Bacterial infections after bone marrow transplant

What every physician needs to know about bacterial infections after bone marrow transplant:

Bacterial infections are common after bone marrow transplant (BMT). If not promptly treated, these infections have the potential to rapidly progress, with significant risk for morbidity and mortality. For this reason, antibacterial prophylaxis during the highest risk periods, and early initiation of therapy when an infection is suspected, are imperative and potentially lifesaving. However, the critical role of widespread use of antibiotics in this population comes with costs. These includes toxicities directly related to the drugs themselves and indirect effects related to alterations in patients’ microbiome. The latter may manifest as increased carriage of multi-drug resistant bacteria and Candida, and infection due to Clostridium difficile.

Invasive bacterial infections occur when impaired host defenses permit organisms present at mucosal or cutaneous surfaces (e.g., oropharynx, gastrointestinal tract, upper respiratory tract, and skin) to gain access to deeper sites. The specific infecting organism depends upon what is colonizing those surfaces. In that regard, colonization with organisms that are antibiotic resistant and/or have high pathogenic potential is an important factor.

The risk and site of infection as well as infecting organism are heavily influenced by presence of anatomical and physiological impairment in host defenses. For example, mucositis due to cytotoxic chemotherapy or graft-versus-host disease (GVHD) predisposes patients to bloodstream infections with oral and enteric organisms. Disruption of skin by catheters can lead to localized and bloodstream infections with organisms found on cutaneous surfaces. Impaired clearance of respiratory secretions, structural lung disease, or damaged airway epithelium from a viral infection can all predispose patients to pulmonary infections due to organisms colonizing the upper airways and oropharynx.

Continue Reading

The highest risk periods are during times of neutropenia and when anatomical barriers are breached. Infections due to gram-negative bacilli, cutaneous gram-positive organisms, and streptococci and enterococci of enteric origin predominate during these periods. Additional contributing factors include defects in lymphocyte number and function (including hypogammaglobulinemia). Infections with Streptococcus pneumoniae take on an important role under such circumstances.

Antibacterial prophylaxis during times of highest risk is an important facet of BMT care. These include the period between stem cell infusion and recovery from neutropenia and during episodes of GVHD requiring treatment with high doses of systemic corticosteroids. To be useful, a prophylactic agent must target the most likely organisms, possess favorable toxicity and drug interaction profiles, and be available in an oral formulation.

For these reasons, antibacterial prophylaxis is typically with a quinolone (e.g., moxifloxacin 400 mg/day, levofloxacin 500 mg/day, or ciprofloxacin 500 mg twice daily). Potential problems with this class of antibiotics include tendon, joint, and muscle injury, neurological side effects, and alteration of microbial flora favoring C. difficile infection. For patients who are unable to tolerate quinolones, trimethoprim sulfamethoxazole (TMP/SMX) may be considered. Patients who are expected to have a relatively short period of neutropenia (e.g., those receiving non-myeloablative conditioning regimens and autologous transplant recipients) may not always require antibacterial prophylaxis in the pre-engraftment phase. Increasing rates of quinolone resistance at some centers is a worrisome development that may require alteration in prophylaxis approaches.

BMT patients are also at risk for Pneumococcus. In addition to pneumococcal vaccination beginning at 3-6 months after transplant, prophylaxis should also be given for at least a year after transplant and longer if receiving immunosuppressive therapy for chronic GVHD. Penicillin is generally suitable, but if daily TMP/SMX is already being given for the prevention of other opportunistic infections, that usually suffices.

A very short discussion of the microbiology is presented here, as these concepts are important to understand in order to craft effective prevention and treatment strategies.

Aerobic bacteria generally come in two “flavors”, depending on how they appear on Gram stain. Gram-positive bacteria appear purple, secondary to internalized crystal violet that remains after alcohol-based decolorization; gram-negative bacteria appear pink, as the decolorization step more effectively destroys their cell walls. This tells you something about the type of drugs that we use; the peptidoglycan-rich cell wall in gram-positive bacteria is a target for some old (vancomycin) and new drugs (daptomycin), and changes in the cell wall can render resistance. Likewise, in gram-negative bacteria, where the target of the drug is frequently intracellular, bacteria can do crafty things like increase expression of efflux pumps to become resistant over time.

Important gram-positive bacteria in bone marrow transplant patients include Coagulase-negative Staphylococci, Staphylococcus aureus, Enterococcus faecalis and Enterococcus faecium and Viridans group streptococci. Important gram-negatives are the Enterobacteriaceae (e.g., Escherichia coli, Citrobacter, Klebsiella, and Enterobacter), Pseudomonas aeruginosa, Acinetobacter, and Stenotrophomonas.

