Transfusion practices in patients with advanced cancer: a retrospective study in a palliative care service

Abstract

Background:

Anemia is highly prevalent in patients with advanced cancer and adversely affects the quality of life. There are limited data on the frequency, clinical utility, and effectiveness of red blood cell (RBC) transfusions, and no randomized controlled clinical trials or clinical practice guidelines are available. The aim of this study was to evaluate clinician practices on RBC transfusion in an oncologic palliative care service and its impact on patients' symptoms, adverse events, and overall survival.

Methods:

This is a retrospective analysis of all patients with advanced cancer who received RBC transfusions admitted for 3 years. Preblood counts, the reason for transfusion, subjective benefit, and objective outcomes were listed.

Results:

We identified 179 patients who underwent RBC transfusions. The mean age was 67 years, and 60% were male. We found a total of 435 RBC units in 301 transfusion episodes. Asthenia/fatigue was the most frequent symptom (68%). The mean pretransfusion hemoglobin (Hb) was 6.85 g/dL, and 48% of patients had a Hb above 7 g/dL. The symptomatic benefit was achieved in 36% of patients. Adverse events were reported in 4%, with a 30-day survival rate of 57%. A statistically significant association was found between Eastern Cooperative Oncology Group performance status (ECOG-PS) and the symptomatic benefit (P = .005). Hb level pretransfusion, ECOG-PS, and symptomatic benefits with transfusions were significantly associated with survival.

Conclusion:

This study suggests that patients with advanced cancer with a higher functioning level may benefit more from RBC transfusion. Post-transfusion symptomatic benefits and pretransfusion ECOG-PS and Hb levels are independent predictors of survival. Further studies are needed to develop validated measures of objective functional changes to evaluate transfusions' clinical impact and identify patients most likely to benefit from it.

Introduction

The World Health Organization defines anemia as a hemoglobin (Hb) level <12 g/dL in women and <13 g/dL in men.¹ Anemia is common in palliative care (PC) patients, affecting 77% of men and 68% of women² and 50% of inpatients at any admission and 90% of inpatients in the last admission before death.³

In this context, anemia is often multifactorial. Causes include iron deficiency that can result from chronic blood loss due to gastrointestinal and gynecological malignancies or surgery, chronic inflammation, erythropoietin deficiency, bone marrow infiltration, hematinic depletion, and treatment effects.³ Less frequently, anemia can also derive from nutritional deficiencies due to cancer-induced anorexia.

Anemia causes several symptoms that can influence the patient's physical and functional status, negatively affecting their quality of life.⁴ The main symptoms that can be associated with anemia are fatigue and dyspnea. However, in advanced cancer, those symptoms may have multiple causes, and anemia can be one of them. Nevertheless, those symptoms may exist without anemia, and the correction of anemia may not improve them. Therefore, the relation between anemia and fatigue or dyspnea is unclear in many cases.⁴

Treatment options for anemia include red blood cell (RBC) transfusion; supplements when there is iron, folic acid, or vitamin B12 deficiencies; and erythropoiesis-stimulating agents.

Restrictive blood transfusion strategies (Hb level 7 and 8 g/dL) have been gradually adopted over the past few years because of the lack of clinical evidence demonstrating an improved outcome when compared with more liberal practices. The European Society for Medical Oncology (ESMO) advocates a threshold of 7–8 g/dL and recommends that transfusions are only used in anemic patients with severe symptoms in need of rapid Hb improvement.⁵

National Cancer Comprehensive Network (NCCN) guidelines state that the decision to offer RBC transfusion should not be made based on whether the Hb level of the patient has reached a certain threshold or “trigger.” Balance between transfusion risks and benefits should be evaluated on an individual basis.⁶

For patients with advanced cancer, there are no clinical practice guidelines or randomized controlled trials to assess the usefulness of transfusions or to identify groups of patients who are more likely to benefit.

