ORIGINAL ARTICLE Scand J Infect Dis 29: 479-483, 1997

Purpura Fulminans in Pneumococcal Sepsis: Case Report and Review CHACE T. CARPENTER and ALLEN B. KAISER From the Depurtineni of Medicine, Vanderbilt Uttiuersity School of Medicine, ~ushuille, TN, USA

Purpura fulminans is classically defined by ecchymotic skin lesions, fever, and hypotension. The majority of cases occur in association with bacterial sepsis, and disseminated intravascular coagulation (DIC) is usually present. Prompted by our experience with a patient with pneumococcal sepsis and purpura fulminans in whom hypotension was never observed, we evaluated the important parameters of sepsis in reports of this syndrome. 42 additional cases of pneumococcal bacteremia and purpura fulminans were identified. Hypotension was present in only 51%. Although DIC was present in 85% of patients, hypofibrinogenemia was documented in only 26%. By contrast, both hypotension and hypofibrinogenemia are present in the vast majority of patients described with purpura fulminans in association with meningococcal sepsis. These data confirm that hypotension is not a necessary feature of the syndrome of purpura fulminans associated with pneumococcal sepsis and suggest further that qualitative or quantitative differences exist in the DIC cascade of pneumococcal vs meningococcal sepsis.

A . B. Kaiser, MD, Department of Medicine, 0-3100 MCN. Vandevbilt University School of Medicine, Nashville, TN 37232-2358, USA

20 years have passed since the classic monograph on pur- pura fulminans by Spicer and Rau was published (1). In that review, 3 primary clinical features of the syndrome were noted: (i) large purpuric lesions of the skin; (ii) fever; and (iii) hypotension. Hematologic and/or pathologic evi- dence of disseminated intravascular coagulation (DIC) was characteristically present as well. We present here a case report and review of the literature of purpura fulminans occurring in pneumococcal sepsis. The purpose of this review is to summarize what is known about the patho- genesis of pneumococcal purpura fulminans and to deter- mine the necessity of hypotension as a feature of the syndrome.

CASE REPORT A 46-year-old previously healthy white male presented to his local hospital approximately 18 h after the sudden onset of fever, chills, and malaise. 17 years earlier he had undergone a splenectomy following a gunshot wound.

The patient was alert but appeared acutely ill with complaints of generalized malaise and severe pain in his legs and feet. His blood pressure was 110/60 mmHg; pulse, 104/minute and regular; respiratory rate, I8jminute; and oral temperature, 3823°C. The extremities were cool to the touch, and mottled cyanosis of his nose, lips, ears, and hands was present. He had scattered, non-blanching petechial lesions over his trunk. There were confluent purpuric lesions on his legs and feet. Arterial pulses were intact. The remainder of the physical exam was unremarkable, including clear lungs and a supple neck. Blood cultures were obtained and he was treated with ceftriaxone and gentamicin for presumed sepsis. Methylene blue was given intravenously for concern of possible methemoglobinemia. He was transferred to our facility 24 h later because of increasing cyanosis, hypoxemia, and rapidly rising serum creatinine.

Physical examination at our facility was unchanged. Laboratory examination revealed: white blood cell count, 26.9 x 109/l; hematocrit, 0.46; platelet count, 68 x 109/l; serum creatinine, 433 pmol/l; fibrin split products, 154 (normal range = 0- 1:4);

fibrinogen, 510 mg/dl (normal range = 190-400); prothrombin time, 15 s (normal range 10-13); and activated partial thromboplastin time, 36 s (normal range 25-40). Serum TNF-a level was elevated at 56 pg/ml.

The pulmonary capillary wedge pressure was 10 mmHg. Urine output improved with hydration. The patient was treated with ceftriaxone and penicillin. On the second hospital day, blood cultures from his local hospital were reported as positive for S. pneumoniae, and ceftriaxone was discontinued. Skin biopsy of one of the petechial lesions on his trunk revealed dermal thrombosis and hemorrhage, as typically seen in purpura fulminans. His serum creatinine rose to a maximum of 645 prnol/i, but over the course of his hospitalization. his renal function slowly returned to normal. The purpuric lesions and cyanosis resolved on his ears, nose, lips, and upper extremities. He had necrosis of the tissue on several of his toes and subsequently had autoamputation of the distal portion of eight toes. He was discharged after a 21-day hospital course with plastic surgery follow-up. Pneumococcal vaccine was administered prior to discharge.

