Skip to main content
NCBI home page
BJS Open logoLink to BJS Open
. 2025 Aug 14;9(4):zraf087. doi: 10.1093/bjsopen/zraf087

Association of chronic low-grade inflammation with adverse outcomes after gastrointestinal surgery: observational and Mendelian randomization study

Doruk Orgun 1,✉, Christina Ellervik 2,3,4,5, Henrik Enghusen Poulsen 6,7,8, Børge Grønne Nordestgaard 9, Ismail Gogenur 10,11,2, Ask Tybjærg Nordestgaard 12,2
PMCID: PMC12351454  PMID: 40810384

Abstract

Background

Although overt systemic inflammation immediately before gastrointestinal surgery has been associated with postoperative complications and mortality, it remains unclear whether baseline low-grade inflammation measured by high-sensitivity C-reactive protein (hs-CRP) in a non-surgery-related state is associated with the same outcomes.

Methods

This study included a subset of individuals from the Copenhagen General Population Study (CGPS) who underwent any type of gastrointestinal surgery between 2003 and 2015 after enrolment in the CGPS. Exposures were baseline hs-CRP levels (used in observational analyses) and two genetic variants that affect hs-CRP levels, namely interleukin 6 receptor (IL6R) rs4537545 and CRP rs1130864 (used in Mendelian randomization analyses), all of which were tested routinely at CGPS enrolment. Primary outcomes were 30-day complications and 90-day and 5-year mortality after the index surgery. Associations between exposures and outcomes were assessed using multivariable Cox regression models.

Results

Of the 107 536 individuals in the CGPS, 12 803 were included in the present study. Of these individuals, 1236 (9.7%) experienced 30-day complications, 865 (6.8%) died within 90 days, and 2789 (21.8%) died within 5 years. Adjusted hazard ratios for the highest hs-CRP quartile (hs-CRP ≥ 2.73 mg/l) versus the lowest quartile (hs-CRP < 1.04 mg/l) were 1.19 (95% confidence interval (c.i.) 1.02 to 1.40; P = 0.029) for 30-day complications, 1.29 (95% c.i. 1.07 to 1.57; P = 0.009) for 90-day mortality, and 1.17 (95% c.i. 1.06 to 1.31; P = 0.003) for 5-year mortality. Sensitivity analyses restricted to those with hs-CRP measurements within 1 year before surgery had larger point estimates. Genetically predicted elevations in hs-CRP were not associated with any outcomes.

Conclusion

Baseline hs-CRP levels ≥ 2.73 mg/l, consistent with chronic low-grade systemic inflammation, were associated with higher risk of 30-day complications, 90-day mortality, and 5-year mortality after gastrointestinal surgery.

Keywords: high-sensitivity C-reactive protein, postoperative adverse outcomes, single nucleotide polymorphisms

This study investigated associations of postoperative complications and short- and long-term mortality with chronic low-grade systemic inflammation at baseline, measured by high-sensitivity C-reactive protein (hs-CRP), and variations in interleukin 6 receptor (IL6R) rs4537545 or C-reactive protein (CRP) rs1130864, as genetic markers of hs-CRP levels, patients who underwent gastrointestinal surgery and were enrolled in the Copenhagen General Population Study, a large prospective study in Denmark. hs-CRP concentrations ≥ 2.73 mg/l were associated with a higher relative risk of postoperative complications and both short- and long-term mortality, but genetically predicted hs-CRP elevations, as determined by IL6R rs4537545 and CRP rs1130864 variants, separately or combined into an allele score, were not, possibly because they only explained a minor proportion of the total CRP variance in the cohort.

Introduction

Overt systemic inflammation in the immediate preoperative period is associated with an amplified postoperative inflammatory response and adverse outcomes after major surgery1–4. Specifically, preoperative elevations of inflammation markers, such as C-reactive protein (CRP) and interleukin (IL)-6, have been associated with postoperative complications or death5–10. However, it is unclear whether these associations also apply to chronic low-grade systemic inflammation in an asymptomatic, non-surgery-related baseline state.

Low-grade inflammation is described as the chronic production, but a low-grade state, of inflammatory factors11. Traditional CRP assays are not sensitive to minor variations in CRP concentrations associated with low-grade inflammation12. In contrast, sensitive CRP assays, such as the high-sensitivity CRP (hs-CRP) assay, can assess chronic low-grade inflammation13. Importantly, hs-CRP concentrations remain stable over time, and thus one hs-CRP measurement at any given time is a reliable marker of future hs-CRP levels if the individual is not experiencing an overt inflammatory event, such as infection, at the time of measurement14.

The immune–inflammatory response to surgery varies among individuals, and it has been hypothesized that genetic factors contribute to this variation15. Both baseline IL-6 and CRP levels are affected by heritable genetic variants16–18. The minor frequency allele (T) of the rs4537545 single-nucleotide polymorphism (SNP) in the IL-6 receptor (IL6R) gene disrupts IL-6 receptor signalling, leading to higher levels of IL-6 and lower levels of CRP16,17. In contrast, the minor frequency allele (T) of the rs1130864 SNP in the CRP gene leads to increased CRP levels in the body18,19. A Mendelian randomization study design can use these genetic variants as surrogate markers of baseline hs-CRP levels in an asymptomatic non-disease state by taking advantage of the fact that germline polymorphisms are inherited randomly at gamete formation. Thus, associations between hs-CRP levels predicted by these genetic variants and outcomes are less likely to be affected by confounders and cannot be explained by reverse causation20.

The primary aim of this study was to investigate whether elevated baseline hs-CRP levels, as a marker of chronic low-grade inflammation, are associated with postoperative complications and mortality following gastrointestinal surgery in a large Danish general population cohort from the Copenhagen General Population Study (CGPS). A secondary aim was to investigate whether variations in IL6R rs4537545 or CRP rs1130864, as genetic markers of hs-CRP levels and thus chronic low-grade inflammation, are associated with postoperative adverse outcomes in the same cohort.

Methods

The study was approved by Region Zealand, by an institutional review board at Copenhagen University Hospital—Herlev and Gentofte (Reference no. H-KF-01-144/01) and was conducted in accordance with the Declaration of Helsinki. The Danish Data Protection Agency approved access to the Danish registries. Written informed consent was obtained from all participants. The results are reported in compliance with the STROBE guidelines and MR-STROBE21.

Data sources

The CGPS, an ongoing prospective cohort study of individuals of Danish descent living in the Greater Copenhagen metropolitan area since 2003, was used as the primary data source22,23. At enrolment, participants complete health-related questionnaires and undergo physical examination and blood sampling for biochemical and genetic analyses. Participants are then followed using the Danish health registries. A list of variables extracted from the CGPS is presented in Table S1.