The decision must be individualized and respect the bioethical principles: respect for autonomy, justice, nonmaleficence, and beneficence. Before deciding on transfusion, always inquire about ethical, cultural, and/or religious concerns that may lead patients to refuse blood products, and then, informed consent should be obtained. When a clinical deterioration is eventually expected, discuss expectations along with clinical signs that would suggest an appropriate timing to discontinue them.⁷ RBC is a scarce and expensive resource. The decision of using it for a particular patient requires consideration of who else might need the resource and how much of it is available.⁸ According to a recent review,⁹ adverse events occur in a minority of participants (3%–7%), but RBC transfusion may result in adverse reactions, fluid overload, and infection.¹⁰ In addition, RBC transfusions require frequent visits to health care facilities, which can affect quality of life.⁷ Studies on the beneficial effects of transfusions in patients with advanced cancer are vastly inconclusive, and blood transfusion is highly individualized. Cochrane's review⁹ reported that 31%–70% of patients with advanced cancer showed an improvement in their fatigue, breathlessness, or general well-being, with a limited duration of response of approximately 14 days or less. Timothy et al³ evaluated PC patients 7 days after receiving RBC transfusion and found that in 49% of cases the primary target symptom improved and 78% of transfusions improved at least one of the target symptoms. Neoh et al¹¹ found that 18% of patients had a clinical benefit maintained at 30 days, 31% had a transitory benefit for less than 14 days, and 11% had no benefit from transfusion.

The main aim of this study was to assess clinician practices around RBC transfusion in an oncologic palliative care service (PCS). The impact of RBC transfusion on patients' symptoms, adverse events, and post-transfusion overall survival were also studied.

Methods

Patients and study design

A retrospective cohort study was performed in a PCS from an oncological center, from January 1, 2016, to December 31, 2018. All data were collected from electronic medical records.

Eligible population consisted of adults (18 years or older) with a histologically confirmed malignant tumor, admitted to the PCS for symptomatic control as inpatients or as ambulatory, who received at least 1 RBC transfusion during the study period. No patients were receiving anticancer treatment. There were no exclusion criteria. Cancer staging was set in accordance with Tumor Node Metastasis Classification (American Joint Committee on Cancer 7th edition).

Because patients could have received RBC transfusions at different times, we designated each date associated with the transfusion of one or more units a transfusion episode (TE).

Patients and disease characteristics as demographic data (including age and sex), Charlson Comorbidity Index (CCI), and primary cancer location were collected. History of RBC transfusion in the previous 6 months, treatments used to treat or prevent anemia in the previous 6 months, and number of TEs were recorded.

For each TE, drugs with impact on bleeding events (anticoagulants, antiplatelets, anti-inflammatory drugs, and selective serotonin reuptake inhibitors) administered in the previous 72 hours, pretransfusion patient symptoms, pretransfusion patient Eastern Cooperative Oncology Group performance status (ECOG-PS), pretransfusion report of active bleeding, pretransfusion Hb, local and professional responsible for the decision, number of RBC units administered, symptomatic benefit in the 15 days after TE, adverse events, and survival were collected.

A safety end point was to assess the number of adverse events.

Statistical analysis

Categorical variables are presented as frequencies and percentages and continuous variables as means and standard deviations (SDs) or medians and interquartile range (IQR) for variables with skewed distributions. Normal distribution was checked using skewness and kurtosis.

Variables (ECOG-PS, Hb level, and symptomatic benefit) were compared using the chi-square test or the Fisher exact test. Survival curves were calculated using the Kaplan–Meier estimator and compared using the log-rank test. For multivariable analysis, Cox regression was used. The level of significance was deemed to be 0.05. Statistical analysis of the results was performed using IBM SPSS Statistics version 22.0.0.1 (IBM Corp, Armonk, NY) and MedCalc version 17.7.2 (MedCalc Software, Ostend, Belgium). Missing data were dealt with by listwise deletion.

Compliance with ethical standards

This study was reviewed and approved by the ethics review committee from Instituto Português de Oncologia do Porto (IPO-Porto; Ref. CES. 15/2021). The study involving human participants was in accordance with the ethical standards of the institutional and national research committee and with the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards. Because the study did not imply any contact with patients and all the rules of confidentiality and respect for those involved were assured, no informed consent was required.

