Race and Ethnicity, Stage-Specific Mortality, and Cancer Treatment in Esophageal and Gastric Cancers: Surveillance, Epidemiology, and End Results (2000–2018)

Differences in mortality due to gastroesophageal cancers persist among groups defined by race and ethnicity,1Rhome R.M. et al.J Surg Oncol. 2019; 119: 737-748Crossref PubMed Scopus (8) Google Scholar, 2Lemini R. et al.Anticancer Res. 2020; 40: 881-889Crossref PubMed Scopus (4) Google Scholar, 3Chen Z. et al.Oncotarget. 2017; 8: 47037-47051Crossref PubMed Scopus (34) Google Scholar and individuals of Asian ancestry have more favorable survival outcomes.4Huang R.J. et al.Cancer Epidemiol Biomarkers Prev. 2020; 29: 903-909Crossref PubMed Scopus (13) Google Scholar,5Kim J. et al.Ann Surg Oncol. 2009; 16 (2433–2341)Google Scholar To better understand this health disparity, we conducted a comprehensive assessment investigating associations of race and ethnicity with esophageal and gastric cancers considering a variety of prognostic factors. Cancer cases were identified from the Surveillance, Epidemiology, and End Results (SEER)-18 registries, 2000–2018.6National Cancer Institute. https://seer.cancer.gov/data-software/documentation/seerstat/nov2020/.Google Scholar Sociodemographic and clinical factors were assessed at diagnosis (Supplementary Table 1). Multivariable Cox proportional hazards models estimated hazard ratios for associations between mortality and the following race and ethnicity groups: non-Hispanic White (NHW), non-Hispanic Black (NHB), Hispanic, American Indian/Alaska Native, Asian/Pacific Islander (PI), and other/unknown. SEER-defined subgroups for people of Asian/PI ancestry included Indian/Pakistani, Chinese, Filipino, Hawaiian, Japanese, Korean, Vietnamese, PI/other Asian people, and other/unknown. To evaluate the role of treatment in mortality outcomes, we used the SEER-18 (2007–2011) specialized data set7National Cancer Institute.https://seer.cancer.gov/data-software/specialized.htmlGoogle Scholar containing additional information on medical insurance, census tract–level socioeconomic status, rurality index, and time from diagnosis to treatment. All individuals were followed for 5 or more years. The associations were further stratified by levels of a variable representing the county-level percentage of foreign-born individuals residing in the county of a case’s residence at diagnosis. All analyses were performed in SAS software, version 9 (SAS Institute) and a P value <.05 was considered statistically significant. From a total of 38,200 esophageal cancer cases, 34,352 deaths were reported (median follow-up time, 10 months). Compared with NHW individuals, NHB individuals showed higher risk of mortality (hazard ratio, 1.08; 95% CI, 1.04–1.12). Compared with NHW individuals, large reductions in risk of mortality were seen for individuals in the other/unknown race and ethnicity category (44%), followed by Asian individuals (Indian/Pakistani, 25%; Chinese, 14%; Figure 1A and Supplementary Table 2). Mortality risks were not significantly different for Hispanic, Korean, Filipino, Japanese, and PI/other Asian individuals compared with NHW. Similar results were observed for risk of esophageal cancer–specific death (Supplementary Table 2). When assessed by stage (Figures 1B–D), reduced mortality risk remained for Indian/Pakistani (distant, 31%) and Chinese (localized, 29%) individuals and increased risk remained for NHB