American Journal of Respiratory and Critical Care Medicine

Rationale: Purpura fulminans in adults is a rare but devastating disease. Its pathophysiology is not well known.

Objectives: To understand the pathophysiology of skin lesions in purpura fulminans, the interplay between circulating blood and vascular alterations was assessed.

Methods: Prospective multicenter study in four intensive care units. Patients with severe sepsis without skin lesions were recruited as control subjects.

Measurements and Main Results: Twenty patients with severe sepsis and purpura fulminans were recruited for blood sampling, and skin biopsy was performed in deceased patients. High severity of disease and mortality rates (80%) was observed. Skin biopsies in purpura fulminans lesions revealed thrombosis and extensive vascular damage: vascular congestion and dilation, endothelial necrosis, alteration of markers of endothelial integrity (CD31) and of the protein C pathway receptors (endothelial protein C receptor, thrombomodulin). Elevated plasminogen activating inhibitor-1 mRNA was also observed. Comparison with control patients showed that these lesions were specific to purpura fulminans. By contrast, no difference was observed for blood hemostasis parameters, including soluble thrombomodulin, activated protein C, and disseminated intravascular coagulation markers. Bacterial presence at the vascular wall was observed specifically in areas of vascular damage in eight of nine patients tested (including patients with Streptococcus pneumoniae, Neisseria meningitidis, Escherichia coli, and Pseudomonas aeruginosa infection).

Conclusions: Thrombi and extensive vascular damage with multifaceted prothrombotic local imbalance are characteristics of purpura fulminans. A “vascular wall infection” hypothesis, responsible for endothelial damage and subsequent skin lesions, can be put forward.

Scientific Knowledge on the Subject

Purpura fulminans (PF) is a devastating disease both in adults and children. The presence of thrombi in the blood vessels of the skin and derangement of hemostasis parameters have been described.

What This Study Adds to the Field

Thrombi and extensive vascular damage with multifaceted prothrombotic local imbalance are characteristics of PF. The presence of a bacterial pathogen in the blood vessels is observed at sites of maximal vascular damage, indicating that PF may be due to direct bacterial interaction with the blood vessels.

Adult patients admitted to the intensive care unit (ICU) for severe sepsis occasionally suffer from progressive purpura, which can potentially evolve toward symmetrical acral gangrene (1). Large studies have shown that this condition occurs infrequently but is devastating with high mortality (1, 2) and a frequent requirement for limb amputation (3). Neisseria meningitidis is frequently reported as the causal pathogen; however, staphylococci and Streptococcus pneumoniae have also been described (1, 4). In children this clinical entity is unequivocally referred to as purpura fulminans (PF). In adults, several other terms are used, such as symmetrical peripheral gangrene or ischemic skin lesions (5, 6), probably reflecting uncertainty regarding the underlying pathophysiology. As PF is the term most frequently used in the literature, it will be retained throughout this article (2).

The association between PF and profound abnormalities of hemostasis is constant in reported cases (1, 4, 5, 7). Some authors have attributed PF to thrombotic manifestations of sepsis-related disseminated intravascular coagulation (DIC) (1, 4). Few data exist regarding the pathologic skin changes in patients with PF, but occasional studies confirm the presence of thrombi in small dermal vessels (8, 9). However, describing coagulation abnormalities in patients with PF may not be sufficient to account for its pathophysiology: indeed, not all patients with DIC develop PF. A deleterious interplay between circulating blood coagulation activation and skin endothelial alterations may be hypothesized. Accordingly, one report in children with N. meningitidis PF showed a defective endothelial conversion of protein C to its activated form because of diminished thrombomodulin (TM) and endothelial protein C receptor (EPCR) endothelial expression (8). Finally, no prospective studies in adults regarding PF pathophysiology beyond circulating blood analysis have been performed to our knowledge.

Working on the hypothesis that PF is dependent on interplay between circulating blood and endothelial alterations, we conducted a prospective multicenter study to explore blood and skin vascular changes in adult patients with PF. As endothelial damage is thought to be a major component of all forms of severe sepsis even in the absence of PF, patients suffering from severe sepsis without PF were concomitantly recruited to determine the specificity of our observations toward PF.