Information on surgical procedures, postoperative complications, and co-morbidities with corresponding admission/discharge dates was extracted from the Danish National Patient Registry, which includes information on all hospital contacts in Denmark since 197724. Dates and causes of death were extracted from the Danish National Causes of Death Registry25. Data from Danish registries are accurate and complete, and the risk of misclassification of procedure and diagnosis codes is low24,25. Data from all registries and the CGPS were combined using the Danish personal identification number, which is unique for every individual in Denmark26.

Study cohort

This study included participants aged at least 20 years in the CGPS who underwent a gastrointestinal surgical procedure, excluding thoracic oesophageal and perianal procedures, at any time point after their inclusion in the registry. Gastrointestinal surgical procedures were defined as procedures that were registered in the Danish National Patient Registry with a Nordic Medico-Statistical Committee (NOMESCO) Classification of Surgical Procedures 2005 code of KJ. For patients who underwent more than one procedure after enrolment in the CGPS, the first recorded non-reoperation procedure was defined as the index surgery. Patients who did not undergo genotyping for IL6R rs4537545 and CRP rs1130864, as well as patients who did not have a baseline hs-CRP measurement at the time of enrolment into the CGPS, were excluded. Individuals were followed up from CGPS enrolment (November 2003–April 2015) until death or the end of the follow-up period in December 2021.

Exposures (hs-CRP and genetic variants)

Plasma hs-CRP concentrations at enrolment in the CGPS were measured using a high-sensitivity assay with internal and external quality controls, as described previously27. These hs-CRP measurements were performed without a clinical indication and are therefore incidental, according to the individual’s enrolment in the CGPS. In this context, hs-CRP levels reflect an asymptomatic, non-surgery-related baseline state for each individual.

IL6R rs4537545 and CRP rs1130864 genotyping was done from whole-blood DNA with TaqMan assays27. These were then verified by DNA sequencing, and call rates following reruns were > 99%. For both SNPs, the allele known to be associated with CRP elevations (the C allele for IL6R rs4537545 and the T allele for CRP rs113086416–19,27) was chosen as the exposure allele and was referred to as the CRP-increasing allele. Conversely, the allele associated with lowest CRP levels was chosen as the reference. Therefore, for each SNP, a person could have from zero to two CRP-increasing alleles (one for each locus on the two homologous chromosomes), and from zero to four CRP-increasing alleles when the two SNPs were combined into an allele score of CRP-increasing alleles.

Outcomes

The primary outcomes of interest were any event of a postoperative complication within 30 days after the index surgery, 90-day postoperative mortality, and 5-year postoperative mortality. Postoperative complications included infections, bleeding, venous thromboembolism, reoperation, stroke, acute myocardial infarction, acute pancreatitis, or a novel event of atrial fibrillation. Infections included surgical site infections, pneumonia, urinary tract infections, and sepsis. All complications except reoperation were defined by at least one occurrence of the hospital-based International Classification of Diseases, 10th Revision (ICD-10) (Danish version) diagnosis code in the Danish National Patient Registry (Table S2)24. Reoperation was defined by the NOMESCO Classification of Surgical Procedures 2005 code of KJW, and a novel event of atrial fibrillation was defined as the first occurring diagnosis code entry for the condition.

Co-variables

Co-variables included sex, income level, body mass index (BMI), and smoking status assessed at the time of enrolment in the CGPS, with the assumption that these variables did not change considerably over the time span between inclusion in the CGPS and surgery. Information on income level and smoking status was gathered using self-reported questionnaires, whereas BMI was measured objectively at the time of CGPS enrolment.

Other included co-variables were patient age at the time of surgery, surgery type according to the subdivisions of gastrointestinal system by procedure code (appendix, biliary, gastroduodenal, hepatic, intestinal/colorectal, pancreatic, abdominal wall/hernia, splenic), open versus minimally invasive surgery (according to procedure code), emergency surgery (according to the ‘admission for procedure’ variable in the Danish National Patient Registry), perioperative gastrointestinal cancer diagnosis, time between enrolment into the CGPS and date of surgery, and co-morbidities at the time of surgery. Co-morbidities included ischaemic heart disease, previous acute myocardial infarction, heart failure, atrial fibrillation, previous stroke, chronic obstructive pulmonary disease, asthma, diabetes, previous deep vein thrombosis/pulmonary embolism, hypertension, cirrhosis, cancer history, chronic inflammatory disease/autoimmune disease/vasculitis, chronic pancreatitis, interstitial lung disease, and atherosclerotic peripheral arterial disease. All co-morbidities were captured by the hospital-based ICD-10 (Danish version) diagnosis codes before surgery in the Danish National Patient registry or by self-reported questionnaires at the time of CGPS enrolment.

Statistical analyses

The distributions of co-variables according to baseline hs-CRP quartiles (Q1–Q4) or allele status for the IL6R rs4537545 or CRP rs1130864 SNPs were compared using analysis of variance (ANOVA) or the Kruskal–Wallis test for continuous variables and Pearson’s χ2 test for categorical variables. Two-sided P < 0.050 was considered statistically significant.

For the two SNPs, deviation from Hardy–Weinberg equilibrium was assessed using a χ2 test. Then, a maximum-likelihood estimation method was used to assess for linkage disequilibrium between the SNPs. The correlation coefficient was expressed as r2, where r2 = 0 indicates no correlation and |r2|=1 indicates a perfect correlation between the SNPs28,29.

Observational analyses

Spearman’s rank correlation coefficient test was used to examine whether time between enrolment in the CGPS and the date of surgery correlated with baseline hs-CRP concentrations to indirectly examine whether the results could be confounded by this time span. Because very small correlation coefficients can still be statistically significant in large data sets, correlation coefficients < 0.10 were considered negligible clinical correlations a priori30.

Then, the study cohort was stratified by baseline hs-CRP concentration quartiles and Kaplan–Meier curves were used to examine absolute risk between hs-CRP quartiles for each outcome of interest. Pairwise comparisons of the curves were performed using the log-rank test with the Bonferroni–Holm method for P value adjustment. Adjusted hazard ratios (HRs) with corresponding 95% confidence interval (c.i.) values were calculated for the outcomes using multivariable Cox regression models, with the lowest hs-CRP quartile (Q1) as the reference. All Cox models were adjusted for all patient- and operative-specific co-variables known at the time of surgery because these were possible confounders.

Mendelian randomization analyses

First, to examine whether the genetic variants predicted hs-CRP levels in the study population, multiple linear regression models adjusted for sex and age were used to calculate changes in baseline hs-CRP concentrations for the IL6R rs4537545 and CRP rs1130864 genotypes.

