Granulocyte transfusions

INTRODUCTION — In spite of modern antimicrobials and supportive therapy, bacterial, fungal, and viral infections are still major complications in patients with prolonged disease-related or therapy-related neutropenia. An increasing number of clinical disorders are being treated with aggressive chemotherapy and bone marrow or hematopoietic stem cell transplantation. Neutropenia is one of the most frequent side effects of these aggressive treatments, and the risk of infection increases rapidly when the granulocyte count falls below 500 cells/microL (table 1).

The spectrum of infections in neutropenic patients has shifted, with multidrug-resistant bacterial infection and fungal infection such as Aspergillus, Fusarium, and Zygomyces emerging as the main cause of morbidity and mortality [1]. Since these infections are the direct result of neutropenia, granulocyte transfusion (GTX) as replacement therapy is a logical therapeutic approach and has been available for the last 40 years.

The history, indications, clinical efficacy, and complications of GTX will be reviewed here, as well as donor selection, qualification, stimulation, granulocyte collection, processing, storage, and transfusion.

Granulocyte transfusion is not a frequent procedure. A survey of United States hospitals estimated that approximately 2200 granulocyte transfusions were administered in 2019 [2].

Separate topic reviews discuss evaluation and general management of patients with fever and neutropenia.

●(See "Overview of neutropenic fever syndromes".)

●(See "Fever in children with chemotherapy-induced neutropenia".)

●(See "Management of children with non-chemotherapy-induced neutropenia and fever".)

●(See "Evaluation of children with non-chemotherapy-induced neutropenia and fever".)

●(See "Diagnostic approach to the adult cancer patient with neutropenic fever".)

●(See "Treatment of neutropenic fever syndromes in adults with hematologic malignancies and hematopoietic cell transplant recipients (high-risk patients)".)

●(See "Treatment and prevention of neutropenic fever syndromes in adult cancer patients at low risk for complications".)

HISTORY OF GRANULOCYTE TRANSFUSION DEVELOPMENT — Although granulocyte transfusion (GTX) to neutropenic patients makes intuitive sense, just as does red blood cell (RBC) transfusion to anemic patients or platelet transfusion to thrombocytopenic patients, it remains, in our opinion, an underutilized therapy. This is largely related to the difficulties in collecting adequate numbers of viable, functional granulocytes.

Although normal individuals produce approximately 6 x 10¹⁰ granulocytes/day [3], their half-life in the peripheral blood is only six to seven hours. In addition, these cells cannot be stored for more than a few hours, as they rapidly lose viability. Thus, granulocytes need to be collected and transfused on a daily basis.

Collection of adequate numbers of viable, functional granulocytes was almost impossible before the invention of continuous flow apheresis blood collection devices in the late 1960s. Even so, such devices permitted the collection of only 1 x 10¹⁰ granulocytes (10 percent of normal daily production) from a donor during a six- to seven-hour apheresis collection. To increase the number of granulocytes collected by apheresis, corticosteroids were used to increase the donor's circulating granulocyte count and hydroxyethyl starch (HES) added to the donor's blood as it entered the centrifuge bowl to create better separation between the RBC and white blood cell (WBC) layers in the centrifuge.

With these improvements, collection of 2 to 3 x 10¹⁰ cells/unit was achieved [4]. Although the steroid and HES use and time spent on the apheresis process created a huge burden for the donor, it could be solved by having more than one donor to support a single patient.

Initially, GTX was met with great enthusiasm. A number of clinical trials were performed from 1972 to 1982, but in terms of efficacy, these trials yielded mixed results. Six reports showed favorable responses to GTX, with 59 to 88 percent survival in patients receiving GTX, compared with 20 to 36 percent in control subjects [4-8].

In 1982, a randomized controlled clinical trial of therapeutic GTX was conducted in patients with gram-negative sepsis, which showed no benefits among treated patients [9]. Although this study appeared to be well-designed, the dosages of granulocytes used in the study were far too low, averaging one-half of the minimally acceptable dose of 1 x 10¹⁰ cells per transfusion. Any expected benefit of GTX would have been precluded by such low granulocyte dosage. GTX gradually disappeared from clinical use from 1985 to 1995 due to improvements in antimicrobials and general supportive care, combined with the difficulties in collecting and transfusing granulocytes.

