- Email: [email protected]
Overwhelming postsplenectomy sepsis in childhood
Overwhelming PostsplenectomySepsis in Childhood David L. Posey, MD, New Orleans, Louisiana Charles Marks, MD, PhD, FRCS, New Orleans, Louisiana
In 1952 King and Schumacher [I] emphasized the risk of overwhelming sepsis in children who had undergone splenectomy for hematologic disorders. Since that time there has been a preponderance of publications that have reviewed the incidence of postsplenectomy sepsis. Horan and Colebatch [2] reviewed the incidence of infections, such as osteomyelitis, mastoiditis, and subphrenic abscess, in splenectomized patients as evidence of increased susceptibility to serious infections; they reported a rather high incidence (up to 12 percent). Singer [3] established stricter criteria in an attempt to establish the reality of postsplenectomy sepsis. He described a syndrome of overwhelming sepsis characterized by extreme rapidity of onset, a rapid course, and death within a few hours of onset. The incidence of fulminating and catastrophic postsplenectomy infection reported in his series is naturally lower than that of Horan and Colebatch, with an average of 4.25 percent [3]. In addition, there appeared to be a predilection for pulmonary infection with fulminating pneumonia or after meningitis, with adrenal hemorrhage representing the WaterhouseFriderichsen syndrome. The present study endorses Singer’s criteria and reviews the incidence of overwhelming sepsis, such as occurs with meningitis or life-threatening pneumonia in children who have undergone splenectomy. An attempt has been made to relate the indication for splenectomy with postsplenectomy susceptibility to the syndrome of overwhelming sepsis.
Material and Methods During the decade from 1969 through 1979, 393 splenectomies were performed within the Charity Hospital system. Fifty-six of the 393 operations were performed in children under 16 years of age and they form the basis of From the Department of Surgery, Louisiana State University Medical Center, New Orleans, Louisiana. Request for reprints should be addressed to Charles Marks, MD, PhD, Department of Surgery, Louisiana State University Medical Center, 1542 Tulane Avenue, New Orleans, Louisiana 70112.
318
this review of the risk of infection in children who have undergone splenectomy. Also included are guidelines regarding the role of splenectomy in both emergency and elective conditions. Chart review and personal contact provided information on each child’s age at the time of splenectomy, the indications for splenectomy, and the incidence of serious infection in either the immediate postoperative or follow-up period since the patients received routine health care within the hospital system over the ensuing years. The definition of serious infection was reserved for such diagnoses as pneumonia, meningitis, and any other severe infective illness that required hospitalization and was perceived to be life-threatening. Follow-up data was provided by chart review and telephone contact with the patient or the next of kin. The ages of the children in this series ranged from 16 months to 15 years, with a mean of 9.27 years. The followup period was from 3 to 12 years, with an average of 5.1 years. The indications for splenectomy (Table I) included traumatic rupture in 27 patients, hematologic disorders in 19 patients, and a staging procedure for Hodgkin’s disease in 7 patients; splenectomy was incidental to other procedures in 3 patients. Spherocytosis was the most common hematologic indication for elective splenectomy. Sickle cell disease with life-threatening hemolytic crises and anemia secondary to splenic sequestration provided the indication for splenectomy in six patients, whereas idiopathic thrombocytopenic purpura and hypersplenism were the indications in four patients (two patients each). Autoimmune hemolytic anemia provided the indication for splenectomy in one patient. One patient, a 9 year old child with severe head trauma and splenic rupture who died 3 days after splenectomy from brain death without evidence of sepsis, was obviously excluded from further consideration. The remaining 55 patients have all been carefully reviewed and included in this series.
Results Eight episodes of serious infection were documented in 4 of 55 children who underwent splenectomy. A 6 year old black boy who had undergone the operation for hypersplenism had previously suffered three documented bouts of pneumococcal pneumoThe American Journal of Surgery
Postsplenectomy
TABLE I
Sepsis in Childhood
Incidence of PostsplenectomySepsis In Childhood
Patients (n)
Diagnosis Trauma Hodgkin’s disease Spherocytosis Sickle cell anemia Incidental Hypersplenism ITP Autoimmune hemolytic ITP = idiopathic
26 7 6 6 3 2 2 anemia
thrombocytopenic
1
Average Age at Splenectcmy (yr) 9.6 12 11 7.6 4.11 4 6.5 5
1 2 0 0 0 1 0 0
% 3.9 26.6 0 0 0 50 0 0
0 14 0 0 0 0 0 0
Sepsis
Average Age Interval Between Splenectomy and Sepsis 9mo 2.2 yr
. . l.S;r
. ...
