Influence of blood transfusion on outcomes in patients with gastric cancer

Abstract

Chen et al's research provides valuable data supporting the cautious use of transfusions during gastric cancer surgery. However, to interpret causality, it must be acknowledged that recent tend-adjusted studies have consistently shown that the independent effect of transfusions may be smaller than that shown in unadjusted analyses. Future research should employ the following approaches: (1) Extended temporal characterization; (2) Functional immunological assessment; (3) Prospective designs incorporating detailed transfusion data; (4) Machine learning methods; and (5) Mechanistic studies. The relationship between transfusions and cancer treatment outcomes goes far beyond simple immunosuppression or inflammation. It reflects a complex interplay between patient vulnerability, surgical factors, and immune responses, requiring a comprehensive study across multiple biological levels and temporal dimensions.

Key Words: Blood; Cancer; Gastric; Immune response; Transfusion; Prognosis

Core Tip: The relationship between blood transfusion and cancer outcomes extends far beyond simple immunosuppression or inflammation. It represents a complex interplay of evolutionary biology, immunology, and modern medicine that demands sophisticated investigation. Only through comprehensive approaches integrating multiple biological scales and temporal dimensions can we transform blood transfusion from a necessary evil to a precision therapeutic intervention in surgical oncology, ultimately improving outcomes for the thousands of gastric cancer patients who undergo surgery each year.

TO THE EDITOR

We read with great interest the article by Chen et al[1] examining the impact of perioperative blood transfusion on inflammatory responses and long-term outcomes in gastric cancer patients. In their retrospective analysis of 200 patients, transfused patients demonstrated significantly lower overall survival (OS): 44.74 ± 4.25 months vs 49.52 ± 2.94 months (P < 0.001) and disease-free survival (DFS): 38.54 ± 5.54 months vs 43.85 ± 2.33 months (P < 0.001). Multivariate Cox regression confirmed transfusion as an independent predictor of worse OS [hazard ratio (HR) = 1.876, 95% confidence interval (95%CI): 1.293-2.723, P = 0.004] and DFS (HR = 1.644, 95%CI: 1.134-2.388, P = 0.014). Transfused patients also exhibited elevated inflammatory markers, including interleukin-6 (IL-6; 16.52 pg/mL vs 14.46 pg/mL, P = 0.023), tumor necrosis factor-alpha (TNF-α; 11.09 pg/mL vs 10.86 pg/mL, P = 0.003), and C-reactive protein (CRP; 8.22 mg/L vs 7.45 mg/L, P = 0.004) on postoperative day 1, with changes in these markers (CRP, IL-6, TNF-α) independently predicting survival. While these findings provide important clinical data, several aspects warrant further discussion to contextualize these results within the broader gastric cancer transfusion literature.

Contextualizing within gastric cancer-specific evidence

Chen et al's findings[1] align with the growing body of gastric cancer-specific literature examining transfusion-related outcomes. Agnes et al's comprehensive meta-analysis of 38 non-randomized studies reported pooled HR = 1.34 (95%CI: 1.23-1.45) for OS, 1.48 (95%CI: 1.18-1.86) for DFS, and an odds ratio (OR) of 3.33 (95%CI: 2.10-5.29) for postoperative complications in transfused patients[2]. Wang et al[3] conducted the largest meta-analysis to date, encompassing 51 studies with 41864 patients, and reported that perioperative blood transfusion was associated with worse 5-year OS (multivariate HR = 1.43, 95%CI: 1.24-1.63) and DFS (multivariate HR = 1.45, 95%CI: 1.16-1.82), alongside increased postoperative complications (OR = 2.30, 95%CI: 1.78-2.97) and severe Clavien-Dindo grade III-V complications (OR = 2.50, 95%CI: 1.71-3.63). More recently, Zhang et al[4] conducted a meta-analysis of 12 propensity-adjusted studies encompassing 17607 patients and found that while perioperative allogeneic blood transfusion was associated with worse OS in both gastric cancer (HR = 1.20, 95%CI: 1.08-1.32) and colorectal cancer, it did not correlate with DFS after propensity score adjustment. Importantly, their subgroup analyses suggested that unbalanced major postoperative complications may explain much of this survival difference.