Several early studies showed the importance of granulocyte dose in achieving favorable outcomes. As early as 1975, it had been reported that patients with good clinical responses received four times as many granulocytes as patients with poor responses [10]. Meta-analyses of clinical trials of GTX suggested that higher doses were provided in the studies showing clinical efficacy [11,12]. Therefore, when collection of larger doses of granulocytes became possible after the discovery and availability of recombinant granulocyte colony-stimulating factor (G-CSF), interest in GTX was renewed [13].

In humans, administration of G-CSF causes a rapid increase in the granulocyte count that begins within two hours and peaks at approximately 12 hours after administration [14]. By stimulating donors with a combination of G-CSF and dexamethasone, collection of 5 to 10 x 10¹⁰ granulocytes at one sitting became possible [15-17]. With these larger granulocyte doses, one can achieve significant increments in absolute neutrophil count (ANC) in severely neutropenic and infected patients [18].

In addition, neutrophils collected after the administration of G-CSF appear to be of better quality. They exhibit prolonged survival after transfusion, which may be due, in part, to the fact that these cells are released from the bone marrow early, thus representing a younger population. They also appear to experience delayed apoptosis and improved in vitro function (eg, respiratory burst, chemotaxis, and bactericidal activity) [16,19-21]. Evidence of in vivo granulocyte migration and activity has been demonstrated using multiple techniques, including buccal neutrophil accumulation, radioisotope labeled neutrophil migration, and the skin window test [16,18,22,23]. Thus, we can now collect adequate numbers of viable, functional cells from a single donor after cytokine and steroid administration.

The Resolving Infection in people with Neutropenia with Granulocytes (RING) trial was a randomized, controlled clinical trial designed to test the efficacy of transfusing G-CSF-stimulated granulocytes in patients with neutropenia and severe infection. However, the trial was unable to accrue sufficient numbers of patients to determine whether outcomes were improved with GTX. It is unlikely that another randomized trial will be performed to answer this question. Thus, our approach is based on our experience with GTX and transfusion strategies for other diseases with low blood counts (eg, platelets for severe thrombocytopenia, RBCs for severe anemia). (See 'Chemotherapy or HCT-induced neutropenia' below.)

INDICATIONS AND CLINICAL EFFICACY

Minimal criteria — We recommend that the following minimal criteria be met before initiation of granulocyte transfusion (GTX), regardless of the cause of the patient's neutropenia:

●Absolute neutrophil count (ANC) <500 cells/microL, except in the case of chronic granulomatous disease (see 'Chronic granulomatous disease' below).

●Evidence of bacterial or fungal infection (ie, clinical symptoms of infection, positive cultures, pathological diagnosis of infection from biopsies, radiographic evidence of pneumonia).

●Unresponsiveness to antimicrobial treatment for at least 48 hours (except in extreme circumstances with life-threatening infection) [24].

The rationale for these criteria includes the extreme difficulty in recruitment of donors, the huge burden on the donor, and other costs and risks associated with GTX. (See 'Donor issues' below and 'Complications' below.)

Chemotherapy or HCT-induced neutropenia — Neutropenia from chemotherapy and hematopoietic cell transplantation (HCT) is the most common use of GTX, although the use of GTX in this population remains rare. Most retrospective clinical studies and prospective clinical trials of GTX were conducted in this patient population, including adult and pediatric patients [25-33]. Chemotherapy and HCT can severely suppress bone marrow production of all cell lines and cause severe pancytopenia.

While anemia and thrombocytopenia can be supported with red blood cell (RBC) and platelet transfusion without significant difficulties, severe neutropenia and associated infection remain the most important complication and limiting factor of these therapies. Granulocyte colony stimulating factor (G-CSF) may be used to stimulate patients' bone marrow production of granulocytes, but response to G-CSF in this patient group is usually poor.

Treatment of infection — Most bacterial infections and some fungal infections can be controlled with the modern antimicrobials and supportive therapies, but multidrug-resistant bacterial infection and fungal infection in patients with neutropenia remain a major cause of morbidity and mortality. It is for these patients that GTX should be considered.

Determining the efficacy of GTX in patients with neutropenia and resistant bacterial or fungal infections has been extremely challenging, with several lines of evidence showing a trend towards greater efficacy, but no trial establishing an unequivocal benefit.