See the online supplement for full details. This study was conducted in four medical ICUs from May 2008 to September 2010. All patients admitted for new onset of community-acquired severe sepsis (10) were screened during the first 3 days after admission for rapid onset of progressive purpuric lesions defining PF (1). Patients with PF were included after written informed consent was obtained from the patients or from their relatives. This study was approved by the persons protection committee Paris Ile de France II (2008-129).

Characteristics of Studied Patients

Initial severity, characteristics of infection, use of organ support, ICU and 3-month outcome were registered. DIC was evaluated daily, based on the scoring system of the International Society on Thrombosis and Haemostasis (ISTH), using measurements performed on a routine basis in each hospital laboratory (11).

Blood Sampling and Analysis

Blood was collected for analyses specific to this study in the 24 hours after observation of purpura. Measurements of protein C and protein S (PC and PS, respectively), antithrombin (AT), soluble TM (sTM), and activated protein C (PCa) were performed in the hematology laboratories of Ambroise Paré Hospital (Boulogne-Billancourt, France).

Skin Sampling and Analysis

In case of death, skin sampling was performed immediately, after family members were questioned and the national file for refusal of postmortem sampling was investigated to determine whether the deceased had previously disclosed refusal to allow procedures to be performed for research purposes. Two sets of biopsies were taken for each patient: one set in a purpuric lesion (PF purpuric samples), and one in proximity to a purpuric lesion in normal-appearing skin (PF nonpurpuric samples). A standardized histopathological analysis of the skin biopsies assessed a variety of lesions after hematoxylin–eosin and acid–Schiff staining. Immunohistochemistry was performed to assess vascular expression of CD31, CD61, TM, EPCR, and specific bacterial antigens. Transmission electron microscopy was used to describe endothelial alterations. All of the sections were reviewed by two experienced pathologists (A.C., P.B.) blinded to the clinical and laboratory data. Tissue factor (TF) and plasminogen-activating inhibitor-1 (PAI-1) gene expression levels were studied and expressed as a percentage of expression of glyceraldehyde-3-phosphate dehydrogenase (GAPDH).

Control Patients

Two groups of control patients participated. First, 20 patients with severe sepsis but without purpuric lesions, matched with patients with PF for sex, age ± 7 years, and Simplified Acute Physiology Score II (SAPSII) ± 5, were extracted from a concomitant prospective study performed in Angers Hospital ICU (SAPSII-control). Second, 15 patients with severe sepsis-related DIC (based on ISTH criteria) without PF who could be matched for sex and age ± 7 years with patients with PF were also recruited from the same prospective study (DIC-control). Plasma aliquots sampled on Day 1 of ICU admission were analyzed as described for patients with PF. Last, nine random patients (six from the SAPSII-control group and three from the DIC-control group) who died of septic shock without skin lesions underwent postmortem skin biopsies (control skin samples).

Statistical Analysis

The χ2 test and the Kruskal-Wallis nonparametric test were used as needed.

Extreme Initial Severity and Poor Outcome

Twenty patients with severe sepsis presented with skin lesions evocative of PF during the study period and were included. A few samples of patients’ photographs are given in Figure E1 in the online supplement. Concomitantly, 1,264 patients had been admitted for severe sepsis over the four participating ICUs, establishing a PF frequency of 1.6% (95% confidence interval, 0.9–2.3%).

Patients with PF were characterized by high acute severity scores, requirement for multiple organ support in almost all cases, and a high mortality rate (see Tables 1 and 2). No improvement in skin lesions was observed during the ICU stay in the 16 patients who eventually died. ICU length of stay for these patients averaged 2 days (interquartile range, 1–3 days). Among the four ICU survivors, skin lesions completely resolved in the weeks after admission in two patients who exited the ICU after 4 days; in the other two patients, skin lesions evolved toward limb necrosis requiring multiple limb amputation. ICU length of stay was 49 and 100 days, respectively. All four patients were alive 3 months after ICU admission.


