Abstract
Background:
Increased cancer risk has been reported in Parkinson’s disease (PD) patients carrying the leucine rich repeat kinase 2 (LRRK2) G2019S mutation (LRRK2-PD) in comparison with idiopathic PD (IPD). It is unclear whether the elevated risk would be maintained when compared with unaffected controls.
Methods:
Cancer outcomes were compared among 257 LRRK2-PD patients, 712 IPD patients, and 218 controls recruited from 7 LRRK2 consortium centers using mixed-effects logistic regression. Data were then pooled with a previous study to examine cancer risk between 401 LRRK2-PD and 1946 IPD patients.
Results:
Although cancer prevalence was similar among LRRK2-PD patients (32.3%), IPD patients (27.5%), and controls (27.5%; P = 0.33), LRRK2-PD had increased risks of leukemia (odds ratio [OR] = 4.55; 95% confidence interval [CI], 1.46–10.61) and skin cancer (OR = 1.61; 95% CI, 1.09–2.37). In the pooled analysis, LRRK2-PD patients had also elevated risks of leukemia (OR = 9.84; 95% CI, 2.15–44.94) and colon cancer (OR = 2.34; 95% CI, 1.15–4.74) when compared with IPD patients.
Conclusions:
The increased risks of leukemia as well as skin and colon cancers among LRRK2-PD patients suggest that LRRK2 mutations heighten risks of certain cancers.
Keywords: LRRK2 gene, G2019S mutation, Parkinson’s disease, leukemia, colon cancer, pooled analysis
Although Parkinson’s disease (PD) and cancer have seemingly opposite biological mechanisms, recent studies indicate a potential link.1–3 Epidemiological and family-based studies have reported that PD patients have increased risk of melanoma4–7; however, they have 40% and 25% lower relative risks for smoking-related and other cancers,8,9 respectively, when compared with the general population. Recent studies have reported increased risks of various cancers among PD patients carrying leucine rich repeat kinase 2 (LRRK2) mutations.10–15 The LRRK2 G2019S mutation, with a prevalence ranging from 1% to 40% depending on ethnicity and age at PD onset,16 has been associated with increased risks of nonskin cancers11,14,15 and breast cancer,11,14,15 whereas the R1441C/G mutation, which has a higher prevalence in the Basque population,17 has been associated with colon cancer13 and hematologic cancers.10
We recently performed a pooled meta-analysis among 1549 PD patients (11.4% were LRRK2 G2019S carriers) and reported statistically significant increased risks of nonskin cancer (odds ratio [OR] = 1.62; 95% confidence interval [CI], 1.04–2.52), hormone-related cancers (OR = 1.87; 95% CI, 1.07–3.26), and breast cancer (OR = 2.34; 95% CI, 1.05–5.22) among LRRK2-PD in comparison with idiopathic PD (IPD).15 However, it is unclear whether the increased risk among LRRK2-PD patients would be observed when compared with unaffected controls who are noncarriers of the G2019S mutation.
Therefore, the goal of this analysis was to compare the prevalence of cancer outcomes among LRRK2-PD patients, IPD patients, and controls using a standardized questionnaire across several international LRRK2-PD research centers. Furthermore, we also combined data collected in this study with previously published data from our meta-analysis15 and examined the associations of LRRK2 G2019S mutation with various cancers among 2365 PD patients.
Materials and Methods
Participants included in this analysis were 969 PD patients (257 LRRK2-PD and 712 IPD) and 218 genetically unrelated controls aged 35 years or older who were recruited from 7 Michael J. Fox Foundation LRRK2 sites in Europe, Israel, and the United States. All of the participants completed a detailed health questionnaire, which collected information on demographic, lifestyle, and reproductive factors; self-reported cancer history; personal health-related histories; and family histories of PD and cancer. (Details about participants’ recruitment and data collection methodology are provided in the supporting information materials.) Genetic testing for LRRK2 G2019S mutation was performed in all PD patients and controls. Genotyping for other LRRK2 mutations: R1441G/C and I2020T was performed only in 2 centers in Germany and Spain. Therefore, the focus of this analysis is only on G2019S mutations (carriers vs. noncarriers). The study was approved by the respective institution review boards of all participating sites, and all participants provided written informed consent.