●The Resolving Infection in Neutropenia with Granulocytes (RING) trial was a randomized trial to address the efficacy of G-CSF-mobilized granulocytes in patients with neutropenia (ANC <500/microL) due to chemotherapy or HCT who had a proven or probable bacterial or fungal infection [34]. The entry criteria were expanded to include patients with aplastic anemia due to slow accrual; the target accrual of 236 patients was not reached. A total of 114 patients were enrolled, and 97 completed the trial. The primary endpoint was a composite of survival plus microbial response six weeks after randomization. This was reached in 20 of 48 individuals in the GTX arm and 21 of 49 in the control arm (42 versus 43 percent). On subgroup analysis, there was no effect of baseline patient factors, interval to first GTX, or post-transfusion neutrophil count. A post-hoc analysis of the GTX arm demonstrated improved outcomes in those who received high dose GTX (≥0.6 x 10⁹ granulocytes per kilogram) versus low dose (<0.6 x 10⁹ granulocytes per kilogram), with efficacy in 59 versus 15 percent, respectively. However, since the primary endpoint was reached in 37 percent of the control group who did not receive GTX, it is possible that chance may explain the difference in outcomes found with granulocyte dose.

●Observational studies conducted before RING showed mixed results, with several suggesting a benefit of GTX in treatment of infection and/or survival [18,26,27,29-31]. A notable study included 30 patients with infections not controlled with administration of antibiotics and G-CSF, in whom GTX led to resolution of infections in 82 percent of patients with bacterial infection and 38 percent with fungal infection; the overall survival at 100 days was 82 and 54 percent for these groups, respectively [29].

Additional randomized trials are unlikely to occur in the near future, and we continue to believe that GTX has value in the appropriate clinical setting. (See 'Minimal criteria' above.)

Infection prophylaxis — Prophylactic GTX remains controversial due to the potential adverse effects of GTX, and we do not use GTX for infection prophylaxis outside of a clinical trial. (See 'Complications' below.)

A 2015 Cochrane meta-analysis evaluated 11 randomized (or quasi-randomized) trials involving 653 patients with neutropenia from chemotherapy or HCT [35]. There was no difference in 30 day all-cause mortality or overall infection rate in those who received prophylactic GTX versus those who did not. However, there was a decreased risk of bacteremia and fungemia in those given prophylactic GTX, and the dose of granulocytes appeared to correlate with the risk of infection, with a dose of 10 x 10¹⁰ granulocytes per day being more effective. However, the overall quality of the evidence was considered low.

Aplastic anemia — Aplastic anemia (AA) is a rare acquired disorder in which there is a failure of the bone marrow to produce sufficient blood cells, resulting in pancytopenia. The failure of the stem cells to produce mature blood cells can vary from partial to almost complete, thus producing a disease of varying severity in different patients. Symptoms may be slow to emerge because the loss of stem cell function is gradual. However, patients are often diagnosed at a later stage of the disease with severe complications such as infections and bleeding. (See "Aplastic anemia: Pathogenesis, clinical manifestations, and diagnosis".)

GTX may be appropriate for patients with severe AA and neutropenia when bacterial or fungal infection is unresponsive to maximal antibiotic and/or antifungal therapy [36,37]. A significant issue in the use of GTX for AA is the high incidence of HLA alloimmunization due to transfusions of RBCs and platelets, which can in turn increase the risk for transfusion-related acute lung injury (TRALI). An 11-year experience at the National Institutes of Health in using GTX as supportive therapy in 32 patients with AA has been presented; infections were evenly divided between invasive bacterial and fungal infections [38,39]. One-quarter of the patients had HLA alloimmunization before the initiation of GTX, but no TRALI was reported and there was no difference in mean post-transfusion ANC between the patients divided according to HLA alloimmunization status. Overall survival to hospital discharge was 58 percent, while for patients with invasive fungal infection, survival to hospital discharge was 44 percent.

Chronic granulomatous disease — Patients with chronic granulomatous disease have a normal granulocyte count but have a defect in granulocyte function, leaving them susceptible to recurrent bacterial and fungal infection. Modern antimicrobials can prevent and control most of the bacterial and fungal infections in these patients. GTX has been used in patients who continue to deteriorate under maximum antimicrobial therapy. A number of small studies and case reports have documented the success of GTX in treating patients with chronic granulomatous disease with invasive aspergillosis [40-42]. (See "Chronic granulomatous disease: Treatment and prognosis", section on 'Treatment'.)