VariablePatients with Purpura Fulminans (n = 20)SAPSII-Control Subjects (n = 20)DIC-Control Subjects (n = 15)P Overall
Age, yr, median (IQR)67 (58–79)70 (63–81)72 (67–81)0.9
Male, n (%)12 (60%)15 (75%)11 (73%)0.54
SAPSII, median (IQR)71 (65–75)74 (64–81)61 (49–78)0.3
SOFA score, median (IQR)16 (12–18)11 (10–12)10 (10–13)0.003
Vasopressor infusion, n (%)*19 (95%)15 (75%)14 (70%)0.12
Maximum catecholamine infusion rate, μg/kg/min, median (IQR)1.9 (0.9–3.4)1.5 (0.7–2.2)1.4 (0.3–1.9)0.33
Mechanical ventilation, n (%)*19 (95%)18 (90%)15 (100%)0.43
Renal replacement therapy, n (%)19 (95%)15 (75%)11 (73%)0.16
ICU mortality, n (%)16 (80%)8 (40%)8 (53%)0.03

Definition of abbreviations: DIC = disseminated intravascular coagulation; ICU = intensive care unit; IQR = interquartile range; SAPSII = Simplified Acute Physiology Score II; SOFA = Sequential Organ Failure Assessment.

*Requirement during the 24 hours after ICU admission.

Epinephrine or norepinephrine, during the first 3 days after ICU admission.

Requirement at any time during ICU stay.


VariablePatients with Purpura Fulminans (n = 20)
Extent of skin lesions 
 Limbs bilaterally17
 Superior limbs11
Pathogen identified 
Neisseria meningitidis2
Streptococcus pneumoniae4
 Other Streptococcus5
Staphylococcus aureus2
 Gram-negative bacilli3
Site of infection 

In comparison with the SAPSII- and DIC-control patients, no difference was observed for SAPSII (due to matching for SAPSII-control). However, the SOFA score was significantly different between these three groups, with higher values for patients with PF. The maximum catecholamine infusion rate was not different between the groups. Cocci (positive or negative) were more frequently identified in patients with PF than in SAPSII- and DIC-control patients (13 of 20, 6 of 20, and 5 of 15, respectively; P < 0.05). No difference in site of infection or frequency of bacteremia was observed (data not shown). Last, mortality rate was significantly different between the groups, being lower in the two control groups.

PF presented as a rare but particularly dramatic disease with frequent multiple organ failure, high mortality rate, and frequent extensive sequelae in survivors.

Skin Biopsy Reveals Extensive Vascular Damage and Thrombi

A better knowledge of PF lesion pathology in adults, and its specificity relating to other forms of sepsis, is necessary to understand its pathophysiology. Of the 16 deceased patients with PF, 13 underwent skin biopsy. A comparison between PF purpuric samples, PF nonpurpuric samples, and control skin samples was performed (Figures 1 and 2).

The presence of thrombi in dermal vessels (mostly capillaries) was ascertained in almost all PF purpuric samples, occasionally in PF nonpurpuric samples, and in none of the control samples. CD61, a platelet marker, showed intravascular platelet aggregates in all PF purpuric samples, whereas they were only occasionally seen in PF nonpurpuric samples and absent in control samples.

PF purpuric samples showed severe to massive vascular congestion and vascular dilation in all samples, these lesions likely being responsible for the purpuric appearance of the skin. Capillary wall necrosis without fibrinoid deposits was observed in five PF purpuric samples. Although vascular dilation was also common in PF nonpurpuric samples, congestion was occasional and moderate. Few vascular abnormalities were observed in control samples. Red cell extravasation was observed in only two PF purpuric samples. Occasional and minimal leukocytic infiltration was observed around capillaries only in PF purpuric samples. Endothelial expression of CD31 (platelet endothelial cell adhesion molecule, PECAM), a marker of endothelial integrity, was altered in PF purpuric samples: the labeling was discontinuous or mottled; complete loss of CD31 expression was observed in some samples. In PF nonpurpuric samples, CD31 labeling was variable, from normal to moderately discontinuous whereas it was normal, that is, linear and continuous, in control samples. Diffuse endothelial cell damage in PF purpuric samples was confirmed by transmission electron microscopy showing endothelial cell nuclei with condensed chromatin suggestive of cell death (Figure 3).

