Immune Thrombocytopenia: Surgical Therapy and Predictors of Response By Peter W. Davis, David A. Williams, and Robert C. Shamberger Boston, Massachusetts 0 We have reviewed 40 patients with immune thrombocytopenia purpura (ITP) to assess current methods of preparation for surgery and to evaluate perioperative complications and response to splenectomy. Twenty-one patients had chronic ITP (> 1 year duration) and 19 patients had severe acute thrombocytopenia (platelet counts < 10,000). A progression of methods of pretreatment was seen in the lo-year period reviewed. Seventeen patients received no treatment before admission for surgery, and 10 of these received platelet transfusions. Seventeen patients received steroids immediately preceding surgery; 16 of these responded and 1 received a platelet transfusion. Recently, 5 patients received intravenous gamma globulin (IgG) preceding surgery with all patients responding and none receiving platelet transfusions. One patient received a combination of steroids and IgG with good response and did not require platelet transfusion. No major postoperative complications occurred (ie, pancreatitis. small bowel obstruction, or sepsis) except for one patient requiring a secondary exploration for an accessory spleen and recurrent thrombocytopenia. Eight patients (20%). 6 with severe ITP and 2 with chronic ITP (5 males and 3 females) developed recurrence of thrombocytopenia following surgery up to 1% years after splenectomy. These patients all required further medical therapy. Three additional patients (2 chronic and 1 severe) developed thrombocytopenia following viral illnesses, but required no further therapy. Gf the 6 surgical failures, 4 failed to respond to prior treatment with steroids, 1 to IgG, and 2 failed to respond to combination therapy, while one surgical failure responded to both steroid and combination therapy. Df the responders to splenectomy (32 patients), only 3 failed to respond to prior treatment with steroids. In conclusion, response to steroids (20 patients), IgG (3 patients), or both (6/7 patients) was almost predictive of response to splenectomy (97%). whereas failure to respond to steroids or IgG predicted a 70% failure rate from splenectomy (7/10 patients) (Fisher’s exact test, P < 601). IgG and steroids were equally effective in pretreatment, allowing less frequent use of platelet transfusion. Eighty percent of all patients achieved long-term response to splenectomy. Copyright o 1991 by W.B. Saunders Company INDEX WORDS: Immune thrombocytopenia purpura; pathic thrombocytopenia purpura; splenectomy.
I
idio-
MMUNE (or idiopathic) thrombocytopenia purpura (ITP) is a disease of uncertain etiology. Platelet destruction in this disorder is an immunologic phenomenon that is felt to be humorally mediated in many chronic patients by gamma globulin (IgG) antibody directed against a platelet-associated antigen (Fig 1). Platelet-associated IgG is present in up to 85% of children with ITP.’ Coating of the platelet surface by immunoglobulins results in increased platelet destruction by the reticuloendotheJournalof Pediatric Surgery, Vol 26, No 4 (April), 1991: pp 407-413
lial (RE) system in the bone marrow, liver, and spleen. The majority (75% to 90%) of childhood patients with acute ITP undergo spontaneous remission with or without medical therapy using agents such as corticosteroids or intravenous (IV) IgG.2-7 Resolution is generally early, with 74% to 85% recovery within 6 months of presentation.“.“.8 Splenectomy in the management of ITP is a wellaccepted therapy for the treatment of those patients who fail to obtain remission or develop serious complications. Patients with severe thrombocytopenia unresponsive to maintenance medical regimens and those who have developed chronic ITP have been candidates for splenectomy. Response rates to splenectomy of 50% to 89% have been reported.‘.‘.R-‘S We have undertaken a review of our experience at Children’s Hospital in Boston over the past 10 years to assess predictors of response to splenectomy, perioperative complications, and recent trends in the preoperative preparation of patients undergoing elective splenectomy for thrombocytopenia. MATERIALS AND METHODS We reviewed all available records between 1979 and 1989 of patients undergoing splenectomy for ITP. Forty patients were identified (20 males). Nineteen were operated on within a year of onset of symptoms (mean, 5.3 months) for severe thrombocytopenia ( < 10,000 platelets/mm3) with rapidly recurrent thrombocytopenia after completion of therapeutic courses of medication. Twentyone patients failed to respond to appropriate medical management and underwent splenectomy for chronic ITP (duration of illness over 1 year; mean, 2.25 years). One patient included in the chronic group underwent elective splenectomy at 11 months due to psychiatric complications from steroid administration. A response to medical therapy is defined as an increase of platelet counts to greater than 100,000/mm~ and to splenectomy as platelet counts greater than 150,00O/mm’, which was sustained on follow-up. The mean age at the time of splenectomy was 8 years 9 months (range, 5 months to 19 years; Fig 2). Statistical analysis using the Fisher’s exact test was performed. The diagnosis of immune thrombocytopenia was based on clinical presentation, platelet count. and bone marrow biopsy.