METHODS The literature of purpura fulminans associated with pneumococcal sepsis was reviewed. Case reports were included if the patients had documented pneumococcal bacteremia, temperature greater than 38°C (100.4 F), and a clinical diagnosis of purpura fulminans. Purpura fulminans was considered present if so stated by the authors, or if ‘gangrene’ or ‘purpura’ were used to describe the skin and soft tissue abnormalities. While hypotension was not considered a necessary feature of the syndrome, the presence or absence of hypotension at the time of onset of the skin lesions was noted. Hypotension was considered present if the systolic blood pressure was < 90 mmHg, the patient was described as being in ‘shock,’ or vasoconstrictor agents were employed to support the blood pressure. Coagulation studies were also reviewed. DIC was considered to be present if two of the following abnormalities were present: platelets were described as ‘low’ or were < 100 x 109/l; fibrin degradation products were elevated; fibrinogen was described as ‘low’ or was < 2g/l; or prothrombin time was ‘prolonged,’ > 1 s from control, or > 14 s. An ‘absent spleen’ was assumed if specifically mentioned, if the patient had sickle cell anemia, or if asplenia was documented at autopsy.

0 1997 Scandinavian University Press. ISSN 0036-5548

Scan

d J

Infe

ct D

is D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y SU

NY

Sta

te U

nive

rsity

of

New

Yor

k at

Sto

ny B

rook

on

10/2

7/14

For

pers

onal

use

onl

y.

480 C. T. Carpenter and A . B. Kaiser Scand J Infect Dis 29

Table I. Clinical and laborutory findings in patients presenting with pneumococcemia, fever, and purpura fulminans

DIC Values Case No. Spleen Clinical (Ref) Age/Sex Hypotension absent DIC"? Abnormal Normal gangrene Outcome

1 (2) 2 (3) 3 (4) 4 (4) 5 ( 5 ) 6 (6) 7 (7) 8 (8) 9 (8)

10 (9) 11 (10) 12 (11) 13 (11) 14 (12) 15 (13) 16 (14) 17 (15) 18 (16) 19 (17) 20 (17) 21 (18) 22 (19) 23 (20) 24 (21) 25 (22) 26 (23) 27 (24)

37 (26) 38 (27) 39-41 (28)*** 42 (29) 43

28-36 (25)**

9 1 ~ 60/M 61/F 37/M 32/F

19/M 16/M 70/M 34/F 22/M 22/F 70/F 31/M 63/F 65/M 80/M 62/F 2/F

8 mo/M 34/M 33/F 40/F 31/M 34/M 23/M

1 IF Adults 43/F 67/F 1 5 y*** 60/F 46/M

5 1 ~

No No No Yes* N o Yes Yes Yes Yes Yes No Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No Yes No Yes Yes Yes Yes Yes No No No No ID Yes ID Yes No Yes No Yes Yes Yes No No No Yes No Yes Yes No 3/9** 4/9** Yes No Yes No 3/3*** 0/3*** Yes No No Yes

ID IDb ID ID Yes Yes No Yes No Yes Yes ID Yes Yes Yes Yes Yes ID Yes ID Yes Yes Yes No Yes Yes Yes 7/9** Yes Yes ID Yes Yes

PT, F

P, F

p, F

F

F F F P F PT, F

F

p, F F

F

F

Yes Yes No No Yes Yes Yes No ID Yes Yes Yes Yes Yes Yes Yes Yes Yes No No Yes Yes No No Yes Yes No 9/9** Yes Yes ID No Yes

Survived' Survived'.d Expired Expired Survivedc Survived Survived Expiredd Expiredd Expired Expired Survived' SurvivedC Survived Unknown Expired Expired Expired Expiredd Survivedd SurvivedC Survived Survivedd Expired" Survivede Survived" Expired" 9/9** Survived Expiredd Survived" 2/3*** Survived Expired Survived"

a Disseminated intravascular coagulation defined as present if so stated by the authors or if abnormalities occurred in at least two of the following laboratory tests: platelets (P), defined as low (JP), if so stated or if reported as < 100 x 109/1; fibrin degradation products (FDP), defined as elevated (TFDP), if so stated; prothrombin time (PT), defined as prolonged (TPT), if so stated or if z 1 s from control or > 14 s; fibrinogen (F), defined as low (JF), if so stated or <200 mg/dl. ID, insufficient data available to make a determination. Surgical or auto-amputation of tissue (usually on toes or fingers) occurred. Identifies patients who also had pneumococcal meningitis. Patient known to be HIV-positive.