Results

During the study period, a total of 179 patients received RBC transfusions. The mean age of the participants was 67.0 years (SD ± 12.9), and 60% were male (n = 107). The mean CCI was 8.9 (SD ± 2.3; Table 1).

Table 1.

Patients and disease characteristics

	n = 179
Sex, n (%)
Male	107 (60)
Female	72 (40)
Age, years
Mean ± SD	67.0 ± 12.9
Min-max	30–93
Charlson Comorbidity Index, score
Mean ± SD	8.9 ± 2.3
Min-max	3–15
Primary tumor location, n (%)
Colorectal	30 (17)
Gastric	30 (17)
Prostate	25 (14)
Gynecological	22 (12)
Others	14 (8)
Head and neck	11 (6)
Kidney	11 (6)
Esophagus	9 (5)
Lung	8 (4)
Breast	7 (4)
Hematologic	7 (4)
Skin	5 (3)

SD, standard deviation.

Most patients had gastrointestinal (42%) and genitourinary (35%) primary tumors. The most common were colorectal cancer (17%), gastric cancer (17%), and prostate cancer (14%). Records of TEs in the previous 6 months were found in 56% of patients. In addition, in the previous 6 months, at least 43% (n = 77) were under other therapies to treat or prevent anemia as oral iron supplements (n = 33), aminocaproic acid (n = 28), folic acid supplement (n = 22), and hemostatic radiotherapy (n = 16).

Drugs with impact on bleeding events administered in the 72h prior every TE were also analyzed and found in 59% (n = 179) of TEs. Anti-inflammatories (mainly corticosteroids) were used by 49% (n = 146), anticoagulants (mostly low molecular weight heparin) in 9% (n = 27), antiplatelets in 2% (n = 6), and selective serotonin reuptake inhibitors in 8% (n = 25).

Over the 3-year period, 301 TEs were recorded, with a median of 1.0 (1–9; Table 2). Pretransfusion symptoms were recorded in 67% (n = 202) of TEs. Asthenia/fatigue was the most frequent symptom, reported in 138 (68%) TEs, followed by dizziness/lipothymia/syncope (n = 29, 14%) and dyspnea (n = 24, 12%; Table 2). Before transfusions, the majority (n = 169, 73%) had an ECOG-PS >2 (Table 2). In 68 TEs, the ECOG-PS was not recorded.

Table 2.

Data collected by transfusion episode

	Transfusion episodes
Total, n	301
Per patient
Median	1.0
Min-max	1–9
Pretransfusion symptoms, n (%)
Fatigue/asthenia	138 (46)
Dizziness/lipothymia/syncope	29 (19)
Dyspnea	24 (8)
Other	11 (4)
None	48 (16)
Unknown	51 (17)
Pretransfusion ECOG-PS, n (%)
1	13 (4)
2	51 (17)
3	105 (35)
4	64 (21)
Unknown/not evaluated	68 (23)
Active bleeding, n (%)
Yes	100 (33)
No	201 (66)
Pretransfusion hemoglobin, g/dL
Mean ± SD	6.85 ± 1.15
Min-max	3.40–9.00
RBC units received
Total, n	435
Per transfusion episode
Median	1.0
Min-max	1–3

ECOG-PS, Eastern Cooperative Oncology Group performance status; RBC, red blood cell; SD, standard deviation.

Evidence of active bleeding pretransfusion was recorded in 100 (33%) TEs. The mean pretransfusion Hb was 6.85 g/dL (SD ± 1.15) with a range of 3.40–9.00 g/dL, and 145 patients (48%) had a Hb above the trigger threshold of 7 g/dL (Table 3). The decision for transfusion was made by a PC physician in 86% of TEs. In 13% of cases, the physician on-call, usually not a PC doctor, was responsible for the decision on transfusion. Fifty-eight percent of TEs concerned patients hospitalized in the PCS while the remaining were relative to outpatients observed in the emergency room or in consultations. The median RBC units given per TE was 1.0, ranging between 1 and 3, and a total of 435 RBC units were transfused (Table 2). Symptomatic benefit in the 15 days after transfusion was achieved in 36% (n = 107), observed in 55 TEs with fatigue/asthenia, 16 with dizziness/lipothymia/syncope, and 9 with dyspnea (Table 4). In 31% (n = 93), there was no benefit (Table 4). The symptomatic benefit was not associated with the number of RBC units given (P = .537). No serious adverse events were recorded. In 13 (4%) TEs, mild adverse events were documented: 11 cases of fever and 2 cases of volume overload (Table 4).