purpura.
nia. Each of these episodes responded rapidly to antibiotic therapy. Within 12 months after undergoing splenectomy this patient suffered a fourth episode of severe pneumococcal pneumonia complicated by pneumococcal meningitis; it fortunately resolved rapidly with antibiotic therapy. He was readmitted to the hospital at 8 years of age with a second episode of pneumococcal meningitis which again responded satisfactorily to antibiotic therapy. A 13 year old black boy underwent a staging splenectomy for stage IVB Hodgkin’s disease complemented with appropriate chemotherapy. Four months after splenectomy a virulent pneumococcal meningitis developed which responded to antibiotic therapy and respiratory support. Five years later the patient was readmitted to the hospital with all the clinical features of meningitis associated with a cutaneous herpetic infection on the left side of the face. Central spinal fluid analysis was negative for organisms and no bacteria were grown on a culture specimen. The patient’s condition responded rapidly to therapy with ampicillin. A 10 year old black boy who underwent splenectomy for stage IIIB Hodgkin’s disease subsequently required chemotherapy and irradiation for control of his disease. He presented with cryptococcal meningitis 3.5 years after splenectomy. Two months later he presented with proved enterococcal sepsis as well as disseminated herpes zoster. He recovered from these episodes but died 5 years after splenectomy from uncontrollable Proteus mirabilis sepsis. A 7 year old girl recovered satisfactorily after undergoing splenectomy subsequent to a gunshot wound to the abdomen. Nine months later she was readmitted to the hospital with clinical features suggestive of meningitis and headache and fever of 105.8’F. Lumbar puncture at this time revealed 22 red blood cells and 16 white blood cells per mm3 represented predominantly by lymphocytes, but no organisms were found. She responded rapidly to intravenously administered penicillin and chloramphenicol without any subsequent neurologic sequelae. In retrospect this patient was considered to have Voknno 145. March 1933
Patients-With n
Mortality Secondary to Sepsis (%I
suffered an episode of aseptic meningitis; subsequent follow-up demonstrated no other infective episodes. In this series of 55 patients, one or more episodes of significant postsplenectomy infection developed in 4 children (7.3 percent). The infective complications occurred from 8 months to 6 years after splenectomy, (average lag time 3.7 years). Specific organisms encountered in this series include pneumococcus, herpes zoster, enterococcus, cryptococcus, unclassified virus, and Proteus mirabilis. On only one occasion in one patient was the typical pattern of overwhelming postsplenectomy infection as described by Krivit et al [4] encountered. The patient’s parents noticed that he was slightly lethargic, and within 24 hours he had lapsed into a confusional stage with tachypnea, high fever, and finally, death 7 hours after admission to the hospital despite vigorous antibiotic therapy. The mortality rate from overwhelming postsplenectomy sepsis in this series of 55 children is thus 1.8 percent; however, it should be emphasized that the development of serious postsplenectomy sepsis in our series resulted in a 25 percent mortality rate. Comments The new surgical orthodoxy that stresses splenic conservation at all costs and under all circumstances cannot, in the light of our experience, be permitted to go unchallenged. We believe that the technical feasibility of splenectomy is certainly not an indication for its performance, and the fortuitous “bagging” of a spleen during surgical maneuvers in the left hypochondrium needs to be roundly condemned. No one questions the functional importance of this trabecular organ, in which destruction, formation, and storage of blood cells takes place, and its importance in providing immunologic competence and resistance to infection is generally endorsed. It is clear that the body can function normally without the spleen under most circumstances because the liver, bone marrow, and lymphatic system can usually compensate by protecting against infection. Contemporary infor319
Posey and Marks