From the Department of Surgery and the Division of Hematology, Department of Pediatrics, The ChildrenS Hospital, and Harvurd Medical School, Boston, MA. Presented at the 21st Annual Meeting of the American Pediattic Surgical Association, Vancouver, British Columbia, May 19-22, 1990. Address reprint requests to Roben C. Shamberger, MD, Children S Hospital, 300 Longwood Ave, Boston, MA 02115. Copyright o 1991 by W.B. Saunders Company 0022-3468/91/2604-0009$03.00/O
407
408
DAVIS, WILLIAMS, AND SHAMSERGER
Platelet
Reticuloendothelial
A total of 30 patients (75%) had demonstrated a response to some type of therapy within the 6 months prior to splenectomy, achieving a platelet count of greater than 100,000 platelets/mm”. Twenty-seven patients received prednisone therapy alone (1 to 2 mg/kg/d for 7- to 1Cday courses) and 20 responded to therapy. Four patients received IV IgG therapy alone (400 mg/kg/d for 1 to 5 days) and three responded to therapy. Nine patients received combination therapies using separate as well as combined courses of prednisone and IV IgG and seven of these patients responded to IV IgG or steroid therapy alone or in combination.
RESULTS
phagocytosis
Platelet associated
IgG
Fig 1. A proposed mechanism for destruction of platelets in ITP. Platelet-associated IgG (PAlgG) is bound to platelets. IgG is then attached to the Fc receptor on the macrophage and the platelet destroyed by phagocytosis.
Symptoms included purpura 90%, petechial rash 2O%, and epistaxis 32.5%. Additional presenting symptoms included gastrointestinal bleeding 7.5%, mucosal bleeding 7.5%, and vaginal bleeding or menorrhagia 7.5%. Five patients presented with less typical features, including intracerebral hemorrhage (l), abdominal pain following blunt abdominal trauma (I), or were identified as incidental findings at the time of phlebotomy for unrelated problems (3). Platelet counts at presentation ranged from 1,000 to 38,000/mm3 (median, 8,000/mms). Bone marrow aspirates were obtained on 36 patients, demonstrating normal myeloid and erythroid lines with slightly decreased, slightly increased, or normal numbers of megakaryocytes. Some smears demonstrated slightly enlarged megakaryocytes. These represented normal or near-normal bone marrow aspirates. Nine patients (22.5%) had a viral illness preceding the onset of their symptoms. One of these children also had associated use of aspirin. Au exposure to an insecticide in the home was noted in the history of one patient. Two patients had associated illnesses (systemic lupus erythematosa and Evans syndrome-hemolytic anemia associated with ITP) and two a positive family history of bleeding or thrombocytopenia. 15
l-4
5-8
9-12
13-16
17-20
Age at Operation Fig 2.
Age at occurrence of thrombocytopenia
by year.
There were very few complications to splenectomy and only two complications to the medical management of the patients in the study. One child came to early operation for psychiatric complications from steroid therapy. High fever occurred in a second patient and was attributed to too rapid a withdrawal from previous steroid therapy. After slowing the steroid taper the patient had an uneventful postoperative course. There were no complications from IV IgG therapy. No major perioperative complications of small bowel obstruction, hemorrhage, pneumonia, wound infection, or pancreatitis occurred. One patient had a transient increase of serum amylase to 256 mU/mL on the fourth postoperative day, which returned to normal (56 mU/mL) the following day without specific treatment. No patient experienced postsplenectomy sepsis. One patient developed recurrent thrombocytopenia 8 months after primary splenectomy. A radionuclide scan showed an accessory spleen (Fig 3). She responded to removal of the accessory spleen. There was a single blood transfusion given in this series for a hematocrit of 22.6% in a patient who was otherwise stable and whose hematocrit before surgery was 27.4%. Thirty-two patients responded to splenectomy (80%). The mean follow-up was 1 year 3 months. Three patients who responded to splenectomy demonstrated transient thrombocytopenia during viral illnesses (recurrent ITP) but did not require further therapy. In the group of 8 failures to surgical therapy, thrombocytopenia recurred up to 1% years postsplenectomy. All of these patients developed significant thrombocytopenia requiring further medical management. Nineteen of 21 patients (90%) with chronic ITP responded to splenectomy whereas only 13 of 19 (68%) with severe thrombocytopenia responded (P > .05). Of the 30 patients who responded to any therapy, only 1 (3.3%) failed to respond to splenectomy, whereas only 3 patients of 10 (30%) who had failed medical management responded to splenectomy (P < .OOl). Four of 7 patients who failed to respond to steroid therapy did not respond to splenec-
409
ITP: SURGICAL THERAPY AND RESPONSE
course of prednisone and IV IgG with a preoperative platelet count of 201,000/mm3. In contrast the median platelet count on admission for 17 patients not receiving some form of preoperative treatment was 13,000/mm3 (range, 1,000 to 85,000/ mm3). Ten of these patients received perioperative platelet transfusions. DISCUSSION
Fig 3. Radionuclide scan of patient with recurrent thrombocytopenia 8 months after splenectomy. Scan demonstrates area of uptake in the left upper quadrant (arrow) consistent with accessory spleen.
