Hi, good morning. My name is Leon Chen. I'm an advanced practice provider with the Department of Anesthesiology and Clinical Care Medicine at Memorial Sloan Kettering Cancer Center. I'm also a clinical assistant professor at NYU Rory Meyers College of Nursing. Today I have the pleasure of discussing febrile neutropenia. I have no relevant financial disclosure. Febrile neutropenia is actually a complex medical condition and these patients at times can be super sick. About 50% of these patients progress to neutropenic sepsis and they actually require ICU level of care. Now, some of us as critical care clinicians may have already encountered some of these patients. Of course, based on your practice setting, some of us may have encountered more of these patients than others, but I have no doubt that we have all taken care or likely will be taking care of these patients as cancer treatments for patients start to expand more and more. So, it's important for us to all be familiar with some fundamentals of caring for these patients. So, in this brief lecture, I want to define the term febrile neutropenia. I want to help everyone understand how to risk stratify these patients, what are some of the principles of diagnostic decision-making, what are guideline-based treatment options, and we'll finish by summarizing some key learning points. So, first, I'd like to define febrile neutropenia, which seems pretty straightforward. Febrile neutropenia, as the name indicates, is the patient having a fever while neutropenic, although we should still clarify some definitions. When we say having a fever in this population, we mean either a single oral temperature of equal or greater than 30.3 degrees Celsius or an oral temperature of equal or greater than 30 degrees Celsius over an hour period. We emphasize oral temperature because it's a more accurate reflection of the patient's core temperature. And patients are considered neutropenic when their absolute neutrophil count is less than 1,000. As we know, neutrophil is a part of our innate immunity and is a key component of our acute inflammatory process. So, if it's compromised, we won't be able to mount an adequate immune response to potential infections. So, logically, with the increase in severity of neutropenia, the sicker the patient may get. In general, neutropenia is a result of cancer treatments like chemotherapy, and treatment for some cancers tend to be more cytotoxic than others, meaning that those patients tend to have a greater degree of neutropenia and for a longer duration, making them more at risk for clinical deterioration. For reference, I also included the formula to calculate absolute neutrophil count. But let's be honest, majority of the time, if the computer system, the eMAR, doesn't automatically give you the ANC count along with the CBC with differential, you can use one of many calculators available. For my critical care colleagues who are used to caring for fibronutropenic septic patients, the idea of sending these patients home after a short observation seems, frankly, crazy. But this is something our ER colleagues often have to decide who to admit and who to discharge. If they're fortunate, they might have an oncologist colleague to provide consultation and support. But if you're left to your own devices, I do want to give you some key items to consider to help you make that decision. Let's look at a couple of clinical examples. The first patient that we're going to look at is patient 001. He is a 75-year-old male with history of hypertension, lung cancer, undergoing chemotherapy currently. He had a fever at home and followed instruction from his provider to come to the ER. Aside from the fever, he was feeling actually okay. He's able to carry out his activities of daily living. He's able to eat and drink adequately. And he lives with his spouse. In the ER, his vital signs are fairly stable, as you can see, the blood pressure, heart rate, respiratory rate, and saturation on room air. His lab is significant for ANC count of less than 500, which is considered severe neutropenia. Otherwise, his liver and kidney functions both seem to be okay. He's not generating any lactate. And it's debatable whether or not this patient needed an X-ray, a chest X-ray in the first place. But the ER team did get one, and the result is no gross abnormalities. He did a physical exam, which was more or less unrevealing, and he doesn't have a tunnel catheter or a port that might be concerning for Lyme infection. Overall, you can see in the picture, he looks pretty spry for his age and condition. Now, this is something you have to consider. Would you admit this patient for further treatment, or would you discharge him after a bit of sprucing up a little treatment? And how do you decide that? How do we actually decide if a neutropenic patient is low-risk enough to be observed and discharged? So here are some key things to consider. First, you have to consider the type of cancer. Patient one has a solid organ tumor, and in general, these