The proliferation of multidrug-resistant gram-negative and gram-positive bacteria is a major challenge in BMT patients. Infections due to extended-spectrum beta-lactamases (ESBL) and Klebsiella pneumoniae carbapenemase (KPC)-producing Enterobacteriaceae, Stenotrophomonas spp., Acinetobacter spp., and Vancomycin-resistant Enterococcus faecium (VRE) are frequently refractory to typical first-line antibacterial regimens. Multifaceted approaches that include standard infection control practices to prevent cross-contamination between patients and antimicrobial stewardship programs that assist in the proper use of precious antibiotics have been only partially effective in keeping this problem at bay. The importance of getting the right antibiotic to patients in a timely manner has even led to an increased use of drugs previously thought of as those of last resort (e.g., carbapenems, daptomycin, linezolid, and colistin) in empirical therapy situations.

There are several key concepts that BMT physicians should be aware of, focused largely on risks and prevention, appropriate management in an era of antimicrobial drug resistance, and classic presentations of typical syndromes. These are discussed below, with a syndromic approach.

Gram-negative bacteria, such as Enterobacteriaceae and P. aeruginosa, were historically the most common cause of bloodstream invasion, until effective prevention strategies changed our epidemiology in favor of infection caused by gram-positive bacteria. All of these organisms usually come from the skin, GI tract, or respiratory tract before invading blood. Bacteremia risks are particularly high during the first month after BMT, in association with GI tract mucositis. However, development of acute GVHD confers extended risks for bacteremia after allogeneic BMT.

When dealing with bloodstream infection the following concepts must be kept in mind:

  • Delays in appropriate therapy can prove catastrophic, so initial therapy must take into account the most likely infecting organisms.

  • The most likely sources are the intravenous catheter, the GI tract (especially when there is mucositis), the lungs, and the urinary tract.

  • Treatment duration and complications differ for each source, so it’s important to identify it if possible with cultures and radiographs.

The initial regimen:

An initial regimen of vancomycin (20-25 mg/kg loading dose, followed by 15-20 mg/kg every 8-12 hours) AND either cefepime (2 grams IV every 8 hours) or piperacillin/tazobactam (4.5 grams every 6 hours) can serve as a basic starting point.

Modifications of the basic regimen should be made based on suspected source of infection, concern for resistant organisms, and clinical status of the patient. If a clinically evident site of infection (e.g., lungs, GI, or GU tract) is present, knowledge of the organisms likely to be found at those sites can serve as a guide for initial adjustment. For example, infections from GI sites may also require anaerobic coverage (hence favoring piperacillin/tazobactam or addition of metronidazole to a cefepime-based regimen).

If a patient is known to be colonized with an ESBL-producing gram-negative organism and/or has had recent exposure to a broad-spectrum beta-lactam antibiotic, empiric therapy may require the use of an anti-pseudomonal carbapenem (e.g., meropenem 1gram IV every 8 hours, imipenem/cilastatin 500 mg IV every 6 hours, or doripenem 500 mg IV every 8 hours). The addition of a second agent with good gram-negative activity, such as an aminoglycoside (e.g., gentamicin, tobramycin, amikacin) or a quinolone (e.g., ciprofloxacin or levofloxacin, but not moxifloxacin) so as to broaden the spectrum of coverage, may be considered in patients with hypotension or who are requiring pressors.

Please note that all regimens may require adjustment in patients with abnormal renal function and that selection of antibiotics must take into account local susceptibility patterns. Once additional information is available, such as results of microbiology tests, response to therapy, and new clinical findings, the initial regimen should be adjusted accordingly.

The catheter:

Bacteremia isn’t always from the catheter and is frequently from the gut, especially with gram-negative bacteria. The catheter doesn’t always have to be removed.

Catheters should be removed in the following circumstances:

  • Infection at the insertion site (e.g., port pocket or tunnel infection).

  • Clinical instability despite appropriate resuscitation efforts.

  • Septic thrombophlebitis.

  • Persistent bacteremia despite appropriate antibiotics.

Removal of the device should also be strongly considered when bacteremia is due to the following bacteria: S. aureus, Corynebacterium jeikeium, atypical mycobacteria, Bacillus species, vancomycin-resistant enterococci, P. aeruginosa, and S. maltophilia.