individuals (regional, 9%; distant, 10%). Mortality risk estimates among those having received treatment (eg, chemotherapy, radiotherapy, or surgery) with complete covariate information (n = 10,143) were comparable with those seen in the total population, except for NHB, Korean, Chinese, and Asian Indian/Pakistani individuals, in whom it was no longer significant (data not shown). From a total of 71,914 gastric cancer cases, 59,327 deaths were reported (median follow-up time was 12 months). Compared with NHW individuals, nonsignificantly higher mortality risk was seen among American Indian/Alaskan native individuals (2%) (Figure 1E, Supplementary Table 2), and no differences were apparent for NHB, Native Hawaiian, or Japanese individuals. Individuals in the other/unknown category had the largest difference in risk (67% lower), followed by people of various Asian ethnicities (Indian/Pakistani, 32%; Korean, 25%; Chinese, 17%; Filipino, 13%; Vietnamese, 15%; PI/other Asian individuals, 11%) and Hispanic individuals (8%). All associations were similar for gastric cancer–specific death, except for Japanese individuals, who had a 9% significantly lower risk (Supplementary Table 2). In stage-stratified models (Figures 1F–H), NHB individuals diagnosed at the localized stage had 9% higher mortality risk. Mortality risk was significantly lower among Hispanic individuals (localized, 9%; regional, 13%), Asian individuals (Japanese [regional, 12%; distant, 10%], Filipino [localized, 27%; distant, 12%], Vietnamese [localized, 27%; regional, 17%], Chinese [localized, 12%; regional, 24%; distant, 12%], Korean [localized, 39%; regional, 33%], Indian/Pakistani [localized, 43%; distant, 18%], and PI/other Asian [regional, 17%]). Mortality risk estimates among those who received treatment (n = 17,100) were comparable with those seen in the total population, except those differences compared with NHW were no longer significant for Filipino and Vietnamese individuals and were now significantly inverse for NHB individuals (data not shown). We did not observe any differences in gastroesophageal cancer mortality by levels of percentage of foreign-born individuals (data not shown). Our study was based on large numbers of cancer cases (n = 110,114) with limited missing data on race and ethnicity. Notably, the detailed SEER-18 specialized data set enabled adjustment of several factors. Our results showed that, compared with NHW individuals, those with Asian ancestry had the lowest mortality risk after an esophageal cancer diagnosis and NHB individuals had a modest increase in mortality risk. Asian individuals with gastric cancer also had the lowest mortality risk, followed by Hispanic individuals. American Indian/Alaskan Native, and NHB individuals had a nonsignificantly higher risk. Within the Asian subgroups, the associations were strongest for Indian/Pakistani individuals, showing marked risk reduction for esophageal and gastric cancers compared with the other subgroups. Mortality risk varied across the stage and was lowest among Indian/Pakistani individuals with esophageal cancer in advanced stages and with localized and distant staged gastric cancer. A larger proportion of Indian/Pakistani individuals were younger at cancer diagnosis (Supplementary Table 1), but mortality was still significantly lower after adjusting for age and other prognostic factors. Indian/Pakistani individuals were not more likely to