We compared demographic and lifestyle characteristics as well as prevalence of various cancers among LRRK2-PD patients, IPD patients, and controls using one-way analysis of variance models (for continuous normally distributed variables) and χ2 tests (for categorical variables); all tests were 2-sided (P < 0.05). The associations of various cancer outcomes among LRRK2-PD patients, IPD patients, and controls were examined using mixed-effect logistic regression to estimate ORs and 95% CIs adjusting for age, sex, and Ashkenazi Jewish (AJ) ethnicity as fixed effects and study center as the random effect.18 Multivariate models for all cancers combined, smoking-related cancers, and colon and kidney cancers were additionally adjusted for smoking status and body mass index, whereas body mass index and reproductive factors were included in multivariate models for hormone-related cancers and breast cancer. We also investigated whether the associations between LRRK2 G2019S mutation and cancer outcomes varied by ethnicity (AJ vs. others) in stratified analyses.19
Finally, we pooled data collected in this study with our previously published data15 and compared the risk of various cancers between 401 LRRK2-PD and 1964 IPD patients using mixed-effect logistic regression models adjusted for age, sex, and AJ ethnicity as fixed effects and study center (random effect). For 83 over-lapping PD patients between the 2 datasets, the most recent data were used. We excluded 70 LRRK2-PD patients from this analysis who carried only the R1441G/C mutation as screening for this mutation was not performed in all participating centers. All statistical analyses were performed using STATA version 15 (College Station, TX).
Results
A total of 257 LRRK2-PD patients, 712 IPD patients, and 218 genetically unrelated controls aged 35 years or older were included in this analysis (Table 1). The controls were on average younger, more likely to be women, and less likely to be of AJ ethnicity in comparison with both LRRK2-PD and IPD patients. With regard to lifestyle factors, when compared with controls, both the LRRK2-PD and IPD patients were significantly less likely to be cigarette smokers and alcohol drinkers, but there was no difference in body mass index among the 3 groups (Table 1). In comparison with female patients with LRRK2-PD or IPD, unaffected female controls had on average a fewer number of pregnancies (1.4 vs. 2.1 and 2.6; P < 0.0001), and a slightly higher proportion were in menopause (94% vs. 90% and 83%; P = 0.004).
TABLE 1.
Demographic and lifestyle factors and cancer outcomes among Parkinson’s Disease patients with LRRK2 G2019S mutation (LRRK2-PD), idiopathic PD patients (IPD) and genetically unrelated controls
| Characteristic | Controls, N = 218 | LRRK2-PD, N = 257 | IPD, N = 712 | P* | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Age, y; mean (SD) | 64.0 | (11.8) | 68.6 | (10.3) | 67.8 | (10.5) | <0.0001 | |||
| Age at PD Diagnosis, y; mean (SD) | – | – | 57.6 | (11.0) | 60.1 | (11.1) | 0.002 | |||
| Duration of PD, y; mean (SD) | – | – | 11.1 | (7.5) | 7.8 | (6.1) | <0.0001 | |||
| Sex, n (%) | <0.0001 | |||||||||
| Male | 70 | (32.1) | 123 | (47.9) | 441 | (61.9) | ||||
| Female | 148 | (67.9) | 134 | (52.1) | 271 | (38.1) | ||||
| Race/ethnicity, n (%) | <0.0001 | |||||||||
| Caucasian/Ashkenazi Jews | 110 | (50.5) | 205 | (79.8) | 603 | (84.7) | ||||
| Caucasian/non-Ashkenazi Jews | 82 | (37.6) | 40 | (15.6) | 88 | (12.3) | ||||
| Hispanic or mixed/other | 26 | (11.9) | 12 | (4.7) | 21 | (3.0) | ||||
| Study center, n (%) | <0.0001 | |||||||||
| Beth Israel/Mount Sinai | 85 | (39.0) | 72 | (28.0) | 220 | (30.9) | ||||
