Pneumococcal purpura fulminans in asplenic or hyposplenic patients: a French multicenter exposed-unexposed retrospective cohort study

Upon ICU admission and during ICU stay, data pertaining to demographics, comorbidities, clinical examinations, laboratory findings, microbiological investigations, and therapeutic management were collected. Simplified Acute Physiology Score II (SAPS II) [9] and Sequential Organ Failure Assessment (SOFA) [10] scores were computed using the worst values recorded within the first 24 h of admission. Missing data were retrieved from queries to the investigators. In asplenic/hyposplenic patients, the following data were retrieved from the ICU discharge reports: cause of asplenia (congenital or post-splenectomy in case of trauma, hematological diseases or cancer) or hyposplenia (splenic irradiation, sickle cell disease), time interval between asplenia/hyposplenia and ICU admission, presence or absence of antibiotic prophylaxis, and pneumococcal vaccination status.

Data were compared between asplenic/hyposplenic and eusplenic patients. Continuous variables were reported as median [25th–75th interquartile range] (IQR) or mean ± standard deviation (SD) and compared between groups using the Student t test or Mann-Whitney test, as appropriate. Categorical variables were reported as numbers and percentages (95% confidence interval, CI) and compared using χ² test or Fisher’s exact test, as appropriate. All statistical analyses were conducted using the SPSS Base 21.0 statistical software package (SPSS Inc., Chicago, IL.). A p value < 0.05 was considered statistically significant.

This observational, non-interventional analysis of medical records was approved by the Institutional Review Board of the French Society of Intensive Care (FICS) in March 2016. As per French law, no informed consent was required for this type of study.

Among the 306 patients admitted to the ICU for purpura fulminans, 67 (22%) had a pneumococcal purpura fulminans, of whom 34 (51%) had asplenia (n = 29/34, 85%) (Table 1) or hyposplenia (n = 5/34, 15%) and 33 (49%) had eusplenia. Age, gender, and ICU severity scores did not differ between asplenic/hyposplenic and eusplenic patients (Table 2). The prevalence of pneumococcal purpura fulminans was seven times higher in asplenic or hyposplenic patients compared to eusplenic patients with purpura fulminans (n = 34/39, 87% vs. n = 33/267, 12%; p < 0.001) (Fig. 1). The main causes of asplenia were post-splenectomy (n = 26/29, 90%) and congenital (n = 3/29, 10%, two cases being diagnosed during ICU stay) (Table 1). Five patients had hyposplenia related to a splenic irradiation for lymphoma. The median time interval between asplenia/hyposplenia and ICU admission was 20 [9–32] years, with only 2/34 (6%) patients having pneumococcal purpura fulminans within the 2 years following splenectomy/hyposplenia (Table 1). Only one patient (n = 1/32, 3%) (2 missing data) had received long-term antibiotic prophylaxis (time interval between splenectomy and ICU admission < 1 year). Pneumococcal vaccination had been performed in 11/31 (35%) patients (3 missing data) (Table 1). Among the 67 patients with pneumococcal purpura fulminans, 58 (87%) had positive blood cultures, 25 (37%) had a positive pneumococcal urinary antigen testing, and 11 of the 29 (38%) patients who had lumbar puncture had a positive CSF culture (Table 2).

Clinical presentation did not significantly differ between asplenic/hyposplenic and eusplenic patients, except for digestive symptoms (n = 25/34, 73% vs. n = 16/33, 48%; p = 0.04) and lower limb pain (n = 13/34, 38% vs. n = 4/33, 12%; p = 0.01), which were more frequently reported in the former than in the latter. Purpura was also more frequently reported before ICU admission in asplenic/hyposplenic patients than in their counterparts (n = 25/34, 73% vs. n = 13/33, 39%; p = 0.01) (Table 2). Biological and microbiological data did not differ between asplenic/hyposplenic and eusplenic patients. The rate of bacteremia was > 80% but did not significantly differ between hyposplenic/asplenic and eusplenic patients (n = 31/34, 91% vs. n = 27/33, 82%; p = 0.26). There was no significant difference between groups regarding antibiotic therapy before ICU admission (n = 25/35, 73% vs. n = 21/33, 64%, p = 0.38), and all patients received a beta-lactam antibiotic in the ICU. Details regarding comorbidities, ICU scores, clinical presentation, and biological and microbiological data were provided in Table 2.