* History of persistent LUQ pain following an automobile accident and exploratory laparotomy 2 y prior to presentation. ** Reported as a series of 10 patients, 1 of whom was described in detail [No. 23 (20)]. *** Reported from the CDC in a series of 3 patients.

RESULTS

Including this report, 43 cases with pneumococcal sepsis and purpura fulminans were identified (2-29) (Table I). Out-

Of the 24 survivors for whom outcome was reported in

amputation.

The spleen was determined to be absent in 27 patients (63%). Additional patients may have had impaired splenic function due to alcohol-related liver disease. The absence

relationship to survival (60%) vs 56%, respectively).

cases (51Y0). Two of the patients who were reported to be

come was documented in 42 cases; 27 patients survived. presumed presence Of a 'PIeen had no apparent

detail, only 7 survived without losing limbs or digits to Hypotension was present in Only 21 Of the 41 evaluable

Scan

d J

Infe

ct D

is D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y SU

NY

Sta

te U

nive

rsity

of

New

Yor

k at

Sto

ny B

rook

on

10/2

7/14

For

pers

onal

use

onl

y.

Scdnd J Infect Dis 29 Purpura Fulminans 48 1

hypotensive had adrenal hemorrhage (1 documented by autopsy and the other by ultrasound). A trend of higher mortality among hypotensive patients was observed; 50% of patients presenting with hypotension and purpura died compared to only 20% of those presenting with purpura alone (p = 0.097, x2 with Yates’ correction). Coagulation studies were reported for 33 patients, 28 (85%) of whom had a profile consistent with DIC. Seven patients had skin biopsies, 5 of which revealed the classically described lesion of purpura fulminans, i.e. dermal thrombosis and hemorrhage. Of the 19 patients for whom plasma fibrinogen levels were reported, 14 (74%) had normal or elevated levels. However, fibrin degradation products were elevated in all 12 patients for whom these values were reported. Survival had no apparent correlation with any of the parameters of DIC.

DISCUSSION

This review reveals several interesting features of purpura fulminans in pneumococcal sepsis, the first being that hy- potension is not a necessary feature of the syndrome, occurring in only 51% of reported cases. In comparison, literature reviews of meningococcal sepsis indicate that purpura fulminans and shock almost always occur to- gether. On one hand, in several instances hypotension was included in the definition of the syndrome, as in the 35 cases reported by Giraud et al. (30). Similarly, Powars et al. noted that a subset of patients with meningococcal disease exhibited both ‘rapidly progressive purpura fulminans’ and ‘cardiovascular collapse or shock’, features which helped to define a population of severely ill (‘Stage 3’) patients (31). Similarly, McGehee et al., in their study of 19 patients with meningococcal infection, noted that ‘The amount of pur- pura varied but was most marked in those patients with severe shock’ (32). In contrast, Toews and Bass, in a report on the various skin manifestations of meningococcal infec- tion, provide some evidence that purpura fulminans need not always be accompanied by shock; they found that shock was present at the time of hospitalization in only 62% of patients with a ‘bleeding diathesis’ and skin lesions consistent with purpura fulminans (33). However, these investigators did note that patients with purpura were significantly more likely to present with shock than patients with meningococcal infection and no purpura (62% vs 8%, respectively.)

The observation that hypotension was not required in the development of purpura fulminans in pneumococcal sepsis led us to review the pathogenesis of DIC and purpura fulminans in sepsis. Most studies have focused on the effects of endotoxin, the initiator of the inflammatory cas- cade in Gram negative sepsis. Endotoxin has been demon- strated to cause increased levels of cytokines, such as tumor necrosis factor-alpha (TNF-a), which are now widely rec- ognized as the mediators of sepsis (34, 35). TNF-a has been

shown to play a pivotal role in the development of DIC in sepsis through several mechanisms. In brief, TNF-a ini- tiates a procoagulant cascade, inhibits the fibrinolytic sys- tem, and prepares the endothelium to be receptive to neutrophils. The neutrophils then mediate further endothe- lial damage with resultant increased tissue factor expression and, thus, the accelerated coagulation of DIC proceeds (36). Microthrombi form and occlude the microvasculature leading to hypoperfused, ischemic tissue, which may infarct and become gangrenous. Red blood cells escape from the damaged endothelium resulting in local hemorrhage.

