Ann Surg Oncol (2014) 21:3422–3428 DOI 10.1245/s10434-014-3774-7
ORIGINAL ARTICLE – GASTROINTESTINAL ONCOLOGY
Incidence of Additional Primary Malignancies in Patients with Pancreatic and Gastrointestinal Neuroendocrine Tumors Rondi M. Kauffmann, MD, MPH1, Li Wang, MS2, Sharon Phillips, MS2, Kamran Idrees, MD3, Nipun B. Merchant, MD3, and Alexander A. Parikh, MD, MPH3 Department of Oncologic Surgery, City of Hope National Medical Center, Duarte, CA; 2Department of Biostatistics, Vanderbilt University School of Medicine, Nashville, TN; 3Department of Surgery, Vanderbilt University School of Medicine, Nashville, TN 1
ABSTRACT Background. The incidence of secondary malignancies is increased in patients with malignant and premalignant conditions. Although neuroendocrine tumors (NET) are uncommon, their incidence is increasing. We evaluated the rate of additional malignancies in patients with NET. Methods. Using the Surveillance, Epidemiology, and End Results (SEER) database, we identified a cohort of patients with pancreatic NET (PNET) or gastrointestinal NET (GINET). We determined the incidence of additional cancers diagnosed either before or after the diagnosis of PNET or GINET, by comparing these rates with the general population. Using multivariable regression, we evaluated factors that increased the risk of an additional malignancy. Results. A cohort of 9,727 NET patients was identified. A total of 3,086 additional cancers occurred in 2,508 patients (25.8 %). The most common sites of additional malignancies included colorectal (21.1 %), prostate (14.5 %), breast (13.3 %), and lung (11.6 %). Among patients with PNET, the incidence of breast, lung, uterine, lymph, and pancreatic cancers was less than expected in the general population, whereas in patients with GINET, the observed incidence of nearly all malignancies exceeded that expected. Increasing age, marital status, and localized NET were associated with increased risk. Conclusion. Our study shows that the incidence of additional malignancies in patients with PNET and GINET is 25.8 %. Patients with GINET are at increased risk of additional malignancies, whereas patients with PNET have Ó Society of Surgical Oncology 2014 First Received: 28 February 2014; Published Online: 25 July 2014 A. A. Parikh, MD, MPH e-mail: [email protected]
a decreased risk compared with the general population. More vigilant surveillance for secondary malignancies should be performed in patients with GINET. Studies investigating potential etiologic oncogenic pathways are warranted.
Neuroendocrine tumors (NET) are a heterogeneous group of neoplasms that include pancreatic NETs (PNET) and gastrointestinal NETs (GINET). They are characterized as functional or nonfunctional depending on the secretion of hormones. Although these tumors are usually indolent, they can invade and metastasize.1 The incidence of NET is increasing, with an estimated incidence of 1 in 100,000 people in 1975 compared with 5.4 in 100,000 in 2010.2 Although this may be in part due to improved diagnostic modalities and heightened awareness, several studies suggest that a true increase in incidence exists.1 The risk of additional cancers is higher in patients with a history of cancer.3 These additional cancers can be in the same organ or in other organs as part of a genetic syndrome, such as the multiple endocrine neoplasia syndromes I or II, Lynch syndrome, or Von Hippel–Lindau disease.4 More recent evidence suggests that even in the absence of a genetic syndrome, patients with one type of cancer are at a higher risk of other cancers.5–7 In a recent series of 500 patients with GINET and PNET, 13.7 % of patients were found to have a second malignancy, and the presence of a NET was associated with an increased incidence of synchronous malignancies.8 To date, this study represents the largest population-based study to have been performed evaluating the incidence of additional malignancies in patients with NET. Our objective was to use the Surveillance, Epidemiology, and End Results (SEER) tumor registry to determine
Add’l Cancers and Neuroendocrine Tumors
the incidence and site of secondary malignancies in patients with GINET or PNET. Timing of and risk factors associated with the development of additional malignancies were also evaluated. METHODS This study was approved by the Institutional Review Board of Vanderbilt University Medical Center and considered exempt by the City of Hope Medical Center.
3423 TABLE 1 Demographic comparison of the PNET and GINET cohorts Demographics and tumor PNET group characteristics (n = 974)
GINET group (n = 8,753)
p value
Age at diagnosis (years) 59
63
\0.001a
Statistical Analysis The cohort was divided into two groups: those with NET only and those with at least one additional primary malignancy. Patients were analyzed according to the order of occurrence of the second malignancy relative to the NET, as well as whether their NET was a PNET or GINET. Descriptive statistics were performed to determine the most common site of the additional cancers. Univariate analyses including Wilcoxon rank-sum, Kruskal–Wallis, and Pearson Chi square tests were performed. The overall rates of additional malignancies by location were compared with the general population for both groups by using a previously
85 9
81 13
Other
6
6 \0.001b
Male
55
46
Female
45
54 0.012b
Marital status (%) Married
68
64
Unmarried
32
36 \0.001b
Tumor stage
Cohort Selection Using the International Classification of Disease for Oncology (2nd edition) codes provided in SEER, we selected patients aged 18–100 years with a diagnosis of GINET or PNET between 1983 and 1998. The codes used include the following: islet cell (8150), insulinoma (8151), glucagonoma (8153), gastrinoma (8153), mixed islet cell/ exocrine carcinoma (8154), VIPoma (8155), carcinoid (8240), enterochromaffin cell carcinoid (8241), goblet cell carcinoid (8243), composite carcinoid (8244), adenocarcinoid (8245), and neuroendocrine carcinoma (8246). Our cohort was limited to tumors with these ICD codes and occurring in the pancreas or gastrointestinal tract (stomach, small bowel, colon, or rectum). All other sites of NET were excluded. The presence of additional primary tumors in this cohort was evaluated between 1973 and 2008, providing a 10-year window before and after the diagnosis of the NET. Patients with [4 primary tumors were excluded.
White Black Gender (%)
Data Source The SEER registry is maintained by the National Cancer Institute and provides data on demographics, tumor site, histology, number of primary tumors, cancer markers, stage, treatment, and survival status.9 The 2010 submission was used.10
\0.001b
Race (%)
Localized
11
53
Regional
20
18
Distant
61
14
Localized/regional
0
0
Unstaged
8
15
Demographics were based on the first neuroendocrine tumor diagnosis during the study period a
Wilcoxon test
b
Pearson test
described algorithm.11 These calculations were based on the 2008 US Cancer Statistics and the 2010 estimated US population for the US Census Bureau of 308,745,538 people.12,13 Multivariable logistic regression models were built to assess the probability of having additional malignancies after NET. Age, gender, race, marital status, year of diagnosis, histologic stage, and the occurrence of additional cancers before NET were included in the model. Previous malignancy was included as baseline risk factor. The development of an additional cancer was included as a time-varying covariate. Only the first additional cancer after NET was considered. All analyses were done with the statistical programming language R, version 2.13.1 (R Development Core Team, Vienna, Austria). The level of statistical significance was set at p \ 0.05. RESULTS Patient Demographics Between 1983 and 1998, 9,727 patients were diagnosed with NET and met inclusion criteria. A total of 974 PNET patients and 8,753 GINET patients were included. The