mation regarding the postsplenectomy syndrome does, however, impose an obligation to conserve the spleen whenever possible, but to remove it if life is threatened by exsanguinating hemorrhage. Splenectomy remains a technically simple operative procedure, but its role and its long-term risks impose a challenge to surgical judgement. Postsplenectomy infection is generally seen most often in those patients who are already predisposed to infection by their underlying disease [3,5-71. King and Schumacker [I] in 1952 startled the surgical community with their paper that documented the risks of fulminant bacterial infection in infants within a year of undergoing splenectomy. This report described five children who died from overwhelming sepsis subsequent to undergoing splenectomy for spherocytosis [3]. Although their report emphasized that the risk of infection and overwhelming sepsis after splenectomy was confined to young children, it has become apparent that this complication may effect older children and occasionally adults as well. The implicated pathogens are generally encapsulated bacteria: pneumococci, Hemophilus influenzae, and meningococci. The danger of postsplenectomy infection is greatest when the patient has an underlying immunodeficient state, such as the Wiscott-Aldrich syndrome, characterized by thrombocytopenia and dual system immunodeficiency. At one end of the spectrum is the risk factor in patients who undergo splenectomy for traumatic rupture; intermediate in risk are those patients undergoing the staged splenectomy for assessment of Hodgkin’s disease [8], but equally at risk are those children with sickle cell disease who suffer functional hyposplenia. Those individuals who have undergone splenectomy for trauma and hematologic conditions such as idiopathic thrombocytopenic purpura or spherocytosis generally have a much lower incidence of postsplenectomy sepsis than do those individuals who are immunologically compromised [7]. In this regard, Laski and McMillan [9] compared the incidence of postoperative sepsis in two groups of patients who were of similar ages. One group consisted of those who underwent appendectomy and the other group those who underwent splenectomy. They found no difference in the incidence of postoperative sepsis between these two categories. A further observation worthy of merit is the study by Horan and Colebatch [2] who demonstrated that of 142 patients who underwent splenectomy, 17 had had significant presplenectomy infective episodes, for an incidence of sepsis equivalent to that encountered in the postsplenectomy group. The mortality rate, however, was unquestionably higher in the postsplenectomy group PI. There is little doubt that the injured spleen can be conserved at surgical exploration and, indeed, many such patients do not require laparotomy. The patient with a ruptured spleen who is stable and does not
320
continue to bleed may well be observed; noninvasive investigative procedures validate the nonoperative approach. If laparotomy has to be carried out because of clinical indications, local hemostasis may be obtained with suturing or with topical hemostatic agents such as Avitenea. Partial splenectomy provides an alternative technique. It would be an expression of folly to attempt to conserve the spleen, however, if there are associated contaminating injuries which involve the small intestine or colon, if blood replacement be unavailable, or if excessive blood loss threatens the patient’s clinical status. There are elective surgical indications for splenectomy in congenital spherocytosis, idiopathic thrombocytopenic purpura, and associated hematologic conditions characterized by hemolytic or aplastic crises. In children it is wise to defer operation for as long as possible, particularly if the acute episodes are self-limiting or mild. Such watchful waiting may, of course, result in spontaneous remission in some of the milder forms of these conditions and in this way the definitive splenectomy may be deferred until increased immunologic capacity has developed. If splenectomy has proved necessary, either to prevent exsanguinating hemorrhage from trauma or as a definitive indication for elective operation, a prophylactic antibiotic regimen throughout the course of childhood has been recommended [2,10], complemented by administration of the polyvalent pneumococcal serum Pneumovaxe to reduce the risk of overwhelming sepsis [11]. This regimen is generally recommended for all children as well as for young adults; however, it does introduce the patient to the psychologic spectre of a fatal infection that may never occur. For this reason and in light of the statistical data in our series, an alternative strategy appears reasonable to us. The patient or parents are advised that in the event illness or fever develops after splenectomy, the patient should be brought to the hospital for clinical evaluation complemented by blood culture and administration of both an immediate loading dose of penicillin, or another similar antibiotic, and polyvalent serum. Such an individualized approach provides a reasonable alternative to the administration of sera and antibiotics in a “shotgun” method to all patients who have undergone splenectomy. Another dilemma has been posed regarding the role of splenectomy in staging Hodgkin’s disease. Before recommending the procedure a balance needs to be struck between the benefits that such information provides and the risk of fulminant sepsis in the asplenic individual. The role of subtotal splenectomy or hemisplenectomy merits consideration under such circumstances [8]. Splenectomy may be responsible for alterations in host defense because it can result in a low serum immunoglobulin M level, with possible failure to