The multicenter Puértolas et al[5] study specifically examined the synergistic effects of transfusion and infectious complications on inflammatory activation, demonstrating that the combination of both factors resulted in significantly greater neutrophil-to-lymphocyte ratio elevation and HR = 2.85 (95%CI: 1.64-4.95) for DFS. This suggests that Chen et al's observed inflammatory marker elevations may represent a compounding effect rather than transfusion alone[1]. Furthermore, the POWER4 cohort study by Ripollés-Melchor et al[6], which prospectively followed 386 gastrectomy patients across 72 Spanish hospitals, found that while anemia and transfusion were associated with adverse unadjusted outcomes (DFS event rates ranging from 13% in patients without anemia or transfusion to 38% in those with both), neither remained independently significant after multivariable adjustment. The authors concluded that preoperative anemia should be interpreted as a marker of patient vulnerability rather than a directly modifiable risk factor for recurrence.

The immunomodulation paradox: A nuanced interpretation

The authors interpret elevated inflammatory markers as detrimental inflammation. However, this may oversimplify the phenomenon of transfusion-related immunomodulation (TRIM). The simultaneous elevation of pro-inflammatory markers alongside stress hormones suggests a state of inflammatory-immunosuppressive dissociation, where systemic inflammation coexists with localized immunosuppression. This duality may explain why transfused patients experience both increased infection rates and potential tumor progression.

Recent work has clarified the mechanisms underlying TRIM. Goubran et al[7] demonstrated that transfusion carries multiple immunomodulatory mediators including residual leukocytes, apoptotic cells, soluble cytokines, and metabolically active extracellular vesicles that may directly stimulate tumor growth and angiogenesis. Supporting this, Wu et al's propensity-matched analysis of 4030 colorectal cancer patients showed transfusion remained an independent risk factor for both recurrence (HR = 1.41) and mortality (HR = 1.97), with a clear dose-response relationship[8].

The temporal dynamics and storage lesion question

While Chen et al[1] measured markers at days 1, 3, and 7 post-operatively, this temporal resolution may miss critical immunological windows. The period 10-14 days post-transfusion has been proposed as a time when TRIM effects may peak, though direct human evidence in gastric cancer remains limited. The absence of data during this period potentially underestimates the true impact of transfusion on tumor microenvironment remodeling and circulating tumor cell survival. Future studies should consider extended follow-up of inflammatory markers.

Importantly, the authors did not account for the storage duration of transfused red blood cell (RBC) units. The RBC storage lesion phenomenon, characterized by accumulation of pro-inflammatory cytokines, microparticles, and cell-free hemoglobin, follows a non-linear progression. Sut et al[9] reviewed how RBC storage induces physicochemical changes that affect transfused cell quality, functional integrity, and in vivo survival, noting that changes occurring in the first two weeks are partially reversible but become irreversible with extended storage. These storage-induced changes include accumulation of inflammatory mediators and extracellular vesicles that may contribute to TRIM. Tzounakas et al[[10] further elaborated that the functional interplay between donation-associated factors and recipient tumor biology, inflammation, and immune activation state may synergistically define the clinical impact of each transfusion. Units stored 14-21 days vs beyond 28 days may have fundamentally different immunological impacts, potentially confounding the observed associations in Chen et al's study[1].

Methodological considerations: The confounding cascade

The study's transfusion trigger (blood loss exceeding 15% or systolic blood pressure below 90 mmHg) represents a relatively liberal threshold. The 2023 AABB International Guidelines, based on 45 randomized trials encompassing over 20000 participants, now strongly recommend restrictive strategies with hemoglobin thresholds below 7 g/dL for hemodynamically stable adults, including oncologic patients[11]. This restrictive approach reduces transfusion exposure without increasing mortality. The discrepancy raises a critical question: Are the observed adverse outcomes attributable to transfusion itself, or do they reflect the underlying physiological derangement necessitating transfusion?