The rates of invasive mechanical ventilation, renal replacement therapy, systolic myocardial dysfunction, and platelet or plasma transfusions did not differ between asplenic/hyposplenic and eusplenic patients (Table 3), nor did the rate of limb amputation (n = 9/34, 26% vs. 15/33, 45%; p = 0.11) or hospital mortality (n = 20/34, 59% vs. n = 15/33, 45%; p = 0.27) (Fig. 2). Among asplenic/hyposplenic patients, hospital mortality did not differ between vaccinated and unvaccinated patients (n = 4/9, 44% vs. n = 14/22, 63%; p = 0.60). Outcome did not differ between hyposplenic and asplenic patients regarding limb amputations (n = 1/5, 20% vs. 8/29, 28%; p = 0.78) and hospital mortality (n = 3/5, 60% vs. 17/29, 59%; p = 0.98). More details regarding management, organ supports, and outcomes are provided in Table 3.

Our study certainly suffers from several limitations. This was a retrospective study with inherently associated bias, as well as missing data and possible associated errors in data abstraction. However, due to the extreme rarity of pneumococcal purpura fulminans, a prospective study would be hardly feasible. A previous prospective multicenter study on overwhelming post-splenectomy infection in 173 German ICUs prematurely ended due to slow recruitment [2]. Abdominal CT scan and detection of Howell-Jolly bodies in peripheral blood smears or counting of pitted erythrocytes by phase-interference microscopy was not routinely performed. Therefore, we cannot exclude that patients included in the eusplenic group had an unknown asplenia potentially related to congenital asplenia or other diseases (severe celiac disease, inflammatory bowel disease, Whipple’s disease, amyloidosis, Sjögren’s syndrome, HIV infection, cirrhosis) associated with hyposplenism or splenic atrophy [5, 20]. However, only 15% (n = 5/33) of the eusplenic patients had pre-existing diseases potentially associated with hyposplenism (ulcerative colitis n = 1, Sjögren’s syndrome n = 1, HIV infection n = 2, cirrhosis n = 1). Because the patients were recruited over a 17-year period in 55 centers, ICU procedures were inevitably heterogeneous. The serotypes of the Streptococcus pneumoniae strains involved was not available due to the retrospective nature of the study, precluding any comparison to be made between both groups of patients. Last, the absence of significant outcome differences observed between asplenic/hyposplenic and eusplenic patients may be due to a lack of power related to a relatively small number of included patients.

Our study also has some strengths, including the large number of centers and patients included for a very rare infectious disease [6], the comparison between asplenic and non-asplenic patients, with a paucity of previously published data due to the rarity of overwhelming post-splenectomy infection [2], the collection of pneumococcal vaccination status which is poorly reported, and the fact that the definitions used for inclusion of the patients were well-standardized, rendering the comparison of the two groups relevant.

Asplenic or hyposplenic patients admitted to the ICU for a purpura fulminans seem to have a high risk of pneumococcal purpura fulminans. In patients surviving pneumococcal purpura fulminans, unknown congenital asplenia should be searched with abdominal CT-scan. Detecting Howell-Jolly bodies in peripheral blood smear and counting pitted erythrocytes by phase-interference microscopy may also help diagnosing hyposplenia [5]. In such patients, pneumococcal vaccination seems welcomed in order to avoid recurrences of overwhelming post-splenectomy infection. As recommended by the Centers for Disease Control and Prevention 2019 guidelines [16], patients with asplenia or hyposplenia should receive one dose of the 13-valent pneumococcal conjugate vaccine together with one (for persons aged ≥ 64 years) or two doses (for persons aged 19–64 years) of the 23-valent pneumococcal polysaccharide vaccine. Furthermore, continuing education is required to advise patients with a risk of severe infection inherent to asplenia and the need for immediate antibiotic therapy in case of fever, especially since it has been reported that 40 to 84% of splenectomized individuals were not aware of the infectious risks associated with their condition [18, 21]. In adult asplenic/hyposplenic patients, long-term antibiotic prophylaxis [5] is still debated, except in those who survived overwhelming post-splenectomy infection [20], a targeted subgroup of patients in whom such a strategy seems appropriate, although supported by a low evidence.