be diagnosed at localized stage or receive surgery, thus it could not explain the lower mortality observed. However, NHB individuals were less likely to receive surgery for esophageal adenocarcinoma (Supplementary Table 1), which may explain the higher mortality rates in this group. Indian/Pakistani individuals showed the largest mortality risk difference among individuals diagnosed with esophageal adenocarcinoma and gastric cardia (data not shown). In previous mortality assessments, Asian population groups had a better survival advantage from gastroesophageal cancers compared with other US racial and ethnic groups.5Kim J. et al.Ann Surg Oncol. 2009; 16 (2433–2341)Google Scholar,8Kim J. et al.Ann Oncol. 2010; 21: 152-260Abstract Full Text Full Text PDF PubMed Scopus (108) Google Scholar Our study revealed that the survival advantage differed across racial and ethnic subgroups. Heterogeneity in access to treatment, comorbidities, and individual-level socioeconomic status not fully captured in the SEER may explain these disparities. Differences in tumor biology and immune responses to treatment observed between Asian and non-Asian subgroups may contribute to the survival advantages.9Lin S.J. et al.Gut. 2015; 64: 1721-1731Crossref PubMed Scopus (156) Google Scholar In conclusion, our study presents evidence that Asian individuals, particularly Indian/Pakistani individuals, have a more favorable prognosis after esophageal or gastric cancer diagnosis compared with other racial and ethnic groups. Further research is needed to better understand gastroesophageal physiopathology among Asian subgroups. Our results also highlight the importance of disaggregating racial groups into specific ethnicities to identify the unique differences that exist within these subpopulations. Omonefe O. Omofuma, PhD (Conceptualization: Lead; Formal analysis: Lead; Methodology: Lead; Visualization: Lead; Writing – original draft: Lead; Writing – review & editing: Lead). Michael B. Cook, PhD (Conceptualization: Equal; Formal analysis: Supporting; Methodology: Supporting; Writing – original draft: Supporting; Writing – review & editing: Supporting). Christian C. Abnet, PhD (Methodology: Supporting; Visualization: Supporting; Writing – original draft: Supporting; Writing – review & editing: Supporting). M. Constanza Camargo, PhD (Conceptualization: Equal; Formal analysis: Supporting; Methodology: Equal; Supervision: Lead; Visualization: Supporting; Writing – original draft: Supporting; Writing – review & editing: Lead). SEER∗stat case listing sessions extracted the first primary cases of esophageal and gastric cancers with known age diagnosed from 2000–2013 (to allow for a minimal of 5 years of follow-up time) and were followed-up until death or censoring through December 31, 2018. Cancer morphology was defined by the International Classification of Diseases for Oncology 3rd edition (ICD-O-3)e1World Health Organization https://apps.who.int/iris/bitstream/handle/10665/96612/9789241548496_eng.pdfGoogle Scholar: 8140–8575 (for esophageal adenocarcinoma) and 8050–8084 (for esophageal squamous cell carcinoma). Gastric cancer anatomic sites were defined by International Classification of Diseases, 10th revisione2World Health Organization https://apps.who.int/iris/bitstream/handle/10665/246208/9789241549165-V3-eng.pdf?sequence=3&isAllowed=yGoogle Scholar for cardia (16.0), noncardia (16.1–16.6), and overlapping/unspecified (16.8–16.9). ICD-O-3 classified gastric cancer histology as intestinal (8143, 8144, 8210, 8211, 8221, 8260, 8261, 8262, 8263), diffuse (8141, 8142, 8145, 8490), and overlapping/unspecified. Cases of leukemia, lymphoma, mesothelioma, or Kaposi sarcoma (ICD-O-3: 9050–9055, 9140, and 9590–9989) were excluded. Sociodemographic factors assessed include age at diagnosis, race and ethnicity, sex, marital status, medical insurance, census tract–level socioeconomic status, rurality index (rural-urban commuting area), geographic area (Alaska, East, Northern plains, Pacific coast, and Southwest), and percentage of foreign-born persons in county of residence. Clinical factors include cancer stage (localized, regional, distant, or unknown) derived from collaborative stage for 2004+ and extent of disease from 1998–2003. Other factors include histology (intestinal, diffuse, or other or unspecified for gastric cancer, and adenocarcinoma or squamous cell carcinoma for esophageal cancer), anatomic site (for gastric cancer: cardia, noncardia, overlapping, or unspecified), cancer-directed surgery, year of cancer diagnosis, and time from diagnosis to treatment. Covariates included in multivariable regression model were selected through literature review, knowledge on factors influencing cancer mortality and forward stepwise selection using P values <.2 for entry and ≥.3 for exit. Using the SEER-18 specialized data sets from 2018 submission, we selected individuals who had received 1 or more of the following first course of treatment therapies: chemotherapy, radiotherapy, or cancer-directed surgery. A categorical treatment variable was created (chemotherapy, radiotherapy, cancer-directed surgery, or 1 or more treatments) and included in the multivariable-adjusted model along with age at diagnosis, sex, marital status, year of cancer diagnosis, rurality index, census tract–level socioeconomic status, geographic area, cancer stage, medical insurance, time from diagnosis to treatment, histology, and anatomic site (for gastric cancer). Proportional hazards assumption was assessed using the Schoenfeld’s residual test and a stratified Cox model was fitted for the categorical covariates violating the assumption.Supplementary Table 1Baseline Characteristics by Race and Ethnicity for Esophageal and Gastric Cancers Diagnosed From 2000 to 2013CharacteristicNHW (n = 29,044)NHB (n = 4,553)Hispanic (n = 2,681)AI/AN (n = 211)Asian Indian/Pakistani (n = 220)Chinese (n = 351)Esophageal cancer (n = 38,200) Age, y, median676265636670 Follow-up, mo, median117991411 Age, 15–49 y at diagnosis, n (%)2150 (7.4)469 (10.3)331 (12.3)25 (11.8)34 (15.4)21 (5.9) Male, n (%)23,129 (79.6)3193 (70.1)2190 (81.6)155 (73.4)132 (60)271 (77.2) Married, n (%)16,934 (58.3)1317 (28.9)1434 (53.4)89 (42.1)161 (73.1)247 (70.3) Year of diagnosis, 2010–2013, n (%)8751 (30.1)1182 (25.9)936 (34.9)62 (29.3)71 (32.2)102 (29) Localized stage, n (%)6635 (22.8)895 (19.6)492 (18.3)49 (23.2)38 (17.2)62 (17.6)SEER geographic area, n (%) Pacific coast12,023 (41.3)929 (20.4)1867 (69.6)96 (45.4)112 (50.9)317 (90.3) Southwest1223 (4.2)9 (0.2)232 (8.7)30 (14.2)4 (1.8)0 Northern plains3602 (12.4)525 (11.5)40 (1.5)6 (2.8)7 (3.2)5 (1.4) East12,196 (42.0)3090 (67.9)542 (20.2)13 (6.2)97 (44.1)29 (8.3)Cancer-directed surgery, n (%)8695 (29.9)679 (14.9)572 (21.3)53 (25.1)53 (24)68 (19.3)HistologyaICD-O-3 codese1: adenocarcinoma: 8140–8575; squamous cell carcinoma: 