Second, Kaplan–Meier curves were used to assess the rates of outcomes for each IL6R rs4537545 and CRP rs1130864 genotype separately and combined as an allele score of zero to four hs-CRP-increasing alleles. The curves were compared using the same methods as described above. Then, HRs with corresponding 95% c.i. values for each number of hs-CRP-increasing alleles for the IL6R rs4537545 and CRP rs1130864 SNPs and the combined allele score were calculated using multivariable Cox regression models adjusted for sex and age and with the lowest genetically predicted hs-CRP level (that is, zero hs-CRP-increasing alleles) as the reference.

Sensitivity analyses

To assess whether the associations between baseline hs-CRP levels and the outcomes of interest were affected by individuals who had concurrent overt inflammation at the time of enrolment in the CGPS, adjusted HRs were calculated following the exclusion of patients with either an hs-CRP measurement of > 10 mg/l or a diagnosis of infection up to 14 days before CGPS enrolment.

Then, to test whether the findings of the main analyses were influenced by the time between enrolment in the CGPS and surgery, adjusted HRs were calculated in a subgroup of individuals who underwent surgery within 1 year of CGPS enrolment (hs-CRP measurement).

Finally, post hoc sample size calculations were performed to estimate the number of individuals in hs-CRP Q1 and Q4 needed for each of the outcomes of interest using Fleiss’s formula with continuity correction.

All statistical analyses were performed using the R version 4.2.1 (R Foundation for Statistical Computing, Vienna, Austria). Linkage disequilibrium between the two SNPs was assessed using the LDpair Tool (National Institutes of Health, Bethesda, MD, USA)31.

Results

Among 107 536 individuals in the CGPS, 13 683 underwent gastrointestinal surgery at any time point after enrolment. In all, 264 individuals who did not have a baseline hs-CRP measurement and 616 who did not undergo genotyping for IL6R rs4537545 or CRP rs1130864 were excluded. The final study cohort included 12 803 individuals (median age 68 years, 46.9% male; Fig. 1), of whom 1236 (9.7%) had at least one postoperative complication within 30 days after surgery, 865 (6.8%) died within 90 days after surgery, and 2789 (21.8%) died within 5 years after surgery. The distribution of co-variables did not vary across genotypes for the two SNPs (Table 1), but did vary across measured hs-CRP quartiles (Table S3).

Fig. 1.

Fig. 1

Open in a new tab

Flow chart showing the study inclusion process

CGPS, Copenhagen General Population Study; hs-CRP, high-sensitivity C-reactive protein; IL6R, interleukin 6 receptor; CRP, C-reactive protein.

Table 1.

Descriptive statistics and co-variable distributions of the study cohort stratified by IL6R rs4537545 and CRP rs1130864 genotypes