The American Journal of Surgery
Postsplenectomy
switch from immunoglobulin M to immunglobulin G synthesis [12]. Impaired opsonization may be the result of deficient antibody response and diminished alternative complement pathway activity [13,14], and may cause a reduction in circulating T cells. It has also been demonstrated that the human spleen has a role in activating a peptide that promotes phagocytosis [15] and absence of this substance called tuftsin may compromise some asplenic hosts. These various factors may need to be studied preoperatively in patients scheduled for elective splenectomy and may provide guidelines regarding the proportion of individuals who would become vulnerable to overwhelming sepsis after the operation. Summary Three hundred ninety-three splenectomies were performed within the Charity Hospital system during the decade from 1969 through 1979. This number included 56 operations in children under 16 years of age, which formed the basis of this report of the risk of infection in young splenectomized patients and provided guidelines for the role of splenectomy under emergency and elective conditions. Eight episodes of serious infection were documented in four patients. There was only one case of overwhelming postsplenectomy infection which resulted in death, for a mortality rate of 1.8 percent (1 of 56 children) for death due to overwhelming postsplenectomy sepsis. It is emphasized, however, that the development of serious postsplenectomy sepsis in our series resulted in a 25 percent mortality rate. The risk of postsplenectomy sepsis is much greater in those individuals who are immunologically compromised, such as those who undergo staging splenectomy for Hodgkin’s disease. If splenectomy is indicated for a hematologic disorder, it is wise to defer operation for as long as possible, especially if the acute episodes are self-limiting or mild. Rather than the promiscuous use of polyvalent serum and antibiotic therapy after splenectomy in children, it is recommended that parents be advised to bring the child to the hospital anytime an illness or fever develops that might re-
Volume 145, March 1993
Sepsis in Childhood
quire an immediate loading dose of an appropriate antibiotic. The role of subtotal splenectomy or hemisplenectomy merits consideration in staging Hodgkin’s disease. Preoperative study of certain immunologic parameters may provide guidelines as to the proportion of individuals who may be vulnerable to overwhelming sepsis after splenectomy. References 1. King H, Shumacker HB. Splenic studies: I. Susceptability to infection after splenectomy performed in infancy. Ann Surg 1952;136:239-42. 2. Horan M, Colebatch JH. Relation between splenectomy and subsequent infection. Arch Dis Child 1962;37:398-414. 3. Singer DB. Postsplenectomy sepsis. In: Rosenberg HS, Bolande RP. eds. Perspectives on pediatric pathology. Chicago: Year Book Medical, 1973;285-3 11. 4. Krivit W, Giebink GS, Leonard A. Overwhelming postsplenectomy infection. Surg Clin North Am 1977;59:223-33. 5. Eraklis AJ, Kevy SV, Diamond LK, Gross RE. Hazard of overwhelming infection after splenectomy in childhood. N Engl J Med 1967;276:1225-9. 6. Francke EL, Neu HC. Postsplenectomy infection. Surg Clin North Am 1981;61:135-55. 7. Robinson TW, Sturgeon P. Postsplenectomy infection in infants and children. Pediatrics 1960;25:941-51. 8. Dearth JC, Gilchrist GS, Telander RL, O’Connell MI, Weiland LH. Risk of false-negative results. N Engl J Med 1978;299: 345-6. 9. Laski B, MacMillan A. Incidence of infection in children after splenectomy. Pediatrics 1959;24:523-7. 10. Smith CH, Erlandson ME, Stern G, Hilgartner MW. Postsplenectomy infection in Cooley’s anemia: an appraisal of the problem of this and other blood disorders, with a consideration of prophylaxis. N Engl J Med 1962;266:737-43. 11. Amman AJ, Addiego J, Wara DW, Lubin B, Smith WB, Mentzer WC. Pofyvalent pneumococca l-polysaccharii immunization of patients with sickle-cell anemia and patients with splenectomy. N Engl J Med 1977;297:897-900. 12. Schumaker MJ. Serum immunoglobulin and transferrin levels after childhood splenectomy. Arch Dis Child 1970;45: 114-7. 13. Johnston RB, Newman SL, Struth AG. An abnormality of the alternative pathway of complement activation in sickle-cell disease. N Engl J Med 1973;288:803-8. 14. Saslaw S. Bouroncle BA, Well RL. Doan CA. Studies on the antibody response in splenectomized persons. N Engl J Med 1959;261:120-5. 15. Spirer Z, Zakuth V, Diamant S, et al. Decreased tuftsin concentrations in patients who have undergone splenectomy. Br J Med 1977;2: 1574-6.