Table 3.

Pretransfusion hemoglobin levels (n = 301)

Hb (g/dL)	n (%)
⩽7.0	152 (50.5)
7.1–8.0	107 (35.5)
8.1–9.0	38 (12.6)
Not recorded	4 (1.4)

Hb, hemoglobin.

Table 4.

Transfusion episodes assessment by symptomatic benefit, adverse events, and survival

	n = 301
Symptomatic benefit, n (%)
Yes	107 (36)
No	93 (31)
Unknown	101 (34)
Adverse events, n (%)
No	288 (96)
Yes	14 (4)
Survival, days
Median	41.0
IQR	30.6–51.4
Min-max	0–651

IQR, interquartile range.

The median time to death was 41 days (IQR 30.6–51.4 days), with a 15-day survival rate of 70% and a 30-day survival rate of 57% (Table 4).

There was a statistically significant association between pretransfusion ECOG-PS and post-transfusion symptomatic benefit (P = .005), although there were only 9 patients with ECOG-PS ≤1, all with symptomatic benefit. No statistically significant correlation was found between Hb level before transfusion and symptomatic benefit (P = .151).

In an univariable analysis, pretransfusion ECOG-PS and Hb level as well as post-transfusion symptomatic benefit were significantly associated with survival (P<.001; Table 5). In the multivariable analysis, the same variables tested in the univariable analysis (pretransfusion ECOG-PS and Hb levels and post-transfusion symptomatic benefit) were significantly associated with survival (Table 6).

Table 5.

Univariable analysis of factors influencing survival

Factors	Median (days)	P value
ECOG-PS
1	146	<.001
2	67
3	37
4	23
Hemoglobin
>7	51	.001
≤7	35
Symptomatic benefit
Yes	64	<.001
No	14

ECOG-PS, Eastern Cooperative Oncology Group performance status.

Table 6.

Multivariate analysis of factors associated with survival

Factors	HR	95.0% CI
		Lower	Upper
ECOG-PS
1	0.18	0.08	0.40
2	0.16	0.09	0.28
3	0.53	0.35	0.79
4	1
Hemoglobin
>7	0.56	0.39	0.81
≤7	1
Symptomatic benefit
Yes	0.59	0.42	0.83
No	1

CI, confidence interval; ECOG-PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio.

Discussion

Clinical practices

To the best of our knowledge, this study is one of the largest to analyze transfusion practices in a PC setting. Almost half of the patients analyzed had gastrointestinal malignancies because bleeding events are common in these tumors. Although hematologic malignancies are among the most reported advanced cancer diagnoses leading to RBC transfusion,¹² we report only a few cases (4%) of oncohematologic patients. This is in accordance with the literature where oncohematologic patients can be nearly 7% of all oncologic patients.^13,14 There is evidence supporting that these patients behave as a special group in PC when compared with other tumors: There is a smaller number of referrals, with more advanced and symptomatic disease status, and less time between the last treatment date and the referral date or death.¹⁴

We report that 56% of patients received a RBC transfusion in the previous 6 months. Thus, most of the patients were already receiving RBC before being admitted by the PC team, which could have affected the decision to repeat the procedure. Only 43% were under other therapies to treat or prevent anemia, perhaps reflecting an inadequate investigation and treatment before transfusion in some cases.

More than half of the patients were under drugs with impact on bleeding events, which might have affected the decision to transfuse. Anti-inflammatory drugs were the most frequent. These drugs should be used carefully or suspended in patients with bleeding risk.

