John Patlan, M.D.
Associate Professor
General Internal Medicine
The University of Texas MD Anderson Cancer Center
Welcome! My name is John Patlan. I am a general internist here at The University of Texas MD Anderson Cancer Center in Houston and I am going to be talking to you today about oncologic emergencies. This is Part 2 of a two-part series. In Part 1, we talked about structural and metabolic problems in cancer patients.
And, in this section, we will talk about hematologic and infectious problems in cancer patients.
So, among the hematologic problems, we see a wide range of problems, as you can imagine. Patients with leukemia, for instance, may have hyperleukocytosis and symptoms related to that, which we call leukostasis. Very commonly patients have very low platelet counts. They have thrombocytopenia and have bleeding complications from that, sometimes very serious such as intracranial bleeding. A very, very common problem among cancer patients is cancer-related anemia, which happens for lots of reasons, frequently bone marrow suppression due to myelosuppressive chemotherapy, sometimes hemorrhage, other reasons. You can have patients who have disseminated intravascular coagulation either as a manifestation of sepsis of some significant acute problem or sometimes as a complication of the cancer itself, for instance acute myeloid leukemia. And very commonly we'll see patients with transfusion reactions. So, we see this whole spectrum of hematologic problems. We are really going to focus our attention on the most specific of these, which is the first one, the hyperleukocytosis and leukostasis.
So, hyperleukocytosis just means very elevated white blood cell count. So this, as you would expect, occurs in leukemia patients, usually patients with acute myeloid leukemia more than acute lymphoblastic leukemia. And it is more common in subtypes M4 and M5. And basically, in leukemia patients, their white blood cell count can rise very rapidly. Within a period of hours or days, they can have a doubling of their white blood cell count. When the counts get very, very high, 100,000 or higher, they can start to develop symptoms because of this. The symptoms are really because these blast cells, the myeloid blasts in particular, are very large, they are poorly deformable, they have a hard time fitting through the microcirculation, they increase the viscosities of the blood, and they have a very high metabolic activity, so they trigger some cytokine production, some endothelial damages, some local inflammatory response there, which helps produce the symptoms.
So, because this is a microcirculatory problem, the symptoms are really where you would expect a microcirculatory problem to produce symptoms. That is mainly CNS symptoms and pulmonary symptoms. It can be kind of nonspecific. Patients can have headaches, mental status changes. They can have cough, some shortness of breath. They may be hypoxic, may or may not have an infiltrate on chest x-ray. Most of these patients are febrile. And it is very difficult to try to sort out whether this is because of the local inflammatory response or because of infection. Rarely patients can have symptoms in other organs: myocardial ischemia, renal insufficiency, limb or bowel ischemia. I just list that even though it is rare because you should be prepared to look for that. Now, the diagnosis really is not a challenge. A patient that comes in that has acute leukemia, very, very high white blood cell count and some of the symptoms that I mentioned, the treatment is what becomes a challenge. Because this is a problem because of the very high white blood cell count, ultimately the treatment is going to be a rapid reduction of the white blood cell count. Now, if, depending on where you are, this may be accomplished by leukophoresis. For instance, if the patient is in a large referral center, like here at MD Anderson, this involves placement of a large bore catheter and removal of the white blood cells in a leukophoresis machine and then delivering back the red blood cells and plasma to the patient. In other places, it may be accomplished by induction chemotherapy. If you are in a smaller community hospital for instance and you are waiting to transfer the patients to a larger hospital, to a referral center, you can try to stabilize the patient with oral hydroxyurea, large doses, 4-8 gm/day that can at least stabilize the white blood cell count and hopefully stave off the more serious complications until you can get them to a more definitive treatment center. A lot of these patients will be anemic and require some transfusion support. If it's required, please try to do it slowly, because you want to try to avoid raising the blood viscosity as much as you can. That could potentially make the symptoms worse. Try to make sure patients are adequately supported otherwise: hydration, oxygenation, antibiotic therapy if you suspect infection.
Now, we are also going to talk about infectious problems. This is probably the biggest problem that we encounter in the emergency setting among cancer patients. This is both because of the underlying disease, many cancer patients, leukemia patients, lymphoma patients for instance have sort of underlying immune defects at baseline, and partly because of the treatment that we give them. When we give them chemotherapy, which can be myelosuppressive, it will lower their white blood cell count to make them more susceptible to infection. So, in the short term, infection is by far the biggest threat to the health of the cancer patient. So, this is --- this is, for instance, the number one problem that we see in the Emergency Center at MD Anderson Cancer Center is fever and infection, a variety of infectious problems: neutropenic fever, which we will discuss at length in the next section, central venous catheter infections, which happen very commonly because they are so commonly used, pneumonias whether that is post-obstructive from a thoracic malignancy or just de novo, biliary obstruction or cholangitis in patients with GI malignancies and every other kind of infectious complication that you can imagine happens in the cancer center.