321
In 1952 King and Schumacher [I] emphasized the risk of overwhelming sepsis in children who had undergone splenectomy for hematologic disorders. Since that time there has been a preponderance of publications that have reviewed the incidence of postsplenectomy sepsis. Horan and Colebatch [2] reviewed the incidence of infections, such as osteomyelitis, mastoiditis, and subphrenic abscess, in splenectomized patients as evidence of increased susceptibility to serious infections; they reported a rather high incidence (up to 12 percent). Singer [3] established stricter criteria in an attempt to establish the reality of postsplenectomy sepsis. He described a syndrome of overwhelming sepsis characterized by extreme rapidity of onset, a rapid course, and death within a few hours of onset. The incidence of fulminating and catastrophic postsplenectomy infection reported in his series is naturally lower than that of Horan and Colebatch, with an average of 4.25 percent [3]. In addition, there appeared to be a predilection for pulmonary infection with fulminating pneumonia or after meningitis, with adrenal hemorrhage representing the WaterhouseFriderichsen syndrome. The present study endorses Singer’s criteria and reviews the incidence of overwhelming sepsis, such as occurs with meningitis or life-threatening pneumonia in children who have undergone splenectomy. An attempt has been made to relate the indication for splenectomy with postsplenectomy susceptibility to the syndrome of overwhelming sepsis.
Material and Methods During the decade from 1969 through 1979, 393 splenectomies were performed within the Charity Hospital system. Fifty-six of the 393 operations were performed in children under 16 years of age and they form the basis of From the Department of Surgery, Louisiana State University Medical Center, New Orleans, Louisiana. Request for reprints should be addressed to Charles Marks, MD, PhD, Department of Surgery, Louisiana State University Medical Center, 1542 Tulane Avenue, New Orleans, Louisiana 70112.
318
this review of the risk of infection in children who have undergone splenectomy. Also included are guidelines regarding the role of splenectomy in both emergency and elective conditions. Chart review and personal contact provided information on each child’s age at the time of splenectomy, the indications for splenectomy, and the incidence of serious infection in either the immediate postoperative or follow-up period since the patients received routine health care within the hospital system over the ensuing years. The definition of serious infection was reserved for such diagnoses as pneumonia, meningitis, and any other severe infective illness that required hospitalization and was perceived to be life-threatening. Follow-up data was provided by chart review and telephone contact with the patient or the next of kin. The ages of the children in this series ranged from 16 months to 15 years, with a mean of 9.27 years. The followup period was from 3 to 12 years, with an average of 5.1 years. The indications for splenectomy (Table I) included traumatic rupture in 27 patients, hematologic disorders in 19 patients, and a staging procedure for Hodgkin’s disease in 7 patients; splenectomy was incidental to other procedures in 3 patients. Spherocytosis was the most common hematologic indication for elective splenectomy. Sickle cell disease with life-threatening hemolytic crises and anemia secondary to splenic sequestration provided the indication for splenectomy in six patients, whereas idiopathic thrombocytopenic purpura and hypersplenism were the indications in four patients (two patients each). Autoimmune hemolytic anemia provided the indication for splenectomy in one patient. One patient, a 9 year old child with severe head trauma and splenic rupture who died 3 days after splenectomy from brain death without evidence of sepsis, was obviously excluded from further consideration. The remaining 55 patients have all been carefully reviewed and included in this series.
Results Eight episodes of serious infection were documented in 4 of 55 children who underwent splenectomy. A 6 year old black boy who had undergone the operation for hypersplenism had previously suffered three documented bouts of pneumococcal pneumoThe American Journal of Surgery
Postsplenectomy
TABLE I
Sepsis in Childhood
Incidence of PostsplenectomySepsis In Childhood
Patients (n)
Diagnosis Trauma Hodgkin’s disease Spherocytosis Sickle cell anemia Incidental Hypersplenism ITP Autoimmune hemolytic ITP = idiopathic
26 7 6 6 3 2 2 anemia
thrombocytopenic
1
Average Age at Splenectcmy (yr) 9.6 12 11 7.6 4.11 4 6.5 5
1 2 0 0 0 1 0 0
% 3.9 26.6 0 0 0 50 0 0
0 14 0 0 0 0 0 0
Sepsis
Average Age Interval Between Splenectomy and Sepsis 9mo 2.2 yr
. . l.S;r
. ...