3424
R. M. Kauffmann et al.
TABLE 2 Demographics and tumor characteristics by number and order of additional malignancies Demographics and tumor characteristics
Neuroendocrine tumor only (n = 7,219)
Additional nonneuroendocrine malignancy before neuroendocrine tumor (n = 1,194)
Additional nonneuroendocrine malignancy after neuroendocrine tumor (n = 1,099)
Additional non-neuroendocrine p value malignancy before and after neuroendocrine tumor (n = 215)
Age at diagnosis, years (25th/ 75th)
60 (48/71)
71 (63/77)
64 (55/72)
69 (63/74)
\0.001a
0.013b
Race (%) White
81
82
80
86
Black
13
14
14
11
Other
6
4
5
3 0.002b
Gender (%) Male
46
48
49
56
Female
54
52
51
44 \0.001b
Marital status (%) Married
64
62
70
68
Unmarried
36
38
30
32 \0.001b
Tumor stage (%) Localized
46
51
59
62
Regional
18
21
18
24
Distant
21
15
12
8
Unstaged
15
13
11
6
a
Kruskal–Wallis test
b
Pearson test
demographic characteristics of both cohorts are displayed in Table 1.
NET. More patients with locoregional primary NET were present in those with an additional malignancy after or before and after NET diagnosis.
Incidence of Additional Primary Malignancies Sites of Additional Malignancies Of the 9,727 patients in the cohort, a total of 2,508 patients (25.8 %) were diagnosed with an additional nonNET malignancy at a median time interval of 57 months (interquartile range, 14–113 months). A total of 144 additional malignancies occurred in 123 PNET patients, and 2,942 additional malignancies occurred in 2,385 GINET patients. Of these 2,508 patients, 1,194 patients (47.6 %) were diagnosed with an additional malignancy before their NET tumor, and 1,099 patients (43.8 %) after. In 215 (8.6 %) patients, additional malignancies were diagnosed both before and after the NET. Demographics of Patients with NET-Only Versus Additional Non-NET Malignancies Table 2 summarizes the demographics of patients with NET only compared with patients with additional non-NET malignancies before, after, or before and after diagnosis of
There were a total of 3,086 additional non-NET malignancies occurring in 2,508 patients, with some patients having more than one additional cancer. Colorectal cancer was the most frequent site, comprising 21.1 % of additional malignancies. Other common sites included prostate (14.5 %), breast (13.3 %), and lung/bronchus (11.6 %). PNET Versus GINET We then examined the incidence of additional malignancies both before and after the diagnosis of NET, stratified by PNET and GINET (Tables 3, 4). In the PNET group, there were a total of 144 additional malignancies in 123 patients (14.8 %). The most common site was the prostate (27.8 %), followed by colorectal (11.8 %), breast (11.1 %), and lung (6.9 %). There were 2,942 additional malignancies in 2,385 patients in the GINET group
Add’l Cancers and Neuroendocrine Tumors
3425
TABLE 3 Additional primary cancers by tumor type and timing relative to NET in the PNET group
TABLE 5 Observed and expected rates of additional cancers by type: PNET group
Tumor site
Before (n = 89), n (%)
After (n = 55), n (%)
Combined (n = 144), n (%)
Cancer type
Annual US incidence per 100,000a
Expected cases
Observed cases
Observed/ expected ratio
Prostate
25 (28.1)
15 (27.3)
40 (27.8)
Prostate
143
50
40
0.8
11 (12.4)
6 (10.9)
17 (11.8)
Breast
87.5
30.6
16
0.52a
8 (9)
8 (14.5)
16 (11.1)
Lung/bronchus
68.5
24
10
0.42a
Colorectal
a
Breast Lung and bronchus
6 (6.7)
4 (7.3)
10 (6.9)
Colorectal
47
16
17
1.06
Urinary bladder
6 (6.7)
1 (1.8)
7 (4.9)
Corpus uteri
28
10
5
0.5
Corpus uteri
4 (4.5)
1 (1.8)
5 (3.5)
Lymph nodes
22.7
8
2
0.25
Stomach
1 (1.1)
1 (1.8)
2 (1.4)
Urinary bladder
21.5
7.5
7
0.93
Lymph nodes
1 (1.1)
1 (1.8)
2 (1.4)
Pancreas
12.5
4
2
0.5
Pancreas
1 (1.1)
1 (1.8)
2 (1.4)
Stomach
6.8
2.4
2
0.83
Esophagus
0 (0)
2 (3.6)
2 (1.4)
Esophagus
5.2
1.8
2
1.1
Other
26 (29.2)
15 (27.3)
41 (28.5)
a
There were 144 additional non-NET and NET primary malignancies in the PNET group a
Includes appendix
Statistically significant
TABLE 6 Observed and expected rates of additional cancers by type: GINET group Cancer type
Annual US incidence per 100,000a
Expected cases
Observed cases
Observed/ expected ratio
TABLE 4 Additional primary cancers by tumor type and timing relative to NET in the GINET group
Prostate
143
450
406
0.9a
Breast
87.5
276
395
1.43a
Tumor site
Lung/bronchus
68.5
216
349
1.6a
Colorectal
47
148
635
4.3a
Corpus uteri
28
88
95
1.08
Lymph nodes
22.7
71.5
52
0.73
Urinary bladder
21.5
68
111
1.63a
Pancreas
12.5
39
58
1.49a
Stomach
6.8
21
86
4.1a
Esophagus
5.2
16.4
23
1.4
(33.6 %). The most common site was colorectal (21.6 %),
Colorectal
Before (n = 1,516), n (%) 429 (28.3)
After (n = 1,426), n (%) 206 (14.4)
Combined (n = 2,942), n (%) 635 (21.6)
Prostate gland
179 (11.8)
227 (15.9)
406 (13.8)
Breast
196 (12.9)
199 (14)
395 (13.4)
Lung and bronchus
153 (10.1)
196 (13.7)
349 (11.9)
Urinary bladder
59 (3.9)
52 (3.6)
111 (3.8)
Corpus uteri
63 (4.2)
32 (2.2)
95 (3.2)
Stomach
49 (3.2)
37 (2.6)
86 (2.9)
Lymph nodes
22 (1.5)
30 (2.1)
52 (1.8)
Pancreas
19 (1.3)
39 (2.7)
58 (2)
Esophagus
7 (0.5)
16 (1.1)
23 (0.8)
Other
340 (22.4)
392 (27.5)
732 (24.9)
There were 2,942 additional non-NET and NET primary malignancies in the GINET group
prostate (13.8 %), breast (13.4 %), and lung (11.9 %). Comparison with the General Population The overall rates of additional malignancies by location were compared with the general population for each of these two groups. For selected common additional cancer sites, Tables 5 and 6 show the annual US incidence and expected and observed rates of cancer in patients with
Based on the 2010 estimated US population of 308,745,538: all races, male and female a
Statistically significant
PNET and carcinoid. In the PNET group, the observed rates of colorectal and esophageal malignancies were similar to those expected in the general population (Table 5), whereas the observed rates of pancreatic, breast, lung, uterine, and lymph node malignancies were below those expected. In the GINET group, the observed rates of stomach and colorectal malignancies were particularly high, with the rates of stomach and colorectal cancer 4.1 and 4.3 times higher than the general population, respectively (Table 6). Logistic Regression Analysis A logistic regression model was used to model the probability of having an additional non-NET after the diagnosis
