Koki Nakanishi, Mitsuro Kanda, Yasuhiro Kodera

Abstract Gastrectomy with radical lymph node dissection is the most promising treatment avenue for patients with gastric cancer. However, this procedure sometimes induces excessive intraoperative blood loss and requires perioperative allogeneic blood transfusion. There are lasting discussions and controversies about whether intraoperative blood loss or perioperative blood transfusion has adverse effects on the prognosis in patients with gastric cancer. We reviewed laboratory and clinical evidence of these associations in patients with gastric cancer. A large amount of clinical evidence supports the correlation between excessive intraoperative blood loss and adverse effects on the prognosis. The laboratory evidence revealed three possible causes of such adverse effects： anti-tumor immunosuppression, unfavorable postoperative conditions, and peritoneal recurrence by spillage of cancer cells into the pelvis. Several systematic reviews and meta-analyses have suggested the adverse effects of perioperative blood transfusions on prognostic parameters such as all-cause mortality, recurrence,and postoperative complications. There are two possible causes of adverse effects of blood transfusions on the prognosis： Anti-tumor immunosuppression and patient-related confounding factors (e.g., preoperative anemia). These factors are associated with a worse prognosis and higher requirement for perioperative blood transfusions. Surgeons should make efforts to minimize intraoperative blood loss and transfusions during gastric cancer surgery to improve patients'prognosis.

Key words: Gastric cancer; Blood loss; Prognosis; Transfusion; Adverse effect;Immunosuppression; Mortality; Recurrence; Complication

INTRODUCTION

Surgical resection is still the most promising avenue for patients with resectable gastric cancer. However, it sometimes induces excessive intraoperative blood loss(IΒL) and requires perioperative allogeneic blood transfusion (ΒTF), especially when gastrectomy with systematic lymph node dissection is performed. Furthermore, as IΒL becomes more excessive, the need for ΒTF further increases; thus, IΒL is closely associated with ΒTF.

There are lasting discussions and controversies about whether IΒL or ΒTF has adverse effects on the prognosis in patients with gastric cancer[1-3]. Patients requiring ΒTF often have severe illness, advanced cancer, a poor general condition, and a higher prevalence of comorbidities, and these confounding factors themselves induce postoperative complications, surgical death, and a worse prognosis[4-6]. The same is true of IΒL. Thus, it is difficult to evaluate whether IΒL or ΒTF itself has adverse effects on the prognosis. However, clinical trials that determine these effects in surgical treatment are not ethically permissible. We therefore reviewed clinical and laboratory evidence to explore the effects of IΒL and ΒTF on the prognosis of patients with gastric cancer.

MECHANISM OF ADVERSE EFFECT OF IBL

There are three possible causes of the adverse effect of IΒL on the prognosis： Antitumor immunosuppression induced by IΒL, an unfavorable postoperative condition induced by IΒL, and spillage of microscopic cancer cells in the pelvic cavity via the blood lost during IΒL. These mechanisms are summarized in Figure 1.

First, many studies have shown that the main cause of the adverse effect of excessive IΒL is anti-tumor immunosuppression via the loss of plasma constituents[7-10]. In support of this concept, several studies have revealed that the prevalence of hematogenous recurrence, which is correlated with immunosuppression, was significantly higher in patients with excessive IΒL[9,10]. However,these studies did not demonstrate the mechanism. Βruns et al[11]reported that IΒL of ＞700 mL during gastrointestinal surgery was associated with a significant decrease in natural killer cell activity, leading to an unfavorable prognosis. Miki et al[12]reported that interleukin (IL)-6 and tumor growth factor triggered by IL-6 were increased in patients with colorectal cancer receiving ΒTF due to excessive IΒL. Thus, the mechanism of immunosuppression in gastric cancer surgery is speculative, and laboratory evidence is lacking. A validation study is therefore needed.

Second, excessive IΒL may lead to an unfavorable postoperative condition, such as the development of postoperative complications, thus adversely affecting the prognosis[9,13]. Postoperative complications occur may lead to severe tissue damage caused by local and generalized inflammatory reactions, resulting in more severe immunosuppression[13].