Some of the more common “problem bugs” are discussed here briefly. This list is not complete!

P. aeruginosa is a gram-negative bacterium that is commonly found in the environment and in water supplies. Because it can create biofilms readily, the most common infections that it causes are associated with catheters and the lungs.

It can readily become resistant to drugs using multiple different mechanisms. Occasionally, you will see an organism with pan-antibacterial susceptibility; this is a tip that the bacteria likely came from the environment and sneaked into the bloodstream through a catheter rather than as part of endogenous (GI or respiratory tract) flora. Other environmental gram-negatives do this too. It’s not just an academic issue, as this tip can be helpful when considering whether the catheter needs to be removed.

There’s controversy regarding whether severe P. aeruginosa infections require combination drug therapy, with no definitive data giving us guidance. We recommend a basic initial treatment with cefepime (2 grams IV every 8 hours), ceftazidime (2 grams IV every 8 hours), piperacillin/tazobactam (4.5 grams IV every 6 hours), or anti-pseudomonal carbapenem (if resistance to one of the other drugs is suspected). The addition of an aminoglycoside (e.g., gentamicin, tobramycin, amikacin) or a quinolone (e.g., ciprofloxacin or levofloxacin) should be considered based upon severity of illness, initial source of infection, local susceptibility patterns, and the patient’s organ function. Doses may require adjustment with abnormal renal function.

We favor combination therapy in patients that are unstable, have infection at a site with poor antibiotic penetration, such as the lungs, are colonized with multidrug resistant Pseudomonas, or are at an institution where such organisms are prevalent. We try to avoid aminoglycoside use in patients at risk for renal toxicity, but this is not always feasible. Once the susceptibility profile of the organism is known, a switch over to monotherapy can be considered. However, when someone has a clear focus of infection that might be difficult to remove – the perfect example is pneumonia – longer therapy with two drugs may be better as it may help decrease the likelihood of emergent resistance and poor responses. However, even in such cases, consideration should be given to stopping one of the agents (particularly the aminoglycoside, if used) after 5-7 days of therapy.

There are data that suggest that pneumonia caused by P. aeruginosa should be treated longer than the standard 2-week regimen, as BMT recipients have a higher rate of recurrent pneumonia and poor outcomes. In the setting of pneumonia, we extend therapy to at least 3 weeks.

Resistance to beta-lactam antibiotics can be exhibited by many different mechanisms, and is common. However, certain Enterobacteriaceae have acquired mechanisms to become resistant to the extended-spectrum beta-lactam antibiotics, by producing these enzymes – “ESBLs”. This is most common with E. coli and K. pneumoniae, and it has become quite a problem in certain hospitals, and especially in certain parts of the world, such as in Asia.

In BMT patients, ESBL-associated infections carry a high mortality, largely due to difficulties in choosing the most appropriate initial drug. Keep in mind that antibiotic susceptibility tests can be confusing with these organisms, as results may show susceptibility to extended-spectrum beta-lactam antibiotics (such as piperacillin-tazobactam); the laboratory typically identifies which organisms appear as if they can become resistant, and does another test to indicate that the organism is an ESBL-producer.

In the setting in which this is suspected, carbapenems i.e., imipenem/cilastatin, meropenem, doripenem, and ertapenem (1 gram IV every 24 hours in normal renal function) are the appropriate first-line therapy. Keep in mind that if infection with P. aeruginosa is also suspected, ertapenem should not be used as it is NOT active against that organism.

It is wise to treat people who are known to be colonized (or previously infected) with ESBL-producing organisms with carbapenems first-line, even with suspected infection during fever and neutropenia. Risks should be judged according to prior infections, colonizing organisms (if known), as well as hospital epidemiology.

Despite the name, these carbapenemases can be produced by a range of bacteria including Enterobacteriaceae (including. Klebsiella spp., E. coli, and Enterobacter spp.), as well as some P. aeruginosa and Acinetobacter spp. isolates. Such organisms are often resistant to carbapenems and multiple other beta-lactam antibiotics.

Infection with Klebsiella pneumoniae carbapenemase (KPC)-producing bacteria is typically associated with poor outcomes, again due to the difficulty in choosing the appropriate first-line drug, and the fact that these organisms can demonstrate resistance to many different classes of drugs simultaneously. These infections can be difficult to treat, and may require a combination of different classes of drugs, use of drugs with narrow therapeutic windows (e.g., colistin and aminoglycosides), and extended or continuous infusions of beta-lactam or carbapenems; therapy should be guided by susceptibility testing, and Infectious Diseases (ID) consultants should be involved.