3426
R. M. Kauffmann et al.
TABLE 7 Logistic regression model predicting risk of additional non-NET malignancy after diagnosis of PNET Variable
OR (95 % CI)
Age (72:51)
2.08 (1.09–3.93)a
Race (black/white)
2.19 (0.85–5.67)
Race (other/white)
1.20 (0.35–4.17)
Marital status (married/unmarried)
4.0 (1.59–10.0)a
Histologic stage (localized/regional)
2.56 (1.03–6.25)a
Histologic stage (localized/distant)
3.45(1.61–7.14)a
Histologic stage (localized/unstaged) Date of diagnosis (before/after 1990)
2.70 (0.71–10.0) 1.23 (0.66–2.29)
Gender (M/F)
1.15 (0.61–2.14)
Malignancy before PNET (Y/N)
0.93 (0.35–2.49)
a
Statistically significant
TABLE 8 Logistic regression model predicting risk of additional non-NET malignancy after diagnosis of GINET Variable
OR (95 % CI)
Age (72:51)
1.60 (1.32–1.93)a
Gender (M/F)
1.27 (1.11–1.45)a
Race (black/white)
1.08 (0.89–1.31)
Race (white/other)
1.39 (1.02–1.89)a
Marital status (married/unmarried)
1.18 (1.02–1.37)a
Histologic stage (localized/regional)
1.23 (1.04–1.47)a
Histologic stage (localized/distant) Histologic stage (localized/unstaged)
2.27 (1.82–2.78)a 2.04 (1.61–2.50)a
Date of diagnosis (before/after 1990)
1.19 (1.04–1.37)a
Malignancy before NET (Y/N)
0.91 (0.76–1.08)
a
Statistically significant
of either PNET or GINET. Covariates included age, gender, race, marital status, stage, year of diagnosis (before or after 1990), and whether the patient had an additional primary malignancy before the diagnosis of NET. Logistic regression model results for the PNET group are displayed in Table 7. Increased age at diagnosis, marital status, and localized disease were independent predictors of increased risk of an additional malignancy after diagnosis of NET. Logistic regression model results for the GINET group are displayed in Table 8. Increasing age, male gender, diagnosis before 1990, white race, marital status, and localized stage were associated with increased risk of an additional non-NET after the diagnosis of GINET. DISCUSSION The incidence of additional cancers in patients with PNET or GINET is approximately 26 %. The incidence of additional malignancies in patients with PNET is 15 %, whereas the incidence in patients with GINET is 34 %.
The pattern of observed additional malignancies correlates with the pattern of observed cancers in the United States. For patients with PNET, the incidence of several common cancers was lower than that expected. In patients with GINET, however, the incidence of colorectal, stomach, breast, lung, uterine, bladder, pancreatic, and esophageal malignancies exceeded the incidence seen in the general population, particularly for colorectal and gastric cancer. The increased risk of additional malignancies in patients with GINET compared with PNET may be explained by the indolent course of GINET compared with PNET, in that survival rates for patients with GINET are longer than that of patients with PNET, allowing more time to develop additional malignancies. Median survival in our PNET group was 2.58 years (95 % confidence interval (CI) 2.17–2.92 years), compared with 10.9 years (95 % CI 10.4–11.4 years) in the GINET group.14,15 It is likely that patients with previous GINET are maintained on a surveillance program in which they undergo regular esophagogastric duodenoscopy and colonoscopic evaluation. Although enrollment in surveillance programs or adherence to the recommended guidelines is impossible to ascertain from the SEER database, such surveillance would confound our results to the degree that these patients are monitored more closely than the general population and therefore have their second malignancies diagnosed more readily. This is particularly relevant to the increased rate of gastric and colorectal cancer observed in our GINET cohort, but it would not completely explain the increase in breast, lung, uterine, or pancreatic malignancies in this population, because patients with GINET would not necessarily be expected to have closer surveillance for these malignancies. The observation that more than half of the additional non-NET malignancies were diagnosed before the NET strengthens our conclusion that there is a true increase in the risk of non-NET malignancies in patients with GINET that is not related solely to closer surveillance. Furthermore, in multivariate regression analysis, a diagnosis of malignancy before NET was not associated with increased risk of subsequent non-NET cancer, thus suggesting that closer monitoring alone is not driving the association. It is also possible that the increased risk of a second cancer in patients with GINET is related to common genetic or tumorigenic pathways. The relationship between GINET and other malignancies in the setting of a family history of multiple endocrine neoplasia I, neurofibromatosis type 1, and von Hippel–Lindau disease is well known. There is a growing body of literature reporting an increased incidence of other malignancies in patients with GINET in the absence of a known genetic syndrome’s. This may indicate yet-unknown genetic mutations predisposing to multiple malignancies. These additional malignancies are
Add’l Cancers and Neuroendocrine Tumors
more commonly synchronous and gastrointestinal in origin but can be metachronous and occur anywhere.5–7 Tichansky et al.7 used the SEER database to evaluate the incidence of additional malignancies in 2,086 patients with colorectal carcinoid tumors and found an increased incidence of cancers of the colon/rectum, small bowel, esophagus/stomach, lung/bronchus, urinary tract, and prostate in these patients. Additional gastrointestinal malignancies were usually synchronous, whereas additional nongastrointestinal malignancies were more likely metachronous. A recent study from Kamp et al.8 reported that the incidence of synchronous, but not metachronous, malignancies was increased in patients with NET of the gastrointestinal tract and pancreas. The authors observed an increased risk of melanoma and cancers of the breast, colon/rectum, and prostate in their study population. However, their population was small, comprising only 459 patients, and this may account for the differences from our study. Tsai et al.16 used the Taiwan Cancer Registry to study the rate of second cancers in patients with NET of various sites and found an increased risk of second cancers after NET, particularly in patients diagnosed with NET after the age of 70 years. They concluded that close monitoring for a second malignancy is warranted in patients with a history of NET. However, their study did not distinguish risk of second cancers by site of primary NET.16 In multivariable regression analysis of patients with PNET or GINET, common factors associated with increased risk of developing additional non-NET malignancies included increased age, marital status, and localized disease. In patients with GINET, age, male gender, diagnosis before 1990, marital status, and localized disease were associated with increased risk. In PNET patients, increased age, marital status, and localized disease were associated with increased risk. As would be expected, the risk of malignancy increases with age, because the patient has had a longer lifespan to develop additional malignancies. It is unclear why marital status would increase risk of additional malignancies. Several studies have reported marital status to be associated with increased survival after cancer.17–19 Although this finding has not been completely explained, psychosocial support and the spouse’s influence in monitoring and directing healthrelated behavior have been suggested. If marital status translates to a higher likelihood of undergoing cancer screening, this may explain why married patients in our cohort had an increased risk of additional non-NET malignancies. Higher pathologic stage of the NET was associated with decreased risk of additional malignancies. This is likely because patients with higher-grade NET are more likely to die of their neuroendocrine disease and thus lack sufficient time in which to develop an additional non-NET malignancy. Similarly, those diagnosed before 1990 had a longer