| CUMC/New York | 22 | (10.1) | 50 | (19.5) | 131 | (18.4) | ||||
| TA Sourasky/Israel | 3 | (1.4) | 80 | (31.1) | 248 | (34.8) | ||||
| Barcelona/Spain | 39 | (17.9) | 36 | (14.0) | 44 | (6.2) | ||||
| Norway | 20 | (9.2) | 3 | (1.2) | 28 | (3.9) | ||||
| Germany | 45 | (20.6) | 6 | (2.3) | 28 | (3.9) | ||||
| PROGENI | 4 | (1.8) | 10 | (3.9) | 13 | (1.8) | ||||
| BMI, kg/m2; mean (SD) | 25.70 | (4.38) | 25.58 | (4.10) | 25.70 | (4.05) | 0.92 | |||
| Smoking status, n (%) | N = 215 | N = 241 | N = 574 | 0.005 | ||||||
| Never | 111 | (51.6) | 132 | (54.8) | 355 | (61.9) | ||||
| Former | 89 | (41.4) | 92 | (38.2) | 203 | (35.4) | ||||
| Current | 15 | (7.0) | 17 | (7.1) | 16 | (2.3) | ||||
| Drinking habits, n (%) | N = 200 | N = 186 | N = 348 | <0.0001 | ||||||
| Never | 56 | (28.0) | 90 | (48.4) | 132 | (37.9) | ||||
| Former | 11 | (5.5) | 20 | (10.8) | 42 | (12.1) | ||||
| Current | 133 | (66.5) | 76 | (40.9) | 174 | (50.0) | ||||
| Reproductive factors-women | N = 148 | N = 134 | N = 271 | |||||||
| Age at menarche, y; mean (SD) | 12.84 | (1.51) | 12.85 | (1.54) | 12.84 | (1.53) | 0.99 | |||
| Ever pregnant, n (%) | 132 | (89.2) | 123 | (91.8) | 252 | (93.0) | 0.40 | |||
| No of pregnancies, mean (SD) | 1.43 | (1.30) | 2.06 | (1.58) | 2.63 | (1.93) | <0.0001 | |||
| Menopausal status, n (%) | 139 | (93.9) | 121 | (90.3) | 226 | (83.4) | 0.004 | |||
| Age at menopause, n (%) | 49.4 | (5.9) | 50.7 | (4.9) | 49.5 | (5.6) | 0.12 | |||
| Ever HRT use, n (%) | 34 | (23.0) | 41 | (30.6) | 91 | (33.6) | 0.15 | |||
| Controls, N = 218 |
LRRK2 PD, N = 257 |
IPD, N = 712 |
||||||||
| Cancer outcomes | n | % | n | % | n | % | P* | |||
| All cancers | 60 | 27.5 | 83 | 32.3 | 196 | 27.5 | 0.33 | |||
| Age at first cancer, mean and SD | 55.3 | 12.6 | 58.9 | 12.9 | 59.3 | 12.9 | 0.22 | |||
| No of all cancers | 0.24 | |||||||||
| 1 | 52 | 23.9 | 61 | 23.7 | 149 | 20.9 | ||||
| 2 | 8 | 3.7 | 18 | 7.0 | 42 | 5.9 | ||||
| 3 or 4 | – | – | 4 | 1.6 | 5 | 0.7 | ||||
| Skin cancer | 30 | 13.8 | 52 | 20.2 | 109 | 15.3 | 0.11 | |||
| Melanoma | 7 | 3.2 | 14 | 5.4 | 27 | 3.8 | 0.40 | |||
| All nonskin cancers | 35 | 16.1 | 43 | 16.7 | 115 | 16.2 | 0.97 | |||
| Head and neck cancer | 1 | 0.5 | 1 | 0.4 | 4 | 0.6 | 0.94 | |||
| Lung cancer | – | – | 2 | 0.8 | 4 | 0.6 | 0.47 | |||
| Esophageal cancer | – | – | 1 | 0.4 | - | - | 0.16 | |||
| Colon cancer | 2 | 0.9 | 6 | 2.3 | 6 | 0.8 | 0.15 | |||
| Liver cancer | – | – | – | – | 1 | 0.1 | 0.72 | |||
| Pancreatic cancer | – | – | 1 | 0.4 | 2 | 0.3 | 0.68 | |||
| Thyroid cancer | 1 | 0.5 | 1 | 0.4 | 7 | 1.0 | 0.55 | |||
| Kidney cancer | 1 | 0.5 | 4 | 1.6 | 6 | 0.8 | 0.43 | |||
| Bladder cancer | 1 | 0.5 | 1 | 0.4 | 3 | 0.4 | 0.99 | |||
| Brain cancer | – | – | 2 | 0.8 | 2 | 0.3 | 0.31 | |||
| Leukemia | – | – | 5 | 1.9 | – | – | <0.0001 | |||
| Lymphoma | 8 | 3.7 | 3 | 1.2 | 5 | 0.7 | 0.41 | |||
| Hormone-related cancersa | 25 | 11.5 | 22 | 8.6 | 77 | 10.8 | 0.52 | |||
| Womenb | N = 148 | N = 134 | N = 271 | |||||||
| Breast cancer | 14 | 9.5 | 11 | 8.2 | 24 | 8.9 | 0.93 | |||
| Ovarian cancer | – | – | – | – | 3 | 1.1 | 0.21 | |||
| Endometrial cancer | 4 | 2.7 | 1 | 0.7 | 3 | 1.1 | 0.31 | |||
| Cervical cancer | 2 | 1.4 | – | – | 2 | 0.7 | 0.41 | |||
| Menb | N = 70 | N = 123 | N = 441 | |||||||
| Prostate cancer | 6 | 8.6 | 10 | 8.1 | 46 | 10.4 | 0.70 | |||
| Testicular cancer | 1 | 1.4 | 0 | 0.0 | 2 | 0.5% | 0.38 | |||
P values from one-way analyses of variance (continuous) and χ2 tests (categorical variables) comparing LRRK2-PD mutation carriers, IPD patients, and controls. Student T-test was used to compare age at PD diagnosis and duration of PD between LRKK2-PD and IPD patients.