The initiating factor in Gram-positive sepsis is less well defined. However, coat mucopolysaccharides of Gram-posi- tive bacteria, as well as the enterotoxins and exotoxins released by these organisms, have been shown to initiate cytokine release. Thus, sepsis may be initiated in the ab- sence of endotoxin (37). In the case of the pneumococcus, meningitis studies have shown that the cell wall, though less important than the capsule for antigenicity, is responsible for the inflammatory reaction of the host (38). This reac- tion is characterized by activation of the complement cas- cade, release of TNF-a, interleukin-1 (IL-l), and prostaglandins. This activation of cytokines initiates the procoagulant cascade (39). TNF-alpha serum levels were documented as being elevated in our patient 48 h after the onset of his illness.

TNF-a provides the common pathway for the initiation of DIC in both Gram-positive and Gram-negative sepsis. Tuomanen et al., in a recent review on the pathogenesis of pneumococcal infection cite evidence that differences in the binding of endotoxin vs Gram-positive cell walls to cell surface receptors results in different induction patterns for the cytokine cascade (39). These subtle differences may eventually prove to result in a lesser incidence of hypoten- sion in pneumococcal vs meningococcal purpura fulminans.

Another striking observation was the predominantly nor- mal or elevated plasma fibrinogen levels. The classic criteria for the syndrome of DIC have included thrombocytopenia, elevated fibrin split products, and decreased plasma fibrino- gen level. However, more recently, DIC has been recog- nized as including a spectrum of consumptive coagulopathies in which plasma fibrinogen levels may be low, normal, or elevated. Giraud et al. reported that 44% of the patients in their series of meningococcal purpura fulmi- nans had depressed plasma fibrinogen levels (30). This finding was an independent predictor of poor outcome. That the vast majority of patients with pneumococcal pur- pura fulminans have normal or elevated plasma fibrinogen levels suggests that quantitative or qualitative differences exist for the initiation of DIC in these syndromes as they apparently do for the initiation of vasodilatation and subse- quent hypotension.

Finally, an absent spleen was usually noted in reported cases. While bias may be introduced by possible over-re- porting of this finding, the high occurrence (63%) suggests

Scan

d J

Infe

ct D

is D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y SU

NY

Sta

te U

nive

rsity

of

New

Yor

k at

Sto

ny B

rook

on

10/2

7/14

For

pers

onal

use

onl

y.

482 C. T. Carpenter and A. B. Kaiser Scand J Infect Dis 29

that an absent spleen is a risk factor for developing purpura fulminans in the setting of pneumococcal sepsis. The reticu- loendothelial system is responsible for clearance of pneu- mococci from the bloodstream. After splenectomy the remaining macrophages of the reticuloendothelial system require increased amounts of antibody to efficiently clear organisms (40). The spleen was also found to have a graded role in clearance depending on the virulence of the invading pneumococcus, with the greatest splenic clearance occur- ring with the most virulent organisms (41). The unique function of the spleen may be the clearance of organisms that evade hepatic sequestration. The absence of a func- tional spleen may also allow acceleration of DIC. The reticuloendothelial system also has the capacity to sequester some components of the coagulation cascade (4). Thus, coagulation, which proceeds virtually unchecked in DIC, may progress even more rapidly in the absence of the spleen.

In summary, we have demonstrated that hypotension is not a universal feature of purpura fulminans in pneumo- coccal sepsis, occurring in approximately half of all iden- tified cases. In contrast, hypotension does appear to be a dominant feature of meningococcal purpura fulminans. Additionally there appear to be differences in the patho- physiology of DIC in purpura fulminans caused by each of these organisms. Fibrinogen levels, which tend to be nor- mal or elevated in pneumococcal purpura fulminans, are more likely to be depressed in the meningococcal counter- part. These observations suggest there are differences among the pathways that lead to purpura fulminans in Gram-positive compared to Gram-negative sepsis. Recent investigative research suggests the explanation for these findings lies a t the level of the cytokines mediating the syndrome.

ACKNOWLEDGEMENT

We are indebted to Dr Ed Settle, whose referral of the patient presented in our case report provoked our interest in the literature of pneumococcal purpura fulminans.