3427
follow-up and therefore more time to develop another malignancy. It is unclear why white race was associated with an additional risk of a secondary malignancy. Whites have improved survival rates for many cancers, and therefore these patients may have a longer lifespan, thus allowing more time to develop additional cancers.20,21 Additionally, white individuals are more likely to undergo routine health screening for malignancy (i.e., colonoscopy or mammogram) than other races.22,23 There are several limitations to this study. Patients in the SEER database may move between regions. Therefore, our analysis could have missed patients who were diagnosed with a NET in one SEER region and then moved out of the region and developed another malignancy. Additionally, although SEER captures relatively specific data on primary cancers, other potential confounding factors, such as comorbidities, other risk factors for developing a cancer (such as dietary and environmental factors, presence of a genetic syndrome, smoking history, and previous radiation therapy), and details on screening and treatment, are not available. CONCLUSION In conclusion, the incidence of additional primary malignancies in patients with PNET and GINET approaches 26 %. Patients with GINET are at increased risk of developing cancers in nearly every site, whereas those with PNET actually have a lower risk of developing additional cancers, likely because of their decreased survival. These findings underscore the importance of increased surveillance for other malignancies and prompt the question of whether shorter screening intervals or earlier screening age for other malignancies (i.e., stomach and colorectal) in patients with GINET and no history of genetic syndrome are warranted. Prior to a recommendation for shorter screening intervals, large-scale studies are needed to determine feasibility, effectiveness, and cost. Additional study of the incidence of additional malignancies in patients with GINET should further investigate whether risk is related to anatomic location (foregut/midgut/hindgut) of the primary GINET. Finally, future studies are needed to discern common pathophysiologic signaling pathways involved in both NET and non-NET malignancies in the hopes of developing cost-effective screening procedures and common targets for treatment. ACKNOWLEDGMENT CTSA Grant RR024975.
Source of support: Vanderbilt NIH
REFERENCES 1. Yao JC, Hassan M, Phan A, et al. One hundred years after ‘‘carcinoid’’: epidemiology of and prognostic factors for
3428
2.
3.
4.
5. 6. 7.
8.
9. 10.
11.
12.
neuroendocrine tumors in 35,825 cases in the United States. J Clin Oncol. 2008;26:3063–72. Oberg K, Castellano D. Current knowledge on diagnosis and staging of neuroendocrine tumors. Cancer Metastasis Rev. 2011;30(Suppl. 1):3–7. Trentham-Dietz A, Newcomb PA, Nichols HB, et al. Breast cancer risk factors and second primary malignancies among women with breast cancer. Br Cancer Res Treat. 2007;105:195–207. Lubensky IA, Pack S, Ault D, et al. Multiple neuroendocrine tumors of the pancreas in von Hippel–Lindau disease patients. Am J Pathol. 1998;153:223–31. Habal N, Sims C, Bilchik AJ. Gastrointestinal carcinoid tumors and second primary malignancies. J Surg Oncol. 2000;75:310–6. Aggarwal G, Obideen K, Wehbi M. Carcinoid tumors: what should increase our suspicion? Clevel Clin J Med. 2008;75:849–55. Tichansky DS, Cagir B, Borrazzo E, et al. Risk of second cancers in patients with colorectal carcinoids. Dis Colon Rectum. 2002;45:91–7. Kamp K, Damhuis RA, Felders RA, de Herder WW. Occurrence of second primary malignancies in patients with neuroendocrine tumors of the digestive tract and pancreas. Endocr Relat Cancer. 2012;19:95–9. Surveillance, Epidemiology, and End Results. http://seer.cancer. gov. Accessed 15 April 2012. Surveillance, Epidemiology, and End Results (SEER) Program. http://www.seer.cancer.gov. Research data (1973–2008). National Cancer Institute, DCCPS, Surveillance Research Program, Surveillance Systems Branch, released April 2011, based on November 2010 submission. Riall TS, Stager VM, Nealon WH, et al. Incidence of additional primary cancers in patients with invasive intraductal papillary mucinous neoplasms and sporadic pancreatic adenocarcinomas. J Am Coll Surg. 2007;204:803–13. US Cancer Statistics Working Group. http://www.apps.nccd.cdc. gov/uscs/cancersbyraceandethnicity.aspx. United States Cancer
R. M. Kauffmann et al.
13. 14.
15.
16. 17.
18.
19.
20.
21.
22. 23.
Statistics: 1999–2008 Incidence and Mortality Web-based Report. Atlanta: US Department of Health and Human Services, Center for Disease Control and Prevention and National Cancer Institute. Accessed 27 November 2012. US Census. http://www.quickfacts.census.gov/qfd/states/00000. html. Accessed 14 June 2012. Shebani KO, Souba WW, Finkelstein DM, et al. Prognosis and survival in patients with gastrointestinal tract carcinoid tumors. Ann Surg. 1999;229:815–21. Fischer L, Kleef J, Esposito I, et al. Clinical outcome and longterm survival in 118 consecutive patients with neuroendocrine tumours of the pancreas. Br J Surg. 2008;95:627–35. Tsai HJ, Wu CC, Tsai CR, et al. Second cancers in patients with neuroendocrine tumors. PLoS One. 2013;8:e86414. Wang L, Wilson SE, Stewart DB, et al. Marital status and colon cancer outcomes in US Surveillance, Epidemiology, and End Results registries: does marriage affect cancer survival by gender and age? Can Epidemiol. 2011;35:417–22. Baine M, Sahak F, Lin C, et al. Marital status and survival in pancreatic cancer patients: a SEER based analysis. PLoS One. 2011;6:E21052. Zhu K, Weiss NS, Schwartz SM, et al. Assessing the relationship between marital status and cancer incidence: methodologic considerations. Cancer Causes Control. 1994;5:83–7. Silber JH, Rosenbaum PR, Clark AS, et al. Characteristics associated with differences in survival among black and white women with breast cancer. JAMA. 2013;310:389–97. Clegg LX, Li FP, Hankey BF, Chu K, Edwards BK. Cancer survival among US whites and minorities: a SEER (Surveillance, Epidemiology, and End Results) program population-based study. Arch Intern Med. 2002;161:1985–93. DeBarros M, Steele SR. Colorectal cancer screening in an equal access healthcare system. J Cancer. 2013;4:270–80. Peek ME, Han JH. Disparities in screening mammography. J Gen Intern Med. 2004;19:184–94.