tomy whereas all patients who responded to steroids responded to splenectomy (P < .002). Three of four patients who responded to IV IgG therapy responded to splenectomy. The single failure to IV IgG therapy also failed splenectomy (P > .OS). Six of 7 patients responding to combination therapy responded to splenectomy. The 2 patients who failed combination therapy also failed splenectomy (P > .05). Preoperative treatment with steroids or IV IgG was generally successful in elevating platelet counts and avoiding the need for platelet transfusions. Twentythree patients (57.5%) received preoperative treatments with steroids or IV IgG to avoid perioperative platelet transfusions or bleeding achieving a median platelet count on admission of 127,000/mm3. Only one of these patients received a perioperative platelet transfusion. Seventeen patients received steroid therapy preoperatively to elevate their platelet counts or were on a course of steroids at the time of surgery. Platelet counts in this group were greater than 100,000/mm3 except for 2 patients with platelet counts of 13,000/mm3 and 54,000/mm3, which were elevated from pretreatment platelet counts of < 10,000/mm3. The median platelet count at the time of admission for splenectomy was 127,000/mm3. Five patients received preoperative courses of IV IgG with platelet counts raised to greater than 50,000/mm3. The median platelet count at time of splenectomy was 97,000/mm3. One patient received a preoperative
The role of the spleen and splenectomy in the pathophysiology and treatment of ITP has long been accepted. Despite a successful response to splenectomy in the majority of patients (50% to 89%) there still remain a significant number of patients who fail surgical therapy. The response to splenectomy in this series was 80%, which is consistent with other studies. Clearly the ability to predict response to splenectomy would be desirable in light of the minimal but not insignificant risk of overwhelming postsplenectomy sepsis. Investigators have assessed various factors potentially influencing the response to surgery. The level of platelet-associated immunoglobulins did not predict response to splenectomy in previous studies.1”,‘6The sites of platelet sequestration and destruction have been assessed by chromium (“Cr)- and indium (“‘In)labeled platelet radionuclide scans. Najean and Ardai110~ assessed platelet sequestration by chromiumlabeled isologous platelet infusions.” Proportion of uptake in the spleen versus liver was predictive of response to splenectomy in that 90% of patients with splenic sequestration responded to splenectomy whereas only 30% responded who had primarily hepatic sequestration. Recently, Gemsheimer et al studied platelet sequestration with labeled autologous platelets.‘” Hepatic sequestration was normal (15% ? 2% of total body activity) in the 5 patients who responded to splenectomy but was increased in 2 of 3 patients who failed to respond to splenectomy. Total platelet sequestration in the liver and spleen was decreased by more than half compared with preoperative studies in the patients who responded to splenectomy but was decreased by less than 25% in those who failed, demonstrating the role of platelet destruction by the liver as the cause of failure of splenectomy. We evaluated the relationship between the response to medical management and response to splenectomy and demonstrated a clear relationship in these patients. Twenty-nine of 30 patients (97%) who had responded to medical therapy responded to splenectomy. In contrast, only 3 of 10 patients (30%) who had failed medical therapies responded to splenectomy. Researchers have previously differed regard-
410
ing the significance of response to glucocorticoid therapy as a predictor of response to surgery. In a review of 202 adult and pediatric patients undergoing splenectomy for ITP, Coon et al demonstrated that the likelihood for a positive response to splenectomy could be suggested by the degree of response to steroid treatment where 24% of splenectomy failures and 72% of splenectomy successes had had a platelet response of greater than 50,000 platelets/mm3 to steroid therapy.’ Brennan et al similarly found in an adult population that all 15 steroid responders had an excellent response to splenectomy while only 8 of 1.5 steroid failures responded to splenectomy.” In a study of 36 children undergoing splenectomy for ITP, Weinblatt found a 100% response rate to surgery among 20 steroid responders with a failure rate of 85% among children not responding to steroids.14 In contrast, studies by Wilde et al,” Block et al,” and Akwari et al” failed to find a predictive relationship between steroid responsiveness and surgical outcome. The present results support the findings of Weinblatt and others that steroid responsiveness is a reasonable predictor of surgical response in pediatric patients with a 100% response rate among steroid responders. IV IgG has been used in recent years in the treatment of ITP. It has in general been used as a means to boost platelet counts in the acute situation and preoperatively.‘2.‘9 Its mechanism of action has been postulated to be one of competitive inhibition of Fc receptor sites on the macrophages in the RE Evidence suggests that the macrosystem (Fig 4).6,20.21 phage can not distinguish between bound and free IgG. Other possibilities include an antidiotypic suppression of autoantibodies or a passive antibody-
Fig 4. Hypothesis proposed by Fehr and others of the mechanisms of IV IgG treatment of ITP. Exogenous IgG saturates the Fe receptors on the macrophage by competitive inhibition decreasing the binding and subsequent phagocytosis of IgG-coated platelets. A second proposed mechanism is the entiidiotypic binding of exogenous gammaglobulin to platelet-associated immunoglobulin (PAIgG).
