OVERVIEW: What every practitioner needs to know

Neutropenia refers to an absolute neutrophil count (ANC) less than<1500/mm3. Under certain circumstances, the risk of bacterial infection due to neutropenia is related to the ANC but depends also upon the cause of the neutropenia. Adequacy of the marrow reserve pool is the most critical determinant of propensity to infection in a neutropenic patient.

Are you sure your patient has neutropenia? What are the typical findings for this disease?

Definition: Neutropenia means insufficient numbers of circulating neutrophils. The diagnosis is based on calculation of the absolute neutrophil count or ANC. A patient has neutropenia if the ANC is less than 1500/mm3.

While there are age-related normal values and ranges, 1500/mm3 is generally used for all ages. For example, the ANC is higher in infants. It is important to remember that the word “neutropenia” refers to the total neutrophil count and not the % neutrophils in the differential. In children less than 5 years old, the percent PMN is lower than the percent lymphs. This ratio changes around 5 years of age to the normal adult polymorphonuclear leukocyte (PMN) predominance of about 60%.

Continue Reading

Synonyms for PMN are SEGS, POLYS (Table I).

Table I.n

Abbreviations for neutrophils

Under certain circumstances, the risk of bacterial infection due to neutropenia is related to the ANC (Table II) but depends also upon the cause of the neutropenia (Table III).

Table II.n

Clinically significant neutrophil counts

Table III.n

Infection risk by type of neutropenia

Calculating ANC:

ANC = WBC x (PMN + Bands ) where WBC = total white blood cell count expressed as WBC/mm3, PMN = percent mature polymorphonuclear neutrophils and Bands = the percent band neutrophils.

Risk of Infection

Only about 3% of the neutrophils are circulating in the peripheral blood. The vast majority of neutrophils are in the bone marrow reserve pool, and the remainder is in the tissue and in the marginated pool attached to the lining of blood vessels. Thus, a standard complete blood count (CBC) done on peripheral blood is sampling a very small compartment of neutrophils and does not accurately reflect the capacity to protect against bacterial infection.

While you can infer tissue supply of neutrophils from clinical history and certain physical findings (see
Table IV), there is no good clinical laboratory test available to quantitate tissue neutrophil delivery. Adequacy of the marrow reserve pool is the most critical determinant of propensity to infection in a neutropenic patient, although having a normal marrow does not guarantee adequate protection against infection (see below). We have summarized the risk of infection based on ANC in Table II.

Table IV.n

Important clinical signs and symptoms in neutropenia

If the neutropenic patient is febrile, management decisions must be made without information about the adequacy of the bone marrow reserve. Bone marrow aspiration and biopsy are the only ways to directly evaluate marrow reserve. However, there are clinical signs that can be helpful. Table IV lists clinical signs that infer marrow reserve. Items marked with ** suggest normal reserve and those with * suggest low reserve.

If the bone marrow reserve pool is significantly depleted, there is a rough relationship between the absolute neutrophil count and propensity to infection (Table II).

However, and importantly, if the bone marrow reserve pool is completely adequate, there is no relationship between the degree of neutropenia and propensity to infection. Patients with an ANC of zero may be at no increased risk of serious infection because of the neutropenia if the marrow is normal.

What other disease/condition shares some of these symptoms?

Most physicians are aware of the extreme danger present in patients with significant fever and very low absolute neutrophil counts, based upon their experiences during training with patients who have received chemotherapy or who have bone marrow failure syndromes. These patients have no bone marrow reserve. Physicians are usually quite unfamiliar with the other end of the spectrum of patients with immune-mediated neutropenia but normal bone marrow reserve who are at no increased risk of infection because of the neutropenia. Note that we are careful to say “no increased risk of infection because of the neutropenia” and not “no increased risk of infection.”

Neutropenia can be the presenting finding in common variable immunodeficiency, other immunodeficiency states, and collagen vascular diseases; all of which can have adequate marrow reserve yet put the patient at increased risk for infection for reasons other than the neutropenia.

What caused this disease to develop at this time?