Neonatal sepsis — Neonates with sepsis may become neutropenic due to an immature granulopoietic system. This population may also benefit from GTX. Three studies compared survival rates of septic neonates receiving GTX versus no transfusion or intravenous immune globulin (IVIG) therapy alone [43-45]. They concluded that there may be some survival benefit with GTX. However, a Cochrane meta-analysis of randomized and quasi-randomized studies on the use of GTX to treat neonatal neutropenic sepsis concluded that there is still insufficient evidence for its benefit in this patient population [46].

DONOR ISSUES — With the introduction of granulocyte colony stimulating factor (G-CSF) and glucocorticoid-stimulated granulocyte collection, it is now possible to collect enough granulocytes from a single donor to produce a substantial increment in absolute neutrophil count (ANC) in a neutropenic patient.

Donor selection — Both community donors and directed donors, such as family members and friends of the patient, can be utilized as granulocyte donors. Family members and friends tend to be highly motivated and cooperative, but most of them have not donated blood before. Therefore, they have not been screened with blood donor questionnaires and have not been tested for transfusion-transmitted infectious disease markers. It generally takes two to three days to screen and test these potential donors. Therefore, utilizing directed donors alone will delay the initiation of granulocyte transfusion (GTX).

Community donors, taken from an established donor pool, may have previously been tested for infectious disease markers and have a known blood type. Since community donors are likely to have demonstrated good venous access from prior donation, granulocyte collections may be more successful.

One study has suggested that there may be benefit to utilizing community donors rather than directed donors [27]. The authors found that the time interval between the request for granulocytes and transfusion was significantly less in patients receiving granulocytes collected from community donors. These patients also benefited from higher ANC increments than those receiving donations from relatives.

From our own experience, we would recommend the use of both groups of donors to maximize the benefits, by initiating the GTX with community donors to get started, and then continuing GTX with family members and friends of the patient as they qualify after screening and testing. These family members and friends of the patient may be recruited as community donors for future GTX for other patients. Since they have seen that GTX can save their loved one's life, they are more motivated to help others with similar needs.

Donor qualifications — Not everyone who is willing to donate granulocytes will be qualified to do so. Just as with whole blood or plateletpheresis donors, granulocyte donors have to meet the minimum standards in donor history and physical examination set by the US Food and Drug Administration and possibly other country, state, and professional regulatory agencies. In addition to being qualified as a whole blood donor, granulocyte donors may have to meet the following requirements at our institution:

●ABO and Rh matched to the recipient. Due to the difficulties of separating the granulocyte layer and red blood cell (RBC) layer during the apheresis collection procedure, granulocyte concentrates are usually heavily contaminated with RBCs. Therefore, patient and donor need to be cross-match compatible for RBCs. This is a limiting factor in directed donor recruitment, particularly for patients who are blood type O. When ABO-mismatched granulocytes need to be used in certain circumstances, RBC depletion techniques can be used to remove the unwanted RBCs from the granulocyte product [47].

●Tested negative for all blood transfusion-associated infectious disease markers within 30 days of granulocyte donation.

●Female donors <50 years old must have tested negative for pregnancy or have a history of hysterectomy or tubal ligation surgery.

●Good vascular access. Granulocyte collection is performed through apheresis collection, which demands good venous access. The donor should be evaluated by the apheresis collection staff.

●No history of allergies to steroids or starch.

●A history of hypertension, diabetes, gastrointestinal ulcers, glaucoma, tuberculosis, or any fungal infections may be a contraindication to steroid administration.

Donor stimulation — There is variation in the dose of G-CSF across studies, ranging from 200 to 600 mcg. One study compared the efficacy of a standard 8 mg dexamethasone dose with either 450 or 600 mcg of G-CSF [48]. The granulocyte yields and side effects were comparable with both regimens. From an economic perspective, the lower G-CSF dose results in significant cost savings for the blood center.

At our center, we administer G-CSF (300 mcg subcutaneously) and dexamethasone (8 mg orally) on the day prior to each collection. If the donor's precollection WBC count increases to >50,000/microL, stimulation with G-CSF is suspended until the count falls below that level; however, collections continue as scheduled. Circulating neutrophil counts in G-CSF- and dexamethasone-stimulated donors are maximal at 12 hours after treatment [15]. Accordingly, we stimulate the donor in the evening and start granulocyte collection the next morning.