Hormone-related cancers include prostate cancer in men and breast, endometrial and ovarian cancers in women.
Percentages of gender-specific cancers are based on the number of men and women.
Abbreviations: PD, Parkinson’s disease; LRRK2-PD,Parkinson’s Disease patients with LRRK2 G2019S mutation; IPD, idiopathic PD patients; BMI, body mass index; CUMC, Columbia University Medical Center; TA, Tel-Hashomer; PROGENI, Parkinson’s Research Organized Genetics Initiative Study; HRT, hormone replacement therapy.
Overall self-reported cancer prevalence was similar among LRRK2-PD patients (32.3%), IPD patients (27.5%), and controls (27.5%; P = 0.22; Table 1). However, the LRRK2-PD patients reported a higher prevalence of leukemia (1.9%) in comparison with both IPD patients and controls, where no leukemia was reported (P < 0.0001). Interestingly, when compared with IPD patients and controls, the LRRK2-PD patients also reported a higher proportion of multiple cancers: 8.6% vs. 6.6% and 3.7%, respectively. However, these differences were not statistically significant (Table 1).
Although cancer prevalence was similar among the 3 groups, the IPD patients had an overall lower cancer risk (OR = 0.68; 95% CI, 0.45–1.01) in comparison to controls in multivariate-adjusted models (Table 2). With regard to specific cancers, there was a significant increased risk of leukemia (OR = 4.55; 95% CI, 1.46–10.61) when comparing LRRK2-PD patients to either the IPD patients or controls. Skin cancer was also significantly higher among the LRRK2-PD patients when compared with IPD patients (OR = 1.61; 95% CI, 1.09–2.37), but there was no difference when compared with controls (OR = 0.99; 95% CI, 0.57–1.71; Table 2). There was also suggestive evidence of increased risks of colon and kidney cancers among LRRK2-PD patients; however, these associations were not statistically significant. Because 77% of all participants were of AJ ethnicity, we carried out a separate analysis in this group; the overall results were similar to the main analysis (see Supporting Information Table S1).
TABLE 2.
Comparisons of cancer outcomes among LRRK2-PD patients, IPD patients, and controls in newly collected data (N = 1187)
| Cancer Outcomes | Controls, N = 218 |
LRRK2-PD, N = 257 |
LRRK2-PD vs. Controls |
IPD, N = 712 |
IPD vs. Controls |
LRRK2-PD vs. IPD |
||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| N | % | N | % | OR* | 95% CI | N | % | OR* | 95% CI | OR* | 95% CI | |
| All cancers | 60 | 27.5 | 83 | 32.3 | 0.86 | 0.55–1.35 | 196 | 27.5 | 0.68 | 0.45–1.01 | 1.29 | 0.91–1.81 |
| Skin cancer | 30 | 13.8 | 52 | 20.2 | 0.99 | 0.57–1.71 | 109 | 15.3 | 0.62 | 0.37–1.03 | 1.61 | 1.09–2.37 |
| Melanoma | 7 | 3.2 | 14 | 5.4 | 1.22 | 0.47–3.14 | 27 | 3.8 | 0.80 | 0.33–1.94 | 1.54 | 0.79–3.04 |
| Nonskin cancer | 35 | 16.1 | 43 | 16.7 | 0.87 | 0.51–1.47 | 115 | 16.2 | 0.88 | 0.55–1.40 | 1.00 | 0.66–1.50 |
| Smoking-related cancersa | 2 | 0.9 | 4 | 1.6 | 1.26 | 0.22–7.18 | 10 | 1.4 | 1.21 | 0.24–6.01 | 1.02 | 0.31–3.40 |
| Colon cancer | 2 | 0.9 | 6 | 2.3 | 2.15 | 0.41–11.31 | 6 | 0.8 | 0.75 | 0.14–4.09 | 2.71 | 0.85–8.61 |
| Kidney cancer | 1 | 0.5 | 4 | 1.6 | 2.59 | 0.28–23.91 | 6 | 0.8 | 1.10 | 0.1–9.87 | 2.20 | 0.58–8.29 |
| Leukemiab | 0 | – | 5 | 1.9 | 4.55 | 1.46–10.61 | 0 | – | – | 4.55 | 1.46–10.61 | |