ORIGINAL ARTICLE – GASTROINTESTINAL ONCOLOGY
Incidence of Additional Primary Malignancies in Patients with Pancreatic and Gastrointestinal Neuroendocrine Tumors Rondi M. Kauffmann, MD, MPH1, Li Wang, MS2, Sharon Phillips, MS2, Kamran Idrees, MD3, Nipun B. Merchant, MD3, and Alexander A. Parikh, MD, MPH3 Department of Oncologic Surgery, City of Hope National Medical Center, Duarte, CA; 2Department of Biostatistics, Vanderbilt University School of Medicine, Nashville, TN; 3Department of Surgery, Vanderbilt University School of Medicine, Nashville, TN 1
ABSTRACT Background. The incidence of secondary malignancies is increased in patients with malignant and premalignant conditions. Although neuroendocrine tumors (NET) are uncommon, their incidence is increasing. We evaluated the rate of additional malignancies in patients with NET. Methods. Using the Surveillance, Epidemiology, and End Results (SEER) database, we identified a cohort of patients with pancreatic NET (PNET) or gastrointestinal NET (GINET). We determined the incidence of additional cancers diagnosed either before or after the diagnosis of PNET or GINET, by comparing these rates with the general population. Using multivariable regression, we evaluated factors that increased the risk of an additional malignancy. Results. A cohort of 9,727 NET patients was identified. A total of 3,086 additional cancers occurred in 2,508 patients (25.8 %). The most common sites of additional malignancies included colorectal (21.1 %), prostate (14.5 %), breast (13.3 %), and lung (11.6 %). Among patients with PNET, the incidence of breast, lung, uterine, lymph, and pancreatic cancers was less than expected in the general population, whereas in patients with GINET, the observed incidence of nearly all malignancies exceeded that expected. Increasing age, marital status, and localized NET were associated with increased risk. Conclusion. Our study shows that the incidence of additional malignancies in patients with PNET and GINET is 25.8 %. Patients with GINET are at increased risk of additional malignancies, whereas patients with PNET have Ó Society of Surgical Oncology 2014 First Received: 28 February 2014; Published Online: 25 July 2014 A. A. Parikh, MD, MPH e-mail: [email protected]
a decreased risk compared with the general population. More vigilant surveillance for secondary malignancies should be performed in patients with GINET. Studies investigating potential etiologic oncogenic pathways are warranted.
Neuroendocrine tumors (NET) are a heterogeneous group of neoplasms that include pancreatic NETs (PNET) and gastrointestinal NETs (GINET). They are characterized as functional or nonfunctional depending on the secretion of hormones. Although these tumors are usually indolent, they can invade and metastasize.1 The incidence of NET is increasing, with an estimated incidence of 1 in 100,000 people in 1975 compared with 5.4 in 100,000 in 2010.2 Although this may be in part due to improved diagnostic modalities and heightened awareness, several studies suggest that a true increase in incidence exists.1 The risk of additional cancers is higher in patients with a history of cancer.3 These additional cancers can be in the same organ or in other organs as part of a genetic syndrome, such as the multiple endocrine neoplasia syndromes I or II, Lynch syndrome, or Von Hippel–Lindau disease.4 More recent evidence suggests that even in the absence of a genetic syndrome, patients with one type of cancer are at a higher risk of other cancers.5–7 In a recent series of 500 patients with GINET and PNET, 13.7 % of patients were found to have a second malignancy, and the presence of a NET was associated with an increased incidence of synchronous malignancies.8 To date, this study represents the largest population-based study to have been performed evaluating the incidence of additional malignancies in patients with NET. Our objective was to use the Surveillance, Epidemiology, and End Results (SEER) tumor registry to determine
Add’l Cancers and Neuroendocrine Tumors
the incidence and site of secondary malignancies in patients with GINET or PNET. Timing of and risk factors associated with the development of additional malignancies were also evaluated. METHODS This study was approved by the Institutional Review Board of Vanderbilt University Medical Center and considered exempt by the City of Hope Medical Center.
3423 TABLE 1 Demographic comparison of the PNET and GINET cohorts Demographics and tumor PNET group characteristics (n = 974)
GINET group (n = 8,753)
p value
Age at diagnosis (years) 59
63
\0.001a
Statistical Analysis The cohort was divided into two groups: those with NET only and those with at least one additional primary malignancy. Patients were analyzed according to the order of occurrence of the second malignancy relative to the NET, as well as whether their NET was a PNET or GINET. Descriptive statistics were performed to determine the most common site of the additional cancers. Univariate analyses including Wilcoxon rank-sum, Kruskal–Wallis, and Pearson Chi square tests were performed. The overall rates of additional malignancies by location were compared with the general population for both groups by using a previously
85 9
81 13
Other
6
6 \0.001b
Male
55
46
Female
45
54 0.012b
Marital status (%) Married
68
64
Unmarried
32
36 \0.001b
Tumor stage
Cohort Selection Using the International Classification of Disease for Oncology (2nd edition) codes provided in SEER, we selected patients aged 18–100 years with a diagnosis of GINET or PNET between 1983 and 1998. The codes used include the following: islet cell (8150), insulinoma (8151), glucagonoma (8153), gastrinoma (8153), mixed islet cell/ exocrine carcinoma (8154), VIPoma (8155), carcinoid (8240), enterochromaffin cell carcinoid (8241), goblet cell carcinoid (8243), composite carcinoid (8244), adenocarcinoid (8245), and neuroendocrine carcinoma (8246). Our cohort was limited to tumors with these ICD codes and occurring in the pancreas or gastrointestinal tract (stomach, small bowel, colon, or rectum). All other sites of NET were excluded. The presence of additional primary tumors in this cohort was evaluated between 1973 and 2008, providing a 10-year window before and after the diagnosis of the NET. Patients with [4 primary tumors were excluded.
White Black Gender (%)
Data Source The SEER registry is maintained by the National Cancer Institute and provides data on demographics, tumor site, histology, number of primary tumors, cancer markers, stage, treatment, and survival status.9 The 2010 submission was used.10
\0.001b
Race (%)
Localized
11
53
Regional
20
18
Distant
61
14
Localized/regional
0
0
Unstaged
8
15
Demographics were based on the first neuroendocrine tumor diagnosis during the study period a
Wilcoxon test
b
Pearson test
described algorithm.11 These calculations were based on the 2008 US Cancer Statistics and the 2010 estimated US population for the US Census Bureau of 308,745,538 people.12,13 Multivariable logistic regression models were built to assess the probability of having additional malignancies after NET. Age, gender, race, marital status, year of diagnosis, histologic stage, and the occurrence of additional cancers before NET were included in the model. Previous malignancy was included as baseline risk factor. The development of an additional cancer was included as a time-varying covariate. Only the first additional cancer after NET was considered. All analyses were done with the statistical programming language R, version 2.13.1 (R Development Core Team, Vienna, Austria). The level of statistical significance was set at p \ 0.05. RESULTS Patient Demographics Between 1983 and 1998, 9,727 patients were diagnosed with NET and met inclusion criteria. A total of 974 PNET patients and 8,753 GINET patients were included. The