Some donor centers choose not to stimulate their donors before collection due to the difficult logistics in obtaining and administering the G-CSF and dexamethasone. Some donor centers choose to administer dexamethasone only, without using G-CSF. In both of these scenarios, the numbers of granulocytes able to be collected are much lower than collections from donors stimulated by both G-CSF and dexamethasone [4]. As noted above, patients transfused with such inadequate products do not derive the full benefit of GTX therapy. (See 'Indications and clinical efficacy' above.)

There are no published guidelines as to the frequency of collection and total number of donations that are permissible for normal donors. At our institution, using platelet donation limits as a reference, we generally limit collections to twice per week with 48 hours between the two collections and a total of 24 donations per 12-month period.

Side effects of medications — Side effects seen among G-CSF/dexamethasone-stimulated donors seem to be generally mild and self-limited [17,49,50]. Headache, arthralgias, bone pain, fatigue, and insomnia are the predominant symptoms. The vast majority of donors experiencing such adverse effects report a willingness to give further granulocyte donations [17,51].

Typically, circulating neutrophil counts return to normal a few days after cytokine administration is discontinued [52]. Longer term studies of G-CSF-stimulated donors have demonstrated no obvious lasting side effects [53-55]. Although there has been some concern about the development of posterior subcapsular cataracts attributed to the repeated administration of glucocorticoids in granulocyte donors, a statistically significant relationship between glucocorticoid use in such donors and the development of cataracts has not been shown [56].

Hydroxyethyl starch (HES) is routinely used during granulocyte collection for more efficient cell separation. Most studies of the short- and long-term safety of HES involve its use as a volume expander. No long term consequences have been demonstrated in these studies. Therefore, it is reasonable to consider it safe for use in granulocyte collection in normal donors. Side effects of HES include fluid retention, with mild increase of body weight.

Burden on the donor — The complicated process of granulocyte donation may be very burdensome for donors. They must take three trips to the donor center in order to make the first donation (first for donor screening and testing, second for medication administration, third for granulocyte donation), and two trips each for subsequent donations. On the day of donation, the granulocyte collection procedure takes four to five hours.

The unavoidable inconvenience to the donor may prove to be a limiting factor to the success of GTX for a given patient. Donors also must bear the risks associated with G-CSF and dexamethasone, and the risks associated with the apheresis donation procedure itself. (See "Blood donor screening: Overview of recipient and donor protections", section on 'Complications of G-CSF and glucocorticoids (granulocyte donors)'.)

GRANULOCYTE COLLECTION, PROCESSING, AND STORAGE

Granulocyte collection — Granulocyte harvesting is accomplished through the process of apheresis in which only granulocytes are removed from the donor, while red blood cells (RBCs) and plasma are returned to the donor. In our experience, apheresis should be performed for at least 150 to 180 minutes to process 7 to 10 liters of blood in order to maximize the granulocyte yield.

Citrate is commonly used as the anticoagulant, which may result in hypocalcemic symptoms in the donor. Calcium gluconate may be added to the return line to counteract the effect of citrate. Hydroxyethyl starch (HES) is added to the donor's blood as it enters the centrifuge to facilitate separation of granulocytes from RBCs and improve collection. Pain at the venipuncture site, fatigue, tingling around the mouth and lip area, mild vital sign changes are common adverse reactions during the collection process. (See 'Donor stimulation' above and 'Side effects of medications' above and 'Burden on the donor' above.)

Granulocyte concentrate processing — As a blood component, granulocyte concentrates should be labeled properly with unit number, blood type, collection date, expiration date, volume of the product, and collection facilities.

Granulocyte recipients may be immunodeficient and at increased risk for transfusion-associated graft-versus-host disease. To prevent this complication, all granulocyte concentrates should be irradiated with ≥25 gray (Gy), which renders lymphocytes incapable of dividing or attacking host tissues. (See 'Transfusion-associated GVHD' below.)

A sample of the product should be tested for total white blood cell count and differential, which can be combined with product volume to calculate the total number of granulocytes in the concentrates. Another sample of the product should be used to perform a crossmatch with the recipient's plasma.