The differential diagnosis of isolated neutropenia is different than neutropenia associated with low platelets or anemia. The differential diagnosis will also differ depending on whether the patient is encountered in an acute setting associated with infection or if the neutropenia is an incidental finding on a blood count done for other reasons.

The first question in either case is whether the neutropenia is longstanding or is a transient process of relatively short duration. Since neutropenia can be secondary to viral infections or medications, in the absence of other issues, observation for a few weeks to see if the neutropenia resolves spontaneously, or if the neutropenia resolves after stopping a medication, is often the best approach before launching into an extensive and possibly invasive workup.

The presence of the history and physical findings, as shown in Figure 1, can be of help as well. For example, a history of recurrent, unusual infections may suggest that the neutropenia is chronic.

Figure 1.n

Diagnostic approach to neutropenia

Regardless of the specific diagnosis, the main issue that affects the management approach is whether the patient can make neutrophils or not and whether the patient is currently infected or clinically unstable.

In general, isolated neutropenia is rarely, if ever, due to hematological malignancy. Conversely, if more than one cell line is affected, hematologic malignancy or other bone marrow failure process needs to be considered in the differential diagnosis.

Neutropenia can occur because:

1. Neutrophils are not being made.

2. They are being destroyed or sequestered at a rate faster than they can be released from the bone marrow.

We classify neutropenia into “normal marrow reserve” versus “decreased marrow reserve” categories. These groupings are most relevant to the patient’s outcome and management. Disease entities associated with neutropenia are noted in Table III along with their relative propensity to serious infection. They are discussed individually at the end of this text. You will note that infections and drug suppression are listed in more than one category because the marrow reserve status can vary.

Most drug-induced neutropenias are dose-related and reversible. The best approach is to stop the drug and observe. However, some can cause either dose-related or idiosyncratic agranulocytosis and put the patient at severe risk for sepsis.

Presence of mucosal ulcers, cramping abdominal pain and other symptoms, as noted in Table IV, suggest infectious invasions of tissue because neutrophils are not getting to tissue, as one sees with marrow suppression. In contrast, the presence of visible abscess or purulence indicates ability to deliver neutrophils to the tissue and is consistent with normal marrow reserve.

If there is clinical evidence of low marrow reserve in the presence of significant neutropenia (ANC < 300) the possible offending drug must be stopped immediately. If the offending medication is felt to be critical and cannot be stopped, a bone marrow evaluation must be done to determine if there is agranulocytosis. If the marrow reserve is good, the drug can be continued with periodic monitoring of the blood count. Otherwise, there is no choice but to stop the offending agent.

Most immune-mediated neutropenias are associated with circulating IgG antibodies and have normal marrow reserve. The same is true for transient post-infectious neutropenias. However, cytotoxic T-cell related neutropenia, such as seen in the so-called “lymphoproliferative disorders of large granular lymphocytes” or LGL cell disease, can be accompanied by total absence of granulocyte precursors in the marrow. These types of disorders can be seen in association with lupus, rheumatoid arthritis and other primary autoimmune diseases.

Again, the signs in Table IV can be helpful with clinical assessment. If there is evidence of mucosal ulceration, we suspect decreased marrow reserve.

There is one caveat here in the case of autoimmune disease. Rheumatologic disorders can be associated with vasculitis and can have mucosal ulcerations in the absence of neutropenia. The only way to determine the marrow reserve in this case is to do a marrow aspiration.

Malignancy, aplastic anemia and primary marrow failure syndromes can all be associated with neutropenia. In all of these, the marrow reserve is decreased and the propensity to infection is inversely related to the ANC (Table II). Marrow infiltration rarely results in isolated neutropenia. In fact, early in the presentation of leukemia, the ANC may be normal but the patient is still at high risk. The marrow reserve is low because it is replaced, and all of the neutrophils are pushed into the periphery. This phase does not last long, though, and neutropenia rapidly ensues.

Patients with primary agranulocytosis are not as high-risk as patients with aplastic anemia or post-cytotoxic chemotherapy. These patients should be assumed to be at high risk and should only be managed by hematologists with specific experience in these disorders.