Thus our observations confirm the presence of extensive endothelial lesions specific to patients with PF, and predominantly in purpuric samples. Failure of the endothelium to protect against local thrombosis is established by the presence of thrombi and platelets.

Endothelial Damage Coincides with a Prothrombotic Imbalance

To further understand the skin abnormalities associated with PF we assessed the local balance between pro- and antithrombotic factors.

Of the coagulation inhibitors, a defect of the protein C system was investigated by staining for EPCR and TM in biopsies. In accordance with the previous observation that endothelial integrity was altered in PF purpuric samples, EPCR and TM expression was decreased in PF purpuric samples in comparison with PF nonpurpuric samples and control samples (Figure 2). Of note, sTM plasma levels, proposed to reflect endothelium activation, were lower in patients with PF in comparison with DIC- and SAPSII-control subjects (Figure 4). This indicates that a process different from what is observed in the more usual form of septic DIC may occur at the endothelial level in patients with PF.

In addition to staining experiments, the expression profile of two proteins, TF (responsible for coagulation initiation) and PAI-1 (the main fibrinolysis inhibitor), was assessed by mRNA quantification in skin biopsy samples (Figure 5). Although a trend was observed, the difference in TF mRNA between samples was not significant. In contrast, we observed a significant increase (approximately threefold) in PAI-1 expression within skin biopsies (both purpuric and nonpurpuric samples) of patients with PF that may contribute to a fibrinolysis defect.

Thus, the endothelial damage observed on pathological examination in purpuric sample from patients with PF were associated with functional impairment of several hemostatic pathways, all favoring thrombosis.

Purpura Is Not Associated with a More Severe Form of DIC Except for a Profound Reduction in Circulating Natural Anticoagulant

Systemic coagulation activation was assessed, assuming an interplay between vascular lesions and circulating blood abnormalities that could participate in skin lesion development. Patients with PF showed evidence of coagulation activation with diminished platelet counts, elevated prothrombin time, and presence of circulating D-dimer (Figure 4). In 14 of 20 patients with PF overt DIC was diagnosed (70%). Overt DIC was diagnosed in 20% (4 of 20) of SAPSII-control patients (P = 0.004 for comparison with patients with PF) and 100% (15 of 15) of DIC-control patients by definition. Patients with PF had lower platelet counts, higher D-dimer concentrations, and higher ISTH scores in comparison with SAPSII-control subjects but not in comparison with DIC-control subjects (Figure 4).

In all patients with PF, diminished blood PC, PS, and AT concentrations were observed. PS and AT concentrations were found to be lower in patients with PF in comparison with both SAPSII-control subjects and DIC-control subjects, but lower PC levels were observed in comparison with SAPSII-control subjects only (Figure 4). Plasma activated protein C was measured only in PF and SAPSII-control subjects and no significant difference was observed; concentration was detectable in all patients.

Thus, even though profound coagulation activation seems essential to PF, it does not appear to be sufficient to independently induce PF lesions. However, a particular pattern with lower PS and AT, but not PC and aPC, despite the vascular derangements of the PC pathway on skin biopsies, was associated with PF.