But we are really going to focus our attention on neutropenic fever, because it is a very common problem that you need to know how to manage. Now, both body temperature and neutrophil count exist on a continuum. And so, calling someone febrile or neutropenic requires a definition somewhat arbitrary, but these are the guidelines that have been established by the Infectious Diseases Society of America. What we are going to call a fever, for this purpose, is any temperature, any single temperature greater than 38.3°C, which is about 101°F, or a sustained temperature greater than 38°C, which is 100.4°F for more than an hour. So, any patient who meets those criteria will be treated as febrile. And our definition of neutropenia is any neutrophil count, an absolute neutrophil count less than 500, or if they have a neutrophil count less than 1000 at the moment that you are seeing them with fever, but you predict that their nadir may be less 500. The reason that the neutrophil count is important is that the risk of infection rises with the severity and the duration of their neutropenia. So, the lower it goes and the longer it stays down, the higher their risk for infection.
Now, neutropenic fever can happen for lots of reasons. When patients do have serious bacterial infections, they are predisposed to that, because, when we give them chemotherapy, for instance, that can induce mucositis, inflammation of their --- of the lining of their GI tract, you can see that in their mouth frequently. But it can happen anywhere in the GI tract from the mouth to the anus. So that just makes it easier for gut flora to gain access to the blood stream. As I mentioned to you, patients with hematologic malignancies, leukemia, lymphoma, they already have some immune defects. And then when we give cancer patients chemotherapy, it also induces other chemotactic and phagocytic defects, which are over and above their absolute neutrophil count. So, they're, they're at high risk for a lot of reasons.
Now, when a patient comes to the emergency center and they have got fever as we defined it, and neutropenia, you have to go looking for serious bacterial infections. By your history, you ask about any localizing signs or symptoms of infection: respiratory infections, urinary infections, GI infections, abdominal pain. Do a physical exam to look for any signs or symptoms of infection. Look for the kind of mucositis that we mentioned, sometimes very severe mucositis can be the source of infection --- periodontal infections. Do a skin survey, look for any kind of cellulitis or soft-tissue infection. Always consider a perirectal abscess as a potential source of fever in neutropenic fever patients. Students and trainees are always instructed to inspect the perianal area, but we always try to defer a digital rectal exam. You try not to insert your finger or any other kind of instrumentation into the rectum or other orifice of a neutropenic patient, because you can actually cause translocation of bacteria and serious infections. I will tell you, however, that, in real life, perirectal abscesses are not subtle. If you just ask the patient, do you have a lot of pain and swelling in that area, they will tell you, "Oh! Yes, Doctor." Believe me they know it when they have got an abscess down there. If they have central venous catheters, be careful to inspect the area for exit site infections, erythema or discharge, and just ask, by history, if they have any history that sounds like a central venous catheter infection. A good history, for instance, would be, "You know, Doctor, every time my wife flushes the catheter, 30 minutes later I get these shaking chills and just cannot stop shaking, and my fever goes up". That is a good story for a catheter-related infection. So, you will do your history, you will do your physical exam, and then you have to go looking for serious bacterial infection. You do blood cultures. Ideally, if the patient has a central venous catheter, you will obtain a culture from both the catheter itself and from a peripheral site. Do a urinalysis and culture. Do a chest x-ray even in the absence of any significant respiratory symptoms.
Now, when patients do have neutropenic fever, most of the time, even though we go through all of this evaluation I just described to you, we don't find anything. Their cultures are negative. Their chest x-ray is negative. Their urinalysis is normal. You don't see anything on exam. They are just febrile. And, in the large majority, 75% or more, they defervesce and their cultures remain negative and we never find a source. Now, when we do find a source, usually the only thing we find is an isolated positive blood culture. Now, most of the time when patients do have a positive blood culture, it is from their own bacterial flora. You will see, for instance, patients who are neutropenic who are walking around with masks on or have been told to avoid going to the store, avoid being around other people, and there may be some benefit to that. But really, what that will do is try to reduce their risk of catching viral infections, which may cause them to have a fever and then come to the emergency room and undergo this evaluation and treatment. But really it's probably not really going to lower their risk of bacterial infections, because, when they have a real bacterial infection, it's usually from the bacteria that live in or on the patient, and you can't sterilize the patient. Now, 20 to 30 years ago, most of these bacterial infections were gram negatives. So this was gut flora from the patient's own flora by the mechanism that I described to you that you have got some mucositis in the GI tract, and there is translocation of gut flora. Recently though, and for the last 10 to 20 years, these are mostly gram-positive infections, and that is --- the change in microbiology is for two reasons: one, we do a much better job these days of prophylaxing against gram-negative infection in high-risk patients, and, number 2, we use a lot more central venous catheters these days, and so, that is just a route for infection. Those are usually gram-positive infections.