Storage of granulocytes — The utility of granulocyte transfusion is somewhat limited by the inability to store granulocytes for long periods of time. In the past, it was recommended that the granulocytes be stored at room temperature (22°C) and transfused as soon as possible. The improved quality of granulocyte colony stimulating factor (G-CSF) stimulated granulocytes may allow for longer storage times. As an example, subsequent studies showed that granulocyte concentrates may be stored at room temperature for ≥24 hours, with excellent preservation of functional activity [20,57]. Nevertheless, we recommend transfusing granulocyte concentrates within a few hours of collection.

ADMINISTRATION OF THE TRANSFUSION AND DETERMINING RESPONSE

Consent — Consent for blood transfusion must be signed by the patient (or representative) before the granulocyte transfusion (GTX). Since transfusion will take place as soon as possible after collection, results of infectious disease testing of the product will not be available at the time of transfusion (even though the donor tested negative for infectious disease markers within 30 days). A clinician must discuss with the patient or representative the risks of transfusing blood products prior to results of testing, and document consent in a progress note or a consent form.

Administration — We use premedication with acetaminophen (500 mg) and diphenhydramine (25 mg) to prevent or minimize transfusion reactions, which are more likely for GTX than for transfusion of red blood cells.

Granulocytes should only be transfused through a standard blood administration set containing a filter with a pore size of 170 microns. Bedside leukocyte reduction filters should never be used. Transfusion should start slowly (eg, 2 mL/min), with close monitoring of vital signs, oxygen saturation, and signs and symptoms of complications. (See 'Complications' below.)

If no transfusion reaction is observed, the infusion rate can be increased to a rate as fast as the patient can tolerate, but care must be taken to avoid circulatory overload in older adult patients and in patients with congestive heart failure. The product should be transfused in two to four hours.

If the patient develops signs and symptoms of a transfusion reaction, the infusion should be stopped temporarily. The IV line should be kept open with normal saline, but the granulocyte product should not be discarded. The blood bank should be notified, the patient should be evaluated by a physician, and once the patient recovers from the reaction, GTX can be resumed. Any decision to discard the granulocyte concentrate should be made by the patient's attending clinician, not by nursing staff or housestaff.

Determining recipient response — Recipients of GTX should be monitored for changes in absolute neutrophil count (ANC) by monitoring daily morning white blood cell count and differential. The post-transfusion ANC increment can be quite large in patients receiving granulocytes from granulocyte colony stimulating factor (G-CSF)- and corticosteroid-stimulated donors. As an example, one study reported a mean ANC increment of 1000/microL maintained for 1 to 1.5 days following a mean granulocyte dose of 5.1 x 10¹⁰ cells [58].

In addition, patients should be monitored for changes in signs and symptoms of the underlying infection by using microbial culture and imaging studies, as appropriate (eg, chest radiography for pneumonia).

Daily GTX not only can provide a near-normal blood neutrophil count in severely neutropenic patients, but also these transfused granulocytes have normal function in bactericidal, fungicidal, and chemotactic activities and are capable of migrating to extravascular sites and localizing to areas of infection [59,60]. Pictures of a dramatic increase in neutrophil numbers along with pus formation in an arm wound debridement specimen from a patient after just a single GTX have been published [60].

Criteria for stopping granulocyte transfusions — For therapeutic GTX, the decision of how long to continue administering daily GTX should be made by a clinical team including the hematologist and physicians covering the granulocyte collection/transfusion; it sometimes also involves the patient and family members. Generally accepted criteria for discontinuing GTX include the following:

●The clinical infection has been resolved based on clinical signs/symptoms, and laboratory/radiological test results.

●The patient's ANC is above 500 for three days without GTX, which is a sign of bone marrow recovery.

●The patient's clinical condition has worsened (ie, poor response to GTX), and the treatment plan has changed to palliative care with patient and family consent. GTX is generally not considered part of palliative care.

The typical duration of GTX treatment can vary from three days to months, depending on the patient's clinical condition, treatment plan, response to treatment, and donor availability.