IL6R rs4537545 CRP rs1130864 Total (n = 12 803)
C/C (n = 4356) C/T (n = 6221) T/T (n = 2226) P* C/C (n = 6041) C/T (n = 5541) T/T (n = 1221) P*
Baseline variables (at time of enrolment into CGPS)
 Sex
  Male 2062 (47.3%) 2906 (46.7%) 1042 (46.8%) 0.811 2837 (47.0%) 2595 (46.8%) 578 (47.3%) 0.342 6010 (46.9%)
  Female 2294 (52.7%) 3315 (53.3%) 1184 (53.2%) 3204 (53.0%) 2964 (53.2%) 643 (52.7%) 6793 (53.1%)
 Smoking status
  Never smoker 1587 (36.4%) 2265 (36.4%) 834 (37.5%) 0.491 2213 (36.6%) 2032 (36.7%) 441 (36.1%) 0.401 4686 (36.6%)
  Current smoker 907 (20.8%) 1267 (20.4%) 454 (20.4%) 1271 (21.0%) 1119 (20.2%) 238 (19.5%) 2628 (20.5%)
  Former smoker 1862 (42.7%) 2689 (43.3%) 938 (42.1%) 2557 (42.4%) 2360 (43.1%) 542 (44.4%) 5489 (42.9%)
 Annual income (DKK)
  < 200 000 623 (14.5%) 954 (15.6%) 349 (15.8%) 0.210 896 (15.0%) 834 (15.3%) 196 (16.3%) 0.468 1926 (15.2%)
  200 000–400 000 1126 (26.3%) 1683 (27.4%) 624 (28.3%) 1654 (27.7%) 1442 (26.4%) 337 (28.1%) 3433 (27.2%)
  400 000–800 000 1726 (40.2%) 2365 (38.6%) 829 (37.5%) 2299 (38.5%) 2173 (39.8%) 448 (37.3%) 4920 (39.0%)
  ≥ 800 000 814 (19.0%) 1132 (18.5%) 406 (18.4%) 1121 (18.8%) 1012 (18.5%) 219 (18.2%) 2352 (18.6%)
  Unspecified 67 87 18 71 80 21 172
 BMI (kg/m2), mean(SD) 26.9(4.6) 26.8(4.6) 26.9(4.4) 0.553 26.9(4.6) 26.9(4.6) 26.6(4.5) 0.099 26.9(4.6)
Surgery-specific variables
 Age at surgery (years), median (i.q.r.) 68 (58–75) 68 (58–76) 68 (58–76) 0.341 68 (58–75) 68 (58–75) 68 (58–76) 0.338 68 (58–76)
 Surgery type
  Appendix 338 (7.8%) 497 (8.0%) 180 (8.1%) 0.790 476 (7.9%) 441 (8.0%) 98 (8.0%) 0.429 1015 (7.9%)
  Biliary 669 (15.4%) 983 (15.8%) 334 (15.0%) 959 (15.9%) 861 (15.5%) 166 (13.6%) 1986 (15.5%)
  Gastroduodenal 242 (5.6%) 315 (5.1%) 128 (5.8%) 331 (5.5%) 289 (5.2%) 65 (5.3%) 685 (5.4%)
  Hepatic 106 (2.4%) 150 (2.4%) 60 (2.7%) 164 (2.7%) 133 (2.4%) 19 (1.6%) 316 (2.5%)
  Intestinal/colorectal 856 (19.7%) 1188 (19.1%) 414 (18.6%) 1151 (19.1%) 1073 (19.4%) 234 (19.2%) 2458 (19.2%)
  Pancreatic 40 (0.9%) 54 (0.9%) 19 (0.9%) 53 (0.9%) 51 (0.9%) 9 (0.7%) 113 (0.9%)
  Abdominal wall/hernia 2086 (47.9%) 3014 (48.4%) 1076 (48.3%) 2885 (47.8%) 2669 (48.2%) 622 (50.9%) 6176 (48.2%)
  Splenic 19 (0.4%) 20 (0.3%) 15 (0.7%) 22 (0.4%) 24 (0.4%) 8 (0.7%) 54 (0.4%)
 Minimally invasive surgery 2271 (52.1%) 3376 (54.3%) 1154 (51.8%) 0.062 3237 (53.6%) 2913 (52.6%) 651 (53.3%) 0.546 6801 (53.1%)
 Emergency surgery 1444 (33.1%) 2050 (33.0%) 767 (34.5%) 0.424 2006 (33.2%) 1866 (33.7%) 389 (31.9%) 0.470 4261 (33.3%)
 Perioperative gastrointestinal cancer diagnosis 788 (18.1%) 1090 (17.5%) 390 (17.5%) 0.731 1049 (17.4%) 991 (17.9%) 228 (18.7%) 0.498 2268 (17.7%)
 Time between CGPS enrolment and surgery (days), median (i.q.r.) 2339 (1175–3571) 2348 (1204–3564) 2432 (1192–3634) 0.424 2367 (1178–3601) 2350 (1218–3564) 2323 (1176–3575) 0.881 2355 (1196–3577)
Co-morbidities (at time of surgery)
 Ischaemic heart disease 562 (12.9%) 776 (12.5%) 278 (12.5%) 0.794 747 (12.4%) 712 (12.8%) 157 (12.9%) 0.714 1616 (12.6%)
 Previous AMI 192 (4.4%) 282 (4.5%) 93 (4.2%) 0.781 275 (4.6%) 235 (4.2%) 57 (4.7%) 0.663 567 (4.4%)
 Heart failure 202 (4.6%) 276 (4.4%) 94 (4.2%) 0.730 276 (4.6%) 241(4.3%) 55 (4.5%) 0.849 572 (4.5%)
 Atrial fibrillation 413 (9.5%) 585 (9.4%) 217 (9.7%) 0.894 565 (9.4%) 534 (9.6%) 116 (9.5%) 0.873 1215 (9.5%)
 Previous stroke 348 (8.0%) 511 (8.2%) 186 (8.4%) 0.848 513 (8.5%) 440 (7.9%) 92 (7.5%) 0.394 1045 (8.2%)
 COPD 271 (6.2%) 423 (6.8%) 143 (6.4%) 0.479 395 (6.5%) 352 (6.4%) 90 (7.4%) 0.429 837 (6.5%)
 Asthma 212 (4.9%) 351 (5.6%) 129 (5.8%) 0.147 325 (5.4%) 292 (5.3%) 75 (6.1%) 0.472 692 (5.4%)
 T1D 79 (1.8%) 117 (1.9%) 38 (1.7%) 0.872 100 (1.7%) 105 (1.9%) 29 (2.4%) 0.203 234 (1.8%)
 T2D 297 (6.8%) 399 (6.4%) 129 (5.8%) 0.280 403 (6.7%) 341 (6.2%) 81 (6.6%) 0.509 825 (6.4%)
 Previous DVT 178 (4.1%) 234 (3.8%) 82 (3.7%) 0.619 235 (3.9%) 209 (3.8%) 50 (4.1%) 0.857 494 (3.9%)
 Previous PE 98 (2.2%) 124 (2.0%) 43 (1.9%) 0.578 116 (1.9%) 115 (2.1%) 34 (2.8%) 0.150 265 (2.1%)
 Hypertension 1190 (27.3%) 1677 (27.0%) 623 (28.0%) 0.637 1652 (27.3%) 1520 (27.4%) 318 (26.0%) 0.602 3490 (27.3%)
 Cirrhosis 43 (1.0%) 68 (1.1%) 21 (0.9%) 0.780 61 (1.0%) 61 (1.1%) 10 (0.8%) 0.658 132 (1.0%)
 Cancer history 1524 (35.0%) 2186 (35.1%) 751 (33.7%) 0.477 2094 (34.7%) 1929 (34.8%) 438 (35.9%) 0.720 4461 (34.8%)
 Chronic inflammatory disease/autoimmune disease/vasculitis 421 (9.7%) 658 (10.6%) 231 (10.4%) 0.303 630 (10.4%) 531 (9.6%) 149 (12.2%) 0.050 1310 (10.2%)
 Chronic pancreatitis 33(0.8%) 62 (1.0%) 12 (0.5%) 0.102 54 (0.9%) 39 (0.7%) 14 (1.1%) 0.244 107 (0.8%)
 Interstitial lung disease 40 (0.9%) 63 (1.0%) 34 (1.5%) 0.059 60 (1.0%) 63 (1.1%) 14 (1.1%) 0.729 137 (1.1%)
 Atherosclerotic PAD 221 (5.1%) 271 (4.4%) 97 (4.4%) 0.192 283 (4.7%) 258 (4.7%) 48 (3.9%) 0.498 589 (4.6%)
Open in a new tab

Values are n (%) unless otherwise stated. IL6R, interleukin 6 receptor; CRP, C-reactive protein; CGPS, Copenhagen General Population Study; DKK, Danish krone; BMI, body mass index; SD, standard deviation; i.q.r., interquartile range; AMI, acute myocardial infarction; COPD, chronic obstructive pulmonary disease; T1D, type 1 diabetes; T2D, type 2 diabetes; DVT, deep vein thrombosis; PE, pulmonary embolism; PAD, peripheral arterial disease. P-values were generated by the analysis of variance (ANOVA) or the Kruskal–Wallis tests for continuous variables and Pearson’s χ2 test for categorical variables.

Observational analyses

There was no correlation between baseline hs-CRP concentrations and the time from CGPS enrolment to index surgery (r = 0.05). The cohort was stratified into quartiles according to baseline hs-CRP concentration as follows: Q1 (lowest), hs-CRP < 1.04 mg/l; Q2, hs-CRP ≥ 1.04 and < 1.55 mg/l; Q3, hs-CRP≥ 1.55 and < 2.73 mg/l; and Q4 (highest), hs-CRP ≥ 2.73 mg/l (Table S3).

Kaplan–Meier curves for the outcomes of interest according to baseline hs-CRP quartiles are shown in Fig. 2. Compared with Q1 (lowest) and Q2, Q3 and Q4 (highest) were associated with higher absolute risk (cumulative incidence) of 30-day complications (Q1: 8.3%; Q2: 8.1%; Q3: 9.9%; Q4: 12.4%; Fig. 2a), 90-day mortality (Q1: 5.1%; Q2: 5.5%; Q3: 6.8%; Q4: 9.6%; Fig. 2b), and 5-year mortality (Q1: 18.6%; Q2: 18.0%; Q3: 22.7%; Q4: 27.9%; Fig. 2c; log-rank P < 0.001 for all pairwise comparisons between Q3 or Q4 and Q1 and Q2).