Third, Kamei et al[14]reported that excessive IΒL is an independent risk factor for peritoneal recurrence after curative gastrectomy. They suggested the possibility thatblood loss into the peritoneal cavity may promote tumor spillage during surgery,which may be specifically associated with peritoneal recurrence. Arita et al[15]further confirmed the association between IΒL and peritoneal recurrence in the laboratory setting. Although this idea is very interesting, no other study to date has supported this hypothesis. Moreover, it is unclear whether this adverse effect remains when administering S-1 monotherapy which is one of the standard postoperative adjuvant chemotherapies that mainly suppresses peritoneal recurrence[16]. Therefore, further analysis in the clinical practice setting is needed.

Figure 1 The mechanism of adverse effect of intraoperative blood loss and perioperative blood transfusion. IBL: Intraoperative blood loss; BTF: Blood transfusion.

LITERATURE SEARCH OF STUDIES REPORTING EFFECT OF IBL ON PROGNOSIS

Our investigation of the relationship between IΒL and the prognosis was derived from that of patients with colorectal cancer. Heiss et al[17]first reported the possibility that IΒL itself may be beneficial for survival of malignant cells in the host and also found a positive link with tumor recurrence and poor outcomes in patients with colorectal cancer. The adverse effect of IΒL on the prognosis in patients with gastric cancer was first reported by Dhar et al[7]in 2000. We have summarized the studies reporting the effect of IΒL on the prognosis in Table 1. Dhar et al[7]reported that IΒL of＞ 500 mL was an independent predictor of survival in an analysis of 152 patients with transmural (T2N0-T3N2) gastric cancer. They hypothesized that IΒL reduced the body's immunity and its ability to fight cancer cells; this concept was quoted from the report by Βruns et al[11]. However, Dhar et al[7]provided no information on perioperative ΒTF, which is a strong confounding factor for the prognosis. Similar studies were subsequently reported. Kamei et al[14]reported that IΒL of ≥ 475 mL was specifically associated with the development of peritoneal recurrence in 146 patients who underwent curative gastrectomy for advanced gastric cancer. They reported for the first time the relationship between IΒL and the recurrence pattern. Liang et al[8]also reported that IΒL of ≥ 200 mL was an independent prognostic factor in 845 patients who underwent curative gastrectomy. In their study, IΒL of ≥ 200 mL was a prognostic factor even when patients who underwent ΒTF were excluded; however,ΒTF administration was not a prognostic factor. Mizuno et al[9]reported that IΒL of ≥400 mL was a significant predictor of survival and cancer recurrence in 203 patients with stage II/III gastric cancer and was associated with the prevalence of hematogenous recurrence. Their study excluded patients who received ΒTF to eliminate a potential confounding bias caused by the adverse effects of ΒTF. Ito et al[10]reported that IΒL of ＞ 330 mL had an adverse effect on the long-term prognosis in 1013 patients with stage II/III gastric cancer. Their study also excluded patients who received ΒTF and was the largest-scale study, thoroughly eliminating complicated confounding factors. IΒL is closely associated with ΒTF administration, and the prognostic significance of IΒL might be masked by the adverse effect of ΒTF. From this viewpoint, three studies[8-10]excluded this confounding influence, indicating that IΒL itself has an adverse effect on the long-term prognosis in patients with gastric cancer.

Evidence was also found in the field of laparoscopic surgery. Ishino et al[18]reported that IΒL of ≥ 1% body weight was significantly correlated with postoperative complications and was an independent predictor of survival in 214 patients who underwent laparoscopy-assisted gastrectomy for gastric cancer. Conversely, several negative studies of the adverse effects of IΒL on the prognosis have also been published (summarized in Table 1). Ojima et al[19]reported that ΒTF administration was an independent prognostic factor for survival in 856 patients who underwent curative gastrectomy but that IΒL of ≥ 1000 mL was not a prognostic factor. Likewise,two studies showed that ΒTF administration was an independent prognostic factor for survival but that excessive IΒL was not prognostic factor[6,20]. However, the threshold of the IΒL volume in these reports was greatly different, and neither study excluded the confounding influence of ΒTF.

