014, whichever came first. We excluded patients with <2 months of follow-up (including those diagnosed with a subsequent malignancy during this period) and any patients (n = 1) who did not receive radiotherapy for lymphoma, but subsequently received radiotherapy for another cancer before later developing mesothelioma; otherwise, all patients with available data were included. Second primary mesothelioma cases diagnosed after lymphoma diagnosis were identified based on SEER cancer site code 36010. Standardized incidence ratios (SIRs) and Poisson 95% confidence intervals (CIs) were calculated as the number of observed mesothelioma cases divided by the expected number of cases calculated by indirect standardization, applying the number of person-years in each category of 5-year age, sex, race (white/unknown, black, or other), and 5-year calendar period to the corresponding mesothelioma incidence rate. Absolute excess risks (AERs), reported as number of excess cases per 10,000 person-years, were calculated as the number of observed second primary mesothelioma cases minus the number of expected cases, divided by the person-years at risk in the subgroup of interest. All analyses were stratified according to whether beam radiotherapy was used during the first course of treatment; information on subsequent treatment is not available in SEER. In secondary analyses, to ensure the opportunity for a latency period of at least 10 years and to enable comparison with our prior findings in SEER 13 (1973–2003) [6], we restricted HL and NHL patients in the SEER 18 registries to those diagnosed between 1973 and 2003, followed through 2014.

To estimate the independent effect of lymphoma radiotherapy on second primary mesothelioma risk, adjusting for patient and disease characteristics for which information was available in SEER, we used the Fine and Gray [22] adaptation of the Cox proportional hazards regression model to conduct survival analysis accounting for the competing risk of death from any cause after lymphoma diagnosis. Relative risks (RRs, i.e., subdistribution hazard ratios) based on the Fine and Gray [22] subdistribution hazard function accounting for competing risks, with 95% CIs, were adjusted for age at lymphoma diagnosis, sex, race, type of lymphoma, lymphoma nodal status, and calendar period of lymphoma diagnosis. We conducted secondary analyses stratified into two categories by age at diagnosis (<40 vs. ≥40 years, because susceptibility to the carcinogenic effects of ionizing radiation may vary by age [23]); period of diagnosis (before 1995 or 1995 and thereafter, because radiation volumes and doses were reduced with the advent of involved-field and involved-node radiotherapy in the mid-1990s [24]); latency since lymphoma diagnosis (<10 vs. ≥10 years, to evaluate the latency period between radiotherapy exposure, which was assumed to have been initiated shortly after diagnosis, and mesothelioma onset); and nodal status (because recommended radiotherapy regimens differ between extranodal and nodal lymphomas [25, 26, 27]). In all instances, cut-offs between strata were set somewhat arbitrarily based on round numbers, as well as the distribution among study subjects.

Analyses were performed using SAS version 9.4 (Cary, NC).

Results

We identified 47,219 cases of HL diagnosed between 1973 and 2014 in the SEER 18 registries. The average age at HL diagnosis was 38.5 years [standard deviation (SD): 18.9] and differed little between males (55% of cases; 38.8 ± 18.4 years) and females (45% of cases; 38.1 ± 19.5 years) (Table 1); 85% were white. Forty-one percent of HL patients received beam-radiation therapy. In general, HL patients who received radiotherapy were younger (average age 34.4 years) than those who did not (average age 41.3 years). The mean duration of follow-up after HL diagnosis for both sexes combined was 10.2 years (SD: 9.1); median follow-up was 7.8 years [interquartile range (IQR): 3.0–14.4].
Table 1

Characteristics of lymphoma patients in U.S. Surveillance, Epidemiology, and End Results 18 cancer registries, 1973–2014