| Lymphoma | 5 | 2.3 | 3 | 1.2 | 0.44 | 0.1–1.97 | 8 | 1.1 | 0.45 | 0.13–1.52 | 0.51 | 0.22–1.21 |
| Hormonal cancer, women and menc | 25 | 11.5 | 22 | 8.6 | 0.55 | 0.30–1.04 | 77 | 10.8 | 0.74 | 0.43–1.24 | 0.76 | 0.46–1.27 |
| Breast cancer, women | 14 | 9.5 | 11 | 8.2 | 0.61 | 0.26–1.43 | 24 | 8.9 | 0.65 | 0.31–1.35 | 0.86 | 0.40–1.86 |
| Prostate cancer, men | 6 | 8.6 | 10 | 8.1 | 0.92 | 0.29–2.93 | 46 | 10.4 | 1.27 | 0.47–3.47 | 0.73 | 0.34–1.54 |
Bold font represents results that are statistically significant at P < 0.05.
OR and 95% CI were estimated from mixed-effect logistic regression models adjusted for age, sex, and Ashkenazi Jews ethnicity (fixed effect) and study center (random effect). Multivariate models for all cancers, smoking-related cancers, colon cancer, kidney cancer, and hormone cancer also included smoking status and body mass index.
Smoking-related cancers included lung, bladder, and head and neck cancers; these models were also adjusted for smoking status.
Because there were no leukemia cases reported among IPD and controls, the OR = 4.55 and corresponding 95% CI (1.46–10.61) were calculated based on the ratio of observed cases (n = 5) versus expected cases (n = 0.8) among LRRK2-PD.
Hormone-related cancers include prostate cancer in men and breast, endometrial, and ovarian cancers in women. For breast cancer, body mass index and reproductive factors were also included in multivariate models.
Abbreviations: PD, Parkinson’s disease; LRRK2-PD, Parkinson’s Disease patients with LRRK2 G2019S mutation; IPD, idiopathic PD patients; OR, odds ratio; CI, confidence interval.
Finally, the results of the pooled analysis (Table 3), which combined the data from this study with our previously published paper15 and included 401 LRRK2-PD and 1964 IPD patients, showed statistically significantly increased risks of leukemia (OR = 9.84; 95% CI, 2.15–44.94) and colon cancer (OR = 2.34; 95% CI, 1.15–4.72) when comparing the LRRK2-PD with IPD patients.
TABLE 3.
Comparisons of cancer outcomes between LRRK2-PD and IPD patients in the pooled analysis of1 newly collected and previously published data (N = 2365)*
| Cancer Outcomes |
LRRK2-PD G2019S Mutation Carriers, N = 401 |
IPD, N = 1,964 |
LRRK2-PD vs. IPD |
||||
|---|---|---|---|---|---|---|---|
| N | % | N | % | OR† | 95% CI | P | |
| All cancers | 113 | 28.2 | 389 | 19.8 | 1.25 | 0.94–1.64 | 0.12 |
| Skin cancer | 60 | 15.0 | 148 | 7.5 | 1.36 | 0.96–1.95 | 0.09 |
| Melanoma | 18 | 4.5 | 43 | 2.2 | 1.55 | 0.86–2.81 | 0.15 |
| Nonskin cancers | 68 | 17.0 | 270 | 13.7 | 1.12 | 0.81–1.55 | 0.49 |
| Smoking-related cancersa | 6 | 1.5 | 27 | 1.4 | 1.16 | 0.44–3.07 | 0.77 |
| Lung cancer | 4 | 1.0 | 11 | 0.6 | 1.75 | 0.49–6.27 | 0.39 |
| Bladder cancer | 1 | 0.2 | 13 | 0.7 | 0.53 | 0.1–4.37 | 0.55 |
| Colon cancer | 12 | 3.0 | 32 | 1.6 | 2.34 | 1.15–4.72 | 0.018 |
| Kidney cancer | 5 | 1.2 | 13 | 0.7 | 2.22 | 0.74–6.66 | 0.15 |
| Leukemia | 5 | 1.2 | 3 | 0.2 | 9.84 | 2.15–44.94 | 0.003 |
| Lymphoma | 3 | 0.7 | 12 | 0.6 | 0.82 | 0.22–3.10 | 0.77 |
| Brain cancer | 4 | 1.0 | 11 | 0.6 | 2.18 | 0.62–7.69 | 0.23 |
| Hormonal cancerb | 37 | 9.2 | 146 | 7.4 | 1.00 | 0.66–1.51 | 0.99 |
| Breast cancer, women | 21 | 10.0 | 51 | 6.4 | 1.22 | 0.69–2.14 | 0.46 |
| Prostate cancer, men | 16 | 8.4 | 82 | 7.0 | 0.97 | 0.53–1.77 | 0.91 |
Bold font represents results that are statistically significant at P < 0.05.