Acinetobacter baumanii is another potentially problematic gram-negative organism that can cause disease in BMT patients, usually by infected catheters or lungs. These infections can be very severe, as this organism is another that can become resistant to essentially all drugs, through many different mechanisms. Drugs that were previously considered too toxic for widespread use (e.g., colistin) are now coming back off-the-shelf to treat these infections.

Methicillin-resistant Staphylococcus aureus (MRSA) strains are a problem in the hospital, most frequently causing disease in the bloodstream (via catheter) and wounds, and lung infections. By definition, oxacillin and cefazolin, which are the treatment of choice for S. aureus are not effective in MRSA. Vancomycin with target trough levels of 15-20 remains the first-line therapy for serious MRSA infections. Other options are linezolid (600 mg every 12 hours), daptomycin (6-12 mg/kg daily), and ceftaroline (600 mg every 8 hours in normal renal function). A couple of studies have shown linezolid to be relatively safe in non-oncology/BMT patients. Platelet counts can decrease with extended therapy, warranting close monitoring. Daptomycin doesn’t work for pneumonia, but otherwise it is a good option; take care with dosing this drug aggressively enough but according to weight and renal function.

“Nosocomial” MRSA is the organism that has been around for a while; other, “community acquired” MRSA strains (CA-MRSA) have become more common, especially as a cause of skin and soft tissue infections and aggressive pneumonia. Most infections in BMT patients are nosocomial strains, typified by higher level resistance to other drugs as well (e.g., clindamycin). However, patients may develop early pneumonia with CA-MRSA after conditioning, with poor outcomes. In this setting, infection likely progresses from respiratory tract colonization pre-conditioning.

These organisms are “sticky”, and care should be taken to assure that no other sites are infected after bloodstream invasion, using echocardiography and other tests, driven by symptoms. Catheters really should be changed in the setting of bloodstream infection with MRSA. Aggressive source-control management with fluid drainage is necessary.

E. faecium (and less frequently E. faecalis) can become resistant to vancomycin, and it’s creating quite a problem in some transplant centers. Typically, these organisms cause more chronic colonization and infection through the GI tract, urinary tract, and catheters; rarely does VRE actually cause pneumonia, so think about other causes even in people with sputum showing Enterococcus.

Infections can be difficult to clear, and patients are usually sick so outcomes are poor, although there is debate about the proportional attributable mortality vs. disease-related mortality.

Linezolid and daptomycin are treatment options; keep in mind that resistance to both can occur as these organisms start changing the structure of their cell walls. Tigecycline may be also be useful, but not as first-line treatment. Source control needs to be considered priority.

Pneumonia is one of the most common infectious complications in BMT recipients. Risks include immunologic and airway deficiencies, and extend late after allogeneic BMT, largely due to immunoglobulin deficiency. The most common radiographic findings include air-space consolidations, small centrilobular nodules, and ground-glass opacities; distribution most frequently involves the central and peripheral regions of the middle and lower lung zones. Common causative organisms include E. coli, P. aeruginosa, S. pneumoniae, Viridans group streptococci, Haemophilus influenza, and Legionella.

The differential diagnosis of pneumonia includes non-infectious etiologies, especially late after allogeneic BMT (e.g., bronchiolitis obliterans organizing pneumonia [BOOP]). For this reason, and because there are numerous different drug-resistant bacteria, fungi, and viruses that cause infection, diagnosis should be aggressively pursued.

Several studies have shown safety in performing bronchoalveolar lavage (BAL) after transplant, even early pre-engraftment; lavage for culture provides more information for suspected bacterial pneumonia compared to biopsy (with the exception of other more chronic infections such as Mycobacterial infection).

One retrospective study by Shannon et al. of 501 consecutive, adult, non-intubated patients who underwent 598 BALs for evaluation of new pulmonary infiltrates during the first 100 days of BMT reported that the overall yield of BAL for clinically significant pathogens was 55%, with a 2.5-fold higher recovery in procedures performed within the first 4 days of presentation compared to those performed late. However, late antibiotic adjustments were associated with a higher rate of death. We believe that BAL should be performed whenever possible, even when biopsy is not possible due to thrombocytopenia.

Keep in mind that Enterococci and Candida species are poor pulmonary pathogens and may not be the cause of disease even when recovered from airway secretions.