Complicated hospitalized patients often have several things going on at the same time, and the etiology of their neutropenia can be multifactorial. They may have anemia for reasons unrelated to the development of neutropenia. Similarly, thrombocytopenia may be present for reasons unrelated to marrow failure.

The nutritional status of chronically ill hospitalized patients is often significantly neglected. Table V lists nutrient deficiencies that can be associated with marrow failure as well as the tests used to diagnose and monitor them. Note in particular that B12 deficiency can occur with normal serum levels of B12 and is best detected by measuring the downstream metabolite, methylmalonic acid. Similarly, elevated homocysteine can indicate deficiency of B12, folate, or pyridoxine. Low ceruloplasmin due to copper deficiency causes isolated neutropenia or pancytopenia with marrow findings similar to myelodysplasia and is often seen when copper is removed from hyperalimentation fluids because of presumed toxicity.

What laboratory studies should you request to help confirm the diagnosis? How should you interpret the results?

The ESR and to a lesser extent CRP can be very helpful in the approach to patients with neutropenia. The ESR goes up with deep tissue inflammation and can signify occult infection in patients with chronic neutropenia, presumably because neutrophils are not able to clear foci of infection.

It is particularly helpful diagnostically in patients who otherwise appear to be normal. If the ESR is high in a patient with neutropenia and there is no obvious site of infection, you know some systemic process is going on and further evaluation is warranted. If it is normal, you can’t really exclude a more serious form of chronic neutropenia, but observation alone may be a reasonable approach.

Serial monitoring of the ESR can be very helpful in the febrile neutropenic patient. Often, cultures are negative and the patient is treated with empiric antibiotics. A monotonic decrease in ESR obtained every two to three days suggests the choice of antibiotics is working. Failure to normalize or subsequent increase in the ESR can indicate development of resistance or a new organism. Note that the ESR is variable and thus must be obtained fairly frequently to clearly establish a pattern.

The ESR can also help in determination of duration of antimicrobial therapy in patients with chronic neutropenia with low marrow reserve, whether the organism is known or not. These patients cannot not fight bacterial infection normally, and standard durations of therapy based on a fixed number of days treatment really should not be used. We usually treat at least a week or so past normalization of the ESR, depending on the organism.

Confirming the diagnosis

Table III is a classification of neutropenia by risk category. It is intended to apply to neutropenias of all causes. Oncology services often have more specific stratifications that depend on chemotherapy protocol.

Regardless, all post-chemotherapy neutropenias fall within the most severe category in Table III. Note that some types of neutropenia appear in more than one risk grouping. Most post-infectious neutropenias are antibody-mediated and have normal marrow reserve and are at relatively low risk, but some can lead to significant marrow suppression. Most drugs are associated with only mild marrow suppression, but some cause aplasia.

The clinical findings marked with (*,**) in Table IV can help distinguish low reserve from normal reserve neutropenia. Mucosal ulcers without exudate or abdominal symptoms in the presence of a very low ANC (<300 / mm3) suggest decreased reserve. Conversely, the presence of exudate or abscess implies that neutrophils are getting to tissue and there may be more leeway for less aggressive treatment or holding off on treatment until a specific infectious diagnosis can be established.

If you are able to confirm that the patient has neutropenia, what treatment should be initiated?

The treatment required for a febrile or infected neutropenic patient depends on the following:

1. The cause of the neutropenia

2. The magnitude of the neutropenia

3. The nature of the fever or infection

4. Associated clinical symptoms and signs

Based on these points, the clinician can decide if the patient must be immediately treated with parenteral broad-spectrum antibiotics and hospitalized, or whether there is some room for clinical judgment and a more systematic approach to determining the cause of the fever. While there is often a perceived urgency to treat, it is best to establish a correct infectious diagnosis first, if possible.

When an ill, febrile patient with severe neutropenia is first encountered, the cause of the neutropenia is not known. These patients should be assumed to be in a low marrow reserve class and thus at high risk.