Characteristic

Hodgkin lymphoma

Non-Hodgkin lymphoma

All lymphoma

Total patients

Mesothelioma cases

Total patients

Mesothelioma cases

Total patients

Mesothelioma cases

Total

47,219

14

252,090

73

299,309

87

Males

 No radiotherapy

15,666

4

108,230

41

123,896

45

 Radiotherapy

10,285

7

28,249

14

38,534

21

Females

 No radiotherapy

12,015

1

91,406

13

103,421

14

 Radiotherapy

9,253

2

24,205

5

33,458

7

Men and women

 No radiotherapy

27,681

5

199,636

54

227,317

59

 Radiotherapy

19,538

9

52,454

19

71,992

28

Year of lymphoma diagnosis

 Before 1995

13,840

7

53,989

16

67,829

23

 1995 or later

33,379

7

198,101

57

231,480

64

Nodal status

 Nodal

46,257

13

173,295

55

219,552

68

 Extranodal

962

1

78,795

18

79,757

19

Age at lymphoma diagnosis (years)

 Males

Mean: 38.8

SD: 18.4

Median: 35

IQR: 24–51

Mean: 38.2

SD: 20.7

Median: 39

IQR: 18–56

Mean: 59.2

SD: 17.3

Median: 61

IQR: 49–72

Mean: 67.9

SD: 11.2

Median: 69

IQR: 61–75

Mean: 55.9

SD: 19.0

Median: 58

IQR: 43–70

Mean: 63.0

SD: 17.2

Median: 66.5

IQR: 58–75

 Females

Mean: 38.1

SD: 19.5

Median: 32

IQR: 23–51

Mean: 25.0

SD: 3.0

Median: 25

IQR: 22–28

Mean: 63.6

SD: 16.7

Median: 66

IQR: 54–76

Mean: 70.4

SD: 13.1

Median: 73.5

IQR: 66–79

Mean: 59.6

SD: 19.5

Median: 63

IQR: 48–75

Mean: 63.9

SD: 20.3

Median: 72

IQR: 62–78

Duration of follow-up (years)

 Males

Mean: 10.0

SD: 9.0

Median: 7.5

IQR: 2.8–14.1

Mean: 13.9

SD: 10.6

Median: 10.1

IQR: 5.3–20.3

Mean: 5.8

SD: 6.1

Median: 3.9

IQR: 1.2–8.7

Mean: 6.0

SD: 6.1

Median: 4.3

IQR: 2.1–7.9

Mean: 6.5

SD: 6.8

Median: 4.3

IQR: 1.3–9.5

Mean: 7.3

SD: 7.6

Median: 5.3

IQR: 2.1–9.1

 Females

Mean: 10.6

SD: 9.2

Median: 8.2

IQR: 3.2–14.9

Mean: 12.5

SD: 6.4

Median: 15.3

IQR: 5.3–17.1

Mean: 6.3

SD: 6.2

Median: 4.5

IQR: 1.3–9.3

Mean: 5.7

SD: 5.6

Median: 2.7

IQR: 2.1–9.3

Mean: 6.9

SD: 6.9

Median: 4.9

IQR: 1.5–10.1

Mean: 6.6

SD: 6.1

Median: 3.7

IQR: 2.2–9.9

IQR interquartile range, SD standard deviation

We identified 252,090 cases of NHL in the SEER 18 registries from 1973 to 2014. The average age at NHL diagnosis was 59.2 years (SD: 17.3) for males (54% of cases) and 63.6 years (SD: 16.7) for females (46% of cases) (Table 1); 86% were white (Table 1). Overall, 21% of NHL patients received beam-radiation therapy. The mean duration of follow-up after NHL diagnosis for both sexes combined was 6.0 years (SD: 6.1); median follow-up was 4.2 years (IQR: 1.3–8.9).

Through 31 December 2014, 14 second primary mesothelioma cases developed in the HL patient cohort, nine of whom had received radiotherapy and five of whom had not. Among HL cases who subsequently developed mesothelioma, the mean latency between HL diagnosis and mesothelioma diagnosis was 17.5 years (SD: 9.3; median: 17.1; IQR: 10.1–20.3) for those who received radiotherapy. In the NHL cohort, we identified 73 second primary mesothelioma cases, 19 of whom had received radiotherapy and 54 of whom had not. Among NHL cases who later developed mesothelioma, the mean latency between NHL diagnosis and mesothelioma diagnosis was 8.8 years (SD: 7.9; median: 5.2; IQR: 2.8–13.8) among those who underwent radiotherapy. Sixty-eight second primary mesothelioma cases (20 with radiotherapy and 48 without radiotherapy) arose among nodal HL and NHL patients combined. Among extranodal HL and NHL patients, 19 second primary mesothelioma cases were diagnosed during follow-up (eight with radiotherapy and 11 without radiotherapy).