Stenotrophomonas spp. can be resistant to the drugs that we typically use empirically – extended-spectrum beta-lactam antibiotics, and carbapenems. One needs to keep this in mind in people who are failing carbapenem therapy, especially if they are known to be colonized or have a history of prior infection. Treatment is typically with TMP/SMX (15-20 mg/kg/day given in 3 divided doses in patients with normal renal function), but levofloxacin and ceftazidime are other options, depending on susceptibility.

Ticarcillin/tazobactam is also active against some Stenotrophomonas isolates, but is no longer available in the US.

Clinicians should be aware that there are numerous bacterial causes of pneumonia, some of which necessitate additional empirical treatment. For instance, Legionella is a known cause of pneumonia, sometimes severe and associated with outbreaks. As antigen assays only detect a specific serotype of Legionella, negative tests do not rule out infection. This is a basis for a quinolone or macrolide in empirical pneumonia treatment regimens.

Other bacteria cause more chronic pulmonary infections, such as Nocardia species, Actinomycetes, and Mycobacteria (tuberculosis [TB] and non-TB). These may manifest in more nodular lesions and disease involving upper lobes (especially with reactivation TB). Treatment of both TB- and non-TB mycobacterial infections (which can also involve the catheter) is complicated and warrants advice from ID experts.

Neutropenic enterocolitis

Neutropenic enterocolitis occurs when the gut barrier is damaged by conditioning therapy, allowing for the invasion of organisms that inhabit the gut lumen. It typically presents as abdominal pain, fever, and diarrhea and evaluation should include a computed tomography (CT) scan of the abdomen. It most frequently is caused by a mix of aerobic and anaerobic bacteria and should thus be treated with an antibacterial drug with wide-spectrum activity (e.g., piperacillin-tazobactam or a carbapenem). Vancomycin is not usually indicated.

We have seen enterocolitis caused by focal invasion of fungi, especially Candida albicans and even filamentous fungi; keep this in mind especially when treatment is not responding to antibacterial therapy. The role of surgery is a little questionable; most patients can be treated successfully without surgical resection. The need for a more aggressive management should be judged by serial CT scans and clinical examination.

Clostridium difficile colitis

C. difficile is another important cause of colitis, usually manifested by diarrhea. It most frequently occurs early after BMT, both after autologous and allogeneic transplant recipients, although allogeneic BMT patients appear to have a low level of protracted risks, especially with concurrent GVHD. Studies have shown that C. difficile disease occurs both early and late after allogeneic BMT, and people who get C. difficile-associated diarrhea (CDAD) have particularly high risks for GI tract GVHD.

The clinical signs and symptoms of disease do not mimic what is found in the generalized hospital population, as these patients may not mount a significant neutrophil count and severe complications requiring surgical management are actually uncommon.

Different tests are available to diagnose C. difficile disease: these detect the presence of the bacterial toxin in stool. Polymerase chain reaction (PCR) is more sensitive than enzyme immunoassays, and many institutions are moving towards utilizing these advanced molecular tests.

Treatment options include vancomycin (125-500 mg 4 times daily by mouth or nasogastric tube-NOT IV) and metronidazole (500 mg every 8 hours). There is some indication that vancomycin is preferred, with fewer recurrences, and it is our drug of choice. In patients with severe disease, especially those who are not taking oral medication, the combination of the two may be applied.

Recurrence is a major problem in C. difficile infection. Fidaxomicin (200 mg by mouth twice daily) may be associated with fewer recurrences compared to vancomycin, however, it has not yet been studied specifically in the BMT population. Limiting exposure to antibacterials that alter the gut flora is a good idea, when possible; people with recurrent disease should be referred to ID specialists for consideration of more definitive treatments (such as long-tapered vancomycin). Probiotic (e.g., Lactobacillus or Saccharomyces preparations) and fecal transplants are gaining favor, but cannot be recommended in BMT patients at this time due to safety concerns.

The most common causes of cellulitis are streptococci and Staphylococcus species, however, in hospitalized patients and in those with altered immunity, infection can be caused by other organisms, such as gram-negative bacteria. For this reason, we treat cellulitis with a broader spectrum regimen in BMT patients, providing coverage for gram-negatives as well. It is important to assure that no other focus of infection is apparent and that there are not drainable fluid collections underlying the skin, as these usually require drainage.

Skin infections, especially those that involve catheters or non-removable foci, usually get worse even with effective therapy and especially during neutrophil engraftment; for this reason, surgery may become necessary.