Fig. 2.

Fig. 2

Open in a new tab

Kaplan–Meier curves for a 30-day postoperative complications, b 90-day mortality, and c 5-year mortality according to baseline hs-CRP quartiles (Q1–Q4)

hs-CRP, high-sensitivity C-reactive protein.

Compared with hs-CRP Q1 (lowest), Q4 (highest) was associated with an increased relative risk of all outcomes of interest, with corresponding adjusted HRs of 1.19 (95% c.i. 1.02 to 1.40; P = 0.029) for 30-day complications, 1.29 (95% c.i. 1.07 to 1.57; P = 0.009) for 90-day mortality, and 1.17 (95% c.i. 1.06 to 1.31; P = 0.003) for 5-year mortality (Fig. 3a). Furthermore, in the subgroup of individuals who underwent surgery within 1 year after enrolment in the CGPS (n = 1038), HRs for individuals in hs-CRP Q4 (highest) were 1.89 (95% c.i. 1.02 to 3.61; P = 0.039) for 30-day complications, 2.55 (95% c.i. 1.01 to 6.69; P = 0.041) for 90-day mortality, and 1.79 (95% c.i. 1.16 to 2.78; P = 0.009) for 5-year mortality (Fig. 3b).

Fig. 3.

Fig. 3

Open in a new tab

Adjusted hazard ratio (HR) estimates for 30-day postoperative complications, 90-day mortality, and 5-year mortality according to baseline hs-CRP quartiles (Q1–Q4)

a All subjects who underwent surgery after CGPS enrolment (12,803; primary analyses). b Individuals who underwent surgery within 1 year of CGPS enrolment (1038; sensitivity analyses). Cox regression models were adjusted for all patient- and operative-specific co-variables known at the time of surgery for all reported hazard ratio estimates. In all cases, hs-CRP Q1 was the reference group. Values in parentheses are 95% confidence intervals. hs-CRP, high-sensitivity C-reactive protein; CGPS, Copenhagen General Population Study.

Mendelian randomization analyses

Neither IL6R rs4537545 nor CRP rs1130864 deviated from Hardy–Weinberg equilibrium (P > 0.05) and were not in linkage disequilibrium (r2 = 0, P = 0.879). Each T-allele of the IL6R rs4537545 genotype was associated with lower CRP levels (F = 185; R2 = 0.14; P < 0.001), whereas each T-allele of the CRP rs1130864 genotype was associated with higher CRP levels (F = 189; R2 = 0.14; P < 0.001; Table S4). Compared with patients with zero CRP-increasing alleles (wild-type for IL6R rs4537545 and homozygote for CRP rs1130864), hs-CRP levels were estimated to be higher by 15.3% (95% c.i. 8.5 to 20.0%), 25.2% (95% c.i. 17.0 to 28.0%), 35.6% (95% c.i. 24.4 to 36.5%), and 46.3% (95% c.i. 28.4 to 47.6%) for one to four CRP-increasing alleles, respectively (F = 165; R2 = 0.14; P < 0.001; Table S4).

Kaplan–Meier curves for the outcomes according to IL6R rs4537545 genotype, CRP rs1130864 genotype, and CRP-increasing allele scores are shown in Fig. S1. No significant differences were found between the groups for any of the outcomes of interest (log-rank P > 0.05 for all comparisons). In line with this observation, the adjusted HRs for the outcomes did not differ significantly from 1.00 for any of the CRP-increasing genotypes of IL6R rs4537545 or CRP rs1130864, separately or combined into an allele score (P > 0.05 for all; Fig. 4).

Fig. 4.

Fig. 4

Open in a new tab

Adjusted hazard ratio (HR) estimates for 30-day postoperative complications, 90-day mortality, and 5-year mortality for IL6R rs4537545 genotypes, CRP rs1130864 genotypes, and CRP-increasing allele scores

Cox regression models were adjusted for sex and age for all reported HR estimates. The genotype or allele score associated with the lowest median CRP level was used as the reference. Values in parentheses are 95% confidence intervals. HR, hazard ratio; IL6R, interleukin 6 receptor; CRP, C-reactive protein; hs-CRP, high-sensitivity C-reactive protein.

Sensitivity analyses

Adjusted HR estimates were similar when patients with an hs-CRP measurement of > 10 mg/l or infection at the time of CGPS enrolment were excluded (530 patients; Table S5). Post hoc sample size calculations for all outcomes demonstrated that the number of individuals in hs-CRP Q1 and Q4 was sufficient to estimate a statistical power of 80% (two-tailed, α=0.05; Table S6).

Discussion

In this study of 12 803 individuals from the Danish general population who underwent gastrointestinal surgery, hs-CRP concentrations ≥ 2.73 mg/l in a non-surgery-related baseline state were associated with a higher relative risk of complications and mortality after surgery. In contrast, genetically predicted hs-CRP elevations, as determined by the IL6R rs4537545 and CRP rs1130864 variants, separately or combined into an allele score, were not associated with the same outcomes.

CRP is a biomarker of IL-1 and IL-6 mediated inflammation. Although IL-6 measurements are limited to conditions such as sepsis32, CRP is commonly measured in clinical practice32. In previous studies33–37, elevated preoperative CRP levels at admission were shown to be associated with postoperative morbidity and mortality in patients with colorectal or lung cancer, as well as in patients undergoing hip fracture surgery. In addition, a poor Glasgow Prognostic Score, which is based on preoperative CRP and albumin levels, has been associated with infections and poor survival after various cancer resections38–43. Although numerous genetic variants are known to predict CRP levels44, associations between elevated CRP levels and postoperative adverse outcomes have not been tested using genetics. In a study of 604 patients undergoing coronary artery bypass surgery45, the minor allele of the CRP rs1800947 SNP (not investigated in the present study) was correlated with preoperative and postoperative CRP levels, but the associations with postoperative outcomes were not reported.