Pathogens Show a Close Association with Endothelium Damage

Evidence of extensive endothelial damage observed specifically in purpuric lesions of patients with PF, and data showing the ability of certain pathogens to invade the endothelium, prompted us to search for pathogen presence in the blood vessels of skin samples (12). This search was conducted in nine patients with PF, using immunostaining with specific antibodies directed against pathogens identified during the course of the illness (Figures 6 and 7): five patients with S. pneumoniae infection, one with N. meningitidis serotype B, one with Escherichia coli, one with Pseudomonas aeruginosa, and one with Haemophilus influenzae. Bacterial antigen was detected in the purpuric samples of all patients except the patient with H. influenzae infection. Microorganisms were identified typically inside the capillaries in foci with multiple bacteria. Immunostaining was negative in nonpurpuric samples of all these patients except for the patient with N. meningitidis infection. Of note, histological signs of vascular damage were present in this sample. Last, the search for bacteria was negative in all five control patients tested (three patients with E. coli, one with S. pneumoniae, and one with Staphylococcus aureus).

In brief, we observed the presence of pathogen antigen without inflammatory cells specifically in patients with PF at the site of maximal vascular damage.

The present study was dedicated to understanding the pathophysiology of PF in the adult population. We assessed a variety of blood and tissular parameters susceptible to be implicated and especially the balance between pro- and anticoagulant parameters. To determine the specificity of the encountered abnormalities compared with other forms of severe sepsis, skin and blood abnormalities were compared with those of control patients with severe sepsis but without PF lesions. In accordance with previous data, PF prevalence among patients with severe sepsis was low (1, 2, 4). The extreme severity of PF with multiple organ dysfunctions, high mortality rate, and frequent debilitating sequelae in survivors was established by comparison with the two control groups, which confirms and extends previous reports in the literature (1, 2, 4).

Our working hypothesis was that PF was related to a thrombotic process secondary to an interplay between systemic and local factors. Indeed, beyond thrombi, we observed in PF purpuric samples devastating vascular lesions with massive vascular congestion and dilation together with diffuse endothelial damage with loss of endothelial integrity markers. These lesions culminated in frequent endothelial cell necrosis. These vascular lesions retained little attention in the previous literature focusing on meningococcal sepsis (8, 9). Beyond these morphological changes, a multifaceted local process impairing the balance between pro- and antithrombotic factors was observed in purpuric samples: diminished expression of the endothelial receptors EPCR and TM confirmed a previous study, in children with PF, focused on the PC pathway (8); increases in PAI-1 mRNA and in TF mRNA, although the latter was not statistically significant. Of note, even in the absence of increased TF expression, the endothelial necrosis may have uncovered TF normally present beneath the endothelium. Although the specificity of these observations for the pathophysiology of PF was shown by comparison with skin control samples, biopsies obtained in nonpurpuric samples of patients with PF occasionally showed some of the above-mentioned abnormalities, indicating that the process underlying PF lesions may be more diffuse, although at a low grade, than could be inferred only from examination of the skin. It may be debated whether endothelial lesions precede and provoke thrombosis or whether they are, rather, the consequence of thrombus formation through an upstream increase in blood pressure. Although the absence of red cell extravasation argues against the presence of such a mechanism, the two hypotheses are not mutually exclusive. Last, our observations establish that PF is both a thrombotic and vascular disease.

Severe activation of coagulation was evidenced in all patients with PF (1, 4, 5, 7). Comparison with SAPSII-control patients showed a dramatically higher frequency of overt DIC. However, the comparison with DIC patients without skin lesions indicated that the level of coagulation abnormalities observed in patients with PF does not automatically lead to PF. This observation challenges the frequent hypothesis that DIC is the origin of PF lesions through vascular thrombosis, or at least that DIC alone is sufficient to induce PF (1, 4, 5, 7). Additional evidence against DIC being a key initiating factor for PF comes from the observation that DIC is encountered in 30% of severe sepsis, contrasting with the low frequency of PF (13). Special attention is to be paid to AT and PS, which were shown to be lower in patients with PF than in control groups. These natural anticoagulants are known to decrease in severe sepsis, with lower concentrations correlating with enhanced severity (14, 15). It is therefore difficult to infer from our data to what extent these remarkably low AT and PS levels have an active role in the pathophysiology of PF lesions or whether they are only markers of disease severity. Alternatively, in the absence of circulating aPC diminution in patients with PF, a defect in aPC generation alone cannot explain the prothrombotic process, as previously suggested (8).