Table 1 Studies of effects of intraoperative blood loss on prognosis in patients with gastric cancer

SUMMARY OF EFFECT OF IBL ON PROGNOSIS

The accumulation of clinical evidence reveals that excessive IΒL may have adverse effects on the prognosis in patients with gastric cancer by promoting anti-tumor immunosuppression, unfavorable postoperative conditions, and a specific association with peritoneal recurrence by spillage of cancer cells into the pelvic cavity during surgery. However, the laboratory evidence is weak and some issues remain unclear.IΒL thresholds varied, and the results might differ depending on these thresholds. A higher threshold for the amount of IΒL would introduce more confounding factors(e.g., ΒTF, postoperative anemia, and postoperative complications). Another issue is that only two studies have reported the relationship between IΒL and peritoneal recurrence.

MECHANISM OF ADVERSE EFFECT OF BTF

There are two possible causes of the adverse effect of ΒTF on the prognosis： antitumor immunosuppression induced by ΒTF and patient-related confounding factors(e.g., preoperative anemia and postoperative complications). These factors are associated with a worse prognosis and higher requirement for perioperative ΒTF.These mechanisms are summarized in Figure 1.

Gantt[21]was the first to report the possibility of promoting tumor growth by immunosuppression due to ΒTF in 1981. Numerous authors have since considered that ΒTF administration has profound adverse effects on the host's immune system[22-25]. Mechanisms of inhibition of host immunity by ΒTF are diverse and include cytokine-mediated immune responses and suppression of cellular and humoral immunity against cancer cells. ΒTF-induced immunomodulatory effects drive the immune system to inhibit IL-2 production[25], decrease interferon gamma[23],suppress natural killer cell function[24], release immunosuppressive prostaglandins[25],decrease monocyte activity[25], and increase of regulatory T cells (suppressor T cells)[24-27].

ΒTF administration also promotes increases in IL-6[12], vascular endothelial growth factor[28], and hepatocyte growth factor[29], which play fundamental roles in tumor growth, malignant transformation, and invasion of tumor cells[30,31]. Additionally, in patients with gastric cancer, overexpression of these cytokines is reportedly correlated with a poor prognosis[12,32-34]. The immunosuppression caused by ΒTF creates favorable conditions for tumor growth; additionally, ΒTF increases the risk of postoperative complications[3], which also have adverse effects on the prognosis[13].

However, some issues remain unclear. The type of blood products received (e.g.,red blood cells, leukodepleted blood, whole blood) was not constant among studies.The disorder observed after ΒTF administration is caused by the presence of leukocytes and their products, as mentioned above. Current ΒTF products are often leukodepleted, and filtered transfusion is routinely performed; thus, the contamination of cytokines is decreased and the effect is weakened. In contrast to this concept, no difference in the prognosis was found in comparative studies between leukocyte-depleted blood and non-leukocyte-depleted blood[35,36]. The roles of these cytokines and growth factors in current transfusion treatment remain unclear.

Preoperative anemia, which is a patient-related confounding factor, is another possible cause of a worse prognosis in patients with malignancy[17]. Gastrointestinal tumors sometimes bleed due to the passage of intestinal contents and the effect of digestive juices, and this bleeding may lead to anemia. In particular, once anemia has occurred in patients of advanced age, it persists because of these patients'physiological decrease in hematopoietic cells in the bone marrow, decrease in hematopoietic stem cells, and reduced serum erythropoietin levels due to renal impairment[37]. Additionally, continuous anemia causes malnutrition, which also has adverse effects on the prognosis[38]. Numerous prospective and retrospective studies have shown that patients with preoperative anemia have a worse prognosis than patients without anemia[39]. Hence, preoperative anemia is a cause of the requirement for ΒTF, which itself also has adverse effects on the prognosis.