DAVIS, WILLIAMS, AND SHAMBERGER
induced suppression of the humoral immune response.22-24Its utility has been limited primarily by its high cost and transient efficacy. We found a suggestive relationship between response to IV IgG and response to subsequent splenectomy, but the numbers were too small to be statistically significant. A review of a larger group of IV IgG treatment patients undergoing splenectomy may be helpful to clarify this issue. No major complications from surgery and only minimal perioperative morbidity were observed. The population of 40 is small, but the findings are consistent with the observation that elective splenectomy in ITP is a relatively safe procedure. Overwhelming postsplenectomy sepsis is a significant concern with a reported incidence of 5% to 7%.25,26No documented episodes of this entity in any of the patients reviewed was observed. In a review of 78 patients with uncomplicated ITP, Eraklis et alz5also found no incidence of postsplenectomy sepsis. With the current use of routine preoperative immunization as well as antibiotics postsplenectomy, the risk of overwhelming sepsis after splenectomy for ITP will continue to be small as has been reported in an early series of patients postsplenectomy for Hodgkin’s disease.27 In this review, one patient developed recurrent thrombocytopenia due to the presence of an accessory spleen and subsequently responded to its removal. Block et al, in a review of 67 patients, found an incidence of 19% accessory spleens at the time of primary splenectomy.” An incidence of 18% accessory spleens identified at the time of primary splenectomy is noted by Akwari et al in a review of 100 patients.” Accessory spleens were also found in 12 patients following splenectomy by radionuclide scan performed for recurrent (5) or persistent (7) thrombocytopenia. Of nine patients electing to undergo repeat splenectomy, six (66%) had successful outcomes. Wallace et al reported on three patients with recurrent thrombocytopenia that developed up to 11 years after primary splenectomy.w All responded to removal of the accessory spleen. Verheyden et al reported response to accessory splenectomy in only one of six patients.29 Search for accessory splenic tissue should be pursued at primary operation in all patients and by radionuclide scan in any patient who demonstrates recurrence of thrombocytopenia. The results of Akwari et al** suggest that all primary splenectomy failures should also be scanned. In recent years we have noted a trend in the preoperative preparation of patients undergoing splenectomy. There has been a clear effort to avoid transfusion of blood products perioperatively in recog-
411
ITP: SURGICAL THERAPY AND RESPONSE
nition of the risk of transmission of hepatitis and particularly human immunodeficiency virus (HIV) infection. Hence, treatment with steroids, IV IgG, or a combination of both has developed as a means to increase the platelet counts immediately prior to elective splenectomy for ITP. Wilde et al, in a review of 43 patients undergoing splenectomy for ITP, recommended routine steroid therapy perioperatively to “improve capillary fragility and diminish troublesome oozing.“” In addition, they recommended routine platelet transfusions. Steroid therapy alone has clearly been adequate without the need for additional platelet transfusions. IV IgG has also proven useful as a preparatory treatment for splenectomy. Akwari et al reports six patients thus treated with IV IgG with preoperative platelet counts greater than
100,000/mm3.‘* Schmidt et al also recommended IV IgG as a preoperative therapy.” The response to IV IgG is generally as efficacious as steroids and it has the added advantage of having less associated side effects.5 Most adverse responses to IV IgG are minimal, such as fever and local reactions at infusion sites. Major complications such as anaphylaxis and transmission of diseases (HIV, hepatitis) are not reported with current preparations.30.3’ Corticosteroid therapy, particularly high-dose perioperative administration, has many well known side effects of concern to the surgeon, including poor wound healing, increased risk of infection, and gastrointestinal bleeding.32’3 Although serious perioperative complications relating to steroid therapy were not encountered, we feel that IV IgG as a preoperative strategy is warranted.
REFERENCES 1. Cheung NV, Hilgartner MW, Schulman I, et al: Plateletassociated IgG in childhood idiopathic thrombocytopenic purpura. J Pediatr 102:366-370, 1983 2. Lusher JM, Zuelzer WW: Idiopathic thrombocytopenia purpura in childhood. J Pediatr 68:971-979.1966 3. Walker JH, Walker W: Idiopathic thrombocytopenic purpura in childhood. Arch Dis Child 36:649-657, 1961 4. Imbach P, Barandum S, Hirt A, et al: Intravenous immunoglobulin for idiopathic thrombocytopenic purpura (ITP) in childhood. Am J Pediatr Hematol Oncol6:171-174,1984 5. Imbach P, Berchtold W, Hirt A, et al: Intravenous immunoglobulin versus oral corticosteroids in acute immune thrombocytopenic purpura in childhood. Lancet 2:464-468, 1985 6. Bussel JB, Schulman I, Hilgartner MW, et al: Intravenous use of gammaglobulin in the treatment of chronic immune thrombocytopenic purpura as a means to defer splenectomy. J Pediatr 103:651-654, 1983 7. Newton WA, Zwelzer WW: Idiopathic thrombocytopenia purpura in childhood. N Engl J Med 245:879-885,195l 8. Benham ES, Taft LI: Idiopathic thrombocytopenic purpura in children: Results of steroid therapy and splenectomy. Aust Paediatr J 8:311-317, 1972 9. Coon WW: Splenectomy for idiopathic thrombocytopenic purpura. Surg Gynecol Obstet 164:225-229,1987 10. Block GE, Evans R, Zajtchuk R: Splenectomy for idiopathic thrombocytopenic purpura. Arch Surg 92:484-489,1966 11. Wilde RC, Ellis LD, Cooper WM: Splenectomy for chronic idiopathic thrombocytopenic purpura. Arch Surg 95344-350, 1967 12. Akwari OE, Itani KMF, Coleman RE, et al: Splenectomy for primary and recurrent immune thrombocytopenic purpura (ITP). Ann Surg 206:529-541.1987 13. Schwartz SI, Hoepp LM, Sachs S: Splenectomy for thrombocytopenia. Surgery 88:497-506, 1980 14. Weinblatt ME, Ortega JA: Steroid responsiveness: A predictor of the outcome of splenectomy in children with chronic immune thrombocytopenic purpura. Am J Dis Child 136:1064-1066, 1982 15. Cola B. Tonielli E, Sacco S, et al: Surgical treatment of chronic idiopathic thrombocytopenic purpura: Results in 107 cases. Int Surg 71:195-198, 1986 16. Gernsheimer T, Stratton J, Ballem PJ, et al: Mechanisms of response to treatment in autoimmune thrombocytopenic purpura. N Engl J Med 320:974-980,1989