The role of the endothelium at the microcirculatory level is known to be crucial in severe sepsis, but our study adds another level of complexity to the relationship between microcirculation and systemic abnormalities. Few data have been published on ultrastructural vascular changes in PF or even in sepsis, and even less on the relationship between microvascular circulation derangements and systemic changes (1618). In one study on sepsis little correlation could be made between ultrastructural vascular abnormalities and circulating factors considered to be markers of endothelial injury, such as soluble intercellular adhesion molecule (sICAM) and soluble vascular cell adhesion molecule (sVCAM) (19). Circulating endothelial cells have been demonstrated in septic shock, and are taken to be indirect proof of endothelial damage, but the implication of these cells regarding the extent of endothelial damage is unknown (20). In our study, although skin expression of endothelial proteins (TM and EPCR) involved in the aPC pathway was altered in patients with PF, circulating aPC was not diminished. In addition, sTM, an increase in which is generally considered a marker of endothelial damage (21), was paradoxically found to be lower in patients with PF in comparison with DIC- and SAPSII-control subjects despite evidence for extensive endothelial damage in the former. Therefore, it appears that local endothelial abnormalities may be difficult to identify when analyzing circulating parameters. Last, the discrepancy between low sTM values in PF and high values in DIC suggests different pathophysiologies, indicating again that PF cannot be attributed to a severe form of DIC. Identically, extensive cellular damage in our patients with PF contrasts with previous observations of a functional microvascular dysfunction in the “more usual” form of sepsis as assessed by a conserved response to acetylcholine in the sublingual circulation (22).

The role of the pathogen in severe sepsis is generally seen as a trigger to a maladaptive response of innate immunity, which is the key element responsible for organ failure (16, 23). However, the close and specific association between bacterial presence and vascular damage in patients with PF infection is likely to indicate a direct link between these pathogens and PF lesions. Persistence of N. meningitidis DNA or antigen in skin samples of patients with PF despite prior active antibiotic therapy is in accordance with previous reports (24, 25), but we add here new evidence regarding the presence of other pathogens including both gram-positive and gram-negative bacteria. We cannot rule out that the bacterial presence in the skin biopsies is related to entrapment of circulating bacteria or bacterial products in previously injured blood vessels. However, a new humanized mouse model of N. meningitis infection has led to a similar finding with bacterial adhesion to human dermal microvessels seen to be crucial to the development of dermal lesions (26). Taken together, this animal study and our observations lead us to propose a vascular wall infection hypothesis in which bacterial attachment to endothelial cells is an initiating factor for endothelial damage leading to the development of PF lesions.

Our study suffers from several limitations. First, it included only 20 patients. The similarity of clinical presentation, outcome, circulating blood abnormalities, and pathological lesions across all patients reassures us in the belief that our data set is sufficient to describe a common pathophysiological process irrespective of the initiating infection agent. Second, we did not assess vascular lesions and bacterial presence in organs other than the skin. In one study on PF, little thrombosis was noted outside the skin and adrenal glands (25). Alternatively, in a study on kidney lesions in septic shock patients without PF, we observed vascular dilation, congestion, and thrombi in the kidney of some patients (17). This may evoke the possibility of a similar mechanism that the one hypothesized here in various organs in some patients with severe sepsis.

In conclusion, we observed that PF lesions associated both thrombosis and dramatic vascular damage with multifaceted prothrombotic local imbalance. Although coagulation abnormalities in the circulating blood were constant, the relationship between skin and circulating abnormalities was difficult to establish. Last, bacterial presence, specifically at the site of vascular damage, was observed. Putting these different elements into perspective allowed us to propose a primitive “vascular wall infection” hypothesis, responsible for endothelial damage.