patients undergo regimens that are a little bit less cytotoxic than liquid tumors like leukemia. This means that, in general, they tend to be less neutropenic and usually for shorter duration, typically under seven days. And of course, this is more favorable for the patient. Now, this patient has a history of hypertension. We're not clear how well-controlled it is, but he doesn't have many comorbidities apart from having cancer, and this is also a good thing. He's able to take PO, which is important because that means he's likely – he doesn't have leukocytis, which can be a result of chemotherapy and neutropenia, and also this can be a major source of infection. Additionally, his eligibility for home treatment is obviously dependent on his tolerance for PO med because he likely will have to go home with PO antibiotics. He has a relatively benign exam with no obvious source of infection. With neutropenic patients, sometimes their response to infection can be blunted, so he may not have the typical respiratory symptoms even if he were to have pneumonia. So I think it's fair that the team did an X-ray. Usually if the patient doesn't have an obvious source of infection and it's just febrile, which happens to approximately 50% of patients, they may be able to go home. A couple of things to consider aside from medical considerations that relate to social economic status and access, in general we recommend that the patient should have access to ER within 60 minutes. So if they all of a sudden get worse, they can be seen again by a provider in a timely fashion. Now, I used to live in Upper East Side, and we basically live in between hospitals. So that is not an issue for us, and we take that access for granted. But remember, outside of big cities, 60-minute access to ER might actually be considered a luxury depending on how rural the area is. Additionally, not everyone has a ride to ER. Ambulance and EMT is often voluntary in some municipalities and not considered essential, so it might be limited. Finally, the patient should have someone with them for 24 hours a day, a caregiver of some sort or a spouse, and really to keep an eye on them and can call for help if it's necessary. And not everybody has that. Another thing that's important to point out and is very helpful for a clinician to decide who is considered low risk is this MASK risk index. As you can see, MASK risk index is a very well-studied and well-validated tool in multiple types of oncological patients. It's pretty popular and has been used by many clinicians to identify the low-risk febrile neutropenic patients. Another score that comes up in literature every once in a while is the CISNY score, which in terms of performance is comparable to MASK, but it's a bit harder to use and has not been really as well-studied in liquid tumor patients. The MASK score is pretty straightforward, but the CISNY score looks at things like monocytes and also to calculate a mucositis score, which not everybody knows how to do. The MASK score item seems a little bit more user-friendly if you look through the items, but with that being said, I would use MD-Calc or some other tool to calculate this every single time because I have a hard time memorizing every single element. Our patient, going by the scoring, would score 24, and the magic number that you want him to be is over 21. So anybody who has a score of over 21 is considered low-risk. And the only thing that our patient will be losing points on based on the scale is the age. It's important to know that none of the scoring systems and these tools can really supersede a clinician's judgment. So if the patient has a high score, but somehow it just doesn't look right to you, always trust your instinct. Get a colleague, ask a consultant, but don't feel compelled to disqualify the patient based on one score. When we're evaluating these patients who are febrile and neutropenic, it's important to identify a source of infection and source control whenever possible. Going through the literature, the main source of infection is really three systems, the lungs, GI tract, and skin slash central line. For lungs, we would ask for any respiratory symptoms, but again, remember some of these immune responses may be blunted, so we may not get as a clear-cut respiratory symptoms from the patient. But if there are any suspicions, it would be fair to start with a respiratory viral panel and chest X-ray. GI tract is another big source of infection, so asking the patient if they have any painful swallowing, nausea, vomiting, diarrhea are all important questions. Painful swallowing can be from lymphocytes, which is a source of infection. Any diarrhea will warrant a GI PCR and C. diff study. The patients might also complain of rectal pain during bowel movement, and if that's the case, it might be helpful to do a visual inspection looking for ulceration of fistula. In general, we avoid doing rectal exams on people who are neutropenic. You should also ask for and look for any rashes or signs of cellulitis. Our