17. Najean Y, Ardaillou N: The sequestration site of platelets in idiopathic thrombocytopenic purpura: Its correlation with the results of splenectomy. Br J Haematol21:153-164, 1971 18. Brennan MF, Rappeport JM, Moloney WC, et al: Correlation between response to corticosteroids and splenectomy for adult idiopathic thrombocytopenic purpura. Am J Surg 129:490-492, 1975 19. Schmidt RE, Budde U, Broschen-Zywietz C, et al: High dose gammaglobulin therapy in adults with idiopathic thrombocytopenic purpura (ITP) clinical effects. Blut 48:19-25,1984 20. Fehr J, Hofmann V, Kappeler U: Transient reversal of thrombocytopenia in idiopathic thrombocytopenic purpura by high-dose intravenous gamma globulin. N Engl J Med 306:12541258,1982 21. Templeton JG, Cocker JE, Crawford RJ, et al: Fey-receptor blocking antibodies in hyperimmune and normal pooled gammaglobulin. Lancet 1:1337. 1985 22. Stuart MJ, Kelton JG: The platelet: Quantitative and qualitative abnormalities, in Nathan DG, Oski FA (eds): Hematology of Infancy and Childhood. Philadelphia, PA, Saunders, 1987, pp 1348-1373 23. Sultan Y, Maisonneuve P, Kazatchkine MD, et al: Antiidiotypic suppression of autoantibodies to factor VIII (antihaemophilic factor) by high-dose intravenous gammaglobulin. Lancet 2:765-768,1984 24. Gadd SJ, Brogan MP, Ashman LK: Specificity of the passive antibody-induced suppression of the humoral immune response of mice to surface antigens on human cells. Immunology 54:223-231, 1985 25. Eraklis AJ, Kevy SV, Diamond LK, et al: Hazard of overwhelming infection after splenectomy in childhood. N Engl J Med 276:1225-1229,1967 26. Posey DL, Marks C: Overwhelming postsplenectomy sepsis in childhood. Am J Surg 145:318-321, 1983 27. Muraji T, Hays DM, Siegel SE, et al: Evaluation of the surgical aspects of staging laparotomy for Hodgkin’s disease in children. J Pediatr Surg 17:843-848, 1982 28. Wallace D, Fromm D, Thomas D: Accessory splenectomy for idiopathic thrombocytopenic purpura. Surgery 91:134-136, 1982 29. Verheyden CN, Beart RW, Clifton MD, et al: Accessory
412
DAVIS, WILLIAMS, AND SHAMBERGER
splenectomy in management of recurrent idiopathic thrombocytopenic purpura. Mayo Clin Proc 53:442-446,1978 30. Stiehm ER, Ashida E, Kim KS, et al: UCLA conference: Intravenous immunoglobulins as therapeutic agents. Ann Intern Med 107:367-382,1987 31. Centers for Disease Control: Safety of therapeutic immune globulin preparations with respect to transmission of human
T-lymphotropic views type III/lymphadenopathy-associated infection. MMWR 35:231-233, 1986 32. Glenn F, Grafe WR: Surgical complications steroid therapy. Ann Surg 165:1023-1034,1967
virus
of adrenal
33. Vogel HG: Tensile strength of skin wounds in rats after treatment with corticosteroids. Acta Endocrinol64:295-303,197O
Discussion R T. Schaller (Seattle, WA): ITP is characterized by antiplatelet antibodies that mediate the rapid destruction of these cells by the RE system. The spleen is involved in the pathogenesis of ITP in two respects: it is the major site of both autoantibody production and phagocytosis of opsonized platelets. In a climate of evolving pretreatment regimens over the past 10 years, the authors have convincingly demonstrated the predictive value of response to steroids and/or IgG. Twenty-nine of 30 patients who responded to medical management did well after splenectomy whereas only 3 of 10 patients who failed medical management responded to splenectomy. The reason for this difference must be the production of significant amounts of autoantibody against platelets elsewhere in the immune system. However, low serum titers of autoantibodies and the polyclonal nature of human serum make it very difficult to identify platelet target antigens with plasma antibodies. Work is now being done to isolate human monoclonal antiplatelet antibodies from patients with ITP so we can identify which platelet protein is carrying the immunogenic epitope. This may have implications for the future treatment of patients with ITP. Once a blocking antibody is developed against the antiplatelet autoantibody, surgeons may no longer be needed to treat this disease or, more realistically, an immune solution to the splenectomy treatment failures may be developed. The importance of autoantibodies against platelets must be implied by the therapeutic efficacy of removing accessory spleens that have been identified in some relapsed patients. Another factor of prognostic significance described recently has been splenic follicle size. In a recent review, patients with hyperplasia of splenic follicles greater than 500 pm were more likely to relapse or to develop additional autoimmune disorders than patients without hyperplasia of splenic follicles. In fact, if the follicle size was normal in the resected specimens, there was 100% response to splenectomy whereas one of four patients with follicle hyperplasia failed splenectomy. I congratulate the authors on the lack of complications in the series and the clear description of the progression of their