The results from the SIR and AER analysis are summarized in Table 2. Risk of second primary mesothelioma was statistically significantly increased among males and females with HL treated with radiotherapy (nine cases observed vs. 2.06 expected; SIR = 4.37, 95% CI 2.00–8.31), and not among those not treated with radiotherapy (5 cases observed vs. 2.94 expected; SIR = 1.70, 95% CI 0.55–3.97). Results were similar after stratification by sex (Table 2). Among irradiated HL patients, the AER of second primary mesothelioma ranged between 0.13 and 0.42 excess cases per 10,000 person-years.
Table 2

Standardized incidence ratios (SIRs) with 95% confidence intervals (CIs) and absolute excess risks (AERs) for second primary mesothelioma in lymphoma patients by radiotherapy status in U.S. Surveillance, Epidemiology, and End Results 18 cancer registries, 1973–2014

Radiotherapy status

Mesothelioma cases

AER (per 10,000 p-years)

Total lymphoma patients

Person-years (1,000s)

Observed

Expected

SIR

SIR 95% CI

Hodgkin lymphoma

 Males

  No radiotherapy

4

2.44

1.64

0.45–4.19

0.12

15,666

130.0

  Radiotherapy

7

1.64

4.28

1.71–8.81

0.42

10,285

128.8

 Females

  No radiotherapy

1

0.49

2.03

0.05–11.29

0.05

12,015

101.4

  Radiotherapy

2

0.42

4.75

0.57–17.14

0.13

9,253

123.8

 Males and females

  No radiotherapy

5

2.94

1.70

0.55–3.97

0.09

27,681

231.3

  Radiotherapy

9

2.06

4.37

2.00–8.31

0.27

19,538

252.6

Non-Hodgkin lymphoma

 Males

  No radiotherapy

41

39.56

1.04

0.76–1.39

0.02

108,230

607.4

  Radiotherapy

14

11.21

1.25

0.68–2.10

0.14

28,249

193.8

 Females

  No radiotherapy

13

8.50

1.53

0.81–2.62

0.08

91,406

550.1

  Radiotherapy

5

2.46

2.04

0.66–4.74

0.14

24,205

175.3

 Males and females

  No radiotherapy

54

48.06

1.12

0.87–1.46

0.05

199,636

1,157.4

  Radiotherapy

19

13.67

1.39

0.84–2.20

0.14

52,454

369.2

All lymphoma

 Males

  No radiotherapy

45

42.01

1.07

0.78–1.44

0.04

123,896

737.4

  Radiotherapy

21

12.85

1.63

1.01–2.50

0.25

38,534

322.7

 Females

  No radiotherapy

14

8.99

1.56

0.85–2.62

0.08

103,421

651.4

  Radiotherapy

7

2.88

2.43

0.98–5.01

0.14

33,458

299.1

 Males and females

  No radiotherapy

59

51.00

1.16

0.89–1.50

0.06

227,317

1,388.8

  Radiotherapy

28

15.73

1.78

1.18–2.58

0.20

71,992

621.8

Among NHL survivors, the risk of second primary mesothelioma was elevated, but not statistically significantly so, among males and females combined after treatment with radiotherapy (19 cases observed vs. 13.67 expected; SIR = 1.39, 95% CI 0.84–2.20), and not appreciably increased after treatment without radiotherapy (54 cases observed vs. 48.06 expected; SIR = 1.12, 95% CI 0.87–1.46). Again, results were similar after stratification by sex (Table 2). The AER of second primary mesothelioma among irradiated NHL patients was 0.14 excess cases per 10,000 person-years among males, females, and both sexes combined.

When both lymphoma types were combined, the risk of second primary mesothelioma was significantly increased after radiotherapy (28 cases observed vs. 15.73 expected; SIR = 1.78, 95% CI 1.18–2.58), but not without radiotherapy (59 cases observed vs. 51.00 expected; SIR = 1.16, 95% CI 0.89–1.50), with the AER among irradiated patients ranging between 0.14 and 0.25 excess cases per 10,000 person-years (Table 2).