Data from Agalliu et al.15 LRRK2 R1441G mutation carriers (N = 70) were excluded from this analysis.
OR and 95% CI were estimated from mixed-effect logistic regression models adjusted for age, sex, and Ashkenazi Jews ethnicity (fixed effect) and study center (random effect).
Smoking-related cancers included lung, bladder and head and neck cancers.
Hormonal cancers include prostate cancer in men and breast, endometrial, and ovarian cancers in women.
Abbreviations: PD, Parkinson’s disease; LRRK2-PD, Parkinson’s Disease patients with LRRK2 G2019S mutation; IPD, idiopathic PD patients; OR, odds ratio; CI, confidence interval.
Discussion
We report the findings from a primary analysis of cancer outcomes among 257 LRRK2-PD patients, 712 IPD patients, and 218 unaffected controls, which used a standardized questionnaire to collect demographic and lifestyle factors as well as cancer outcomes across 7 participating sites from the largest international LRRK2-PD consortium. The results showed that the LRRK2-PD patients had a statistically significant 4.6-fold increased risk of leukemia in comparison with the IPD patients and controls. In additional support, the findings from the pooled analysis demonstrated a stronger risk of leukemia (OR = 9.84; 95% CI, 2.15–44.94) when comparing LRRK2 G2019S mutation carriers with IPD patients, although this was based on a small number (n = 5) of leukemia reports. The observed positive association with leukemia in this study also supports the finding of Ruiz-Martinez and colleagues,10 who reported an OR = 7.1 for myeloproliferative cancers among their PD patients carrying the R1441G mutation. Both studies also show that IPD patients have a lower frequency of hematologic cancers, which has been previously reported.20,21
In our primary analysis, we also observed a 61% increased risk of nonmelanoma skin cancer when comparing LRRK2-PD patients with IPD patients, although in the pooled analysis the strength of this association was attenuated (OR = 1.36) and was no longer statistically significant. These findings need to be interpreted with caution as nonmelanoma skin cancers tend to be misreported by participants.22,23 Although LRRK2-PD patients had also suggestive evidence of an increased risk of colon cancer in the primary analysis (albeit not statistically significant because of the small numbers), in the pooled analysis we observed a significant 2.34-fold increased risk of colon cancer (95% CI, 1.15–4.72) when comparing LRRK2-PD with IPD patients. The increased risk of colon cancer has also been reported among PD patients with LRRK2 R1441C mutations from a large pedigree of 190 individuals in western Nebraska.13
Despite the associations with the specific cancers mentioned previously, we did not observe increased risks of hormone-related cancers and breast cancer, as reported in our previous study.15 Two other studies14,24 also reported increased risks of breast cancer among LRKK2-PD patients, and these populations were in part included in the pooled analysis. One of the differences is that the numbers of self-reported cancers were much lower in the international LRRK2-PD consortium. Moreover, the effect of LRRK2 mutations on cancer might vary among different populations, and other studies from Italy25 or the United Kingdom26 did not observe an overall increase in hormone-related cancers with LRRK2 mutations.