The risk categories in general relate to how much of the immune system is functioning. Patients who are undergoing cytotoxic chemotherapy not only have significantly reduced marrow reserve, but also have varying degrees of T-cell immune dysfunction as well as possible disruption of mucosal barriers in the gastrointestinal tract that can serve as portals of infection for Gram-negative and enteric organisms. All of these factors place the patient at extreme risk for overwhelming infection. On the other end of the spectrum, patients with benign neutropenia have totally normal marrow reserve and are at no risk despite a very low ANC.

Whether or not it is safe to treat a febrile neutropenic patient as an outpatient depends on the associated clinical findings, the primary type of neutropenia, and the ANC.

We will discuss the approaches based on the neutropenia risk groups defined in Table III. If the cause of the neutropenia is unknown or not certain, one should assume the patient has a severe form of neutropenia with decreased marrow reserve.

Use of granulocyte colony stimulating factor (GCSF)

GCSF is a naturally occurring hormone that can increase the ANC. It has two actions: 1) It stimulates release of PMN from the marrow reserve; and 2) It stimulates production of PMN by the marrow. The release of PMN form the marrow occurs within 24 hours or less, so if you see a significant increase within 24 hours, the marrow is making neutrophils. The increase in production takes longer with counts increasing in 4 to 7 days.

GCSF is really only indicated if there is a problem with production of neutrophils. Its use should be reserved for situations where the marrow reserve is decreased. If the marrow reserve is normal, it is not indicated. Its use in specific diseases is noted below.

In serious infectious situations where the marrow status is not known but there is significant concern that neutrophils are not being produced and the patient is unstable, it is not unreasonable to start daily subcutaneous GCSF at a dose of 5 to 15 mcg/kg/day. However, as soon as the critical period is over, the GCSF should be stopped and a complete evaluation including marrow aspiration should be done after a week or two off the medication.

An immediate response within 24 hours reflects release of neutrophils from the marrow reserve and suggests that the marrow is making neutrophils already. Note that this medication is not without risk. There have been reports of massive splenomegaly with rupture in patients receiving the drug due to overstimulation of myelopoiesis.

Low Risk

If a patient is known to be in a low-risk class of neutropenia (Table II), their fevers and infections can be managed the same as any other patient, regardless of the ANC. Importantly, there is no relation between the degree of neutropenia and propensity to infection. These patients do not need to be hospitalized or even treated with parenteral antibiotics. The key here is that the diagnosis of low-risk neutropenia has been established.

In the face of neutropenia, and in particular an ANC less than 300 /mm3, the presence of definite purulent exudate or abscess formation, such as a bulging tympanic membrane, is strong evidence in support of the ability to get neutrophils to tissue and supports the diagnosis of low-risk neutropenia. This should be clearly documented in the chart, as it could affect future treatment decisions. Note that just because a patient is “low-risk” does not mean the cannot be septic. It just means that the patient is not septic because of being neutropenic.

It is difficult to give clear recommendations regarding antibiotic management for this group. In general, one must guard against admission to the hospital for every fever while protecting against the very low possibility of serious infection. The major decisions depend on clinical signs of sepsis or decompensation. Such patients should be admitted and treated aggressively with broad-spectrum antibiotics covering Gram-positive and Gram-negative organisms. The specific antibiotics depend on the sensitivity at the individual institutions. If the child clinically looks good when the fever is reduced and follow-up is good, the child can be managed as an outpatient.

Moderate Risk

Many of the syndromes in this group are associated with decreased marrow reserve; thus, the propensity to bacterial infection is more likely to be related to the ANC. While some of these diagnoses are associated with normal reserve and the neutropenia itself may not increase risk, the presence of other immunological abnormalities does put the patient at risk for bacterial infection. Immunologically-mediated neutropenia is not an unusual finding in patients with immunodeficiency syndromes or with autoimmune disorders like systemic lupus or rheumatoid arthritis. An ANC over 300 and certainly over 600 to 800 suggests that the neutropenia is not a major contributor to the risk of sepsis. An ANC below 300 should be considered to put patients at increased risk for bacterial disease, particularly from organisms the patients carry on their own bodies.