Results were comparable when restricted to lymphoma cases diagnosed between 1973 and 2003. For males and females with HL or NHL combined, the SIR of second primary mesothelioma was significantly increased after radiotherapy (24 cases observed vs. 12.05 expected; SIR = 1.99, 95% CI 1.28–2.95), but not in the absence of radiotherapy (30 cases observed vs. 32.77 expected; SIR = 0.92, 95% CI 0.62–1.31).

Results of the competing risks survival analysis are shown in Table 3 and Fig. 1. After multivariate adjustment for age at lymphoma diagnosis, sex, race, lymphoma type, year of lymphoma diagnosis, and nodal status, receipt of radiotherapy remained significantly associated with risk of second primary mesothelioma (RR = 1.64, 95% CI 1.05–2.57). This association was augmented after restriction to cases with a latency of at least 5 years (RR = 2.32, 95% CI 1.29–4.17; not shown in tables) or at least 10 years (RR = 4.05, 95% CI 1.65–9.95). Other factors significantly associated with increased risk of second primary mesothelioma were younger age (<25 years) or older age (≥55 years) at lymphoma diagnosis, male sex, and white race (Table 3). Risk did not differ significantly by lymphoma type, year of diagnosis, or lymphoma nodal status.
Table 3

Multivariate adjusteda relative risks (RRs) with 95% confidence intervals (CIs) for second primary mesothelioma in lymphoma patients in U.S. Surveillance, Epidemiology, and End Results 18 cancer registries, 1973–2014

Characteristic

Mesothelioma cases

RR

95% CI

Age at lymphoma diagnosis (years)

 <25

6

1.00

Ref.

 25–34

2

0.34

0.07–1.70

 35–44

6

0.89

0.30–2.65

 45–54

4

0.44

0.12–1.64

 55–64

17

1.53

0.62–3.76

 65–74

27

2.34

1.02–5.36

 ≥75

25

2.27

0.97–5.30

Sex

 Female

21

1.00

Ref.

 Male

66

2.95

1.79–4.84

Race

 Non-white

4

1.00

Ref.

 White

83

2.85

1.04–7.78

Lymphoma type

   

 Non-Hodgkin lymphoma

73

1.00

Ref.

 Hodgkin lymphoma

14

1.30

0.72–2.35

Year of lymphoma diagnosis

 1995 and later

64

1.00

Ref.

 Before 1995

23

0.74

0.47–1.15

Nodal status

 Nodal

68

1.00

Ref.

 Extranodal

19

0.73

0.43–1.24

Radiotherapy for lymphoma

 No

59

1.00

Ref.

 Yes

28

1.64

1.05–2.57

aEstimates are mutually adjusted for all variables shown in table

Fig. 1

Multivariate Cox proportional hazards survival analysis of second primary mesothelioma after lymphoma by time since lymphoma diagnosis, comparing patients who received radiotherapy (dashed gray) with those who did not (solid black), adjusting for age, sex, race, lymphoma type, year of lymphoma diagnosis, and nodal status. U.S. Surveillance, Epidemiology, and End Results 18 cancer registries, 1973–2014

In stratified analyses, radiotherapy for lymphoma was consistently associated with increased risk of second primary mesothelioma among both males and females and patients with both nodal and extranodal lymphoma (Table 4). Second primary mesothelioma risk after radiation was higher for patients diagnosed with lymphoma before 1995 than for those diagnosed later, and after a latency period of at least 10 years. Risk also appeared to be higher for patients diagnosed with lymphoma before age 40 years than those diagnosed at older ages, although risk was not statistically significantly increased in the younger group (based on 11 cases). Risk was higher for non-whites (based on four cases) than whites.
Table 4

Stratified, multivariate adjusteda relative risks (RRs) with 95% confidence intervals (CIs) for radiotherapy vs. no radiotherapy in association with risk of second primary mesothelioma in lymphoma patients in U.S. Surveillance, Epidemiology, and End Results 18 cancer registries, 1973–2014