REFERENCES

1.

2.

3.

4.

5.

6.

Spicer TE, Rau JM. Purpura fulminans. Am J Med 1976; 61: 566-71. Uhr JS. Gangrene of the toes following pneumococcemia, treatment with sulfanilamide. Am J Dis Child 1938; 56: 1317. Bettigole RE. Symmetrical peripheral gangrene in pneumococ- cal sepsis. Ann Intern Med 1961; 54: 335-42. Whitaker AN. Infection and the spleen: Association between hyposplenism, pneumococcal sepsis and disseminated intravas- cular coagulation. Med J Aust 1969; 1: 1213-9. Stossel TP, Levy R. Intravascular coagulation associated with pneumococcal bacteremia and symmetrical peripheral gan- grene. Arch Intern Med 1970; 125: 876-8. Adner MM, Kauff RE, Sherman JD. Purpura fulminans in a child with pneumococcal septicemia two years after splenec- tomy. JAMA 1970; 213: 1681-3.

7. Karz SC, Thompson RE, Depenbusch FL, Hart GB. Hyper- baric oxygen treatment of gangrene in pneumococcemia.

8. Coonrod JD, Leach RP. Antigenemia in fulminant pneumo- coccemia. Ann Int Med 1976; 84: 561-3.

9. Dubouloz F, Domergue R, Francois G, Charrel M, Michaud A, Levy G. Purpura fulminans chez une malade splenectomiste [Letter]. Presse Med 1978; 7: 3459.

10. Tanguy RL, Delalande JP, L’Azou M, Egreteau JP. Purpura fulminans a pneumocoques chez un adulte splenectomisk avec persistance de tissu splenique accessoire [Letter]. Presse Med 1981; 10: 2290.

11. Hautekeete ML, Berneman ZN, Bieger R, Stevens WJ, Bridts C, Buyssens N, Peetermans ME. Purpura fulminans in pneu- mococcal sepsis. Arch Intern Med 1986; 146: 497-9.

12. Rusonis PA, Robinson HN, Lamberg SI. Livedo reticularis and purpura: presenting features in fulminant pneumococcal septicemia in an asplenic patient. J Am Acad Dermatol 1986; 15: 1120-2.

13. Benson PM, Lupton GP, James WD. Purpura and gangrene in a septic patient. Arch Dermatol 1988; 124: 1851-4.

14. Shperber Y, Rudick V, Geller E, Orda R. Purpura fulminans associated with overwhelming sepsis and disseminated in- travascular coagulation following splenectomy. Isr J Med Sci

15. Marcus EL, Hersch M, Sonnenblick M. A rare association of rhabdomyolysis and purpura fulminans in Strepneumoniae septicaeniia [Letter]. Intensive Care Med 1990; 16: 521.

16. Singer RM, Gorosh JE. Purpura fulminans. J Hand Surg 1990;

17. Cohen JR, Lackner R, Keller A, Douglas B. The surgical implications of purpura fulminans. Annals of Vascular Surgery

18. Har-el G, Nash M, Chin NW, Meltzer CJ, Weiss MH. Pur- pura fulminans of the head and neck. Otolaryngol Head Neck Surg 1990; 103: 660-3.

19. Bielory L, Fisher AM, Lyons MM, Holland CL. Palpable purpura in a splenectomized patient with a streptococcal infec- tion. Ann Allergy 1991; 66: 459-63.

20. Johansen K, Murphy T, Pavlin E, Ledbetter D. Digital is- chemia complicating pneumococcal sepsis: Reversal with sym- pathetic blockade. Crit Care Med 1991; 19: 114-6.

21. Barradas MCR, Musher DM, Hamill RJ, Dowell M, Bagwell JT, Sanders CV. Unusual manifestations of pneumococcal infection in human immunodeficiency virus-infected individu- als: The past revisited. CID 1992; 14: 192-9.

22. Shivaram U, Cash M. Purpura fulminans, metastatic endoph- thalmitis, and thrombotic thrombocytopenic purpura in an HIV-infected patient. NY State J Med 1992; 92: 313-4.

23. Murphy CB, Nueller K. Purpura fulminans secondary to pneumococcal sepsis in an asplenic patient. Am J Pediatr Med 1993; 83: 43-6.