3424
R. M. Kauffmann et al.
TABLE 2 Demographics and tumor characteristics by number and order of additional malignancies Demographics and tumor characteristics
Neuroendocrine tumor only (n = 7,219)
Additional nonneuroendocrine malignancy before neuroendocrine tumor (n = 1,194)
Additional nonneuroendocrine malignancy after neuroendocrine tumor (n = 1,099)
Additional non-neuroendocrine p value malignancy before and after neuroendocrine tumor (n = 215)
Age at diagnosis, years (25th/ 75th)
60 (48/71)
71 (63/77)
64 (55/72)
69 (63/74)
\0.001a
0.013b
Race (%) White
81
82
80
86
Black
13
14
14
11
Other
6
4
5
3 0.002b
Gender (%) Male
46
48
49
56
Female
54
52
51
44 \0.001b
Marital status (%) Married
64
62
70
68
Unmarried
36
38
30
32 \0.001b
Tumor stage (%) Localized
46
51
59
62
Regional
18
21
18
24
Distant
21
15
12
8
Unstaged
15
13
11
6
a
Kruskal–Wallis test
b
Pearson test
demographic characteristics of both cohorts are displayed in Table 1.
NET. More patients with locoregional primary NET were present in those with an additional malignancy after or before and after NET diagnosis.
Incidence of Additional Primary Malignancies Sites of Additional Malignancies Of the 9,727 patients in the cohort, a total of 2,508 patients (25.8 %) were diagnosed with an additional nonNET malignancy at a median time interval of 57 months (interquartile range, 14–113 months). A total of 144 additional malignancies occurred in 123 PNET patients, and 2,942 additional malignancies occurred in 2,385 GINET patients. Of these 2,508 patients, 1,194 patients (47.6 %) were diagnosed with an additional malignancy before their NET tumor, and 1,099 patients (43.8 %) after. In 215 (8.6 %) patients, additional malignancies were diagnosed both before and after the NET. Demographics of Patients with NET-Only Versus Additional Non-NET Malignancies Table 2 summarizes the demographics of patients with NET only compared with patients with additional non-NET malignancies before, after, or before and after diagnosis of
There were a total of 3,086 additional non-NET malignancies occurring in 2,508 patients, with some patients having more than one additional cancer. Colorectal cancer was the most frequent site, comprising 21.1 % of additional malignancies. Other common sites included prostate (14.5 %), breast (13.3 %), and lung/bronchus (11.6 %). PNET Versus GINET We then examined the incidence of additional malignancies both before and after the diagnosis of NET, stratified by PNET and GINET (Tables 3, 4). In the PNET group, there were a total of 144 additional malignancies in 123 patients (14.8 %). The most common site was the prostate (27.8 %), followed by colorectal (11.8 %), breast (11.1 %), and lung (6.9 %). There were 2,942 additional malignancies in 2,385 patients in the GINET group
Add’l Cancers and Neuroendocrine Tumors
3425
TABLE 3 Additional primary cancers by tumor type and timing relative to NET in the PNET group
TABLE 5 Observed and expected rates of additional cancers by type: PNET group
Tumor site
Before (n = 89), n (%)
After (n = 55), n (%)
Combined (n = 144), n (%)
Cancer type
Annual US incidence per 100,000a
Expected cases
Observed cases
Observed/ expected ratio
Prostate
25 (28.1)
15 (27.3)
40 (27.8)
Prostate
143
50
40
0.8
11 (12.4)
6 (10.9)
17 (11.8)
Breast
87.5
30.6
16
0.52a
8 (9)
8 (14.5)
16 (11.1)
Lung/bronchus
68.5
24
10
0.42a
Colorectal
a
Breast Lung and bronchus
6 (6.7)
4 (7.3)
10 (6.9)
Colorectal
47
16
17
1.06
Urinary bladder
6 (6.7)
1 (1.8)
7 (4.9)
Corpus uteri
28
10
5
0.5
Corpus uteri
4 (4.5)
1 (1.8)
5 (3.5)
Lymph nodes
22.7
8
2
0.25
Stomach
1 (1.1)
1 (1.8)
2 (1.4)
Urinary bladder
21.5
7.5
7
0.93
Lymph nodes
1 (1.1)
1 (1.8)
2 (1.4)
Pancreas
12.5
4
2
0.5
Pancreas
1 (1.1)
1 (1.8)
2 (1.4)
Stomach
6.8
2.4
2
0.83
Esophagus
0 (0)
2 (3.6)
2 (1.4)
Esophagus
5.2
1.8
2
1.1
Other
26 (29.2)
15 (27.3)
41 (28.5)
a
There were 144 additional non-NET and NET primary malignancies in the PNET group a
Includes appendix
Statistically significant
TABLE 6 Observed and expected rates of additional cancers by type: GINET group Cancer type
Annual US incidence per 100,000a
Expected cases
Observed cases
Observed/ expected ratio
TABLE 4 Additional primary cancers by tumor type and timing relative to NET in the GINET group
Prostate
143
450
406
0.9a
Breast
87.5
276
395
1.43a
Tumor site
Lung/bronchus
68.5
216
349
1.6a
Colorectal
47
148
635
4.3a
Corpus uteri
28
88
95
1.08
Lymph nodes
22.7
71.5
52
0.73
Urinary bladder
21.5
68
111
1.63a
Pancreas
12.5
39
58
1.49a
Stomach
6.8
21
86
4.1a
Esophagus
5.2
16.4
23
1.4
(33.6 %). The most common site was colorectal (21.6 %),
Colorectal
Before (n = 1,516), n (%) 429 (28.3)
After (n = 1,426), n (%) 206 (14.4)
Combined (n = 2,942), n (%) 635 (21.6)
Prostate gland
179 (11.8)
227 (15.9)
406 (13.8)
Breast
196 (12.9)
199 (14)
395 (13.4)
Lung and bronchus
153 (10.1)
196 (13.7)
349 (11.9)
Urinary bladder
59 (3.9)
52 (3.6)
111 (3.8)
Corpus uteri
63 (4.2)
32 (2.2)
95 (3.2)
Stomach
49 (3.2)
37 (2.6)
86 (2.9)
Lymph nodes
22 (1.5)
30 (2.1)
52 (1.8)
Pancreas
19 (1.3)
39 (2.7)
58 (2)
Esophagus
7 (0.5)
16 (1.1)
23 (0.8)
Other
340 (22.4)
392 (27.5)
732 (24.9)
There were 2,942 additional non-NET and NET primary malignancies in the GINET group
prostate (13.8 %), breast (13.4 %), and lung (11.9 %). Comparison with the General Population The overall rates of additional malignancies by location were compared with the general population for each of these two groups. For selected common additional cancer sites, Tables 5 and 6 show the annual US incidence and expected and observed rates of cancer in patients with
Based on the 2010 estimated US population of 308,745,538: all races, male and female a
Statistically significant
PNET and carcinoid. In the PNET group, the observed rates of colorectal and esophageal malignancies were similar to those expected in the general population (Table 5), whereas the observed rates of pancreatic, breast, lung, uterine, and lymph node malignancies were below those expected. In the GINET group, the observed rates of stomach and colorectal malignancies were particularly high, with the rates of stomach and colorectal cancer 4.1 and 4.3 times higher than the general population, respectively (Table 6). Logistic Regression Analysis A logistic regression model was used to model the probability of having an additional non-NET after the diagnosis