The authors attempted multivariate adjustment, but several confounders remain difficult to address. Higher blood loss may indicate more extensive tumor burden or technical difficulty not captured by TNM staging, creating surgical complexity bias. Additionally, hemodynamic instability itself serves as an independent prognostic factor. Surgeon experience also introduces variation in technique and decision-making that statistical adjustments cannot fully address.

The microbiome connection: Mechanistic considerations

The authors note increased infection rates in transfused patients but do not explore the potential role of gut microbiome disruption. While direct evidence linking allogeneic transfusion to microbiome alterations in gastric cancer patients is currently limited, several plausible mechanisms warrant investigation. Transfusion-associated iron delivery, inflammation-mediated alterations in intestinal permeability, and immunomodulatory effects on mucosal immunity may each contribute to dysbiotic changes.

Hajjar et al[12] demonstrated that gut dysbiosis contributes to anastomotic failure and tumor dissemination in colorectal surgery through modulation of mucosal proinflammatory cytokines. Their experimental work showed that mice transplanted with microbiota from patients with anastomotic leak exhibited poor healing, larger anastomotic tumors, and greater peritoneal dissemination. Profiling of gut microbiota revealed correlations between healing parameters and specific bacterial strains, with Alistipes onderdonkii promoting leakage while Parabacteroides goldsteinii improved healing through anti-inflammatory effects. Patients with anastomotic leak presented upregulation of mucosal MIP-1α, MIP-2, MCP-1, and IL-17A/F and higher circulating neutrophil and monocyte counts before surgery. Whether similar mechanisms operate in gastric cancer patients receiving transfusion represents a hypothesis requiring dedicated investigation with metagenomic analysis to elucidate whether transfusion-associated dysbiosis contributes to the inflammatory response and adverse outcomes observed by Chen et al[1].

Therapeutic implications: Toward precision transfusion

Chen et al's findings[1] suggest an opportunity for developing risk-stratified approaches to transfusion decision-making. A precision transfusion framework could incorporate pre-operative inflammatory status (baseline IL-6/CRP ratios), tumor molecular subtypes (microsatellite instability-high vs microsatellite stable), patient immunological phenotypes, and anticipated surgical complexity. Such scoring could identify patients who might benefit from alternative strategies including pre-operative optimization protocols, autologous blood conservation, or emerging hemoglobin-based oxygen carriers that lack immunological activity.

Conclusion and future directions

Chen et al[1] provide valuable data supporting cautious transfusion practices in gastric cancer surgery. However, interpreting causality requires acknowledging the consistent finding across recent propensity-adjusted studies that the independent effect of transfusion may be smaller than unadjusted analyses suggest[4,6]. We propose that future research should employ: (1) Extended temporal profiling: Daily inflammatory marker measurements through at least day 14, with expanded biomarker panels including damage-associated molecular patterns and resolution mediators (resolvins and protectins); (2) Functional immunological assessments: T-cell proliferation assays and NK cell cytotoxicity to provide mechanistic insights beyond biomarker elevation; (3) Prospective designs with granular transfusion data: Including unit storage duration, leukoreduction status, and precise timing relative to surgery; (4) Machine learning approaches: To identify patient subgroups with differential transfusion responses that may guide individualized decision-making; and (5) Mechanistic studies: Using patient-derived organoids to dissect transfusion effects on tumor biology at the cellular level.

The relationship between blood transfusion and cancer outcomes extends beyond simple immunosuppression or inflammation. It represents a complex interplay of patient vulnerability, surgical factors, and immunological responses that demands sophisticated investigation integrating multiple biological scales and temporal dimensions.