preparative regimen during the past 10 years. I have several practical questions for them. What was the long-term outcome of those patients who did not respond to splenectomy and did splenectomy make their medical management easier? How often did you identify accessory spleens at operation and do you have any special techniques for locating them? And finally, what is your current 1990 preference for preoperative preparation with steroids and/or IgG? P. W. Davis (response): As far as the specific pathology of the spleens removed, that was not looked at in the study. In terms of our recent preference for preoperative treatment, we do prefer IV IgG for its transient effect and the limitation of potential perioperative complications from steroid use. Regarding the outcome for patients who had failed splenectomy, who typically would have undergone continued medical courses with steroids, some with danozal, there is one patient who received courses of vincristine. The course of thrombocytopenia in several of these patients was somewhat attenuated. RM. Filler (Toronto, Ontario): Could you clarify one point about the 17 patients who received no treatment before surgery. Do you mean they never received treatment or just on that admission? Had they received treatment at a prior time? P. W. Davis (response): They did not receive any immediate preoperative treatment. All patients had received various types of treatment during their illness. D.M. Hays (Los Angeles, CA): Do you have any idea how many of the patients with an original diagnosis of ITP in Boston turned out to have lupus? P. W. Davis (response): I don’t know in terms of all patients with ITP. We had one patient in the series who had lupus and responded well to splenectomy. R. Soninno (Cleveland, OH): Aside from the fact that I always thought that the “I” in ITP stood for idiopathic, do you have any experience with preoperative nuclear platelet trapping studies and, if so, do you think they help in any way to predict outcome? P. W. Davis (response): In review of the literature for this paper, we did identify several reports that used
ITP: SURGICAL THERAPY AND RESPONSE
platelet trapping studies trying to identify patients who may benefit from splenectomy. Patients were found that had a predominence of either splenic trapping or liver sequestration and in these studies, the patients who had predominantly splenic sequestration seem to do better with splenectomy. However, we have not had any direct experience with these methods. A. Hailer (Baltimore, MD): I am confused by your conclusions in the sense that we continue to use steroids as the primary mode of treatment for ITP and operate on those patients who do not have a response. Your conclusions are that those who have a good response also respond to splenectomy. But what about those patients who do not respond and after they have had their splenectomy have not responded? What’s wrong with them? How are they different? Have you attempted to pick up those patients before so that we can look at another mode of treatment for them? And I guess my question is, are you really suggesting that a patient who responds to steroids still needs to have a splenectomy if they have gotten a good response to it? P. W Davis (response): The question of response to any medical therapy is not implied to be a curative response. Patients remit with or without medical therapies, the medical courses are primarily being used in terms of reversing episodes of thrombocytopenia. So the response we are looking at is in terms of a reversal of the thrombocytopenia that may be temporary. Some of the severe cases have reversal of thrombocytopenia, but after completion of those courses of medical therapy had rapid recurrence of their thrombocytopenia to very low levels and they ended up receiving splenectomy because they required repeated courses of steroids or other medical regimens. As far as those patients who have had no response to medical therapy, we have considered investigating the use of platelet radionuclide scan to try and identify those patients who may have predominantly liver sequestration as a reason for their failure, thereby trying to manage them potentially differently than splenectomy. W.L. Donnellan (Lansing, MI): I just wanted to clear up a point with mathematics. I am sure it is clear in the final paper, but you treated 23 patients preoperatively and 22 got a good response and were
413
operated on and the other 17 didn’t get any prior treatment at all. How do you reach the conclusion that response to preoperative treatment will condition what happens after the operation? P. W. Davis (response): Those numbers specifically are looking at those patients who received an immediate preoperative treatment in an effort to boost their platelet counts to lessen the likelihood of platelet transfusions. So those numbers are specifically looking at that group of 23 patients who received preoperative treatment only one required platelet transfusions, so we are looking at two sort of different categorizations of patients. The overall response rate relative to a response to any type of treatment during the course of disease is what is predictive of a response to splenectomy. J.L. Hill (Baltimore, MD): Did you look at the difference in ages with respect to the responsiveness of either medical treatment or surgical treatment to lump the children that are 1 or 2 years of age together with a teenager? In other problems, you can have quite a difference. For instance, giving ATG to a child versus an adult at a comparative dose, you can get a remarkable thrombocytosis in the child that far exceeds that in a teenager. Second, when did you use your preoperative steroid treatment? There is a test actually for bone marrow function where steroids are given, the response is assessed, and it peaks between 1 and 2 hours and then in fact the platelet count will drop below the baseline. And did you do this in a matter of hours or was it a matter of days in your preoperative management? P. W. Davk (response): Regarding your first question, that is not something that we looked at specifically. The timing of the preoperative treatments were basically within 2 weeks of planned surgery so they were not immediately preoperative. D. Tapper (Seattle, WA): Do you have an idea of the total number of patients with ITP during that lo-year period? In addition to that in the previous years, what’s happening to the incidence of the patients that we are seeing is it less? P. W. Davis (response): I’m afraid I really can’t comment on that. We didn’t look at ITPs that would include those patients with a complete response to steroids or spontaneous recovery who were never considered for splenectomy.