purpura.
nia. Each of these episodes responded rapidly to antibiotic therapy. Within 12 months after undergoing splenectomy this patient suffered a fourth episode of severe pneumococcal pneumonia complicated by pneumococcal meningitis; it fortunately resolved rapidly with antibiotic therapy. He was readmitted to the hospital at 8 years of age with a second episode of pneumococcal meningitis which again responded satisfactorily to antibiotic therapy. A 13 year old black boy underwent a staging splenectomy for stage IVB Hodgkin’s disease complemented with appropriate chemotherapy. Four months after splenectomy a virulent pneumococcal meningitis developed which responded to antibiotic therapy and respiratory support. Five years later the patient was readmitted to the hospital with all the clinical features of meningitis associated with a cutaneous herpetic infection on the left side of the face. Central spinal fluid analysis was negative for organisms and no bacteria were grown on a culture specimen. The patient’s condition responded rapidly to therapy with ampicillin. A 10 year old black boy who underwent splenectomy for stage IIIB Hodgkin’s disease subsequently required chemotherapy and irradiation for control of his disease. He presented with cryptococcal meningitis 3.5 years after splenectomy. Two months later he presented with proved enterococcal sepsis as well as disseminated herpes zoster. He recovered from these episodes but died 5 years after splenectomy from uncontrollable Proteus mirabilis sepsis. A 7 year old girl recovered satisfactorily after undergoing splenectomy subsequent to a gunshot wound to the abdomen. Nine months later she was readmitted to the hospital with clinical features suggestive of meningitis and headache and fever of 105.8’F. Lumbar puncture at this time revealed 22 red blood cells and 16 white blood cells per mm3 represented predominantly by lymphocytes, but no organisms were found. She responded rapidly to intravenously administered penicillin and chloramphenicol without any subsequent neurologic sequelae. In retrospect this patient was considered to have Voknno 145. March 1933
Patients-With n
Mortality Secondary to Sepsis (%I
suffered an episode of aseptic meningitis; subsequent follow-up demonstrated no other infective episodes. In this series of 55 patients, one or more episodes of significant postsplenectomy infection developed in 4 children (7.3 percent). The infective complications occurred from 8 months to 6 years after splenectomy, (average lag time 3.7 years). Specific organisms encountered in this series include pneumococcus, herpes zoster, enterococcus, cryptococcus, unclassified virus, and Proteus mirabilis. On only one occasion in one patient was the typical pattern of overwhelming postsplenectomy infection as described by Krivit et al [4] encountered. The patient’s parents noticed that he was slightly lethargic, and within 24 hours he had lapsed into a confusional stage with tachypnea, high fever, and finally, death 7 hours after admission to the hospital despite vigorous antibiotic therapy. The mortality rate from overwhelming postsplenectomy sepsis in this series of 55 children is thus 1.8 percent; however, it should be emphasized that the development of serious postsplenectomy sepsis in our series resulted in a 25 percent mortality rate. Comments The new surgical orthodoxy that stresses splenic conservation at all costs and under all circumstances cannot, in the light of our experience, be permitted to go unchallenged. We believe that the technical feasibility of splenectomy is certainly not an indication for its performance, and the fortuitous “bagging” of a spleen during surgical maneuvers in the left hypochondrium needs to be roundly condemned. No one questions the functional importance of this trabecular organ, in which destruction, formation, and storage of blood cells takes place, and its importance in providing immunologic competence and resistance to infection is generally endorsed. It is clear that the body can function normally without the spleen under most circumstances because the liver, bone marrow, and lymphatic system can usually compensate by protecting against infection. Contemporary infor319
Posey and Marks