Appropriate cultures should be obtained and parenteral antibiotics given for temperatures over 38°C. Outpatient follow up is appropriate if the patient is clinically totally stable. If clinical signs of mucosal invasion are present (Table IV) admission to the hospital is probably the best approach.

High Risk

The management of this group is similar to the moderate-risk neutropenia group, except viral and fungal causes of fever need to be more prominent in the differential diagnosis in those neutropenias associated with T and B cell dysfunction. Patients with Kostmann and Shwachman syndromes are not at as high-risk as other patients in this group, but they certainly can have overwhelming infection. Again, evidence of breakdown of mucosal barriers puts the patients at even higher risk of sepsis from skin and enteric organisms. Thus, patients with abdominal cramps, diarrhea and fever in this group should be admitted, cultured, and treated with empiric antibiotics. Note that with cyclic neutropenia at least, Clostridia species are the most common organisms.

Incidental Neutropenia

Neutropenia is often discovered as an incidental finding on a CBC done for other reasons. As long as the neutropenia is isolated, the best approach is to observe with periodic CBCs at a weekly interval. Many post-viral neutropenias will resolve spontaneously in a week or two.

Figure 1 is a general schema for the approach to these patients. If there are no significant history or physical findings, periodic monitoring is often all that is required. However, if there is any evidence of underlying pathology, more extensive monitoring is required. If any of the key history and physical points noted in Figure 1are present or occur during the course of monitoring, more extensive evaluation is warranted.

The central part of this diagram suggests continued monitoring of the CBC. Note that monitoring is less focused on changes in the ANC and more so on development of anemia, macrocytosis, thrombocytopenia, or unusual inflammatory symptoms that may suggest that some other more serious process is going on. The frequency of monitoring can be adjusted based on the presumed diagnosis. We have noted specific recommendations in the section on individual diagnoses.

We find monitoring of the erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) to be very helpful in patients with neutropenia. Elevation of the ESR or CRP, in the absence of any other findings of acute infection, may suggest that the neutropenia is more significant and that closer follow-up or further evaluation is indicated. The ESR is a marker of deep tissue inflammation, and the CRP indicates monocyte activation. They do not always parallel each other; thus, we request both when consulted regarding neutropenia. The ESR is probably more indicative of the kinds of deep tissue inflammation seen with chronic neutropenia with decreased tissue delivery of granulocytes.

The presence of chronic inflammation and in particular mucosa ulcerations and gingivitis with neutropenia suggests decreased marrow reserve and that bone marrow aspiration and biopsy are indicated. We usually complete the rest of the evaluations listed in Table V prior to marrow aspiration, if possible, because a positive result may obviate the need for marrow aspiration. A complete discussion of the evaluation of all neutropenias, and particularly rare genetic neutropenia, is beyond the scope of this discussion.

Mild Neutropenia in an Asymptomatic Patient

It is not uncommon to encounter seemingly totally asymptomatic patients who have neutropenia. More commonly, the ANC is between 800 and 1500. Assuming that they have none of the history or physical findings in Figure 1, these are most likely patients with chronic benign neutropenia, chronic non-immune neutropenia or ethnic neutropenia. It is always helpful to try to get blood counts from several years ago to see if the neutropenia is acquired, but these results are often not available. Unfortunately, there is really no way to determine if the neutropenia you are seeing is the beginning of collagen vascular disease or is simply chronic benign or chronic non-immune neutropenia.

If the ANC is above 800 or so, we check the ESR, ANA, C3, C4 and then have the patients follow up wIth their family physician to monitor the CBC and ESR three or four times over the ensuing year. In particular, we are looking for changes in other cell lines, including a rise in MCV, anemia, or a drop in platelets. We tell the patient to return if there are any symptoms of unusual infection or evidence of collagen vascular disease. Our approach to asymptomatic patients with more profound neutropenia follows Figure 1 as well and is outlined in references 1 and 2 below.

Neutropenia in Complex Hospitalized Patients

Hospitalized patients with multiple problems may develop neutropenia. In these cases, the neutropenia is usually secondary to a drug that is being administered or some other underlying disease process. Anemia is often present and may be multifactorial in nature as well. The approaches already mentioned apply. However, we would emphasize that nutritional deficiency, particularly low copper and ceruloplasmin, are probably under-diagnosed in the inpatient setting.