In the present study, rather than associations for elevated preoperative CRP levels per se, the associations for chronic low-grade systemic inflammation, as indicated by elevated baseline hs-CRP levels, were studied. The word ‘baseline’ here signifies a non-surgery-related state because hs-CRP was measured incidentally and not because of clinical indication. Even though an overt inflammatory event immediately before surgery would influence postoperative outcomes, it is not expected that this would affect the associations between baseline inflammation and outcomes because such an event would occur after baseline hs-CRP measurement. Nevertheless, the multivariable Cox regression models were adjusted for variables such as emergency surgery and perioperative GI cancer diagnosis because these could have acted as proxies for overt inflammatory events before surgery. Here, in line with previous findings for overt preoperative CRP elevations, elevated baseline hs-CRP levels were associated with an increased risk of adverse postoperative outcomes. However, genetically predicted hs-CRP elevations were not associated with the same outcomes. Therefore, it seems less likely that chronic low-grade inflammation is a causal risk factor for adverse postoperative outcomes, and the observed associations with elevated measured hs-CRP levels could be explained by residual confounding by chronic diseases that were not fully accounted for in the analyses. In contrast, if low-grade inflammation is indeed causally associated with the risk of adverse postoperative outcomes, hs-CRP measurements could help identify patients at high risk of postoperative mortality and morbidity who could benefit from preventive measures.

The strengths of this study include the analysis of a large cohort from the general population of Denmark who underwent gastrointestinal surgery without any loss to follow-up. The study population was ethnically homogeneous, and the genotype distributions were in Hardy–Weinberg equilibrium; thus, genotyping or population stratification errors are unlikely. Because only individuals of Danish descent were included in the study, the results may not be applicable to other ethnicities; however, there is no current evidence to suggest this.

Study limitations include the fact that baseline hs-CRP measurements were taken at varying time points before surgery. Reassuringly, no correlation between hs-CRP values and time to surgery was observed, and the Cox regression models were adjusted for time between hs-CRP measurement and surgery. Moreover, the results were consistent when participants with > 1 year between CGPS enrolment and surgery were excluded. Furthermore, because the focus was on hs-CRP elevations as a measure of chronic low-grade inflammation, and because hs-CRP is stable over time, the time between blood sampling and surgery is likely of lower importance. Although a sensitivity analysis of patients who underwent surgery within 1 year after hs-CRP measurement demonstrated higher risk increases than the primary analyses, these results were possibly biased by the actual indication of the (upcoming) surgery event. Moreover, the point estimates could have been higher due to the wider confidence intervals in this subgroup analysis, which was possibly due to the smaller number of events. Although there may be residual confounding in the observational analyses, reverse causation is unlikely given that surgery was after inclusion into the CGPS. In addition, the results could have been influenced by individuals with elevated CRP levels due to underlying overt inflammation around the time of hs-CRP measurement, although the results did not change when individuals with baseline hs-CRP levels > 10 mg/l were excluded. In addition, because only diagnoses from hospital contacts are registered in Danish registries, the incidence of mild postoperative complications that were only treated in the primary care sector could have been underestimated.

Finally, although the two SNPs strongly predicted CRP levels (F > 10 for both variants), they only explained a minor proportion of the total variance in CRP (R2 = 14% for the combined SNPs). Thus, it is difficult to exclude the possibility that the Mendelian randomization results are explained by type II errors. To overcome this limitation, future studies may benefit from the inclusion of additional genetic variants that predict hs-CRP levels.

Elevated baseline hs-CRP levels, as a marker of chronic low-grade inflammation, were associated with a high risk of postoperative complications, as well as short- and long-term mortality, after gastrointestinal surgery in a large Danish general population cohort. In contrast, baseline hs-CRP elevations determined by germline variations in IL6R rs4537545 and CRP rs1130864 were not associated with complications and mortality in the same cohort. Further studies based on larger populations or using more genetic variants are needed to examine associations between genetically predicted hs-CRP and postoperative adverse outcomes.

Supplementary Material

zraf087_Supplementary_Data
zraf087_supplementary_data.docx (645.9KB, docx)

Contributor Information

Doruk Orgun, Center for Surgical Science, Department of Surgery, Zealand University Hospital Køge, Køge, Denmark.

Christina Ellervik, Department of Clinical Medicine, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark; Department of Pathology, Harvard Medical School, Boston, Massachusetts, USA; Department of Laboratory Medicine, Boston Children’s Hospital, Boston, Massachusetts, USA; Department of Clinical Biochemistry, Zealand University Hospital, Køge, Denmark.

Henrik Enghusen Poulsen, Department of Clinical Medicine, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark; Department of Endocrinology, Copenhagen University Hospital Bispebjerg and Frederiksberg, Copenhagen, Denmark; Department of Cardiology, Copenhagen University Hospital Hillerød, Copenhagen, Denmark.

Børge Grønne Nordestgaard, Department of Clinical Biochemistry, Copenhagen University Hospital—Herlev and Gentofte, Herlev, Denmark.

Ismail Gogenur, Center for Surgical Science, Department of Surgery, Zealand University Hospital Køge, Køge, Denmark; Department of Clinical Medicine, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark.

Ask Tybjærg Nordestgaard, Department of Clinical Biochemistry, Copenhagen University Hospital—Herlev and Gentofte, Herlev, Denmark.

Funding

D.O. was funded by the Edith and Henrik Henriksens Memorial Fund. C.E. was partially funded by the Laboratory Endowment Fund at Boston Children’s Hospital (Boston, MA, USA).

Author contributions

Doruk Orgun (Conceptualization, Data curation, Formal analysis, Methodology, Visualization, Writing—original draft), Christina Ellervik (Conceptualization, Methodology, Supervision, Writing—review & editing), Henrik Poulsen (Conceptualization, Methodology, Supervision, Writing—review & editing), Børge Nordestgaard (Methodology, Supervision, Writing—review & editing), Ismail Gogenur (Conceptualization, Funding acquisition, Methodology, Supervision, Writing—review & editing), and Ask Nordestgaard (Conceptualization, Formal analysis, Methodology, Supervision, Writing—review & editing)

Disclosure

The authors declare no conflict of interest.

Supplementary material

Supplementary material is available at BJS Open online.

Data availability

Data sharing is not applicable to this study according to Statistics Denmark’s guidelines. The data used in this study are stored on Statistics Denmark’s servers. The users signed special confidentiality and non-disclosure agreements in advance and performed their analyses in a secure analysis environment authorized by Statistics Denmark. For more information, please check Statistics Denmark’s website: (https://www.dst.dk/Site/Dst/SingleFiles/GetArchiveFile.aspx?fi=formid&fo=dataconfidentiality–pdf&ext={2}).