8050–8084., n (%) Adenocarcinoma21,499 (74.1)686 (15.1)1594 (59.5)118 (55.9)65 (29.5)67 (19) Squamous cell carcinoma7545 (25.9)3867 (84.9)1087 (40.5)93 (44.1)155 (70.5)284 (81)Survived ≥5 y, n (%)4886 (16.8)464 (10.1)357 (13.3)38 (18.0)46 (20.9)49 (13.9)Japanese (n = 363)Korean (n = 120)Filipino (n = 222)Other Asian/PI (n = 392)Other/unknown (n = 43) Age, y, median696667.56667 Follow-up, mo, median997919 Age, 15–49 y at diagnosis, n (%)17 (4.6)8 (6.6)22 (9.9)37 (9.4)4 (9.3) Male, n (%)279 (76.8)99 (82.5)167 (75.2)309 (78.8)32 (74.4) Married, n (%)225 (61.9)73 (60.8)145 (65.3)224 (57.1)18 (41.8) Year of diagnosis, 2010–2013, n (%)96 (26.4)44 (36.6)68 (30.6)137 (34.9)21 (48.8) Localized stage, n (%)68 (18.7)20 (16.6)42 (18.9)78 (19.8)14 (32.5)SEER geographic area, n (%) Pacific coast350 (96.4)99 (82.5)205 (92.3)352 (89.7)23 (53.4) Southwest1 (0.3)005 (1.3)1 (2.3) Northern plains2 (0.6)1 (0.8)02 (0.5)3 (7.0) East10 (2.8)20 (16.7)17 (7.7)33 (8.4)16 (37.2)Cancer-directed surgery, n (%)78 (21.4)22 (18.3)37 (16.6)65 (16.5)13 (30.2)Histology,aICD-O-3 codese1: adenocarcinoma: 8140–8575; squamous cell carcinoma: 8050–8084. n (%) Adenocarcinoma96 (26.4)14 (11.7)96 (43.3)101 (25.8)31 (72.1) Squamous cell carcinoma267 (73.6)106 (88.3)126 (56.7)291 (74.2)12 (27.9)Survived ≥5 y, n (%)46 (12.6)21 (17.5)30 (13.5)49 (12.5)19 (44.1)NHW (n = 38,668)NHB (n = 9907)Hispanic (n = 12,402)AI/AN (n = 609)Asian Indian/Pakistani (n = 467)Chinese (n = 2187)Japanese (n = 1878)Gastric cancer (n = 71,914) Age, y, median70666464617176 Follow-up, mo, median1111119211815 Age, 15–49 y at diagnosis, n (%)3502 (9)1445 (14.5)2768 (22.3)119 (19.5)126 (26.9)259 (11.8)80 (4.2) Male, n (%)24,622 (63.6)5513 (55.6)7009 (56.5)373 (61.2)278 (59.5)1,233 (56.3)1,010 (53.7) Married, n (%)22,316 (57.7)3696 (37.3)6938 (55.9)264 (43.3)332 (71)1,511 (69)1,147 (61) Year of diagnosis, 2010-2013, n (%)11,055 (28.5)2865 (28.9)4138 (33.3)202 (33.1)179 (38.3)6,76 (30.9)457 (24.3) Localized stage, n (%)10,038 (25.9)2602 (26.2)2718 (21.9)148 (24.3)130 (27.8)545 (24.9)520 (27.6)SEER geographic area, n (%) Pacific coast15,894 (41.1)2491 (25.1)9296 (74.9)167 (27.4)238 (50.9)1927 (88.1)1800 (95.8) Southwest1471 (3.8)43 (0.4)926 (7.5)196 (32.2)6 (1.3)7 (0.3)16 (0.8) Northern plains4421 (11.4)1166 (11.8)136 (1.1)13 (2.1)39 (8.4)24 (1.1)14 (0.7) East16,882 (43.7)6207 (62.7)2044 (16.5)27 (4.4)184 (39.4)229 (10.5)48 (2.6)Cancer-directed surgery, n (%)18,091 (46.7)4823 (48.6)5989 (48.2)271 (44.4)248 (53.1)1330 (60.8)1106 (58.8)Histology,bICD-O-3 codes: intestinal (8143, 8144, 8210, 8211, 8221, 8260, 8261, 8262, 8263); diffuse (8141, 8142, 8145, 8490). n (%) Intestinal2869 (7.4)948 (9.5)1273 (10.2)62 (10.1)33 (7.0)266 (12.1)501 (26.6) Diffuse7120 (18.4)1867 (18.8)3673 (29.6)147 (24.1)114 (24.4)544 (24.8)451 (24.0) Other/unspecified28,679 (74.1)7092 (71.5)7456 (60.1)400 (65.6)320 (68.5)1377 (62.9)926 (49.3)Subsite, n (%) Cardia15,213 (39.3)1226 (12.3)1777 (14.3)124 (20.3)103 (22)243 (11.1)239 (12.7) Noncardia14,538 (37.5)5709 (57.6)6842 (55.1)311 (51)230 (49.2)1445 (66)1173 (62.4) Overlapping/unspecified8917 (23)2972 (29.9)3783 (30.5)174 (28.5)134 (28.6)499 (22.8)466 (24.8)Survived ≥5 y, n (%)8598 (22.2)2332 (23.5)2781 (22.4)114 (18.7)159 (34.0)651 (29.7)504 (26.8)Korean (n = 2195)Filipino (n = 1088)Vietnamese (n = 944)Hawaiian (n = 282)Other Asian/PI (n = 1124)Other/unknown (n = 163) Age, y, median676867686661 Follow-up, mo, median281315101469 Age, 15–49 y at