patient didn't have a tunnel catheter, but if they do, make sure you look at any port or catheters for signs of infection. If they had invasive procedures done, like an LP or bone marrow biopsy, make sure to also look at the site to make sure that that site is not infected. Urine analysis and culture in general is pretty low yield, and in the absence of symptoms or tumor instrumentation near the GU area, we don't routinely culture or obtain those tests. A lot of times, people end up getting treated for piurea unnecessarily, and that, of course, is a bad thing. We usually ask that two sets of peripheral cultures to be drawn, and if there's an implanted catheter, all the lumens should be cultured along with one peripheral set. The yield is much higher if the patient is in septic shock, but otherwise, only 20-30% of all cultures may come back positive. It won't impact our decision-making on treatment initially, but makes sense for antibiotic stewardship. Of course, these days, we have to check a Q-code that put PCR in a positive. We have to consider things like remdesivir or whatever monoclonal antibody treatment that's still effective. Organ dysfunction could be a sign of severe sepsis, or it just might be a sign the patient is sicker than they appear, so in general, those patients will not be sent home. And of course, any clinical instability, I should say, should be pretty self-explanatory, and those patients will not be going home. So now for these febrile neutropenic patients, we want to start antibiotics pretty early on. The data on speciation during febrile neutropenia are mostly stemming from when it was initially described as an entity back in the 1970s, and the organisms are mostly gram-negative bacteria, including Pseudomonas. They usually have pretty high mortality rate if they weren't treated. So now we're starting to see a little bit more gram-positive organisms like Staph and Strep, especially if there's a Lyme infection or heartworm infection that's suspected. And the patient has been neutropenic for an extended period of time, and without proper prophylactic antifungal, the obviously invasive fungal infection should be considered as well. We do want to start with a broad-spectrum antibiotic with good gram-negative coverage, including Pseudomonas. So that would be either Piperacillin, Tazobactam, or Cefepime in most centers. Cefetazidime is a pretty good anti-pseudomonas agent, but not many centers use that as a single agent because of poor gram-positive coverage. So if you are seeing patient one and you're waiting for the lab results, it's good to actually give them either a dose of Piperacillin or Cefepime while waiting. Unless there's a suspicion of MRSA colonization or Lyme infection, the guideline actually does not recommend routine use of vancomycin in these patients. I know some of us would disagree with that, but from our experience. If the patient looks pretty good after the IV antibiotic remains clinically stable and you categorize them as a low-risk patient based on the mask index with adequate social support and proximity to the ER, then this patient could be discharged with PO antibiotic. And for PO antibiotics, we usually recommend amoxicillin, clavironic acid, with a fluoroquinolone. If they have a penicillin allergy, they should go home on acundamycin and also a fluoroquinolone. And once they go home, if the fever deferred for 72 hours, they could actually stop antibiotics. And of course, if the ANC recovers, they can stop antibiotics altogether. So that was our low-risk patient. Let's move on to the second patient. I know he looks a lot like the first patient. They're actually twin brothers. So this one is 70 years old with a history of leukemia. He's undergoing conditioning therapy in preparation for a bone marrow transplant. He was lethargic at home and was febrile to 39 degrees Celsius, so he was brought into ER by family. Vital sign, as you can see, is significantly worse. Low blood pressure, tachy, tachypneic, and also hypoxic, really seems septic by criteria. Lo and behold, a chest X-ray shows he likely has pneumonia. Maybe he aspirated. If you look at the labs, he's much more neutropenic. He actually categorizes profound neutropenic. Since he got induction chemo in preparation for the transplant, those regimens tend to be much more cytotoxic than the ones for solid tumors, which explains his profound low neutropenic state. Also, you can expect him to be neutropenic for quite a while. Additionally, he seems to have an ATI, thrombocytopenic, and is generating a lactate. His lung sounds correlate with his imaging, and it seems like he has a good lung source of infection. I don't think it'd be controversial that this man needs to be admitted to the hospital for further treatment. For this gentleman, we still start with either epiphthalate or cefepime for gram-negative coverage. As a whole, there's no evidence to support double gram-negative coverage, like adding an inviscolytic site, unless the patient is profoundly unstable. And if