mation regarding the postsplenectomy syndrome does, however, impose an obligation to conserve the spleen whenever possible, but to remove it if life is threatened by exsanguinating hemorrhage. Splenectomy remains a technically simple operative procedure, but its role and its long-term risks impose a challenge to surgical judgement. Postsplenectomy infection is generally seen most often in those patients who are already predisposed to infection by their underlying disease [3,5-71. King and Schumacker [I] in 1952 startled the surgical community with their paper that documented the risks of fulminant bacterial infection in infants within a year of undergoing splenectomy. This report described five children who died from overwhelming sepsis subsequent to undergoing splenectomy for spherocytosis [3]. Although their report emphasized that the risk of infection and overwhelming sepsis after splenectomy was confined to young children, it has become apparent that this complication may effect older children and occasionally adults as well. The implicated pathogens are generally encapsulated bacteria: pneumococci, Hemophilus influenzae, and meningococci. The danger of postsplenectomy infection is greatest when the patient has an underlying immunodeficient state, such as the Wiscott-Aldrich syndrome, characterized by thrombocytopenia and dual system immunodeficiency. At one end of the spectrum is the risk factor in patients who undergo splenectomy for traumatic rupture; intermediate in risk are those patients undergoing the staged splenectomy for assessment of Hodgkin’s disease [8], but equally at risk are those children with sickle cell disease who suffer functional hyposplenia. Those individuals who have undergone splenectomy for trauma and hematologic conditions such as idiopathic thrombocytopenic purpura or spherocytosis generally have a much lower incidence of postsplenectomy sepsis than do those individuals who are immunologically compromised [7]. In this regard, Laski and McMillan [9] compared the incidence of postoperative sepsis in two groups of patients who were of similar ages. One group consisted of those who underwent appendectomy and the other group those who underwent splenectomy. They found no difference in the incidence of postoperative sepsis between these two categories. A further observation worthy of merit is the study by Horan and Colebatch [2] who demonstrated that of 142 patients who underwent splenectomy, 17 had had significant presplenectomy infective episodes, for an incidence of sepsis equivalent to that encountered in the postsplenectomy group. The mortality rate, however, was unquestionably higher in the postsplenectomy group PI. There is little doubt that the injured spleen can be conserved at surgical exploration and, indeed, many such patients do not require laparotomy. The patient with a ruptured spleen who is stable and does not
320
continue to bleed may well be observed; noninvasive investigative procedures validate the nonoperative approach. If laparotomy has to be carried out because of clinical indications, local hemostasis may be obtained with suturing or with topical hemostatic agents such as Avitenea. Partial splenectomy provides an alternative technique. It would be an expression of folly to attempt to conserve the spleen, however, if there are associated contaminating injuries which involve the small intestine or colon, if blood replacement be unavailable, or if excessive blood loss threatens the patient’s clinical status. There are elective surgical indications for splenectomy in congenital spherocytosis, idiopathic thrombocytopenic purpura, and associated hematologic conditions characterized by hemolytic or aplastic crises. In children it is wise to defer operation for as long as possible, particularly if the acute episodes are self-limiting or mild. Such watchful waiting may, of course, result in spontaneous remission in some of the milder forms of these conditions and in this way the definitive splenectomy may be deferred until increased immunologic capacity has developed. If splenectomy has proved necessary, either to prevent exsanguinating hemorrhage from trauma or as a definitive indication for elective operation, a prophylactic antibiotic regimen throughout the course of childhood has been recommended [2,10], complemented by administration of the polyvalent pneumococcal serum Pneumovaxe to reduce the risk of overwhelming sepsis [11]. This regimen is generally recommended for all children as well as for young adults; however, it does introduce the patient to the psychologic spectre of a fatal infection that may never occur. For this reason and in light of the statistical data in our series, an alternative strategy appears reasonable to us. The patient or parents are advised that in the event illness or fever develops after splenectomy, the patient should be brought to the hospital for clinical evaluation complemented by blood culture and administration of both an immediate loading dose of penicillin, or another similar antibiotic, and polyvalent serum. Such an individualized approach provides a reasonable alternative to the administration of sera and antibiotics in a “shotgun” method to all patients who have undergone splenectomy. Another dilemma has been posed regarding the role of splenectomy in staging Hodgkin’s disease. Before recommending the procedure a balance needs to be struck between the benefits that such information provides and the risk of fulminant sepsis in the asplenic individual. The role of subtotal splenectomy or hemisplenectomy merits consideration under such circumstances [8]. Splenectomy may be responsible for alterations in host defense because it can result in a low serum immunoglobulin M level, with possible failure to