Stratum

Mesothelioma cases

RR

95% CI

Age at lymphoma diagnosis <40 years

11

2.88

0.67–12.30

Age at lymphoma diagnosis ≥40 years

76

1.45

0.88–2.41

Males

66

1.66

1.00–2.74

Females

21

1.58

0.61–4.13

Non-whites

4

11.69

1.48–92.62

Whites

83

1.48

0.93–2.35

Lymphoma diagnosed before 1995

23

2.59

1.18–5.69

Lymphoma diagnosed 1995 or later

64

1.32

0.73–2.38

Nodal lymphoma

68

1.61

0.95–2.72

Extranodal lymphoma

19

1.74

0.72–4.22

Latency <10 years

66

1.13

0.61–2.09

Latency ≥10 years

21

4.05

1.65–9.95

aEstimates are adjusted for age, sex, race, lymphoma type, year of lymphoma diagnosis, and nodal status, as appropriate

Discussion

Although radiotherapy has been a mainstay of curative treatment for HL since the 1960s, growing awareness of the long-term toxicities of radiation exposure, combined with improvements in involved-field/site/node radiotherapy delivery and development of combined chemoradiotherapy regimens, has led to progressive decreases in radiotherapy volume and dose [24, 26]. Consequently, typical doses of 40–45 Gy with mantle/extended-field radiotherapy have decreased to typical doses of 20–30 Gy or less with involved-field/site/node radiotherapy [28]. A similar evolution from higher-dose, extended-field radiotherapy to lower-dose, involved-field/site/node radiotherapy also has taken place in the treatment of early-stage NHL [25, 27], with geographic and institutional variation in the implementation of such changes. We observed greater excess risk of second primary mesothelioma among irradiated lymphoma patients diagnosed before 1995, when radiation volumes and doses were higher, than those diagnosed later; thus, our results may indirectly suggest a dose–response effect of ionizing radiation on second primary mesothelioma risk. Alternatively, the greater excess risk among irradiated lymphoma patients diagnosed before 1995 could be due largely to the longer latency to mesothelioma. Even lower-dose radiotherapy may confer an increased risk of second primary malignancy after lymphoma [29], making it imperative to continue surveillance of the long-term sequelae of modern chemoradiotherapy approaches.

Like previous studies that reported a higher risk of second primary mesothelioma with longer latency [3, 4, 5, 7, 8, 10, 11, 13], we found that risk of second primary mesothelioma was substantially increased with a latency of at least ten years. Although based on few mesothelioma cases among women (n = 21; 7 irradiated), our results for HL and NHL combined are consistent with those of other studies showing that, overall, the risk of second primary malignancy after radiotherapy does not differ appreciably between men and women with HL [19, 23].

In HL, the radiotherapy-related relative risk of second primary malignancy in general rises with earlier age at diagnosis, probably due in part to the lower background risk of cancer at younger ages, as well as the greater susceptibility of developing organs to the carcinogenic effect of ionizing radiation [1, 19, 23]. Our results, which show higher (albeit statistically non-significant) risk of second primary mesothelioma in irradiated lymphoma patients diagnosed before age 40 years than in those diagnosed later, are consistent with increased susceptibility to the effect of ionizing radiation at earlier ages on mesothelioma development.

Although our SIR analyses showed a statistically significant increase in risk of second primary mesothelioma after radiotherapy for HL but not NHL, the survival analysis revealed no difference in mesothelioma risk after HL vs. NHL, suggesting that the difference in SIRs might be due to the typically earlier age at diagnosis, longer potential latency, and lower competing risk of death for HL than NHL.

Although extranodal and nodal lymphomas have long been distinguished clinically and therapeutically (e.g., [30]), there has until recently been a lack of formal guidelines for radiotherapy of extranodal lymphomas [27]. Our results showing similar risks of second primary mesothelioma between nodal and extranodal lymphomas, although based on few subsequent mesotheliomas among extranodal lymphoma survivors, suggest that radiotherapy doses may not differ substantially between the two. The indication of a higher risk of second primary mesothelioma after radiotherapy for non-whites than whites is based on a small number of non-white cases, and requires further investigation.

The AER of second primary mesothelioma after HL in our study was substantially lower than that reported in a cohort of long-term HL survivors in the Netherlands (AER = 2.6 per 10,000) [19]. The difference may be largely due to the Dutch study’s earlier treatment period, longer duration of follow-up, higher average attained age, and restriction to survivors for at least 5 years, all of which would be expected to increase the risk of second primary mesothelioma.