There are several soft tissue infections that require more aggressive surgical exploration and drainage: necrotizing fasciitis and Fournier’s gangrene can be caused by numerous different organisms and rapidly progress without aggressive surgical management. In such circumstances, broad antibiotic therapy that includes coverage for aerobic gram-negative organisms, MRSA, and anaerobes is necessary, as well as rapid surgical evaluation.

What features of the presentation will guide me towards possible causes and next treatment steps?


What laboratory studies should you order to help make the diagnosis and how should you interpret the results?


What conditions can underlie bacterial infections after bone marrow transplant:


When do you need to get more aggressive tests?


What imaging studies (if any) will be helpful?


What therapies should you initiate immediately and under what circumstances – even if root cause is unidentified?


What other therapies are helpful for reducing complications?


What should you tell the patient and family about prognosis?


“What if” scenarios.




What other clinical manifestations may help me to diagnose bacterial infections after bone marrow transplant?


What other additional laboratory studies may be ordered?


What’s the Evidence?

Tomblyn, M, Chiller, T, Einsele, H, Gress, R. “Guidelines for preventing infectious complications among hematopoietic cell transplant recipients: a global perspective. Preface”. Bone Marrow Transplant.. vol. 44. 2009. pp. 453-5. (This comprehensive article provides important details regarding strategies, including targeted prophylaxis, for prevention of bacterial infections in BMT.)

Rubin, LG, Levin, MJ, Ljungman, P, Davies, EG. “2013 IDSA clinical practice guideline for vaccination of the immunocompromised host”. Clin Infect Dis.. vol. 58. 2014. pp. 309-18. (This article provides state of the art guidance for utilization of vaccination as a means of preventing a range of infections in immunocompromised patients.)

Bock, AM, Cao, Q, Ferrieri, P, Young, JH, Weisdorf, DJ.. “Bacteremia in blood or marrow transplantation patients: clinical risk factors for infection and emerging antibiotic resistance”. Biol Blood Marrow Transplant.. vol. 19. 2013. pp. 102-108. (This study demonstrates epidemiology, risks, and antibiotic sensitivity profiles of bacteremias in one transplant center over a recent 5-year period.)

Hakki, M, Limaye, AP, Kim, HW, Kirby, KA. “Invasive Pseudomonas aeruginosa infections: high rate of recurrence and mortality after hematopoietic cell transplantation”. Bone Marrow Transplant.. vol. 39. 2007 Jun. pp. 687-93. (This study is informative with consideration of predicted outcomes and duration of antibiotic therapy for pneumonia caused by P. aeruginosa.)

Zuckerman, T, Benyamini, N, Sprecher, H. “SCT in patients with carbapenem resistant Klebsiella pneumoniae: a single center experience with oral gentamicin for the eradication of carrier state”. Bone Marrow Transplant.. vol. 46. 2011. pp. 1226-30. (This is one example of new investigational strategies to decolonize patients in the setting of complicated endogenous drug resistance.)

Mitchell, AE, Derrington, P, Turner, P, Hunt, LP. “Gram-negative bacteraemia (GNB) after 428 unrelated donor bone marrow transplants (UD-BMT): risk factors, prophylaxis, therapy and outcome”. Bone Marrow Transplant.. vol. 33. 2004. pp. 303-10. (This study demonstrates high risks in patients with GVHD and related therapies, also outlining epidemiology.)

Engelhard, D, Cordonnier, C, Shaw, PJ. “Infections Disease Working Party for the European Bone Marrow Transplantation (IDWP-EBMT). Early and late invasive pneumococcal infection following stem cell transplantation: a European Bone Marrow Transplantation survey”. Br J Haematol.. vol. 117. 2002. pp. 444-50. (This large study demonstrates prolonged risks for pneumococcus infection post-allogeneic BMT.)

Shannon, VR, Andersson, BS, Lei, X, Champlin, RE, Kontoyiannis, DP.. “Utility of early versus late fiberoptic bronchoscopy in the evaluation of new pulmonary infiltrates following hematopoietic stem cell transplantation”. Bone Marrow Transplant.. vol. 45. 2010. pp. 647-55. (This large study demonstrates the relative safety and efficacy of bronchoscopy in BMT patients.)

Alonso, CD, Treadway, SB, Hanna, DB, Huff, CA. “Epidemiology and outcomes of Clostridium difficile infections in hematopoietic stem cell transplant recipients”. Clin Infect Dis. vol. 54. 2012. pp. 1053-63. (This retrospective study outlines risks and outcomes of C. difficile disease in a large cohort of BMT recipients.)