I
idio-
MMUNE (or idiopathic) thrombocytopenia purpura (ITP) is a disease of uncertain etiology. Platelet destruction in this disorder is an immunologic phenomenon that is felt to be humorally mediated in many chronic patients by gamma globulin (IgG) antibody directed against a platelet-associated antigen (Fig 1). Platelet-associated IgG is present in up to 85% of children with ITP.’ Coating of the platelet surface by immunoglobulins results in increased platelet destruction by the reticuloendotheJournalof Pediatric Surgery, Vol 26, No 4 (April), 1991: pp 407-413
lial (RE) system in the bone marrow, liver, and spleen. The majority (75% to 90%) of childhood patients with acute ITP undergo spontaneous remission with or without medical therapy using agents such as corticosteroids or intravenous (IV) IgG.2-7 Resolution is generally early, with 74% to 85% recovery within 6 months of presentation.“.“.8 Splenectomy in the management of ITP is a wellaccepted therapy for the treatment of those patients who fail to obtain remission or develop serious complications. Patients with severe thrombocytopenia unresponsive to maintenance medical regimens and those who have developed chronic ITP have been candidates for splenectomy. Response rates to splenectomy of 50% to 89% have been reported.‘.‘.R-‘S We have undertaken a review of our experience at Children’s Hospital in Boston over the past 10 years to assess predictors of response to splenectomy, perioperative complications, and recent trends in the preoperative preparation of patients undergoing elective splenectomy for thrombocytopenia. MATERIALS AND METHODS We reviewed all available records between 1979 and 1989 of patients undergoing splenectomy for ITP. Forty patients were identified (20 males). Nineteen were operated on within a year of onset of symptoms (mean, 5.3 months) for severe thrombocytopenia ( < 10,000 platelets/mm3) with rapidly recurrent thrombocytopenia after completion of therapeutic courses of medication. Twentyone patients failed to respond to appropriate medical management and underwent splenectomy for chronic ITP (duration of illness over 1 year; mean, 2.25 years). One patient included in the chronic group underwent elective splenectomy at 11 months due to psychiatric complications from steroid administration. A response to medical therapy is defined as an increase of platelet counts to greater than 100,000/mm~ and to splenectomy as platelet counts greater than 150,00O/mm’, which was sustained on follow-up. The mean age at the time of splenectomy was 8 years 9 months (range, 5 months to 19 years; Fig 2). Statistical analysis using the Fisher’s exact test was performed. The diagnosis of immune thrombocytopenia was based on clinical presentation, platelet count. and bone marrow biopsy.
From the Department of Surgery and the Division of Hematology, Department of Pediatrics, The ChildrenS Hospital, and Harvurd Medical School, Boston, MA. Presented at the 21st Annual Meeting of the American Pediattic Surgical Association, Vancouver, British Columbia, May 19-22, 1990. Address reprint requests to Roben C. Shamberger, MD, Children S Hospital, 300 Longwood Ave, Boston, MA 02115. Copyright o 1991 by W.B. Saunders Company 0022-3468/91/2604-0009$03.00/O
407
408
DAVIS, WILLIAMS, AND SHAMSERGER
Platelet
Reticuloendothelial
A total of 30 patients (75%) had demonstrated a response to some type of therapy within the 6 months prior to splenectomy, achieving a platelet count of greater than 100,000 platelets/mm”. Twenty-seven patients received prednisone therapy alone (1 to 2 mg/kg/d for 7- to 1Cday courses) and 20 responded to therapy. Four patients received IV IgG therapy alone (400 mg/kg/d for 1 to 5 days) and three responded to therapy. Nine patients received combination therapies using separate as well as combined courses of prednisone and IV IgG and seven of these patients responded to IV IgG or steroid therapy alone or in combination.
RESULTS
phagocytosis
Platelet associated
IgG
Fig 1. A proposed mechanism for destruction of platelets in ITP. Platelet-associated IgG (PAlgG) is bound to platelets. IgG is then attached to the Fc receptor on the macrophage and the platelet destroyed by phagocytosis.
Symptoms included purpura 90%, petechial rash 2O%, and epistaxis 32.5%. Additional presenting symptoms included gastrointestinal bleeding 7.5%, mucosal bleeding 7.5%, and vaginal bleeding or menorrhagia 7.5%. Five patients presented with less typical features, including intracerebral hemorrhage (l), abdominal pain following blunt abdominal trauma (I), or were identified as incidental findings at the time of phlebotomy for unrelated problems (3). Platelet counts at presentation ranged from 1,000 to 38,000/mm3 (median, 8,000/mms). Bone marrow aspirates were obtained on 36 patients, demonstrating normal myeloid and erythroid lines with slightly decreased, slightly increased, or normal numbers of megakaryocytes. Some smears demonstrated slightly enlarged megakaryocytes. These represented normal or near-normal bone marrow aspirates. Nine patients (22.5%) had a viral illness preceding the onset of their symptoms. One of these children also had associated use of aspirin. Au exposure to an insecticide in the home was noted in the history of one patient. Two patients had associated illnesses (systemic lupus erythematosa and Evans syndrome-hemolytic anemia associated with ITP) and two a positive family history of bleeding or thrombocytopenia. 15
l-4
5-8
9-12
13-16
17-20
Age at Operation Fig 2.
Age at occurrence of thrombocytopenia
by year.