Although we have summarized the mechanisms of adverse effects of ΒTF, some unmeasurable and non-excludable confounding factors remain. Surgical damage is one such factor and can also lead to severe immunosuppression. The degree of surgical damage depends on the surgical organ. Surgical damage induced during colon cancer surgery is considered to be relatively mild, and there are many negative reports on the influence of ΒTF on the prognosis of such patients[40,41]. However,surgical damage induced during esophageal cancer surgery is considered to be relatively severe, and there are many positive reports on the influence of ΒTF on the prognosis of such patients[42,43]. In gastric cancer surgery, the degree of surgical damage varies greatly depending on the operation type; therefore, it may be more effective to investigate this issue according to the operation type (total gastrectomy vs distal gastrectomy, open surgery vs laparoscopic surgery).

LITERATURE SEARCH OF STUDIES REPORTING EFFECT OF BTF ON PROGNOSIS

Many studies have been performed to evaluate the adverse effect of perioperative ΒTF on the prognosis in patients with gastric cancer, and we have summarized these studies in Table 2. Kaneda et al[44]first reported that ΒTF administration had an adverse effect on survival in 231 patients who underwent curative gastrectomy. Ojima et al[19]subsequently reported that ΒTF administration was an independent prognostic factor for survival in 856 patients who underwent curative gastrectomy, even when the amount of transfused blood was small. Kanda et al[6]reported that ΒTF administration was an independent prognostic factor for survival and recurrence in patients with stage II/III gastric cancer, regardless of the volume of ΒTF. They also reported that the prognostic impact of ΒTF became less clear after introduction of adjuvant chemotherapy with S-1. Three systematic reviews and meta-analyses support the idea that ΒTF is associated with a worse prognosis, all-cause mortality,cancer-related mortality, and recurrence (summarized in Table 3).

However, some studies have shown that ΒTF does not have an adverse effect on the prognosis (summarized in Table 2). Kampsch?er et al[45]performed a large-scale retrospective study and found that the survival of patients who had undergone ΒTF was shorter than that of patients who had not undergone ΒTF; however, after stratifying patients into stages and applying proportional regression analyses, ΒTF administration did not appear to have any effect on the prognosis but was instead associated with other prognostic features. Pacelli et al[46]conducted a multicenter retrospective study and reported that ΒTF had a slight, but not significant, adverse effect on survival of 927 patients who underwent curative gastrectomy.

Table 2 Studies of effects of blood transfusion on prognosis in patients with gastric cancer

SUMMARY OF EFFECT OF BTF ON PROGNOSIS

The adverse effects of ΒTF have been well verified by clinical and laboratory data;these adverse effects are caused by anti-tumor immunosuppression and patientrelated confounding factors that lead to a requirement for ΒTF. However, this information may not be helpful in the clinical setting because ΒTF is still required in the event of massive bleeding during surgery or preoperative anemia. However, there may be room for consideration such as adjusting preoperative anemia and paying attention so as not to lead postoperative complications.

REMAINING PROBLEMS FOR FUTURE STUDIES

The adverse effects of IΒL or ΒTF were previously ascertained by clinical evidence.However, continuous and untiring efforts to minimize IΒL and surgical damage have been progressing (i.e., development of laparoscopic surgery, improvements in surgical techniques and devices, and enhanced recovery after surgery programs), and the amount of IΒL and frequency of ΒTF administration have been decreasing. In addition, perioperative chemotherapy has been further developed, helping to prolong survival. Further accumulation of data and performance of high-quality studies are required to clarify whether IΒL or ΒTF still have adverse effects on the prognosis.

CONCLUSION

IΒL and ΒTF lead to adverse effects on the prognosis in patients with gastric cancer,and the main causes are anti-tumor immunosuppression and confounding factors such as postoperative complications and preoperative anemia. Surgeons should make efforts to minimize IΒL and ΒTF to improve patients' prognosis.

ACKNOWLEDGMENTS

We thank Angela Morben, DVM, ELS from Edanz Group for editing a draft of this manuscript.

Table 3 Systematic reviews and meta-analyses of effects of blood transfusion on prognosis in patients with gastric cancer