By: Jennifer Southall
Sepsis is a leading cause of preventable death across the globe, and patients with cancer have a 10 times higher likelihood of developing this life-threatening condition than those without cancer.
More than one in five sepsis hospitalizations is cancer-related, according to a study by Hensley and colleagues, who also found significantly higher rates of hospital readmission within 30 days among patients with cancer-related vs. noncancer-related sepsis.
Such hospitalizations and other health care services associated with sepsis can double the cost of cancer care, especially within the first year of cancer diagnosis. The excess cost is highest for patients with hematologic malignancies, who have a significantly greater risk for sepsis than patients with solid tumor cancers.
“Patients with hematologic malignancies, almost by definition, have a cancer of the immune system, which means they are immunocompromised from the moment they walk through the door. Thus, sepsis is common in this patient population compared with the general population and individuals who do not have these cancer types,” Mikkael A. Sekeres, MD, chief of the division of hematology at Sylvester Comprehensive Cancer Center at the University of Miami Health System, told HemOnc Today. “Early recognition of infection is our greatest weapon to reduce the burden of sepsis in patients with cancer. If we can identify infections and treat them appropriately and quickly, then we can prevent a patient from developing sepsis.”
HemOnc Today spoke with hematologist-oncologists, pulmonologists, intensivists and health economists about risk factors for sepsis and the economic burden it creates for patients with cancer, as well as steps that can be taken to prevent sepsis and improve management of these patients.
Incidence and mortality
Patients with cancer are at higher risk for both sepsis and associated mortality than the general population, according to a study by Liu and colleagues.
The researchers used the sepsis-indicator International Classification of Disease codes from the Nationwide Inpatient Sample database between 2006 and 2014 to identify nearly 14 million adults with sepsis, 13.6% of whom had cancer.
Results showed patients with cancer had higher rates of respiratory tract (35% vs. 31.6%), intra-abdominal (5.5% vs. 4.6%), fungal (4.8% vs. 2.9%) and anaerobic (1.2% vs. 0.9%) infections, as well as a higher rate of sepsis. Hospital mortality also was higher among those with cancer after propensity score matching (HR = 1.25; 95% CI, 1.24-1.26). Of note, patients with lung cancer had the lowest survival rate (HR = 1.65).
“Sepsis is a significant problem in critical care and certainly a big problem in patients with cancer,” Ayman O. Soubani, MD, pulmonologist, intensivist and professor at Wayne State University School of Medicine and medical director of the ICU at Karmanos Cancer Institute, told HemOnc Today. “Sepsis is one of the most common diagnoses for patients who become critically ill and are admitted to the ICU, accounting for 30% to 50% of these patients. Historically, the prognosis in patients with cancer who become critically ill, whether from sepsis or other causes of critical illness, has been thought to be very poor. In fact, many times in the past, patients with cancer were denied admission to the ICU because of their poor prognosis when they become critically ill. However, we have seen in the literature that the prognosis of critically ill patients with cancer has improved over the years. These patients overall are having better prognoses when they become critically ill.”
In a retrospective cohort study published in American Journal of Clinical Oncology, Soubani and colleagues reported that mortality declined among more than 19 million patients hospitalized for sepsis between 2008 and 2017.
“We found that around 20% of patients admitted to the ICU with sepsis have a cancer diagnosis,” Soubani said. “What is important and encouraging is the fact that the outcome of patients with cancer and sepsis significantly improved during the decade that we studied. For example, in 2008, the mortality rate for patients with cancer and sepsis was 23% vs. 15% in 2017, which is a significant decrease. Another important finding was that the rate of decline and mortality from sepsis was similar between patients with cancer and without cancer.”
This improvement may reflect better management of sepsis, regardless of the host, Soubani added.
“Whether it’s patients with cancer or without cancer, the management of sepsis in general has significantly improved as a result of multiple studies and guidelines,” he said. “Physicians know exactly what to do with a patient who has sepsis and how quickly to resuscitate with intravenous fluids and give antibiotics. Advances in antibiotics and other approaches in the management of sepsis have improved significantly over the last couple of decades for all patients.”