you have suspicion for a resistant organism because of a prior culture, it's fair to upgrade to an antibiotic like meropenem, which can cover ESBL organisms. Given that this patient seems pretty sick, it's fair to add vancomycin for MRSA coverage, even though we don't have any mucositis or Lyme infection that demonstrate that kind of infection. Although, you could argue that it's worth adding vanco for the off chance that this patient has MRSA pneumonia, which by itself would have pretty bad outcomes. Often, these patients who undergo intense induction therapy are expected to be neutrophilic for a prolonged period of time, so it's common for them to be on a prophylactic antibiotic like a fluoroquinolone, an antifungal like a portoconazole, and usually an antiviral like a cyclavir. And given that a majority of these cases are still caused by bacterial infection, there's not an immediate need to start antifungal therapy, although it's important to send fungal markers such as BD-glucan or galactomannan along with the blood cultures initially. If the patient's persistently neutrophilic and the fever is not defermenting after four to seven days, and he's been on good broad-spectrum antibiotics, or if the fungal marker or imaging is consistent with a fungal infection, then obviously we would start antifungal therapy. The two groups of organisms that can be broadly categorized are the cannula species or molds such as invasive Aspergillus, the old-school trial and true antifungal, if you're not sure which organism it could be and the patient's sick, you would use amphotericin B. But as we all know, it has a pretty significant side effect profile and can be quite toxic. You can also give an echinocanin like a mycofungin, which is pretty well-tolerated and effective against candida species with minimal side effects. But if you suspect mold infection, then an azole such as boriconazole is a good choice. Intuitively, some people would ask, if the lack of neutrophil is the issue, why can't we just give them GCSF and let them produce neutrophils again? So the guideline is a little bit controversial, but as a whole, it's never been proven to have any mortality benefits. It has been, however, been shown to decrease time of neutropenia with no real risk of harm. So depending on the oncologist and how sick the patient is, some of them would recommend giving it for patients who we think are going to do poorly without additional intervention. Based on studies, the patients who we think are going to do poorly are the ones who are older, with more prolonged and profoundly neutropenic patients, those with high disease burden, pneumonia, who have septic shock, proven invasive fungal infection, and those who happen to be hospitalized when they develop febrile neutropenia. In these patients, it's fair to consult and work closely with the oncology team to see if we should start GCSF in addition to conventional antibiotic therapy. 50% of those patients who have febrile neutropenia will eventually develop sepsis, and 10% of those patients will progress to septic shock, needing intensive care unit. Those patients usually do very poorly. The liquid tumor patients who end up with septic shock and end up in ICU has an approximately 46% chance of mortality. Some of the predictors for mortality include sepsis after a stem cell transplant, the presence of graft-versus-host disease, respiratory failure requiring mechanical dilation, positive blood cultures, cardiac failure, renal failure, and hepatic failure. Once the febrile neutropenia progress to sepsis or septic shock, we're back in our critical care element. So the focus should be on adequate and appropriate IV fluid administration, appropriate vasopressor for hemodynamic support. If there's a source of infection that we can control, like an abscess, of course, we should pursue invasive interventions such as drainage. If we're profoundly hypotensive despite vasopressors, it's fair to start stress-still steroids. I discussed the GCSF utility. In rare cases, we have utilized granulocyte transfusion. The data isn't clear to support routine use of this to treat super sick patients. With that being said, it's important that we work with the oncology team. If that's something they want to pursue, I definitely wouldn't fight them on that. The key takeaway I want everybody to have from this lecture is that febrile neutropenic patients could potentially become very, very sick. So they should be seen in a timely fashion. And there's a small subset of low-risk patients that may be able to treat it as an outpatient if they qualify by various validator tools and clinical judgment. Empirical antibiotics should be started as soon as possible, and you should source control whenever you're able to. And substance management should be done per the SECM guideline. And always ask for help from the oncologist specialist for help. And with that, I thank you for your time.

febrile neutropenia

cancer

neutropenic sepsis

ICU care

risk stratification

antibiotic treatment