Blood transfusion has been associated with worse clinical outcomes in patients with various malignancies.3,11,12,16 Inflammatory cells and red cell elements transfected during blood transfusion can mediate an immunosuppression effect leading to impaired immune surveillance and increase risk of tumor progression. 3,11,12,16 In the present study, we found that receiving at least one RBCT significantly increased the risk of disease progression among SCCA patients undergoing definitive ChRT. While transfused patients more frequently had stage III tumors and ECOG performance status 2 or higher, the receipt of RBCT was associated with a 6.7 times higher risk of progression, independently of clinical stage. Additionally, despite the similar rates of persistent tumors after ChRT in both groups, transfused patients experienced nearly 4.5 times more disease recurrences after complete response. Patients who received RBCT maintained anemia, and this may have contributed to inferior oncological outcomes.
Several retrospective studies observed the associations between blood transfusion (mostly red blood cells) and increased rates of bacterial infections after curative resection of colorectal cancer (CRC), with subsequent increased postoperative mortality.11,12,13 Because such higher mortality may reflect higher tumor stage and/or comorbidities, it is unknown whether RBCT is a true prognostic factor or just a confounder for overall survival and morbidity after surgery. Therefore, studies evaluating CRC recurrence following RBCT have been conducted to avoid such confounding. And again, receipt of RBCT was linked to decreased disease-free survival time and high tumor recurrence rates in CRC, pointing to a possible causal effect involving unfiltered red cells, which are overloaded with immunomodulatory leucocytes of donors.15,16
The hypothesis that blood transfusion can induce immunosuppression was seen in early studies with renal transplant.9 In fact, transfusion of stored blood was found to be an effective immunosuppressive method after kidney transplant from unrelated donors. Nowadays, the better understanding and modulation of the immune system resulted in great advances in the treatment of many neoplasms. We previously demonstrated that HIV infection resulted in worse prognosis in patients with locally advanced SCCA.18 As a result, it is believed that immune surveillance plays an important role in combating SCCA tumor cells.17
Anemia is a frequent adverse event during chemotherapy and is associated with poor prognosis.2 Our two groups were similar in terms of type of chemotherapy regimen utilized, with approximately 23% rates of grade 3 or 4 anemia, which is similar to that reported by randomized trials of SCCA managed by ChRT.21 Decreased red blood production is not solely the result of myelotoxicity but it may also reflect a compromised clinical condition and patient frailty. In fact, in our study patients who received RBCT more frequently presented stage III tumors and had an ECOG performance status 2 or higher. Yet, receipt of RBCT was an independent risk factor for disease progression in the multivariable analyses.
It is also possible that persistent anemia after RBCT was a contributing factor that impaired treatment efficacy. In fact, the mean hemoglobin level of the transfused patients was 7.2 g/dl and was elevated to only 8.5 g/dl during treatment – likely because of myelosuppression of ChRT. Based on these results we can conclude that most patients received ChRT with low hemoglobin levels. The best cut-off levels for hemoglobin prior to ChRT for SCCA has not been determined. A multicenter study conducted in Italy, with 161 SCCA patients treated with ChRT showed that the chance of obtaining a complete response was increased by 5.6% for each g/dl in patients with hemoglobin equal to or greater than 11 g/dl up.22 In contrast, another retrospective study found no association between anemia and response to ChRT in SCCA.2 While it is possible that the immunosuppressive effects of RBCT plus persistent anemia both contributed to inferior progression free survival among transfused patients, we did not observe differences in complete response rates according to receipt of RBCT, which were more than 80% in both groups. Interestingly, the rates of progression, after complete response, were significantly higher among transfused patients, pointing to a later, rather than an acute, effect on the risk of cancer recurrence. Therefore, it seems that persistent anemia in the transfused group did not influence response to ChRT. Another possible explanation for worse progression free survival following RBCT is that patients who required RBCT were frailer and needed more treatment-interruptions. Unfortunately, we could not retrieve information of ChRT dose-intensity, an important known prognostic factor for progression.23
Our study has some relevant limitations. As a retrospective analysis, it was not possible to prove causality between RBCT and SCCA progression after ChRT. The number of HIV-infected patients who received RBCT was too small, what prevent us from testing this known prognostic factor in the multivariable analyses. Could not retrieve information on dose-intensity of ChRT, a known and important prognostic factor for localied SCCA.23 Also, the slow increment of hemoglobin levels after RBCT, possibly induced by myelossupression from ChRT, might have confounded (or contribute to) higher progression rates among transfused patients.
Prospective studies are the only way to better evaluate causality between RBCT and cancer recurrence or progression. The prospective evaluation of immune serum markers prior and after RBCT could measure whether (and to which extent) the transfusion of filtered and irradiated pre-storage blood leads to an increase in serum markers associated with immunosuppression. Additionally, detailed analyses of pre-transfusion red cell storage conditions (time, temperature) are important to measure cellular and humoral changes associated with RBCT. However, prospective studies are hard to be performed in this context because other confounders may be associated with immunomodulation, such as concomitant receipt of chemotherapy and radiation, the concurrent diagnosis of cancer, and the intrinsic selection bias, since patients who require RBCT may themselves be at higher risk of tumor progression for unknown biological reasons.