In a separate study published in PLoS One of 119,379 patients newly diagnosed with cancer, Van de Louw and colleagues reported a 3.7% cumulative incidence of sepsis 1 year after cancer diagnosis, with a 35.5% probability of death after sepsis. Both incidence and mortality were associated with baseline patient and cancer characteristics, including age, male sex, Charlson comorbidity index, hematologic malignancies and metastases.
“The risk for sepsis varies greatly depending upon how immunocompromised the patient is, as well as the type of cancer therapy the patient is receiving,” Sekeres said. “For example, if I have a patient with acute myeloid leukemia and I admit that patient to the hospital and give him or her intensive induction chemotherapy, we are further compromising their immune system. We are wiping out their bone marrow and waiting for a functional immune system to grow back.
“During this period, these patients are extremely vulnerable to infections and ultimately to the development of sepsis, which is why we keep them in the hospital for 4 to 6 weeks until their immune systems recover,” he said. “Up to 10% of my patients with acute leukemia receiving intensive induction chemotherapy will go to the ICU, and the overwhelming reason is sepsis.”
Imrana Malik, MD, DABSM, FCCP, pulmonologist and associate professor in the department of critical care at The University of Texas MD Anderson Cancer Center, agreed that immunosuppression from the cancer itself and the cancer treatment are key risk factors for sepsis.
“Patients with cancer also tend to undergo more invasive procedures for long-term central access catheters, urinary catheters or drains,” Malik told HemOnc Today. “Various treatments have also been known to cause damage to the mucosal barrier, including severe mucositis in the back of the throat or in the mouth, that can easily lead to bloodstream infections because of the breakdown of that barrier. Other common risk factors include malnutrition and frequent hospitalizations, which can lead to the increased risk for hospital-acquired infections.”
The increased use of immunotherapy in recent years has also played a role.
“We are using immunotherapy more and more for various cancer types and are learning more about its contributions to sepsis incidence,” Malik said. “These therapies are known to increase the risk for pneumonia, urinary tract infections, skin and soft-tissue infections, as well as gastroenteritis. A lot of attention has been given to multidisciplinary approaches to diagnosing sepsis infections as quickly as possible and treating the sepsis in this patient population to reduce the risk.”
Data are scarce on the economic impact of sepsis among patients with cancer.
Michelle Tew, PhD, health economics research fellow in the department of health services research at Peter MacCallum Cancer Center in Melbourne, Australia, and colleagues sought to better understand the excess costs of care associated with sepsis among patients with cancer.
Results of their case-control, matched cohort study using Canadian population-linked data, published in PLoS One, showed the excess cost of care, in Canadian dollars, among individuals with solid tumor cancers who developed sepsis was $29,081 within the first year, increasing to $60,714 within 5 years of cancer diagnosis. An even higher excess cost was observed among individuals with hematologic malignancies, rising from $46,154 during the first year to $75,931 within 5 years of diagnosis.
“Our study showed that particular groups are associated with higher burden of sepsis — for example, patients with hematologic malignancies — and patients are most at risk within the first year of their cancer diagnosis,” Tew told HemOnc Today. “This helps provide important indicators in recognizing at-risk groups and risk periods to closely monitor patients for sepsis. Clinical pathways should be embedded into hospitals to identify and treat sepsis promptly, as should strategies to prevent infections progressing to sepsis.”
In the U.S., the economic burden of sepsis is astounding, and costs tend to depend upon the severity, according to Malik.
“For hospitalizations for sepsis without organ failure, costs can average around $16,000, whereas sepsis accompanied by organ failure increases to approximately $24,000. With septic shock, the cost increases to $38,000,” Malik said. “When looking at the situation of sepsis diagnosed later during hospitalization, which tends to carry a higher mortality rate, the average cost increases up to $51,000. These numbers are expected to be higher in patients with cancer who develop sepsis.”
Hospital readmissions largely contribute to the economic burden of sepsis in patients with cancer.
“Thirty-day readmission rates for any reason are higher after a cancer-related sepsis hospitalization — 23.3% for cancer-related sepsis hospitalization compared with 20.1% for a noncancer-related sepsis hospitalization. This may seem like a small difference, but it is definitely statistically significant,” Malik said. “Readmissions can also mean fewer days of work for patients who are still actively employed and fewer days of work for caretakers, as well.”
Tew and colleagues plan to continue their research by further studying the impact of sepsis on patients with different types of cancer and its potential spillover effects on treatment pathways, outcomes and associated costs.
“We want to understand how sepsis impacts patients’ pathways of cancer care, the role of different health care services, and the resources required to provide this support,” Tew said. “With such evidence, we can help guide policy design and allocation of health care resources to alleviate both the cost and illness burden of sepsis on the health care system, as well as on patients.”
‘Attention to the basics’
Increased prevention efforts could substantially reduce sepsis-associated deaths among patients with cancer, according to experts with whom HemOnc Today spoke.
“It is extremely important to emphasize prevention. We have learned this with COVID-19, where if we can prevent infection, we can halt so much more in terms of morbidity and mortality, and the same goes for sepsis in patients with cancer,” Malik said. “Attention to the basics is key in prevention and this includes hand hygiene, vaccination, good catheter care and, very importantly, increasing awareness about sepsis. We want early identification, which can lead to early treatment and, ultimately, improved outcomes for these patients.”
A study published in JAMA Network Open suggests prevention efforts must extend beyond better hospital-based care.
The study by Rhee and colleagues included 568 adults admitted to six different hospitals who died in the hospital or were discharged to hospice without readmission. Among them, 52.8% had sepsis, which was directly responsible for 34.9% of deaths — more than due to progressive cancer (16.2%) and heart failure (6.9%). However, only 3.7% of sepsis-associated deaths were deemed definitively or moderately likely preventable.
Another study published in Critical Care Medicine showed the display of an electronic health record-based sepsis early warning system-triggered flag combined with EHR-based pharmacist notification resulted in shorter time to antibiotic administration without an increase in undesirable or potentially harmful clinical interventions among 598 adults presenting to the ED.
“EHR triggers will note that a patient has low blood pressure and/or fast heart rate and low immune system and suggest that this person may have sepsis, though it’s not very specific in our patient population,” Sekeres said. “Another thing that we started to study when I was at Cleveland Clinic Taussig Cancer Institute was whether the use of wearables could predict eventual transition to the ICU. This is where the future is headed — the use of wearables and incorporating data from the EHR into a machine learning approach to predict ICU admission for sepsis.”
Prophylactic antibiotic therapy is commonly prescribed for patients with cancer.
“We typically give prophylactic antibiotics to our patients who are undergoing intensive induction chemotherapy for leukemia, for example, and also to patients who are undergoing chemotherapy as outpatients when we see their immune system start to drop,” Sekeres said. “There are also some good data for use of fluoroquinolones as prophylactic antibiotics in patients with myeloid malignancy who are undergoing therapy. We also give prophylactic antibiotics to help prevent certain viral infections and some fungal infections, although the data supporting their use is not as robust.”
Experts anticipate an update to the general guidelines for sepsis management will be released in the coming months.
“These guidelines emphasize screening for sepsis, appropriate antimicrobial therapy, fluid resuscitation for those who are hypotensive and source control as soon as possible, among other things,” Malik said. “These are the basics that we need to do to reduce the morbidity and mortality from sepsis in our patients. Right now, we do not have cancer-specific methodologies for treating sepsis.”
Awareness is a banner taken up by many organizations, such as the Sepsis Alliance in the U.S. and the international Global Sepsis Alliance. In a Sepsis Alliance white paper from 2018, the institution coined the acronym TIME — Temperature, Infection, Mental decline, and Extremely ill — to improve understanding of the importance of quickly identifying and treating sepsis.
Tew said there is a need to strengthen initiatives for prompt sepsis identification and treatment, particularly within the first year of cancer diagnosis, and management and prevention of subsequent sepsis episodes.
“In addition, by understanding where costs associated with sepsis are accrued along the cancer care continuum, we can help inform how resources are allocated and considerations for how to best prioritize health policies and research,” Tew added. “Knowing where the costs are is useful in helping inform development of sepsis programs and policies across the cancer care continuum, which should include prevention, such as with vaccinations, and early recognition strategies, including screening and prompt treatment.”
Timely communication also is needed among oncologists, infectious disease specialists, ER physicians and intensivists in the ICU to appropriately manage sepsis in patients with cancer.
“We believe very strongly that we should continue to follow our patients on a daily basis when they’re transferred to the ICU, but we also try to ‘stay in our lanes.’ For example, going to an ICU as a hematologist-oncologist and suggesting to the intensivist that they need to replace our patient’s potassium is kind of obnoxious,” Sekeres said. “They know what they’re doing with intensive management. In the same light, we are not going to go to the ICU and make recommendations on ventilator settings or electrolyte repletion because we rely on our expert colleagues to do that. What we can and should do is make recommendations about anticipated recovery of blood counts or whether or not to continue chemotherapy, and whether or not there’s any benefit to use of growth factors.”
The next step is to identify ways to prevent sepsis from occurring and anticipate occurrence, Sekeres added, most likely through the use of wearables, data from EHRs and analyzing the data from both through machine learning approaches.
Soubani said the improvements in diagnosis, management and cancer treatments during the past 2 decades have led to better prognosis in patients with cancer and sepsis.
“This message has to be propagated more so that the attitude toward patients with cancer who become critically ill with sepsis will change and we can continue to provide this patient population with the best opportunities possible and ensure that appropriate aggressive treatment is not withheld because of a cancer diagnosis,” Soubani said. “These views are changing, but they need to be more accepted by practitioners everywhere, not only across the large cancer centers, but across all community and smaller hospitals and practices, as well.”
Malik said continued improvements in management of patients with cancer and sepsis can benefit from more precise molecular definitions and identifying phenotypes of patients more likely to respond to new therapeutics on the horizon.
“The field of oncology has really taken this to heart, and it is far ahead of all other fields in being precise in treatment definitions and in choice of patients for certain therapies,” Malik said. “Right now, sepsis care is homogeneous or occurs in an underlying heterogeneous population, which does not work very well, and we hope this will change with more research. Because of COVID-19, more studies are looking at new and exciting treatments for sepsis. These therapies that were developed for COVID-19 have some potential in sepsis, but we have to find the right patients for their use. Otherwise, it will look like these are not helpful or possibly even harmful. We need to take a page out of the oncologists’ book and get our definitions down precisely and identify the right patients so that we can really hone in on the best appropriate therapies to treat sepsis.”
- Hensley MK, et al. Crit Care Med. 2019;doi:10.1097/CCM.0000000000003896.
- Liu MA, et al. Am J Crit Care. 2021;doi:10.4037/ajcc2021632.
- Rhee C, et al. JAMA Netw Open. 2019;doi:10.1001/jamanetworkopen.2018.7571.
- Sharma A, et al. Am J Clin Oncol. 2021;doi: 10.1097/COC.0000000000000859.
- Tarabichi Y, et al. Crit Care Med. 2021;doi:10.1097/CCM.0000000000005267.
- Tew M, et al. PLoS One. 2021;doi:10.1371/journal.pone.0255107.
- Van de Louw, et al. PLoS One. 2020;doi:10.1371/journal.pone.0243449.
Barbara Jacoby is an award winning blogger that has contributed her writings to multiple online publications that have touched readers worldwide.