EP542 - Postoperative Serum Sodium as a Promising Low-Cost Biomarker for Anastomotic Leakage: More Evidence Needed – A Systematic Review and Meta-Analysis

Image or Table (Do Not add text here):

Purpose/Background: Background
Anastomotic leaks (AL) occur at rates up to 40%, leading to mortality rates of up to 27%. Nonosmotic stimuli may lead to increased vasopressin secretion with subsequent hyponatremia, which may serve as a low-cost biomarker to enable earlier detection of AL.

Methods/Interventions: Methods
A systematic review and meta-analysis was conducted using PRISMA guidelines and registered on PROSPERO (CRD42024522436). Cochrane Library, Embase, Pubmed, and Google Scholar were searched for relevant publications in English irrespective of timeframe. Studies were included if they analyzed cohorts with anastomosis of the small or large bowel that compared serum sodium levels or proportions of patients with hyponatremia with the development of anastomotic leaks. Two independent reviewers screened the studies, appraised quality using the QUADAS-2 tool, and extracted data using standardized tables. A third reviewer resolved conflicts. Meta-Mar v3.5.1 was used for meta-analysis.

Results/Outcomes: Results
The search strategy yielded 121 references, of which five were critically appraised. One study had a high risk of bias regarding patient selection, and the remaining had a low risk of bias in all domains. All five studies were included for review, tabulated in Table 1. The sample sizes ranged between 70 and 1,106 (median, 287), totaling 2,034 patients, with 188 developing anastomotic leakage (9%). The reported rates of AL were 7-37%, with mortality of 1-52%. All five studies recorded preoperative serum sodium levels but were only reported in three. Three studies investigated postoperative serum sodium. These measurements varied in timing and values used for statistical analysis were chosen by the investigators. Meta-analysis showed no significant difference in preoperative sodium levels (SMD: -0.45, 95% CI: -1.97 to 1.07, p = 0.33, I2 = 86.5%, tau = 0.57) or postoperative sodium levels (SMD: -3.61, 95% CI: -10.32 to 3.11, p = 0.15, I2 = 99.3%, tau = 2.69) (Fig. 1). However, postoperative sodium levels were significantly lower when reported in individual studies, and a drop in sodium levels was associated with AL in two of the studies, which concluded that postoperative hyponatremia is an independent predictive factor for AL. Three studies calculated sensitivities of 23-70% and specificities of 72-93% using variable definitions of hyponatremia between 130-139.5 mEq/dl.

Conclusion/Discussion: Conclusions
Further prospective double-blind observational studies with large sample sizes are warranted to conclusively establish whether AL leads to postoperative hyponatremia, as well as the value of tracking the changes in sodium levels as another predictive tool. A drop in serum sodium may be a helpful diagnostic marker even when above the lower limit of normal. Trials comparing CRP and serum sodium may also validate the use of sodium as an alternate biomarker.

Discussion
AL increases mortality by an estimated factor of 8-10 and doubles the length of hospital stay. AL may not become clinically evident until POD 8-12 due to the late development of symptoms that are often nonspecific. On average, AL is diagnosed on POD 6-12, and a delay, particularly after POD 5, is associated with poorer outcomes. The perianastomotic microenvironment is thought to be altered before it becomes clinically apparent, so biomarkers that reflect this change may be useful predictive tools. CRP and procalcitonin have been shown to have high negative predictive values for AL, and serum sodium may be a cheaper alternative.