The American Journal of Surgery
Postsplenectomy
switch from immunoglobulin M to immunglobulin G synthesis [12]. Impaired opsonization may be the result of deficient antibody response and diminished alternative complement pathway activity [13,14], and may cause a reduction in circulating T cells. It has also been demonstrated that the human spleen has a role in activating a peptide that promotes phagocytosis [15] and absence of this substance called tuftsin may compromise some asplenic hosts. These various factors may need to be studied preoperatively in patients scheduled for elective splenectomy and may provide guidelines regarding the proportion of individuals who would become vulnerable to overwhelming sepsis after the operation. Summary Three hundred ninety-three splenectomies were performed within the Charity Hospital system during the decade from 1969 through 1979. This number included 56 operations in children under 16 years of age, which formed the basis of this report of the risk of infection in young splenectomized patients and provided guidelines for the role of splenectomy under emergency and elective conditions. Eight episodes of serious infection were documented in four patients. There was only one case of overwhelming postsplenectomy infection which resulted in death, for a mortality rate of 1.8 percent (1 of 56 children) for death due to overwhelming postsplenectomy sepsis. It is emphasized, however, that the development of serious postsplenectomy sepsis in our series resulted in a 25 percent mortality rate. The risk of postsplenectomy sepsis is much greater in those individuals who are immunologically compromised, such as those who undergo staging splenectomy for Hodgkin’s disease. If splenectomy is indicated for a hematologic disorder, it is wise to defer operation for as long as possible, especially if the acute episodes are self-limiting or mild. Rather than the promiscuous use of polyvalent serum and antibiotic therapy after splenectomy in children, it is recommended that parents be advised to bring the child to the hospital anytime an illness or fever develops that might re-
Volume 145, March 1993
Sepsis in Childhood
quire an immediate loading dose of an appropriate antibiotic. The role of subtotal splenectomy or hemisplenectomy merits consideration in staging Hodgkin’s disease. Preoperative study of certain immunologic parameters may provide guidelines as to the proportion of individuals who may be vulnerable to overwhelming sepsis after splenectomy. References 1. King H, Shumacker HB. Splenic studies: I. Susceptability to infection after splenectomy performed in infancy. Ann Surg 1952;136:239-42. 2. Horan M, Colebatch JH. Relation between splenectomy and subsequent infection. Arch Dis Child 1962;37:398-414. 3. Singer DB. Postsplenectomy sepsis. In: Rosenberg HS, Bolande RP. eds. Perspectives on pediatric pathology. Chicago: Year Book Medical, 1973;285-3 11. 4. Krivit W, Giebink GS, Leonard A. Overwhelming postsplenectomy infection. Surg Clin North Am 1977;59:223-33. 5. Eraklis AJ, Kevy SV, Diamond LK, Gross RE. Hazard of overwhelming infection after splenectomy in childhood. N Engl J Med 1967;276:1225-9. 6. Francke EL, Neu HC. Postsplenectomy infection. Surg Clin North Am 1981;61:135-55. 7. Robinson TW, Sturgeon P. Postsplenectomy infection in infants and children. Pediatrics 1960;25:941-51. 8. Dearth JC, Gilchrist GS, Telander RL, O’Connell MI, Weiland LH. Risk of false-negative results. N Engl J Med 1978;299: 345-6. 9. Laski B, MacMillan A. Incidence of infection in children after splenectomy. Pediatrics 1959;24:523-7. 10. Smith CH, Erlandson ME, Stern G, Hilgartner MW. Postsplenectomy infection in Cooley’s anemia: an appraisal of the problem of this and other blood disorders, with a consideration of prophylaxis. N Engl J Med 1962;266:737-43. 11. Amman AJ, Addiego J, Wara DW, Lubin B, Smith WB, Mentzer WC. Pofyvalent pneumococca l-polysaccharii immunization of patients with sickle-cell anemia and patients with splenectomy. N Engl J Med 1977;297:897-900. 12. Schumaker MJ. Serum immunoglobulin and transferrin levels after childhood splenectomy. Arch Dis Child 1970;45: 114-7. 13. Johnston RB, Newman SL, Struth AG. An abnormality of the alternative pathway of complement activation in sickle-cell disease. N Engl J Med 1973;288:803-8. 14. Saslaw S. Bouroncle BA, Well RL. Doan CA. Studies on the antibody response in splenectomized persons. N Engl J Med 1959;261:120-5. 15. Spirer Z, Zakuth V, Diamant S, et al. Decreased tuftsin concentrations in patients who have undergone splenectomy. Br J Med 1977;2: 1574-6.
321