So, once a patient comes in, you have seen them in the emergency center, they have fever and they have neutropenia. You take your history, you go looking for serious bacterial infections. If they have some clues about what kind of infection they have may have had, for instance, if they have had a positive culture in the past for some organism, you can focus your --- make sure that your treatment covers that organism. If they don't have anything, if you don't have any localizing signs or symptoms of infection, you don't have any prior cultures, all you have is fever, you still need to start empiric antibiotics. Now, this wasn't always the standard practice. Several decades ago, what we used to do was, collect our cultures, go looking for infection, see if something showed up. Well, we soon figured out that, if we didn't use empiric antibiotics, patients would have a higher mortality. Twenty to thirty years ago, probably 75% of the mortality of patients treated with chemotherapy was because of infection. So, what we have learned is that we need to start empiric antibiotics on everybody with this problem. Now, even though I have just told you that gram-negative infections are not the most common these days, they are more virulent. People are much more likely to die of gram-negative sepsis faster than a gram-positive bacteremia. So everybody gets gram-negative infection --- gram-negative coverage up front, including anti-pseudomonal coverage. Now, if somebody --- if you have a reason to think that somebody has a gram-positive infection, they have got horrible mucositis and most of the oral flora are gram positives, if you have got a reason to think they have a catheter-related infection, they told you a story like I suggested, some shaking chills after a catheter being flushed, something like that, then you would add gram-positive coverage. Or, if patients look septic, they are hypotensive or otherwise look bad, then you would, you would kind of throw the book at them. You would give them broad spectrum coverage, gram-negatives and gram-positives upfront.
Now, what about anaerobes? Well, very uncommon for anaerobes to be the cause or to be the identified cause of neutropenic fever. Less than 3% of patients with positive blood cultures are growing anaerobes. So, one place though where you might consider making sure that you have empiric anaerobic coverage is if patients have very, very severe necrotizing mucositis, periodontal abscesses. These could be anaerobic bacteria. If they do have a perirectal abscess, like as I have suggested to you, that is not very subtle to find that can be a gram --- I mean, an anaerobic infection. Or if they have some abdominal pain, you have gotten a CT scan, you see evidence for an intra-abdominal abscess, or they have so called typhlitis or neutropenic enterocolitis, which typically affects the colon, which could be visualized radiographically, that is also high risk for anaerobic infection. In those sort of settings, you would make sure that you include anaerobic coverage. What about antifungals? This really isn't the standard treatment up front for most patients. This is frequently add-on therapy, if patients are persistently febrile for some period of time despite antibiotic --- antimicrobial coverage for bacteria. Sometimes you will suspect fungal infections in patients who are at high-risk, like leukemia patients who have some prior fungal infection or radiographic appearance suggestive of fungal pneumonia like a nodule or infiltrate, something like that.
Now one thing that you will see in a big cancer center like MD Anderson is outpatient management of low-risk patients. I will emphasize to you that this is not the standard of care. This is recognized by the Guidelines of Infectious Diseases Society as a valid option, but only in a place where you have the infrastructure and experience to be able to closely monitor patients. So, people that we would call low-risk would be solid tumor patients, so no patients with hematologic malignancies, no leukemia patients, no lymphoma patients, no bone marrow transplant patients. Patients who don't look septic, so they cannot be hypotensive or otherwise look septic, cannot have major organ system failure, can have only minimal mucositis, so they are able to tolerate oral intake without difficulty, and they have to have adequate social factors. They have to be able to be --- to come back for appointments, to be monitored and to be reliable. So, if they satisfy all those criteria, and, if you are in a setting where they can be monitored closely, it is acceptable to treat patients as an outpatient, but they do require close monitoring. It isn't the standard of care, however, most patients in a community hospital will require admission to the hospital for intravenous antibiotics. There are a variety of antibiotic regimens used for outpatient management. Augmentin® and Ciprofloxacin® are probably the most commonly used and the ones for which we have the longest track record. All the regimens really are based on quinolones, so high-dose Levaquin® and other quinolones can be used as well, but I will refer you to the Infectious Diseases Society Guidelines for other treatment regimens. And, in general, if patients do qualify as low-risk, the outcomes are the same if they are managed as an outpatient or as an inpatient.
So, there are a variety of prophylactic measures that we do to try to prevent this problem because, as I mentioned, this is by far our biggest problem. Patients frequently receive colony-stimulating factors to try to reduce the duration and severity of neutropenia. It cannot prevent patients from becoming neutropenic, but it does prevent them from going down so low or staying down so long, and it reduces --- shortens that window period where they're at risk. Some patients do receive prophylactic antibiotics. It sort of varies depending on their cancer and their treatment regimen. They can diminish instances of fever, but there isn't really a survival benefit as demonstrated for antibiotic prophylaxis.
So, in summary, I will remind you that not all emergency problems in cancer patients are related to the underlying malignancy and its treatment. Remember to keep in mind all the same kind of non-cancer emergency problems that other patients can get. I want you to remember that the signs and symptoms of disease may be blunted or masked in cancer patients for a variety of reasons, which we have discussed in the first lecture. So have a high pre-test probability for whatever problem it is that you are considering and a lower threshold for ordering whatever test is required to try to figure that out. And I want you to remember that --- to try to recognize and respond to these kinds of cancer --- these kinds of emergency problems promptly, because that is essential to try to lowering the morbidity and mortality that we see among cancer patients. And I thank you for your attention.
Oncologic Emergencies: Part II video
©2010 The University of Texas MD Anderson Cancer Center
1515 Holcombe Blvd, Houston, TX 77030
1-800-392-1611 (USA) / 1-713-792-6161