The micronutrient tests noted in Table V should be examined periodically during prolonged hospitalizations, because deficiencies are easily corrected and cytopenias will resolve. Do not assume because the patient is being supplemented parenterally that the nutrients are adequate. There are many factors in critically ill patients that can increase demand and hasten the development of deficiencies.

Often, possibly offending drugs need to be stopped to see if the marrow recovers. If a critical drug is being used, marrow aspiration demonstrating good reserve can support an approach of careful monitoring with continuation of the medication. If there is agranulocytosis, the offending drug(s) must be stopped. Drug-induced agranulocytosis can be fatal.

What are the adverse effects associated with each treatment option?


What causes this disease and how frequent is it?

The following are brief discussions of some of the entities noted in Table III. The reader is referred to the referenced book chapter for more extensive information. In particular, management of rare congenital neutropenia requires special expertise.

Chronic Benign Neutropenia

Chronic benign neutropenia can be seen at all ages. There are several names applied and the literature is somewhat confusing. The syndromes are all characterized by neutropenia, with ANC as low as zero, with an essentially normal marrow or marrow with a late myeloid arrest.

In classical “chronic benign neutropenia of infancy” the child is otherwise totally normal, with no enlarged liver or spleen, no mucosal infection, and a normal to hypercellular marrow with late myeloid arrest. At the same time as these patients have an ANC of zero, they can have clearly visible pus behind the eardrum or on the tonsils, demonstrating that they are able to get neutrophils to the site of infection. This disorder is thought to be antibody-mediated and resolves within a couple of years.

A small percent of these patients can have infections or oral ulcers, and we do perform bone marrows on these patients. We have also seen a few children with significant infection, including Pseudomonas species skin infections, that had totally normal marrows and went on to be perfectly fine. While it is hard to think of such cases as “benign neutropenia”, the connection between the neutropenia and the infection is not clear. These cases are rare; these patients are probably children who encountered invasive organisms and just happened to have neutropenia.

Ethnic Neutropenia

Some individuals of African descent, Yemenite Jews, Falasha Jews, and certain Middle Eastern tribes have middle neutropenia with ANCs in the 800 to 1300 range. All of these groups are dark-skinned. The bone marrows are essentially normal and the individuals have no increased propensity to infection. Other members of the family may also have mild neutropenia. This is a diagnosis of exclusion. In the total absence of any history of infection or rheumatic symptoms and with a normal collagen vascular screen, marrow examination is not necessary.

Chronic Non-immune Neutropenia

These individuals have mild neutropenia in the range of 1000 to 1400/mm3 and totally negative history and physical exam related to infection or rheumatologic findings. Other causes of neutropenia should be screened (Table V). If everything is negative, marrow examination is unlikely to be of any help. The marrow reserve is normal in these patients. A few of these patients have mild thrombocytopenia as well. The mechanism is increased apoptosis of myeloid progenitors. No special follow-up or treatment is required.

Table V.n

Diagnostic laboratory evaluation for neutropenia


Neutropenia is common after viral infections and usually spontaneously resolved in a few weeks or sooner. For this reason, unless there are other mitigating factors, we prefer to delay evaluation of neutropenia discovered in the course of evolution for fever for several weeks to see if spontaneous recovery occurs. These patients usually have good marrow reserve, but this is not always the case. For this reason, we have placed them in the low- and moderate-risk categories.

Drug-induced Neutropenia

Many drugs cause dose-related suppression of the ANC. The mechanism can vary from decreased production to induction of antibody-mediated neutropenia with normal marrow reserve. They can also cause agranulocytosis and have been responsible for septic deaths. For this reason, we place drug-induced neutropenia in the moderate- and high-risk categories.

Optimally, possibly offending medications should be stopped. If this is not feasible, marrow examination can be helpful. If the drug is critical and marrow reserve is good by marrow examination, medications can be continued with careful observation. The clinical findings of low marrow reserve shown in Table IV can be helpful. If there is evidence of low reserve (* in Table IV) the drugs must be stopped immediately and the patient treated aggressively if infected.