1. Davis MDP, Dy KM, Nelson S. Presentation and outcome of purpura fulminans associated with peripheral gangrene in 12 patients at Mayo Clinic. J Am Acad Dermatol 2007;57:944956.
2. Vincent J-L, Nadel S, Kutsogiannis DJ, Gibney RTN, Yan SB, Wyss VL, Bailey JE, Mitchell CL, Sarwat S, Shinall SM, et al. Drotrecogin alfa (activated) in patients with severe sepsis presenting with purpura fulminans, meningitis, or meningococcal disease: a retrospective analysis of patients enrolled in recent clinical studies. Crit Care 2005;9:R331R343.
3. Buysse CMP, Oranje AP, Zuidema E, Hazelzet JA, Hop WCJ, Diepstraten AF, Joosten KFM. Long-term skin scarring and orthopaedic sequelae in survivors of meningococcal septic shock. Arch Dis Child 2009;94:381386.
4. Kravitz GR, Dries DJ, Peterson ML, Schlievert PM. Purpura fulminans due to Staphylococcus aureus. Clin Infect Dis 2005;40:941947.
5. Hayes MA, Yau EH, Hinds CJ, Watson JD. Symmetrical peripheral gangrene: association with noradrenaline administration. Intensive Care Med 1992;18:433436.
6. Dünser MW, Mayr AJ, Tür A, Pajk W, Barbara F, Knotzer H, Ulmer H, Hasibeder WR. Ischemic skin lesions as a complication of continuous vasopressin infusion in catecholamine-resistant vasodilatory shock: incidence and risk factors. Crit Care Med 2003;31:13941398.
7. Fourrier F, Leclerc F, Aidan K, Sadik A, Jourdain M, Tournoys A, Noizet O. Combined antithrombin and protein C supplementation in meningococcal purpura fulminans: a pharmacokinetic study. Intensive Care Med 2003;29:10811087.
8. Faust SN, Levin M, Harrison OB, Goldin RD, Lockhart MS, Kondaveeti S, Laszik Z, Esmon CT, Heyderman RS. Dysfunction of endothelial protein C activation in severe meningococcal sepsis. N Engl J Med 2001;345:408416.
9. Harrison OB, Robertson BD, Faust SN, Jepson MA, Goldin RD, Levin M, Heyderman RS. Analysis of pathogen–host cell interactions in purpura fulminans: expression of capsule, type IV pili, and PorA by Neisseria meningitidis in vivo. Infect Immun 2002;70:51935201.
10. American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference: definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Crit Care Med 1992;20:864874.
11. Taylor FB Jr, Toh CH, Hoots WK, Wada H, Levi M; Scientific Subcommittee on Disseminated Intravascular Coagulation (DIC) of the International Society on Thrombosis and Haemostasis (ISTH). Towards definition, clinical and laboratory criteria, and a scoring system for disseminated intravascular coagulation. Thromb Haemost 2001;86:13271330.
12. Mairey E, Genovesio A, Donnadieu E, Bernard C, Jaubert F, Pinard E, Seylaz J, Olivo-Marin J-C, Nassif X, Duménil G. Cerebral microcirculation shear stress levels determine Neisseria meningitidis attachment sites along the blood–brain barrier. J Exp Med 2006;203:19391950.
13. Dhainaut J-F, Yan SB, Joyce DE, Pettilä V, Basson B, Brandt JT, Sundin DP, Levi M. Treatment effects of drotrecogin alfa (activated) in patients with severe sepsis with or without overt disseminated intravascular coagulation. J Thromb Haemost 2004;2:19241933.
14. Fourrier F, Chopin C, Goudemand J, Hendrycx S, Caron C, Rime A, Marey A, Lestavel P. Septic shock, multiple organ failure, and disseminated intravascular coagulation: compared patterns of antithrombin III, protein C, and protein S deficiencies. Chest 1992;101:816823.
15. Kinasewitz GT, Yan SB, Basson B, Comp P, Russell JA, Cariou A, Um SL, Utterback B, Laterre P-F, Dhainaut J-F; PROWESS Sepsis Study Group. Universal changes in biomarkers of coagulation and inflammation occur in patients with severe sepsis, regardless of causative micro-organism [ISRCTN74215569]. Crit Care 2004;8:R82R90.