Our study has strengths and limitations that should be carefully considered. A strength is the use of a detailed questionnaire that collected demographic, lifestyle, and reproductive factors and cancer outcomes in a standardized manner, which minimized biases as a result of data acquisition across international LRRK2-PD centers. In addition, the pooled analysis included 2365 PD patients with 502 cancer outcomes collected across several sites, which represent the largest cohort of PD patients with genetic screening for LRRK2 mutations. Although self-reports of major cancers (eg, breast, prostate, colon, lung, etc.) have been shown to be valid and reliable,22,23,27 a limitation is that self-reported cancers from PD patients and controls were not validated systematically with medical records or cancer registry reports in participating centers. Another limitation is the relatively small sample size for comparisons of rare cancers among LRRK2-PD patients, IPD patients, and controls. We also did not have information on the tumor grades or stages for various cancers in this study, and because genetic associations might vary by cancer clinical phenotypes, the inclusion of cancer survivors might have affected the results. Finally, the potential for selection bias in using spouse controls and recall bias particularly among elderly PD patients as a result of the increased risk of dementia might have affected the results of this study. Nevertheless, the advantages of using spouse controls are time-and cost-efficiency and their ability to provide proxy lifestyle or health-related data for their affected relatives.28
In conclusion, the results of this study showed significantly increased risks of leukemia and colon cancer among the LRRK2-PD patients when compared with the IPD patients, which suggest that LRRK2 mutations can lead to multiple cancers.
Supplementary Material
Acknowledgments:
Ilir Agalliu and Rachel Saunders-Pullman had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. This work was supported by the Michael J. Fox Foundation, the Bigglesworth Family Foundation, and the National Institutes of Health/National Institute of Neurological Disorders and Stroke (Grants: NS073836 and NS094148 to R.S.P.). The funding agencies had no role in the design and conduct of the study; data collection, management, analysis, and interpretation of the data; as well as in preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication.
Funding agencies: This work was supported by the Michael J. Fox Foundation, the Biggles worth Family Foundation, and the National Institutes of Health/National Institute of Neurological Disorders and Stroke (Grants: NS073836 and NS094148 to R.S.P.).
Footnotes
Relevant conflicts of interests Nothing to report.
financial disclosures: Full financial disclosures and author roles may be found in the online version of this article.
Supporting Data
Additional Supporting Information may be found in the online version of this article at the publisher’s web-site.
References
- 1.Devine MJ, Plun-Favreau H, Wood NW. Parkinson’s disease and cancer: two wars, one front. Nat Rev Cancer 2011;11(11):812–823. [DOI] [PubMed] [Google Scholar]
- 2.D’Amelio M, Ragonese P, Sconzo G, Aridon P, Savettieri G. Parkinson’s disease and cancer: insights for pathogenesis from epidemiology. Ann N Y Acad Sci 2009;1155:324–334. [DOI] [PubMed] [Google Scholar]
- 3.West AB, Dawson VL, Dawson TM. To die or grow: Parkinson’s disease and cancer. Trend Neurosci 2005;28(7):348–352. [DOI] [PubMed] [Google Scholar]
- 4.Gao X, Simon KC, Han J, Schwarzschild MA, Ascherio A. Family history of melanoma and Parkinson disease risk. Neurology 2009; 73(16):1286–1291. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Liu R, Gao X, Lu Y, Chen H. Meta-analysis of the relationship between Parkinson disease and melanoma. Neurology 2011;76(23): 2002–2009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kareus SA, Figueroa KP, Cannon-Albright LA, Pulst SM. Shared predispositions of parkinsonism and cancer: a population-based pedigree-linked study. Arch Neurol 2012;69(12):1572–1577. [DOI] [PubMed] [Google Scholar]
- 7.Inzelberg R, Rabey JM, Melamed E, et al. High prevalence of malignant melanoma in Israeli patients with Parkinson’s disease. J Neural Transm (Vienna) 2011;118(8):1199–1207. [DOI] [PubMed] [Google Scholar]
- 8.Bajaj A, Driver JA, Schernhammer ES. Parkinson’s disease and cancer risk: a systematic review and meta-analysis. Cancer Causes Control 2010;21(5):697–707. [DOI] [PubMed] [Google Scholar]