In this study, the mean pretransfusion Hb level was 6.85 g/dL, which was lower than in the Neoh et al¹¹ study (7.5 g/dL) and in Timothy et al¹⁰ retrospective and prospective³ studies where the levels were 7.8 g/dL and 7.2 g/dL, respectively. On the contrary, that level was higher than in the Sirianni et al¹⁵ study, which was 6.5 g/dL. Neoh et al¹¹ reported that 70% of patients had a Hb above the trigger threshold of 7.0 g/dL, while we found an inferior frequency of 42%. In addition, we found a median of RBC units given per TE (1.0) similar to the Sirianni et al¹⁵ study; however, in the Neoh et al¹¹ study and the Timothy et al¹⁰ retrospective and prospective³ studies, the median number of RBC units was higher: 2.0, 2.3, and 2.1, respectively. These results may reflect doctors' effort to adopt a restrictive blood transfusion strategy in a PC setting.

Impact on symptoms

Symptoms were recorded in 67% of TEs. It is likely that a higher number of patients had symptoms, but they were not registered in patient files. Similar to other studies,^3,10,15 the most reported symptom was asthenia/fatigue.

Considering the overall number of TEs in which it was possible to assess the 15-day symptomatic benefit, we observed a symptomatic benefit in more than half of the cases. However, symptomatic benefit was not assessed in nearly one-third of the TEs, and no standardized tool was used to demonstrate or refute benefit because no specific validated tools to assess the utility of transfusions exist. In addition, symptoms such as fatigue are multifactorial and very difficult to measure, and a single intervention targeting one of these factors is unlikely to have a widespread benefit. Pretransfusion ECOG-PS association with symptomatic benefit suggests that patients with a higher level of functioning may have a bigger benefit from RBC transfusion.

Adverse events

No serious adverse events were documented, and we found a lower frequency of adverse events (4%), comparing with other studies which reported 12%³ and 12.5%¹⁵ adverse events. The low frequency of adverse events may reduce concerns about potential harms.

Survival

Compared with the prospective study by Neoh et al,¹¹ in which 32% of patients died within 30 days, we found an inferior 30-day survival rate (57%). Compared with the retrospective study by Sirianni et al,¹⁵ we found a longer median time to death. Their analyses only considered the first TE date, while we have included each TE, as in our opinion each TE results from an independent decision. Nonetheless, deterioration is expected in this population given the progressive nature of disease and cannot be attributed to RBC transfusions.

The results of multivariable analyses suggest that pretransfusion ECOG-PS and Hb levels as well as symptomatic benefit are independent predictors of survival in patients with advanced cancer who received RBC transfusions.

Limitations

Potential limitations of this study, such as the retrospective design and potential different criteria used by the assistant doctors to decide on transfusions, should be considered. Moreover, because this is a single-center study, it cannot exclude possible selection biases according to our transfusion practices.

Conclusion

In a PC setting, transfusions seem to have a subjective symptomatic benefit and should be offered for symptom relief in patients with a higher level of functioning. They are well-tolerated, but the lack of standardized preassessment/postassessment tools limits any ability to draw conclusions about utility. The duration and magnitude of the symptomatic benefit remain uncertain. According to a recent article, most PC specialists consider RBC transfusion to have a role in symptom management, but many clinical and nonclinical factors influence their decisions to provide or withhold transfusions.¹⁶

Further studies are needed to develop tools to evaluate the clinical impact of transfusion, to identify patients most likely to benefit from transfusion, and to better understand the attitudes of physicians in a PC setting.

Conflicts of interest statement

The authors declare no conflicts of interest.

Previous presentation

Preliminary data of this work were presented at 2021 MASCC/ISOO Annual Meeting.

Authors' contributions

All authors contributed to the study conception and design. Data collection was performed by Sara Marote, Joana Marinho, and Maria Cândida Silva. Data analysis was performed by Sara Marote, Joana Marinho, and José Ferraz Gonçalves. The manuscript was written by Sara Marote, Joana Marinho, and José Ferraz Gonçalves, and all authors commented on the versions of the manuscript. All authors read and approved the final manuscript.