References

  • 1. Matsui  R, Inaki  N, Tsuji  T, Fukunaga  T. Preoperative chronic inflammation is a risk factor for postoperative complications independent of body composition in gastric cancer patients undergoing radical gastrectomy. Cancers (Basel)  2024;16:833. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2. Xiang  J, He  L, Li  D, Wei  S, Wu  Z. Value of the systemic immune-inflammation index in predicting poor postoperative outcomes and the short-term prognosis of heart valve diseases: a retrospective cohort study. BMJ Open  2022;12:e064171. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3. Yoon  J, Jung  J, Ahn  Y, Oh  J. Systemic immune-inflammation index predicted short-term outcomes in patients undergoing isolated tricuspid valve surgery. J Clin Med  2021;10:4147. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Yoneda  A, Ogata  R, Ryu  S, Yoshino  K, Fukui  S, Kugiyama  T  et al.  Prognostic value of systemic inflammation score in patients with esophageal cancer. Ann Med Surg (Lond)  2024;86:3852–3855 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5. Alsaif  SH, Rogers  AC, Pua  P, Casey  PT, Aherne  GG, Brannigan  AE  et al.  Preoperative C-reactive protein and other inflammatory markers as predictors of postoperative complications in patients with colorectal neoplasia. World J Surg Oncol  2021;19:74. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Tian  H, Jiang  X, Duan  G, Chen  J, Liu  Q, Zhang  Y  et al.  Preoperative inflammatory markers predict postoperative clinical outcomes in patients undergoing heart valve surgery: a large-sample retrospective study. Front Immunol  2023;14:1159089. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7. Procházka  V, Lacina  L, Smetana  K  Jr, Svoboda  M, Skřivanová  K, Beňovská  M  et al.  Serum concentrations of proinflammatory biomarker interleukin-6 (IL-6) as a predictor of postoperative complications after elective colorectal surgery. World J Surg Oncol  2023;21:384. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8. Neff  TA, Braun  J, Rana  D, Puhan  M, Filipovic  M, Seeberger  M  et al.  Interleukin-6 is an early plasma marker of severe postoperative complications in thoracic surgery: exploratory results from a substudy of a randomized controlled multicenter trial. Anesth Analg  2022;134:123–132 [DOI] [PubMed] [Google Scholar]
  • 9. Kaufmann  KB, Heinrich  S, Staehle  HF, Bogatyreva  L, Buerkle  H, Goebel  U. Perioperative cytokine profile during lung surgery predicts patients at risk for postoperative complications—a prospective, clinical study. PLoS One  2018;13:e0199807. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10. van der Laan  P, van der Graaf  WTA, Reijers  SJM, Schrage  YM, Hendriks  JJH, Haas  RL  et al.  Elevated preoperative serum interleukin-6 level is predictive for worse postoperative outcome after soft tissue sarcoma surgery. Eur J Surg Oncol  2023;49:106926. [DOI] [PubMed] [Google Scholar]
  • 11. Kim  J, Yoon  S, Lee  S, Hong  H, Ha  E, Joo  Y  et al.  A double-hit of stress and low-grade inflammation on functional brain network mediates posttraumatic stress symptoms. Nat Commun  2020;11:1898. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12. Wolska  A, Remaley  AT. CRP and high-sensitivity CRP: ‘what's in a name?’  J Appl Lab Med  2022;7:1255–1258 [DOI] [PubMed] [Google Scholar]
  • 13. Li  Y, Zhong  X, Cheng  G, Zhao  C, Zhang  L, Hong  Y  et al.  hs-CRP and all-cause, cardiovascular, and cancer mortality risk: a meta-analysis. Atherosclerosis  2017;259:75–82 [DOI] [PubMed] [Google Scholar]
  • 14. Glynn  RJ, MacFadyen  JG, Ridker  PM. Tracking of high-sensitivity C-reactive protein after an initially elevated concentration: the JUPITER study. Clin Chem  2009;55:305–312 [DOI] [PubMed] [Google Scholar]
  • 15. Verdonk  F, Einhaus  J, Tsai  AS, Hedou  J, Choisy  B, Gaudilliere  D  et al.  Measuring the human immune response to surgery: multiomics for the prediction of postoperative outcomes. Curr Opin Crit Care  2021;27:717–725 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16. Ahluwalia  TS, Prins  BP, Abdollahi  M, Armstrong  NJ, Aslibekyan  S, Bain  L  et al.  Genome-wide association study of circulating interleukin 6 levels identifies novel loci. Hum Mol Genet  2021;30:393–409 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17. Sarwar  N, Butterworth  AS, Freitag  DF. Interleukin-6 receptor pathways in coronary heart disease: a collaborative meta-analysis of 82 studies. Lancet  2012;379:1205–1213 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18. Brull  DJ, Serrano  N, Zito  F, Jones  L, Montgomery  HE, Rumley  A  et al.  Human CRP gene polymorphism influences CRP levels: implications for the prediction and pathogenesis of coronary heart disease [published correction appears in Arterioscler Thromb Vasc Biol 2004;24(7):1328]. Arterioscler Thromb Vasc Biol  2003;23:2063–2069 [DOI] [PubMed] [Google Scholar]
  • 19. Delongui  F, Lozovoy  MAB, Iriyoda  TMV, Costa  NT, Stadtlober  NP, Alfieri  DF  et al.  C-Reactive protein +1444CT (rs1130864) genetic polymorphism is associated with the susceptibility to systemic lupus erythematosus and C-reactive protein levels. Clin Rheumatol  2017;36:1779–1788 [DOI] [PubMed] [Google Scholar]
  • 20. Lawlor  DA, Harbord  RM, Sterne  JA, Timpson  N, Davey Smith  G. Mendelian randomization: using genes as instruments for making causal inferences in epidemiology. Stat Med  2008;27:1133–1163 [DOI] [PubMed] [Google Scholar]
  • 21. Skrivankova  VW, Richmond  RC, Woolf  BAR, Yarmolinsky  J, Davies  NM, Swanson  SA  et al.  Strengthening the reporting of observational studies in epidemiology using Mendelian randomization: the STROBE-MR statement. JAMA  2021;326:1614–1621 [DOI] [PubMed] [Google Scholar]
  • 22. Näslund-Koch  C, Nordestgaard  BG, Bojesen  SE. Increased risk for other cancers in addition to breast cancer for CHEK2*1100delC heterozygotes estimated from the Copenhagen General Population Study. J Clin Oncol  2016;34:1208–1216 [DOI] [PubMed] [Google Scholar]