COMPLICATIONS — Transfusion reaction rates are higher in granulocyte transfusion (GTX) compared with red blood cell (RBC) transfusion, with 25 to 50 percent having mild to moderate reactions [61]. There is a 1 percent incidence of severe complications [61]. In general, the most frequently encountered reactions are fever and chills. Slow infusion of the granulocyte concentrate and premedication can reduce the incidence and severity of these side effects. Other complications include pulmonary adverse reactions, transfusion-associated graft-versus-host disease (ta-GVHD), HLA alloimmunization, and transfusion-transmitted infection [61].

Pulmonary adverse reactions — Moderate to severe pulmonary adverse reactions can occur during GTX. Symptoms include varying degrees of cough, dyspnea, hypoxia, and changes on the chest radiograph. In one series, pulmonary reactions occurred in approximately 5 percent of GTX episodes [62]. In one study, no changes were seen in O₂ saturation measured before and after the GTX [38]. Concomitant administration of amphotericin B and granulocyte transfusions is not associated with pulmonary complications [36,37,63-65].

Transfusion-associated GVHD — Transfusion-associated graft-versus-host disease (ta-GVHD) is caused by viable functional donor lymphocytes present in granulocyte concentrates, which may mount an immunologic attack against the recipient. Although usually occurring in immunodeficient patients, ta-GVHD can develop in the presence of intact immunity [66]. (See "Transfusion-associated graft-versus-host disease".)

We routinely irradiate our granulocyte concentrates with 2500 to 3000 gray (Gy) prior to transfusion in order to eliminate this risk. This irradiation does not interfere with granulocyte function.

Alloimmunization — Alloimmunization is another potential risk of GTX [67]. A retrospective review indicated that 17 percent of patients with severe aplastic anemia developed HLA antibodies during the course of GTX, and that patients with detectable HLA antibodies had lower post-GTX white blood cell increments [39].

The development of antibodies to HLA- or granulocyte-specific antigens can reduce in vivo granulocyte survival and cause abnormal migration of granulocytes [68]. In addition, refractoriness to platelet transfusion increases in patients receiving prophylactic GTX, due to the production of anti-HLA antibodies [69]. (See "Refractoriness to platelet transfusion", section on 'Alloimmunization'.)

Infection — Transfusion-transmitted diseases are just as important a consideration in GTX as with the use of any other blood transfusion [70]. This is especially true of granulocytes, which must be transfused as soon as possible after collection, before completion of infectious disease testing [71,72].

The risk of cytomegalovirus (CMV) may be greater in granulocyte concentrates, since CMV is harbored in peripheral blood leukocytes. Accordingly, the standard of care is to select CMV seronegative donors in patients at risk for developing transfusion-transmitted CMV infection [73,74], although two reports have questioned this policy [59,75]. These two latter studies suggest that the risk of transmission is actually very low. Therefore, it may be possible to transfuse a CMV seronegative patient with granulocytes from a CMV seropositive donor, as long as the patient can be closely monitored for CMV viremia. In this manner, antiviral treatment may be implemented early to prevent serious sequelae.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Neutropenic fever in adults with cancer".)

SUMMARY AND RECOMMENDATIONS

●Granulocyte transfusion (GTX) is used infrequently for treatment of septic, neutropenic patients, but prior studies may have been limited due to transfusion with inadequate doses of granulocytes. Use of granulocyte colony stimulating factor (G-CSF) enables GTX with larger granulocyte doses and function. (See 'History of granulocyte transfusion development' above.)

●Clinical applications – The most common use of GTX is unresponsiveness of a documented bacterial or fungal infection to antimicrobial treatment for ≥48 hours in a patient with neutropenia from chemotherapy or hematopoietic cell transplantation (HCT). A randomized trial of GTX did not demonstrate improved outcomes, but the trial did not accrue sufficient patients to adequately assess efficacy. (See 'Indications and clinical efficacy' above.)

●Administration – We premedicate with acetaminophen and diphenhydramine and routinely irradiate granulocyte concentrates to minimize transfusion reactions. Granulocytes should only be transfused through a standard blood administration set containing a filter with a pore size of 170 microns; bedside leukocyte reduction filters should never be used. The typical duration of GTX treatment can vary from three days to months, depending on the patient's clinical conditions, treatment plan, response to treatment, and donor availability. (See 'Administration of the transfusion and determining response' above.)

●Complications of GTX include pulmonary reactions, transfusion-associated graft-versus-host disease, alloimmunization, and infection. (See 'Complications' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Dennis Goldfinger, MD, who contributed to an earlier version of this topic review.