Cyclic Neutropenia

This is an extremely rare inherited form of neutropenia due to mutations in the elastase gene (ELANE, ELA2). There are acquired forms of neutropenia that can cycle as well but should not be considered in this same category. This disorder is relatively benign, and is associated with ANC < 200 for up to 7 days every 21 days. The ANC usually does not go very high during recovery (1500 to 2500 or so) and there is often a reciprocal monocytosis with as many as 50% monocytes when the ANC is low. The diagnosis is established by measuring ANC at least twice a week for eight consecutive weeks to document two complete cycles. Marrow aspiration is not helpful as the results depend entirely on when in the cycle the marrow is examined.

Cyclic neutropenia is dominantly inherited. An ELANE mutation is present in 90% of cases and in 100% of cases with a positive family history. The test for the ELANE mutation can be ordered clinically and is useful to establish the diagnosis. These children usually have flu-like symptoms when neutropenic because of smoldering occult infection but recover as soon as their counts come up.

There have been deaths from overwhelming sepsis. In the reported cases, the children who died had cramping abdominal complaints and had positive cultures for enteric organisms. For this reason we classify the disorder as moderate-risk.

These patients should be treated chronically with GCSF, usually 5 mcg/ kg SQ three times a week, to maintain a nadir ANC > 500 or so. This will decrease the risk of sepsis. The risk of sepsis decreases in the teen years; thus, it probably is acceptable to stop the GCSF then. There are no controlled studies on the management of these patients to provide guidance with this, nor are there likely to be any.

It is hard to regulate the GCSF because of the cycling and difficulty getting counts. We monitor the CBC and ESR monthly. As long as the ANC is above 500 and the ESR is normal, we do not change the dose. A high ESR suggests to us that the ANC is chronically lower than we would like, and we raise the dose.

Nutritional Neutropenia

Nutritional deficiency is not a common cause of neutropenia in the general population. However, in patients with severe chronic illness, significant gastrointestinal surgery (particularly bariatric surgery), and hospitalized patients with multisystem illness and prolonged hospitalization, nutritional deficiency definitely must be considered. It is particularly common in patients with short gut who are on total parenteral nutrition. Physicians have a false sense of security because micronutrients are being given; however, physicians do not realize that the demands may be much higher in critically ill patients.

The nutrients related to marrow failure are listed in Table V. Note that elevation of methylmalonic acid (MMA) indicates B12 deficiency regardless of the B12 level. Homocysteine (HcY) can be elevated in B12 deficiency, folate deficiency, or pyridoxine deficiency. Especially for B12, the downstream metabolites MMA and HcY are the best way to screen and monitor. B12 deficiency can be missed if only serum levels are monitored.

Copper deficiency can cause neutropenia as well as a general marrow failure syndrome indistinguishable from myelodysplasia. Cytopenia including combinations of neutropenia, thrombocytopenia, and macrocytic anemia can develop as the copper binding protein ceruloplasmin drops below normal. Patients with cytopenias recover within about two weeks of normalization of the ceruloplasmin in response to copper supplementation. This problem is seen particularly in patients on parenteral nutrition whose copper supplementation has been stopped because of direct hyperbilirubinemia. In our experience, the cytopenia, as well as elevated direct bilirubin, corrects with copper treatment.

Immune-Mediated Neutropenia

Antibody-mediated neutropenia is usually associated with normal marrow reserve and late myeloid arrest. However, propensity to infection may be increased because of the underlying immune disorder; therefore, we have placed these patients in a moderate-risk category. These children can also be at risk because of immune vasculitis that causes ulcerations in the gastrointestinal track. The clinical signs of mucosal ulcers may not indicate poor marrow reserve in these patients, but may be a direct effect of the autoimmune disorder itself.

Specific treatment of the neutropenia is not required; in fact, GCSF can make the inflammation worse. Treatment of the vasculitis may resolve all symptoms and may not result in correction of the neutropenia. Note that assay of antineutrophil antibodies is not helpful. Each individual antineutrophil antibody assay has about a 20% false negative rate, and the tests are positive in a very high percent of non-neutropenic multiparous females. This makes interpretation of the results clinically not helpful.