Limitations of our study include the rarity of mesothelioma and relatively short average follow-up time; therefore, even with an extended observation period, the number of second primary mesothelioma cases is small, especially in some subgroups. Thus, statistical power was limited to detect any heterogeneity between subgroups. Some radiotherapy received by patients, especially outside of hospital settings in recent years, may have been missed by SEER registries. Such misclassification would have resulted in overestimated SIRs for non-irradiated patients, and underestimated RRs comparing irradiated with non-irradiated patients. SEER also lacks detailed information on radiation doses and treatment fields, thereby precluding detailed dose-response analyses. Our study also lacked information on lymphoma patients’ occupational history, especially asbestos exposure; however, asbestos is unlikely to confound the association between radiotherapy and second primary mesothelioma risk because the excess risk is limited to lymphoma patients who received radiotherapy, and there is no reason to anticipate that asbestos exposure differed systematically by receipt of radiotherapy. Thus, greater opportunity for asbestos exposure among patients diagnosed before 1995 would be unlikely to explain the higher SIR for irradiated than non-irradiated patients during the earlier time period. Moreover, the excess risk was greatest in lymphoma patients diagnosed at an earlier age, and the study cohort was restricted to patients diagnosed with lymphoma in 1973 or later, after the institution of regulations on asbestos use. These patients would have had little opportunity for post-diagnosis asbestos exposure, for which the latency period is typically three decades or more [31]. The higher excess risk for irradiated patients diagnosed in the earlier period also is unlikely to be due to the inclusion of four additional SEER registries from 1992 onward and five more from 2000 onward, since there is no evidence that radiotherapy patterns differed markedly among these geographic regions.

Migration of lymphoma patients, especially younger patients with HL, out of the SEER catchment area may have resulted in underascertainment of subsequent mesotheliomas. However, a recent study of childhood cancer survivors found no evidence of underascertainment of second primary cancers in SEER due to outmigration [32]. Moreover, migration patterns would not be expected to vary distinctly by receipt of radiotherapy, such that comparisons between irradiated and non-irradiated lymphoma survivors should remain unbiased.

Although the causative relationship between receipt of radiation therapy and subsequent development of mesothelioma in cancer survivors has been confirmed by a large body of existing studies, our expanded study population and extended follow-up enhanced our ability to use survival analysis methods adjusting for multiple covariates and accounting for competing risk of death, and to conduct subgroup analyses. The risk of mesothelioma was increased after radiotherapy among both men and women with both NHL and HL, especially with a longer latency, and apparently with younger age at lymphoma diagnosis and treatment. Our results showing higher risk among lymphoma patients treated with radiotherapy in earlier time periods are consistent with a dose–response effect of ionizing radiation on mesothelioma risk. These findings have limited clinical implications in light of the rarity of mesothelioma as an outcome of lymphoma irradiation, as well as the longstanding trend toward lower radiation doses and volumes, such that our findings are unlikely to influence clinical practice. Nevertheless, taken together, our results offer greater insight into how radiotherapy-induced mesothelioma varies by demographic and treatment characteristics, thereby shedding greater light on the epidemiology of this uncommon and lethal cancer.

Notes

Acknowledgments

We thank Dr. Karin Ekström Smedby (Karolinska University Hospital) for her expert guidance on clinical issues. M. Jane Teta and Fionna Mowat have served as expert witnesses and consultants, and Ellen Chang has provided litigation support, involving potential health hazards associated with asbestos in various consumer products. Partial funding for the analyses performed was provided by counsel for CenturyLink, who are involved with asbestos-related litigation and retained M. Jane Teta. This manuscript is exclusively the work product of the authors, and the results and the manuscript were not reviewed or discussed by the study sponsors prior to journal submission.