3426
R. M. Kauffmann et al.
TABLE 7 Logistic regression model predicting risk of additional non-NET malignancy after diagnosis of PNET Variable
OR (95 % CI)
Age (72:51)
2.08 (1.09–3.93)a
Race (black/white)
2.19 (0.85–5.67)
Race (other/white)
1.20 (0.35–4.17)
Marital status (married/unmarried)
4.0 (1.59–10.0)a
Histologic stage (localized/regional)
2.56 (1.03–6.25)a
Histologic stage (localized/distant)
3.45(1.61–7.14)a
Histologic stage (localized/unstaged) Date of diagnosis (before/after 1990)
2.70 (0.71–10.0) 1.23 (0.66–2.29)
Gender (M/F)
1.15 (0.61–2.14)
Malignancy before PNET (Y/N)
0.93 (0.35–2.49)
a
Statistically significant
TABLE 8 Logistic regression model predicting risk of additional non-NET malignancy after diagnosis of GINET Variable
OR (95 % CI)
Age (72:51)
1.60 (1.32–1.93)a
Gender (M/F)
1.27 (1.11–1.45)a
Race (black/white)
1.08 (0.89–1.31)
Race (white/other)
1.39 (1.02–1.89)a
Marital status (married/unmarried)
1.18 (1.02–1.37)a
Histologic stage (localized/regional)
1.23 (1.04–1.47)a
Histologic stage (localized/distant) Histologic stage (localized/unstaged)
2.27 (1.82–2.78)a 2.04 (1.61–2.50)a
Date of diagnosis (before/after 1990)
1.19 (1.04–1.37)a
Malignancy before NET (Y/N)
0.91 (0.76–1.08)
a
Statistically significant
of either PNET or GINET. Covariates included age, gender, race, marital status, stage, year of diagnosis (before or after 1990), and whether the patient had an additional primary malignancy before the diagnosis of NET. Logistic regression model results for the PNET group are displayed in Table 7. Increased age at diagnosis, marital status, and localized disease were independent predictors of increased risk of an additional malignancy after diagnosis of NET. Logistic regression model results for the GINET group are displayed in Table 8. Increasing age, male gender, diagnosis before 1990, white race, marital status, and localized stage were associated with increased risk of an additional non-NET after the diagnosis of GINET. DISCUSSION The incidence of additional cancers in patients with PNET or GINET is approximately 26 %. The incidence of additional malignancies in patients with PNET is 15 %, whereas the incidence in patients with GINET is 34 %.
The pattern of observed additional malignancies correlates with the pattern of observed cancers in the United States. For patients with PNET, the incidence of several common cancers was lower than that expected. In patients with GINET, however, the incidence of colorectal, stomach, breast, lung, uterine, bladder, pancreatic, and esophageal malignancies exceeded the incidence seen in the general population, particularly for colorectal and gastric cancer. The increased risk of additional malignancies in patients with GINET compared with PNET may be explained by the indolent course of GINET compared with PNET, in that survival rates for patients with GINET are longer than that of patients with PNET, allowing more time to develop additional malignancies. Median survival in our PNET group was 2.58 years (95 % confidence interval (CI) 2.17–2.92 years), compared with 10.9 years (95 % CI 10.4–11.4 years) in the GINET group.14,15 It is likely that patients with previous GINET are maintained on a surveillance program in which they undergo regular esophagogastric duodenoscopy and colonoscopic evaluation. Although enrollment in surveillance programs or adherence to the recommended guidelines is impossible to ascertain from the SEER database, such surveillance would confound our results to the degree that these patients are monitored more closely than the general population and therefore have their second malignancies diagnosed more readily. This is particularly relevant to the increased rate of gastric and colorectal cancer observed in our GINET cohort, but it would not completely explain the increase in breast, lung, uterine, or pancreatic malignancies in this population, because patients with GINET would not necessarily be expected to have closer surveillance for these malignancies. The observation that more than half of the additional non-NET malignancies were diagnosed before the NET strengthens our conclusion that there is a true increase in the risk of non-NET malignancies in patients with GINET that is not related solely to closer surveillance. Furthermore, in multivariate regression analysis, a diagnosis of malignancy before NET was not associated with increased risk of subsequent non-NET cancer, thus suggesting that closer monitoring alone is not driving the association. It is also possible that the increased risk of a second cancer in patients with GINET is related to common genetic or tumorigenic pathways. The relationship between GINET and other malignancies in the setting of a family history of multiple endocrine neoplasia I, neurofibromatosis type 1, and von Hippel–Lindau disease is well known. There is a growing body of literature reporting an increased incidence of other malignancies in patients with GINET in the absence of a known genetic syndrome’s. This may indicate yet-unknown genetic mutations predisposing to multiple malignancies. These additional malignancies are
Add’l Cancers and Neuroendocrine Tumors
more commonly synchronous and gastrointestinal in origin but can be metachronous and occur anywhere.5–7 Tichansky et al.7 used the SEER database to evaluate the incidence of additional malignancies in 2,086 patients with colorectal carcinoid tumors and found an increased incidence of cancers of the colon/rectum, small bowel, esophagus/stomach, lung/bronchus, urinary tract, and prostate in these patients. Additional gastrointestinal malignancies were usually synchronous, whereas additional nongastrointestinal malignancies were more likely metachronous. A recent study from Kamp et al.8 reported that the incidence of synchronous, but not metachronous, malignancies was increased in patients with NET of the gastrointestinal tract and pancreas. The authors observed an increased risk of melanoma and cancers of the breast, colon/rectum, and prostate in their study population. However, their population was small, comprising only 459 patients, and this may account for the differences from our study. Tsai et al.16 used the Taiwan Cancer Registry to study the rate of second cancers in patients with NET of various sites and found an increased risk of second cancers after NET, particularly in patients diagnosed with NET after the age of 70 years. They concluded that close monitoring for a second malignancy is warranted in patients with a history of NET. However, their study did not distinguish risk of second cancers by site of primary NET.16 In multivariable regression analysis of patients with PNET or GINET, common factors associated with increased risk of developing additional non-NET malignancies included increased age, marital status, and localized disease. In patients with GINET, age, male gender, diagnosis before 1990, marital status, and localized disease were associated with increased risk. In PNET patients, increased age, marital status, and localized disease were associated with increased risk. As would be expected, the risk of malignancy increases with age, because the patient has had a longer lifespan to develop additional malignancies. It is unclear why marital status would increase risk of additional malignancies. Several studies have reported marital status to be associated with increased survival after cancer.17–19 Although this finding has not been completely explained, psychosocial support and the spouse’s influence in monitoring and directing healthrelated behavior have been suggested. If marital status translates to a higher likelihood of undergoing cancer screening, this may explain why married patients in our cohort had an increased risk of additional non-NET malignancies. Higher pathologic stage of the NET was associated with decreased risk of additional malignancies. This is likely because patients with higher-grade NET are more likely to die of their neuroendocrine disease and thus lack sufficient time in which to develop an additional non-NET malignancy. Similarly, those diagnosed before 1990 had a longer