Large granular lymphocyte disease (Lymphoproliferative Disorders of Large Granular Lymphocytes, LDLGL)

Immune neutropenia due to cytotoxic T-cells results in marked reductions in myeloid precursors. These disorders are collectively called lymphoproliferative disorders of large granular lymphocytes (LDLGL), among other names. These patients are at significant risk of sepsis and are usually quite symptomatic. They can present with a cyclic neutropenia-like picture. The marrow may be infiltrated with large granular lymphocytes. Flow cytometry shows that greater than 20% of the peripheral lymphocytes are positive for CD3 and CD16, 56, or 57. There is overlap with other rheumatologic syndromes associated with neutropenia.

Pure White Cell Aplasia (PWCA)

PWCA is an acquired disorder associated with absence of myeloid precursors. The patients are at high risk because of absent marrow reserve of myeloid cells. It is due to antibodies directed against myeloid progenitors and has been associated with thymoma, as well as secondary to ibuprofen and other drugs. It is very rare. Diagnosis is made by bone marrow aspirate showing no myeloid precursors with other cell lines intact.

Shwachman-Diamond Syndrome

Shwachman syndrome is a very rare congenital disorder comprised of neutropenia, pancreatic insufficiency, and metaphyseal dysplasia. The inheritance is sporadic or recessive and is associated with a specific mutation in the centromeric region of chromosome 11. The SBDS mutations are present in about 80% of the cases. The patients have variable dysmorphic features, short stature, and fat malabsorption. About 30% go on to develop leukemia.

The neutropenia is variable. Patients with ANC greater than 500 to 800 do not require continual treatment with GCSF, but do require close monitoring for inflammation. Patients with lower ANC should be supported with chronic GCSF. The marrows are variably hypocellular; however, there are no characteristic marrow findings. They can develop monosomy 7 and iso chromosome 7 clones in their marrow, often as a prelude to aplastic anemia or leukemia. However, they can harbor these mutations for years and never develop leukemia.

Kostmann Syndrome

Severe congenital neutropenia (SCN) or Kostmann syndrome is very rare. It presents in infancy with severe neutropenia and often with moderate monocytosis. There are no dysmorphic features. The myeloid reserve is very decreased due to an early myeloid arrest at the progranulocyte stage. The marrow findings in this syndrome are characteristic. About 50% have mutations in the elastase gene (ELANE), the same gene as cyclic neutropenia.

While these patients are at risk for infection, the risk is not as severe as post-chemotherapy patients. They should be treated with GCSF. About 90% will response to doses between 5 and 30 mcg/kg/day, although some require more. Patients who require more than 8 to 10 mcg/kg/day should be considered for bone marrow transplantation even if there is no sibling donor, as their risk of leukemic transformation is high. The overall risk of leukemia is about 25% at 15 years.

What is the evidence?

Dinauer, MC, Coates, TD, Hoffman, R, Furie, B, Benz, EJ. “Disorders of phagocyte function and number”. Hematology: basic principles and practice. 2009. pp. 687

Kyono, W, Coates, TD. “A practical approach to neutrophil disorders”. Pediatr Clin North Am. vol. 49. 2002. pp. 929-71.

Ongoing controversies regarding etiology, diagnosis, treatment

Antibiotic treatment in neutropenic patients is not clearly defined. Each institution has guidelines for management of fever and neutropenia. It is clear that the aggressive approach used in patients with post chemotherapy neutropenia or neutropenia from aplastic anemia is not required for patients with other forms of neutropenia. However, there are no published guidelines.

Patients with cyclic neutropenia most commonly get into trouble with Clostridia species, unlike post-chemotherapy patients. Our clinical experience is weighted by oncology patients whom we have seen quickly get into trouble when febrile. However, is it equally problematic to treat low-risk benign neutropenia patients aggressively with parenteral antibiotics.

Patients with rare congenital neutropenias and chronic neutropenia need to be managed in consultation with a hematologist with significant experience with these disorders.