References

  1. 1.
    Ng AK, van Leeuwen FE (2016) Hodgkin lymphoma: late effects of treatment and guidelines for surveillance. Semin Hematol 53:209–215CrossRefPubMedGoogle Scholar
  2. 2.
    Travis LB, Curtis RE, Boice JD Jr, Hankey BF, Fraumeni JF Jr (1991) Second cancers following non-Hodgkin’s lymphoma. Cancer 67:2002–2009CrossRefPubMedGoogle Scholar
  3. 3.
    Tward JD, Wendland MM, Shrieve DC, Szabo A, Gaffney DK (2006) The risk of secondary malignancies over 30 years after the treatment of non-Hodgkin lymphoma. Cancer 107:108–115CrossRefPubMedGoogle Scholar
  4. 4.
    De Bruin ML, Burgers JA, Baas P et al (2009) Malignant mesothelioma after radiation treatment for Hodgkin lymphoma. Blood 113:3679–3681CrossRefPubMedGoogle Scholar
  5. 5.
    Hodgson DC, Gilbert ES, Dores GM et al (2007) Long-term solid cancer risk among 5-year survivors of Hodgkin’s lymphoma. J Clin Oncol 25:1489–1497CrossRefPubMedGoogle Scholar
  6. 6.
    Teta MJ, Lau E, Sceurman BK, Wagner ME (2007) Therapeutic radiation for lymphoma: risk of malignant mesothelioma. Cancer 109:1432–1438CrossRefPubMedGoogle Scholar
  7. 7.
    Berrington de Gonzalez A, Curtis RE, Gilbert E et al (2010) Second solid cancers after radiotherapy for breast cancer in SEER cancer registries. Br J Cancer 102:220–226CrossRefPubMedGoogle Scholar
  8. 8.
    Brown LM, Chen BE, Pfeiffer RM et al (2007) Risk of second non-hematological malignancies among 376,825 breast cancer survivors. Breast Cancer Res Treat 106:439–451CrossRefPubMedGoogle Scholar
  9. 9.
    Deutsch M, Land SR, Begovic M, Cecchini R, Wolmark N (2007) An association between postoperative radiotherapy for primary breast cancer in 11 National Surgical Adjuvant Breast and Bowel Project (NSABP) studies and the subsequent appearance of pleural mesothelioma. Am J Clin Oncol 30:294–296CrossRefPubMedGoogle Scholar
  10. 10.
    Travis LB, Fossa SD, Schonfeld SJ et al (2005) Second cancers among 40,576 testicular cancer patients: focus on long-term survivors. J Natl Cancer Inst 97:1354–1365CrossRefPubMedGoogle Scholar
  11. 11.
    Farioli A, Violante FS, Mattioli S, Curti S, Kriebel D (2013) Risk of mesothelioma following external beam radiotherapy for prostate cancer: a cohort analysis of SEER database. Cancer Causes Control 24:1535–1545CrossRefPubMedPubMedCentralGoogle Scholar
  12. 12.
    Cavazza A, Travis LB, Travis WD et al (1996) Post-irradiation malignant mesothelioma. Cancer 77:1379–1385CrossRefPubMedGoogle Scholar
  13. 13.
    Farioli A, Ottone M, Morganti AG et al (2016) Radiation-induced mesothelioma among long-term solid cancer survivors: a longitudinal analysis of SEER database. Cancer Med 5:950–959CrossRefPubMedPubMedCentralGoogle Scholar
  14. 14.
    Diallo I, Lamon A, Shamsaldin A, Grimaud E, de Vathaire F, Chavaudra J (1996) Estimation of the radiation dose delivered to any point outside the target volume per patient treated with external beam radiotherapy. Radiother Oncol 38:269–271CrossRefPubMedGoogle Scholar
  15. 15.
    Hall EJ (2006) Intensity-modulated radiation therapy, protons, and the risk of second cancers. Int J Radiat Oncol Biol Phys 65:1–7CrossRefPubMedGoogle Scholar
  16. 16.
    Purdy JA (2008) Dose to normal tissues outside the radiation therapy patient’s treated volume: a review of different radiation therapy techniques. Health Phys 95:666–676CrossRefPubMedGoogle Scholar
  17. 17.