24. Ryan CA, Wenman W, Henningsen C, Tse S. Fatal Childhood pneumococcal Waterhouse-Fredrichsen syndrome. Pediatr In- fect Dis J 1993; 12: 250-1.

25. Johansen K, Hansen ST. Symmetrical peripheral gangrene (purpura fulminans) complicating pneumococcal sepsis. Am J Surg 1993; 165: 642-5.

26. Strobel M, Ertlen P, Marie-Saint-Germain E, Lacave 5. Al- coolisme et pneumococcie grave, a propos d’un cas de purpura fulminans a streptococcus pneumoniae. Rev Med Intern 1993; 14: 46-8.

27. Ohnishi M, Skimizu Y, Iwata K, Ookochi Y , Ooe K. Purpura fulminans complicating pneumococcal sepsis: a case report. Kansenshogaku Zasshi (J Jpn Assoc Infect Dis) 1994; 68: 1 117-21.

JAMA 1971; 217: 962-3.

1989; 25: 657-9.

15A: 172-5.

1990; 4: 276-9.

Scan

d J

Infe

ct D

is D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y SU

NY

Sta

te U

nive

rsity

of

New

Yor

k at

Sto

ny B

rook

on

10/2

7/14

For

pers

onal

use

onl

y.

S a n d J Infect Dis 29 Purpura Fulminans 483

28. Centers for Disease Control and Prevention. Hemorrhage and shock associated with invasive pneumococcal infection in healthy infants and children-New Mexico, 1993-94. MMWR Morb Mortal Wkly Rep 1995; 43: 949-52.

29. Murph RC, Matullis WS, Hernandez JE. Rapidly fatal pneu- mococcal sepsis in a healthy adult. Clin Infect Dis 1996; 22: 375-6.

30. Giraud T, Dhainaut JF, Schremmer B, Regnier B, Desjars P, Loirat P, Journois D, Lanore JJ. Adult overwhelming meningococcal purpura: A study of 35 cases, 1977-1989. Arch Intern Med 1991; 151: 310-6.

31. Powars D, Larsen R, Johnson J, Hulbert T, Sun T, Patch MJ, Francis R, Chan L. Epidemic Meningococcemia and purpura fulminans with induced protein C deficiency. Clin Infect Dis

32. McGehee WG, Rapaport SI, Hjort PF. Intravascular coagula- tion in fulminant meningococcemia. Ann Intern Med 1967; 67:

33. Toews WH, Bass JW. Skin manifestations of meningococcal infection, an immediate indicator of prognosis. Am J Dis Child 1974; 127: 173-6.

34. Michie HR, Manogue KR, Spriggs DR, et al. Detection of

1993; 17: 254-61.

250-60.

circulating tumor necrosis factor after endotoxin administra- tion. N Engl J Med 1988; 318: 1481-6.

35. van Deventer SJH, Buller HR, ten Cate JW, Aarden LA, Hack CE, Sturk A. Experimental endotoxemia in humans: analysis of cytokine release and coagulation, fibrinolytic, and comple- ment pathways. Blood 1990; 76: 2520-6.

36. Levi M, tenCate H, vander Poll T, vanDeventer SJH. Patho- genesis of disseminated intravascular coagulation in sepsis. JAMA 1993; 270: 975-9.

37. Bone RC. Gram-positive organisms and sepsis. Arch Intern Med 1994; 154: 26-34.

38. Bruyn GAW, Zegers BJM, van Furth R. Mechanisms of host defense against infection with Streptococcus pneumoniae. Clin Infect Dis 1988; 14: 251-62.

39. Tuomanen EI, Austrian R, Masure HR. Pathogenesis of pneu- mococcal infection. N Engl J Med 1995; 332: 1280-4.

40. Hosea SW, Brown EJ, Hamburger MI, Frank MM. Opsonic requirements for intravascular clearance after splenectomy. N Engl J Med 1981; 304: 245-50.

41. Brown EJ, Hosea SW, Frank MM. The role of the spleen in experimental pneumococcal bacteremia. J Clin Invest 1981; 67: 975-82.

Received May 20, 1997; accepted August 14, 1997

Scan

d J

Infe

ct D

is D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y SU

NY

Sta

te U

nive

rsity

of

New

Yor

k at

Sto

ny B

rook

on

10/2

7/14

For

pers

onal

use

onl

y.