16. Aird WC. The role of the endothelium in severe sepsis and multiple organ dysfunction syndrome. Blood 2003;101:37653777.
17. Lerolle N, Nochy D, Guérot E, Bruneval P, Fagon J-Y, Diehl J-L, Hill G. Histopathology of septic shock induced acute kidney injury: apoptosis and leukocytic infiltration. Intensive Care Med 2010;36:471478.
18. Hotchkiss RS, Swanson PE, Freeman BD, Tinsley KW, Cobb JP, Matuschak GM, Buchman TG, Karl IE. Apoptotic cell death in patients with sepsis, shock, and multiple organ dysfunction. Crit Care Med 1999;27:12301251.
19. Leone M, Boutière B, Camoin-Jau L, Albanèse J, Horschowsky N, Mège J-L, Martin C, Dignat-George F. Systemic endothelial activation is greater in septic than in traumatic-hemorrhagic shock but does not correlate with endothelial activation in skin biopsies. Crit Care Med 2002;30:808814.
20. Mutunga M, Fulton B, Bullock R, Batchelor A, Gascoigne A, Gillespie JI, Baudouin SV. Circulating endothelial cells in patients with septic shock. Am J Respir Crit Care Med 2001;163:195200.
21. Reinhart K, Bayer O, Brunkhorst F, Meisner M. Markers of endothelial damage in organ dysfunction and sepsis. Crit Care Med 2002;30(5 Suppl):S302S312.
22. De Backer D, Creteur J, Preiser J-C, Dubois M-J, Vincent J-L. Microvascular blood flow is altered in patients with sepsis. Am J Respir Crit Care Med 2002;166:98104.
23. Cohen J. The immunopathogenesis of sepsis. Nature 2002;420:885891.
24. Staquet P, Lemee L, Verdier E, Bonmarchand G, Laudenbach V, Michel C, Lemeland J-F, Marret S, Blanc T. Detection of Neisseria meningitidis DNA from skin lesion biopsy using real-time PCR: usefulness in the aetiological diagnosis of purpura fulminans. Intensive Care Med 2007;33:11681172.
25. Guarner J, Greer PW, Whitney A, Shieh W-J, Fischer M, White EH, Carlone GM, Stephens DS, Popovic T, Zaki SR. Pathogenesis and diagnosis of human meningococcal disease using immunohistochemical and PCR assays. Am J Clin Pathol 2004;122:754764.
26. Melican K, Michea Veloso P, Martin T, Bruneval P, Duménil G. Adhesion of Neisseria meningitidis to dermal vessels leads to local vascular damage and purpura in a humanized mouse model. PLoS Pathog 2013;9:e1003139.
Correspondence and requests for reprints should be addressed to Nicolas Lerolle, M.D., Ph.D., Département de Reanimation Médicale et de Medicine Hyperbare, CHU Angers, Rue Larrey, 49933 Angers, France. E-mail:

Supported by a grant from the Société de Réanimation de Langue Française; S. Gandrille was supported by a grant (Contrat d’Interface INSERM/AP-HP) from the Institut National de la Santé et de la Recherche Médicale (INSERM) et Assistance Publique-Hôpitaux de Paris (AP-HP). Electron microscopy was performed with the help of Service Commun d’Imageries et d’Analyses Microscopiques (SCIAM), Université d’Angers, Angers, France.

Author Contributions: N.L. designed research, performed research, collected data, analyzed and interpreted data, performed statistical analysis, and wrote the manuscript; A.C., G.D., K.M., F.A., P.B., and D.B. analyzed and interpreted data and contributed vital analytical tools; M.P., J.-L.D., V.C., and G.H. performed research and collected data; S.G. and C.M. contributed vital analytical tools.

This article has an online supplement, which is available from this issue’s table of contents at

Originally Published in Press as DOI: 10.1164/rccm.201302-0228OC on August 7, 2013

Author disclosures are available with the text of this article at


No related items
American Journal of Respiratory and Critical Care Medicine

Click to see any corrections or updates and to confirm this is the authentic version of record