3427
follow-up and therefore more time to develop another malignancy. It is unclear why white race was associated with an additional risk of a secondary malignancy. Whites have improved survival rates for many cancers, and therefore these patients may have a longer lifespan, thus allowing more time to develop additional cancers.20,21 Additionally, white individuals are more likely to undergo routine health screening for malignancy (i.e., colonoscopy or mammogram) than other races.22,23 There are several limitations to this study. Patients in the SEER database may move between regions. Therefore, our analysis could have missed patients who were diagnosed with a NET in one SEER region and then moved out of the region and developed another malignancy. Additionally, although SEER captures relatively specific data on primary cancers, other potential confounding factors, such as comorbidities, other risk factors for developing a cancer (such as dietary and environmental factors, presence of a genetic syndrome, smoking history, and previous radiation therapy), and details on screening and treatment, are not available. CONCLUSION In conclusion, the incidence of additional primary malignancies in patients with PNET and GINET approaches 26 %. Patients with GINET are at increased risk of developing cancers in nearly every site, whereas those with PNET actually have a lower risk of developing additional cancers, likely because of their decreased survival. These findings underscore the importance of increased surveillance for other malignancies and prompt the question of whether shorter screening intervals or earlier screening age for other malignancies (i.e., stomach and colorectal) in patients with GINET and no history of genetic syndrome are warranted. Prior to a recommendation for shorter screening intervals, large-scale studies are needed to determine feasibility, effectiveness, and cost. Additional study of the incidence of additional malignancies in patients with GINET should further investigate whether risk is related to anatomic location (foregut/midgut/hindgut) of the primary GINET. Finally, future studies are needed to discern common pathophysiologic signaling pathways involved in both NET and non-NET malignancies in the hopes of developing cost-effective screening procedures and common targets for treatment. ACKNOWLEDGMENT CTSA Grant RR024975.
Source of support: Vanderbilt NIH
REFERENCES 1. Yao JC, Hassan M, Phan A, et al. One hundred years after ‘‘carcinoid’’: epidemiology of and prognostic factors for
3428
2.
3.
4.
5. 6. 7.
8.
9. 10.
11.
12.
neuroendocrine tumors in 35,825 cases in the United States. J Clin Oncol. 2008;26:3063–72. Oberg K, Castellano D. Current knowledge on diagnosis and staging of neuroendocrine tumors. Cancer Metastasis Rev. 2011;30(Suppl. 1):3–7. Trentham-Dietz A, Newcomb PA, Nichols HB, et al. Breast cancer risk factors and second primary malignancies among women with breast cancer. Br Cancer Res Treat. 2007;105:195–207. Lubensky IA, Pack S, Ault D, et al. Multiple neuroendocrine tumors of the pancreas in von Hippel–Lindau disease patients. Am J Pathol. 1998;153:223–31. Habal N, Sims C, Bilchik AJ. Gastrointestinal carcinoid tumors and second primary malignancies. J Surg Oncol. 2000;75:310–6. Aggarwal G, Obideen K, Wehbi M. Carcinoid tumors: what should increase our suspicion? Clevel Clin J Med. 2008;75:849–55. Tichansky DS, Cagir B, Borrazzo E, et al. Risk of second cancers in patients with colorectal carcinoids. Dis Colon Rectum. 2002;45:91–7. Kamp K, Damhuis RA, Felders RA, de Herder WW. Occurrence of second primary malignancies in patients with neuroendocrine tumors of the digestive tract and pancreas. Endocr Relat Cancer. 2012;19:95–9. Surveillance, Epidemiology, and End Results. http://seer.cancer. gov. Accessed 15 April 2012. Surveillance, Epidemiology, and End Results (SEER) Program. http://www.seer.cancer.gov. Research data (1973–2008). National Cancer Institute, DCCPS, Surveillance Research Program, Surveillance Systems Branch, released April 2011, based on November 2010 submission. Riall TS, Stager VM, Nealon WH, et al. Incidence of additional primary cancers in patients with invasive intraductal papillary mucinous neoplasms and sporadic pancreatic adenocarcinomas. J Am Coll Surg. 2007;204:803–13. US Cancer Statistics Working Group. http://www.apps.nccd.cdc. gov/uscs/cancersbyraceandethnicity.aspx. United States Cancer
R. M. Kauffmann et al.
13. 14.
15.
16. 17.
18.
19.
20.
21.
22. 23.
Statistics: 1999–2008 Incidence and Mortality Web-based Report. Atlanta: US Department of Health and Human Services, Center for Disease Control and Prevention and National Cancer Institute. Accessed 27 November 2012. US Census. http://www.quickfacts.census.gov/qfd/states/00000. html. Accessed 14 June 2012. Shebani KO, Souba WW, Finkelstein DM, et al. Prognosis and survival in patients with gastrointestinal tract carcinoid tumors. Ann Surg. 1999;229:815–21. Fischer L, Kleef J, Esposito I, et al. Clinical outcome and longterm survival in 118 consecutive patients with neuroendocrine tumours of the pancreas. Br J Surg. 2008;95:627–35. Tsai HJ, Wu CC, Tsai CR, et al. Second cancers in patients with neuroendocrine tumors. PLoS One. 2013;8:e86414. Wang L, Wilson SE, Stewart DB, et al. Marital status and colon cancer outcomes in US Surveillance, Epidemiology, and End Results registries: does marriage affect cancer survival by gender and age? Can Epidemiol. 2011;35:417–22. Baine M, Sahak F, Lin C, et al. Marital status and survival in pancreatic cancer patients: a SEER based analysis. PLoS One. 2011;6:E21052. Zhu K, Weiss NS, Schwartz SM, et al. Assessing the relationship between marital status and cancer incidence: methodologic considerations. Cancer Causes Control. 1994;5:83–7. Silber JH, Rosenbaum PR, Clark AS, et al. Characteristics associated with differences in survival among black and white women with breast cancer. JAMA. 2013;310:389–97. Clegg LX, Li FP, Hankey BF, Chu K, Edwards BK. Cancer survival among US whites and minorities: a SEER (Surveillance, Epidemiology, and End Results) program population-based study. Arch Intern Med. 2002;161:1985–93. DeBarros M, Steele SR. Colorectal cancer screening in an equal access healthcare system. J Cancer. 2013;4:270–80. Peek ME, Han JH. Disparities in screening mammography. J Gen Intern Med. 2004;19:184–94.