diagnosis, n (%)288 (13.1)121 (11.1)162 (17.1)38 (13.4)209 (18.5)32 (19.6) Male, n (%)1274 (58)557 (51.1)555 (58.7)162 (57.4)636 (56.5)90 (55.2) Married, n (%)1494 (68.0)690 (63.4)638 (67.5)144 (51)718 (63.8)58 (35.5) Year of diagnosis, 2010–2013, n (%)622 (28.3)328 (30.1)300 (31.7)78 (27.6)444 (39.5)65 (39.8) Localized stage, n (%)711 (32.3)267 (24.5)198 (20.9)57 (20.2)318 (28.2)69 (42.3)SEER geographic area, n (%) Pacific coast1815 (82.6)996 (91.5)836 (88.5)279 (98.9)931 (82.8)101 (61.9) Southwest12 (0.5)2 (0.2)9 (0.9)029 (2.6)3 (1.8) Northern plains18 (0.8)9 (0.8)16 (1.7)031 (2.8)4 (2.5) East350 (15.9)81 (7.4)83 (8.8)3 (1.1)133 (11.8)55 (33.7)Cancer-directed surgery, n (%)1494 (68.0)565 (51.9)550 (58.2)138 (48.9)598 (53.2)66 (40.4)Histology,bICD-O-3 codes: intestinal (8143, 8144, 8210, 8211, 8221, 8260, 8261, 8262, 8263); diffuse (8141, 8142, 8145, 8490). n (%) Intestinal406 (18.4)106 (9.7)79 (8.3)65 (23)134 (11.9)13 (7.9) Diffuse607 (27.6)291 (26.7)266 (28.1)71 (25.1)273 (24.2)21 (12.8) Other/unspecified1182 (53.8)691 (63.5)599 (63.4)146 (51.7)717 (63.7)129 (79.1)Subsite,cICD-10e2 codes: cardia (16.0); noncardia (16.1–16.6); overlapping/unspecified (16.8–16.9). n (%) Cardia114 (5.1)215 (19.7)80 (8.4)42 (14.8)122 (10.8)33 (20.2) Noncardia1564 (71.2)589 (54.1)644 (68.2)157 (55.6)682 (60.6)63 (38.6) Overlapping/unspecified517 (23.5)284 (26.1)220 (23.3)83 (29.4)320 (28.4)67 (41.1)Survived ≥5 y, n (%)844 (38.4)277 (25.4)255 (27.0)59 (20.9)324 (28.8)98 (60.1)AI/AN, American Indian or Alaskan Native.a ICD-O-3 codese1World Health Organization https://apps.who.int/iris/bitstream/handle/10665/96612/9789241548496_eng.pdfGoogle Scholar: adenocarcinoma: 8140–8575; squamous cell carcinoma: 8050–8084.b ICD-O-3 codes: intestinal (8143, 8144, 8210, 8211, 8221, 8260, 8261, 8262, 8263); diffuse (8141, 8142, 8145, 8490).c ICD-10e2World Health Organization https://apps.who.int/iris/bitstream/handle/10665/246208/9789241549165-V3-eng.pdf?sequence=3&isAllowed=yGoogle Scholar codes: cardia (16.0); noncardia (16.1–16.6); overlapping/unspecified (16.8–16.9). Open table in a new tab Supplementary Table 2Hazards Ratios and 95% CIs for All-Cause and Disease-Specific Mortality by Race and Ethnicity in Persons Diagnosed with Esophageal and Gastric Cancers, Surveillance, Epidemiology, and End Results-18, 2000–2013VariableAll deaths, n (%)HR (95% CI) of mortalityHRasex. (95% CI) of mortality by SEER clinical stageAll-cause mortalityasex.Disease-specific mortalityasex.Localized (n = 8393)bsex and marital status.Regional (n = 12,295)bsex and marital status.Distant (n = 13,645)cvariables in b and age at diagnosis.Unstaged (n = 3867)dage at diagnosis, sex and histology.Esophageal cancer, (n = 38,200) NHW25,970 (75.6)RefRefRefRefRefRef NHB4296 (12.5)1.08 (1.04–1.12)1.07 (1.03–1.12)1.08 (0.99–1.18)1.09 (1.02–1.17)1.10 (1.03–1.17)1.13 (1.01–1.26) AI/AN187 (0.5)1.09 (0.91–1.31)1.16 (0.96–1.41)1.06 (0.72–1.54)1.28 (0.91–1.81)1.02 (0.78–1.34)0.84 (0.47–1.48) Hispanic2399 (7.0)0.97 (0.93–1.02)0.97 (0.92–1.02)0.91 (0.81–1.01)1.00 (0.92–1.08)0.99 (0.93–1.06)0.97 (0.85–1.11) Japanese339 (1.0)0.97 (0.86–1.08)0.97 (0.86–1.09)1.04 (0.80–1.36)1.02 (0.83–1.24)0.89 (0.75–1.05)0.95 (0.67–1.35) Filipino200 (0.6)0.99 (0.86–1.14)0.94 (0.80–1.10)0.76 (0.54–1.08)1.04 (0.80–1.37)1.03 (0.84–1.27)0.97 (0.63–1.49) Korean106 (0.3)0.81 (0.67–0.99)0.90 (0.73–1.10)0.67 (0.40–1.14)0.92 (0.67–1.28)0.76 (0.55–1.05)1.39 (0.85–2.27) Chinese301 (0.9)0.86 (0.77–0.97)0.89 (0.79–1.01)0.71 (0.53–0.96)0.89 (0.73–1.08)0.84 (0.69–1.01)1.01 (0.73–1.40) Asian Indian/Pakistani176 (0.5)0.75 (0.64–0.88)0.71 (0.60–0.84)0.76 (0.51–1.11)0.80 (0.63–1.01)0.69 (0.53–0.91)0.68 (0.45–1.03) PI/other Asian355 (1.0)0.95 (0.86–1.07)0.94 (0.83–1.06)1.00 (0.78–1.28)0.96 (0.79–1.17)1.03 (0.87–1.21)0.67 (0.47–0.94) Other/unspecified23 (0.1)0.56 (0.36–0.87)0.36 (0.20–0.66)0.38 (0.14–1.01)0.55 (0.18–1.73)1.02 (0.51–2.04)0.42 (0.21–0.86)All-cause mortalityeage at diagnosis and sex.Disease-specific mortalityeage at diagnosis and sex.Localized (n = 18,321)fvariables in c, histology and anatomical site.Regional (n = 21,250)gage at diagnosis, sex, histology and anatomic site.Distant (n = 24,652)hage at diagnosis, sex, anatomic site.Unstaged (n = 7691)cvariables in b and age at diagnosis.Gastric cancer (n = 71,914) NHW32,918 (55.5)RefRefRefRefRefRef NHB8276 (14.0)1.00 (0.97–1.03)0.99 (0.96–1.02)1.09 (1.03–1.15)0.97 (0.92–1.02)0.98 (0.94–1.03)0.99 (0.92–1.07) AI/AN529 (0.9)1.02 (0.90–1.15)1.01 (0.88–1.16)0.99 (0.78–1.25)1.06 (0.86–1.31)1.04 (0.87–1.25)0.94 (0.67–1.34) Hispanic9802 (16.5)0.92 (0.90–0.95)0.91 (0.89–0.94)0.91 (0.86–0.97)0.87 (0.83–0.91)0.97 (0.93–1.00)0.87 (0.81–0.94) Hawaiian245 (0.4)0.97 (0.85–1.11)0.95 (0.82–1.10)0.76 (0.54–1.09)1.23 (0.99–1.52)0.90 (0.75–1.09)1.01 (0.59–1.72) Japanese1587 (2.7)0.95 (0.90–1.00)0.91 (0.85–0.97)1.08 (0.96–1.20)0.88 (0.81–0.97)0.90 (0.82–0.99)0.94 (0.79–1.12) Filipino853 (1.4)0.87 (0.81–0.93)0.83 (0.77–0.90)0.73 (0.61–0.87)0.89 (0.79–1.01)0.88 (0.79–0.98)0.82 (0.65–1.02) Korean1521 (2.6)0.75 (0.71–0.79)0.77 (0.72–0.81)0.61 (0.54–0.69)0.67 (0.62–0.74)0.91 (0.83–1.00)0.82 (0.70–0.97) Chinese1659 (2.8)0.83 (0.79–0.88)0.83 (0.79–0.88)0.88 (0.78–0.99)0.76 (0.70–0.83)0.88 (0.81–0.96)0.71 (0.61–0.83) Asian Indian/Pakistani301 (0.5)0.68 (0.61–0.77)0.69 (0.60–0.79)0.57 (0.42–0.77)0.83 (0.68–1.01)0.82 (0.69–0.98)0.36 (0.25–0.52) Vietnamese747 (1.3)0.85 (0.79–0.92)0.87 (0.80–0.95)0.73 (0.59–0.89)0.83 (0.73–0.94)0.91 (0.82–1.03)0.81 (0.64–1.03) PI/other Asian834 (1.4)0.89 (0.82–0.95)0.87 (0.80–0.94)0.85 (0.73–1.00)0.83 (0.73–0.95)0.97 (0.87–1.08)0.75 (0.60–0.92) Other/unspecified55 (0.1)0.33 (0.25–0.44)0.31 (0.22–0.44)0.18 (0.10–0.33)0.55 (0.30–0.99)1.00 (0.65–1.55)0.15 (0.08–0.28)AI/AN, American Indian/Alaskan Native.NOTE. All HRs adjusted for age at diagnosis, sex, marital status (including common law), year of cancer diagnosis, geographic area, stage (for non–stage-specific models), cancer-directed surgery, histology, anatomic site (for gastric cancer). Due to proportional hazards assumption violation all models were stratified by all variables except for:a sex.b sex and marital status.c variables in bsex and marital status. and age at diagnosis.d age at diagnosis, sex and histology.e age at diagnosis and sex.f variables in cvariables in b and age at diagnosis., histology and anatomical site.g age at diagnosis, sex, histology and anatomic site.h age at diagnosis, sex, anatomic site. Open table in a new tab AI/AN, American Indian or Alaskan Native. AI/AN, American Indian/Alaskan Native. NOTE. All HRs adjusted for age at diagnosis, sex, marital status (including common law), year of cancer diagnosis, geographic area, stage (for non–stage-specific models), cancer-directed surgery, histology, anatomic site (for gastric cancer). Due to proportional hazards assumption violation all models were stratified by all variables except for: