This nationwide study found that, of 42,477 adult cancer patients admitted to hospital with septic shock, about 13% experienced dialysis-requiring AKI during hospitalization. Factors associated with dialysis-requiring septic AKI included comorbidities, such as hypertension, diabetes, congestive heart failure, and liver cirrhosis, as well as hematologic cancer. Multivariate Cox-proportional hazard analysis showed that dialysis-requiring AKI had the highest adjusted HR (1.35; 95% CI: 1.31–1.40) for 2-year mortality among cancer patients with septic shock.
The incidence of AKI among cancer patients appears to be increasing due to aggressive treatments and effective critical care. Nationwide studies have reported that the rate of any AKI among oncology patients requiring renal replacement therapy (RRT) ranged from 10–50% 13,14. These differences may be due to differences in the numbers and ethnicities of patients included in these studies, the definition of AKI, the severity of disease, and the stage of cancer. The incidence of AKI was found to be higher in cancer patients with critical illnesses than in non-cancer patients with similar severity of diseases 15,16. Although relatively little is known about the epidemiology of septic AKI in cancer, the present study showed that about 13% of cancer patients with septic shock had dialysis-requiring AKI during hospitalization. The lower incidence in this study than in previous studies may have been due to our exclusion patients who had previously received RRT before ED arrival, as prior RRT may be a risk factor for RRT during hospitalization 17. This study, however, was the largest population-based study of septic AKI among cancer patients in South Korea and may provide evidence on the epidemiology of critically ill cancer patients with a dialysis dependency.
This study also found that male sex, comorbidities, and hematologic cancers were factors associated with more frequent RRT. In contrast to results showing that older age was a risk factor for the development of AKI requiring RRT 18, the present study found that patients with dialysis-requiring AKI were of lower median age than patients not requiring dialysis. One possibility may have been that older patients declined invasive procedures, including RRT. Although we did not determine the numbers of patients with do-not-resuscitate orders, older patients were more likely to decline aggressive intubation, ICU admission, and RRT.
Furthermore, the present study found that the relative incidence of AKI was higher in patients with hematologic than solid cancers, confirming results showing that non-septic patients with hematologic malignancies, such as leukemia, lymphoma, and multiple myeloma, were at highest risk for AKI development 19. Compared with patients with solid cancer, patients with hematologic malignances were more susceptible to infection, had more severe disease, were more frequently treated with nephrotoxins, and more frequently had metabolic disturbances 20.
Our data showed that the overall 2-year mortality rate was 85.1% in cancer patients with dialysis-requiring septic AKI. Although several retrospective studies evaluated the association between septic AKI requiring RRT and short-term outcomes, it may be difficult to determine the contribution of RRT-dependent septic AKI to long-term mortality as cancer itself is a highly burdensome underlying disease. Using nationwide population data with withdrawal of life-sustaining therapy being prohibited, the present study found that dialysis-dependent septic AKI was an independent risk factor for long-term mortality after adjustment for known confounders. Mortality rates 1 year after AKI were found to be higher in cancer patients than in patients with other chronic illnesses 21. Moreover, a trial that included only critically ill patients with dialysis-requiring AKI showed that infection-related factors, such as sepsis, were the most frequent causes of death 22. The negative impact of septic AKI on clinical outcomes in cancer patients was multifactorial. For example, AKI can negatively influence the effectiveness of current or future chemotherapeutic regimens, as well as increasing their side effects or altering their pharmacodynamics and pharmacokinetics. Moreover, AKI would preclude patients from participating in potentially beneficial clinical trials.
This study had both strengths and limitations. One important strength was that we analyzed a recent national database with high coverage of the population. This may provide a more comprehensive understanding of cancer patients with septic shock. One limitation, however, was that the NHIS database did not include detailed clinical and laboratory data, which may have had confounding effects on patient outcomes. Furthermore, the NHIS database did not include specific cancer stage, performance status, or treatment setting. Finally, this study lacked detailed information on septic shock treatment, such as the timing of fluids, antibiotics, and vasopressor administration, all of which could have affected clinical outcomes.
In conclusion, Dialysis-requiring septic AKI occurred in 13% of adult cancer patients with septic shock and was associated with male sex, hematologic cancers, and comorbidities. Moreover, dialysis-requiring septic AKI was significantly associated with increased long-term mortality in the cancer patients. Prospective studies are needed to investigate methods to prevent and properly manage dialysis-requiring septic AKI in adult cancer patients with septic shock.