    Taylor ML, Kron T (2011) Consideration of the radiation dose delivered away from the treatment field to patients in radiotherapy. J Med Phys 36:59–71CrossRefPubMedPubMedCentralGoogle Scholar
  18. 18.
    Neugut AI, Ahsan H, Antman KH (1997) Incidence of malignant pleural mesothelioma after thoracic radiotherapy. Cancer 80:948–950CrossRefPubMedGoogle Scholar
  19. 19.
    Schaapveld M, Aleman BM, van Eggermond AM et al (2015) Second cancer risk up to 40 years after treatment for hodgkin’s lymphoma. N Engl J Med 373:2499–2511CrossRefPubMedGoogle Scholar
  20. 20.
    Andersson M, Wallin H, Jonsson M et al (1995) Lung carcinoma and malignant mesothelioma in patients exposed to thorotrast: incidence, histology and p53 status. Int J Cancer 63:330–336CrossRefPubMedGoogle Scholar
  21. 21.
    Travis LB, Hauptmann M, Gaul LK et al (2003) Site-specific cancer incidence and mortality after cerebral angiography with radioactive thorotrast. Radiat Res 160:691–706CrossRefPubMedGoogle Scholar
  22. 22.
    Fine JP, Gray RJ (1999) A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc 94:496–509CrossRefGoogle Scholar
  23. 23.
    Ng AK, Kenney LB, Gilbert ES, Travis LB (2010) Secondary malignancies across the age spectrum. Semin Radiat Oncol 20:67–78CrossRefPubMedGoogle Scholar
  24. 24.
    Hodgson DC (2011) Late effects in the era of modern therapy for Hodgkin lymphoma. Am Soc Hematol 2011:323–329CrossRefGoogle Scholar
  25. 25.
    Illidge T, Specht L, Yahalom J et al (2014) Modern radiation therapy for nodal non-Hodgkin lymphoma-target definition and dose guidelines from the International Lymphoma Radiation Oncology Group. Int J Radiat Oncol Biol Phys 89:49–58CrossRefPubMedGoogle Scholar
  26. 26.
    Specht L, Yahalom J, Illidge T et al (2014) Modern radiation therapy for Hodgkin lymphoma: field and dose guidelines from the international lymphoma radiation oncology group (ILROG). Int J Radiat Oncol Biol Phys 89:854–862CrossRefPubMedGoogle Scholar
  27. 27.
    Yahalom J, Illidge T, Specht L et al (2015) Modern radiation therapy for extranodal lymphomas: field and dose guidelines from the International Lymphoma Radiation Oncology Group. Int J Radiat Oncol Biol Phys 92:11–31CrossRefPubMedGoogle Scholar
  28. 28.
    Johnson PW (2013) Management of early-stage Hodgkin lymphoma: is there still a role for radiation? Hematology. Am Soc Hematol 2013:400–405CrossRefGoogle Scholar
  29. 29.
    Hoppe RT, Advani RH, Ai WZ et al (2015) Hodgkin lymphoma, version 2.2015. J Natl Compr Cancer Netw 13:554–586CrossRefGoogle Scholar
  30. 30.
    Wong DS, Fuller LM, Butler JJ, Shullenberger CC (1975) Extranodal non-Hodgkin’s lymphomas of the head and neck. Am J Roentgenol Radium Ther Nucl Med 123:471–481CrossRefPubMedGoogle Scholar
  31. 31.
    Lanphear BP, Buncher CR (1992) Latent period for malignant mesothelioma of occupational origin. J Occup Med 34:718–721PubMedGoogle Scholar
  32. 32.
    Kovalchik SA, Pfeiffer RM (2012) Re: assessment of impact of outmigration on incidence of second primary neoplasms in childhood cancer survivors estimated from SEER data. J Natl Cancer Inst 104:1517–1518CrossRefPubMedPubMedCentralGoogle Scholar

Copyright information

© Springer International Publishing AG 2017

Authors and Affiliations

  • Ellen T. Chang
    • 1
    • 2
  • Edmund C. Lau
    • 1
  • Fionna S. Mowat
    • 1
  • M. Jane Teta
    • 3
  1. 1.Center for Health SciencesExponent, Inc.Menlo ParkUSA
  2. 2.Stanford Cancer InstituteStanfordUSA
  3. 3.Center for Health SciencesExponent, Inc.New YorkUSA

Personalised recommendations