  • 23. Nielsen  MB, Çolak  Y, Benn  M, Mason  A, Burgess  S, Nordestgaard  BG. Plasma adiponectin levels and risk of heart failure, atrial fibrillation, aortic valve stenosis, and myocardial infarction: large-scale observational and Mendelian randomization evidence. Cardiovasc Res  2024;120:95–107 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24. Schmidt  M, Schmidt  SA, Sandegaard  JL, Ehrenstein  V, Pedersen  L, Sørensen  HT. The Danish National Patient Registry: a review of content, data quality, and research potential. Clin Epidemiol  2015;7:449–490 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25. Helweg-Larsen  K. The Danish Register of Causes of Death. Scand J Public Health  2011;39:26–29 [DOI] [PubMed] [Google Scholar]
  • 26. Pedersen  CB. The Danish civil registration system. Scand J Public Health  2011;39:22–25 [DOI] [PubMed] [Google Scholar]
  • 27. Pedersen  KM, Çolak  Y, Ellervik  C, Hasselbalch  HC, Bojesen  SE, Nordestgaard  BG. Loss-of-function polymorphism in IL6R reduces risk of JAK2V617F somatic mutation and myeloproliferative neoplasm: a Mendelian randomization study. EClinicalMedicine  2020;21:100280. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28. Hill  WG, Robertson  A. Linkage disequilibrium in finite populations. Theor Appl Genet  1968;38:226–231 [DOI] [PubMed] [Google Scholar]
  • 29. Hill  WG. Estimation of linkage disequilibrium in randomly mating populations. Heredity (Edinb)  1974;33:229–239 [DOI] [PubMed] [Google Scholar]
  • 30. Schober  P, Boer  C, Schwarte  LA. Correlation coefficients: appropriate use and interpretation. Anesth Analg  2018;126:1763–1768 [DOI] [PubMed] [Google Scholar]
  • 31. Machiela  MJ, Chanock  SJ. LDlink: a web-based application for exploring population-specific haplotype structure and linking correlated alleles of possible functional variants. Bioinformatics  2015;31:3555–3557 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32. Molano Franco  D, Arevalo-Rodriguez  I, Roqué i Figuls  M, Montero Oleas  NG, Nuvials  X, Zamora  J. Plasma interleukin-6 concentration for the diagnosis of sepsis in critically ill adults. Cochrane Database Syst Rev  2019;4:CD011811. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33. Lopez-Pastorini  A, Riedel  R, Koryllos  A, Beckers  F, Ludwig  C, Stoelben  E. The impact of preoperative elevated serum C-reactive protein on postoperative morbidity and mortality after anatomic resection for lung cancer. Lung Cancer  2017;109:68–73 [DOI] [PubMed] [Google Scholar]
  • 34. McKechnie  T, Cloutier  Z, Archer  V, Park  L, Lee  J, Heimann  L  et al.  Using preoperative C-reactive protein levels to predict anastomotic leaks and other complications after elective colorectal surgery: a systematic review and meta-analysis. Colorectal Dis  2024;26:1114–1130 [DOI] [PubMed] [Google Scholar]
  • 35. Sun  Y, Peng  HP, Wu  TT. Postoperative C-reactive protein predicts postoperative delirium in colorectal cancer following surgery. Clin Interv Aging  2023;18:559–570 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36. Chen  BK, Liu  YC, Chen  CC, Chen  YP, Kuo  YJ, Huang  SW. Correlation between C-reactive protein and postoperative mortality in patients undergoing hip fracture surgery: a meta-analysis. J Orthop Surg Res  2023;18:182. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37. Pastorino  U, Morelli  D, Leuzzi  G, Rolli  L, Suatoni  P, Taverna  F  et al.  Baseline and postoperative C-reactive protein levels predict long-term survival after lung metastasectomy. Ann Surg Oncol  2019;26:869–875 [DOI] [PubMed] [Google Scholar]
  • 38. Moyes  LH, Leitch  EF, McKee  RF, Anderson  JH, Horgan  PG, McMillan  DC. Preoperative systemic inflammation predicts postoperative infectious complications in patients undergoing curative resection for colorectal cancer. Br J Cancer  2009;100:1236–1239 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39. Ishizuka  M, Kita  J, Shimoda  M, Rokkaku  K, Kato  M, Sawada  T  et al.  Systemic inflammatory response predicts postoperative outcome in patients with liver metastases from colorectal cancer. J Surg Oncol  2009;100:38–42 [DOI] [PubMed] [Google Scholar]
  • 40. Kobayashi  T, Teruya  M, Kishiki  T, Kaneko  S, Endo  D, Takenaka  Y  et al.  Inflammation-based prognostic score and number of lymph node metastases are independent prognostic factors in esophageal squamous cell carcinoma. Dig Surg  2010;27:232–237 [DOI] [PubMed] [Google Scholar]
  • 41. Polterauer  S, Grimm  C, Seebacher  V, Rahhal  J, Tempfer  C, Reinthaller  A  et al.  The inflammation-based Glasgow prognostic score predicts survival in patients with cervical cancer. Int J Gynecol Cancer  2010;20:1052–1057 [DOI] [PubMed] [Google Scholar]
  • 42. Knight  BC, Kausar  A, Manu  M, Ammori  BA, Sherlock  DJ, O'Reilly  DA. Evaluation of surgical outcome scores according to ISGPS definitions in patients undergoing pancreatic resection. Dig Surg  2010;27:367–374 [DOI] [PubMed] [Google Scholar]
  • 43. Nozoe  T, Iguchi  T, Egashira  A, Adachi  E, Matsukuma  A, Ezaki  T. Significance of modified Glasgow Prognostic Score as a useful indicator for prognosis of patients with gastric carcinoma. Am J Surg  2011;201:186–191 [DOI] [PubMed] [Google Scholar]
  • 44. Zacho  J, Tybjaerg-Hansen  A, Jensen  JS, Grande  P, Sillesen  H, Nordestgaard  BG. Genetically elevated C-reactive protein and ischemic vascular disease. N Engl J Med  2008;359:1897–1908 [DOI] [PubMed] [Google Scholar]
  • 45. Perry  TE, Muehlschlegel  JD, Liu  KY, Fox  AA, Collard  CD, Body  SC  et al.  C-Reactive protein gene variants are associated with postoperative C-reactive protein levels after coronary artery bypass surgery. BMC Med Genet  2009;10:38. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

zraf087_Supplementary_Data
zraf087_supplementary_data.docx (645.9KB, docx)

Data Availability Statement

Data sharing is not applicable to this study according to Statistics Denmark’s guidelines. The data used in this study are stored on Statistics Denmark’s servers. The users signed special confidentiality and non-disclosure agreements in advance and performed their analyses in a secure analysis environment authorized by Statistics Denmark. For more information, please check Statistics Denmark’s website: (https://www.dst.dk/Site/Dst/SingleFiles/GetArchiveFile.aspx?fi=formid&fo=dataconfidentiality–pdf&ext={2}).

Articles from BJS Open are provided here courtesy of Oxford University Press

RESOURCES