- 9.Rugbjerg K, Friis S, Lassen CF, Ritz B, Olsen JH. Malignant melanoma, breast cancer and other cancers in patients with Parkinson’s disease. Int J Cancer 2012;131(8):1904–1911. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Ruiz-Martinez J, de la Riva P, Rodriguez-Oroz MC, et al. Prevalence of cancer in Parkinson’s disease related to R1441G and G2019S mutations in LRRK2. Mov Disord 2014;29(6):750–755. [DOI] [PubMed] [Google Scholar]
- 11.Saunders-Pullman R, Barrett MJ, Stanley KM, et al. LRRK2 G2019S mutations are associated with an increased cancer risk in Parkinson disease. Mov Disord 2010;25(15):2536–2541. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Sierra M, Gonzalez-Aramburu I, Sanchez-Juan P, et al. High frequency and reduced penetrance of LRRK2 G2019S mutation among Parkinson’s disease patients in Cantabria (Spain). Mov Disord 2011; 26(13):2343–2346. [DOI] [PubMed] [Google Scholar]
- 13.Strongosky AJ, Farrer M, Wszolek ZK. Are Parkinson disease patients protected from some but not all cancers? Neurology 2008; 71(20):1650; author reply 1650–1651. [DOI] [PubMed] [Google Scholar]
- 14.Inzelberg R, Cohen OS, Aharon-Peretz J, et al. The LRRK2 G2019S mutation is associated with Parkinson disease and concomitant nonskin cancers. Neurology 2012;78(11):781–786. [DOI] [PubMed] [Google Scholar]
- 15.Agalliu I, San Luciano M, Mirelman A, et al. Higher frequency of certain cancers in LRRK2 G2019S mutation carriers with Parkinson disease: a pooled analysis. JAMA Neurol 2015;72(1):58–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Healy DG, Falchi M, O’Sullivan SS, et al. Phenotype, genotype, and worldwide genetic penetrance of LRRK2-associated Parkinson’s disease: a case-control study. Lancet Neurol 2008;7(7):583–590. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Gorostidi A, Ruiz-Martinez J, Lopez de Munain A, Alzualde A, Marti Masso JF. LRRK2 G2019S and R1441G mutations associated with Parkinson’s disease are common in the Basque Country, but relative prevalence is determined by ethnicity. Neurogenetics 2009;10(2):157–159. [DOI] [PubMed] [Google Scholar]
- 18.McCulloch CE, Searle SR, Neuhaus JM. Generalized, Linear, and Mixed Models. 2nd ed Hoboken, NJ: Wiley; 2008. [Google Scholar]
- 19.Kleinbaum DG, Kupper LK, Nizam A, Muller KE. Applied Regression Analysis and Other Multivariable Methods. 4th ed Belmont, CA: Thompson; 2007. [Google Scholar]
- 20.Inzelberg R, Jankovic J. Are Parkinson disease patients protected from some but not all cancers? Neurology 2007;69(15):1542–1550. [DOI] [PubMed] [Google Scholar]
- 21.Becker C, Brobert GP, Johansson S, Jick SS, Meier CR. Cancer risk in association with Parkinson disease: a population-based study. Parkinsonism Relat Disord 2010;16(3):186–190. [DOI] [PubMed] [Google Scholar]
- 22.Cho S, Shin A, Song D, et al. Validity of self-reported cancer history in the health examinees (HEXA) study: A comparison of self-report and cancer registry records. Cancer Epidemiol 2017;50(Pt A):16–21. [DOI] [PubMed] [Google Scholar]
- 23.Loh V, Harding J, Koshkina V, Barr E, Shaw J, Magliano D. The validity of self-reported cancer in an Australian population study. Aust NZJ Public Health 2014;38(1):35–38. [DOI] [PubMed] [Google Scholar]
- 24.Waro BJ, Aasly JO. Exploring cancer in LRRK2 mutation carriers and idiopathic Parkinson’s disease. Brain Behav 2018;8(1):e00858. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Allegra R, Tunesi S, Cilia R, Pezzoli G, Goldwurm S. LRRK2-G2019S mutation is not associated with an increased cancer risk: a kincohort study. Mov Disord 2014;29(10):1325–1326. [DOI] [PubMed] [Google Scholar]
- 26.Mortiboys H, Cox A, Brock IW, Bandmann O. The common PARK8 mutation LRRK2G2019S is not a risk factor for breast cancer in the absence of Parkinson’s disease. J Neurol 2013;260(8):2177–2178. [DOI] [PubMed] [Google Scholar]
- 27.Nash SH, Day G, Hiratsuka VY, Zimpelman GL, Koller KR. Agreement between self-reported and central cancer registry-recorded prevalence of cancer in the Alaska EARTH study. Int J Circumpolar Health 2019;78(1):1571383. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Verhage BA, Aben KK, Straatman H, Verbeek AL, Beaty TH, Kiemeney LA. Spouse controls in family case-control studies: a methodological consideration. Fam Cancer 2003;2(2): 101–108. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.