There were very few complications to splenectomy and only two complications to the medical management of the patients in the study. One child came to early operation for psychiatric complications from steroid therapy. High fever occurred in a second patient and was attributed to too rapid a withdrawal from previous steroid therapy. After slowing the steroid taper the patient had an uneventful postoperative course. There were no complications from IV IgG therapy. No major perioperative complications of small bowel obstruction, hemorrhage, pneumonia, wound infection, or pancreatitis occurred. One patient had a transient increase of serum amylase to 256 mU/mL on the fourth postoperative day, which returned to normal (56 mU/mL) the following day without specific treatment. No patient experienced postsplenectomy sepsis. One patient developed recurrent thrombocytopenia 8 months after primary splenectomy. A radionuclide scan showed an accessory spleen (Fig 3). She responded to removal of the accessory spleen. There was a single blood transfusion given in this series for a hematocrit of 22.6% in a patient who was otherwise stable and whose hematocrit before surgery was 27.4%. Thirty-two patients responded to splenectomy (80%). The mean follow-up was 1 year 3 months. Three patients who responded to splenectomy demonstrated transient thrombocytopenia during viral illnesses (recurrent ITP) but did not require further therapy. In the group of 8 failures to surgical therapy, thrombocytopenia recurred up to 1% years postsplenectomy. All of these patients developed significant thrombocytopenia requiring further medical management. Nineteen of 21 patients (90%) with chronic ITP responded to splenectomy whereas only 13 of 19 (68%) with severe thrombocytopenia responded (P > .05). Of the 30 patients who responded to any therapy, only 1 (3.3%) failed to respond to splenectomy, whereas only 3 patients of 10 (30%) who had failed medical management responded to splenectomy (P < .OOl). Four of 7 patients who failed to respond to steroid therapy did not respond to splenec-
409
ITP: SURGICAL THERAPY AND RESPONSE
course of prednisone and IV IgG with a preoperative platelet count of 201,000/mm3. In contrast the median platelet count on admission for 17 patients not receiving some form of preoperative treatment was 13,000/mm3 (range, 1,000 to 85,000/ mm3). Ten of these patients received perioperative platelet transfusions. DISCUSSION
Fig 3. Radionuclide scan of patient with recurrent thrombocytopenia 8 months after splenectomy. Scan demonstrates area of uptake in the left upper quadrant (arrow) consistent with accessory spleen.
tomy whereas all patients who responded to steroids responded to splenectomy (P < .002). Three of four patients who responded to IV IgG therapy responded to splenectomy. The single failure to IV IgG therapy also failed splenectomy (P > .OS). Six of 7 patients responding to combination therapy responded to splenectomy. The 2 patients who failed combination therapy also failed splenectomy (P > .05). Preoperative treatment with steroids or IV IgG was generally successful in elevating platelet counts and avoiding the need for platelet transfusions. Twentythree patients (57.5%) received preoperative treatments with steroids or IV IgG to avoid perioperative platelet transfusions or bleeding achieving a median platelet count on admission of 127,000/mm3. Only one of these patients received a perioperative platelet transfusion. Seventeen patients received steroid therapy preoperatively to elevate their platelet counts or were on a course of steroids at the time of surgery. Platelet counts in this group were greater than 100,000/mm3 except for 2 patients with platelet counts of 13,000/mm3 and 54,000/mm3, which were elevated from pretreatment platelet counts of < 10,000/mm3. The median platelet count at the time of admission for splenectomy was 127,000/mm3. Five patients received preoperative courses of IV IgG with platelet counts raised to greater than 50,000/mm3. The median platelet count at time of splenectomy was 97,000/mm3. One patient received a preoperative
The role of the spleen and splenectomy in the pathophysiology and treatment of ITP has long been accepted. Despite a successful response to splenectomy in the majority of patients (50% to 89%) there still remain a significant number of patients who fail surgical therapy. The response to splenectomy in this series was 80%, which is consistent with other studies. Clearly the ability to predict response to splenectomy would be desirable in light of the minimal but not insignificant risk of overwhelming postsplenectomy sepsis. Investigators have assessed various factors potentially influencing the response to surgery. The level of platelet-associated immunoglobulins did not predict response to splenectomy in previous studies.1”,‘6The sites of platelet sequestration and destruction have been assessed by chromium (“Cr)- and indium (“‘In)labeled platelet radionuclide scans. Najean and Ardai110~ assessed platelet sequestration by chromiumlabeled isologous platelet infusions.” Proportion of uptake in the spleen versus liver was predictive of response to splenectomy in that 90% of patients with splenic sequestration responded to splenectomy whereas only 30% responded who had primarily hepatic sequestration. Recently, Gemsheimer et al studied platelet sequestration with labeled autologous platelets.‘” Hepatic sequestration was normal (15% ? 2% of total body activity) in the 5 patients who responded to splenectomy but was increased in 2 of 3 patients who failed to respond to splenectomy. Total platelet sequestration in the liver and spleen was decreased by more than half compared with preoperative studies in the patients who responded to splenectomy but was decreased by less than 25% in those who failed, demonstrating the role of platelet destruction by the liver as the cause of failure of splenectomy. We evaluated the relationship between the response to medical management and response to splenectomy and demonstrated a clear relationship in these patients. Twenty-nine of 30 patients (97%) who had responded to medical therapy responded to splenectomy. In contrast, only 3 of 10 patients (30%) who had failed medical therapies responded to splenectomy. Researchers have previously differed regard-
410
ing the significance of response to glucocorticoid therapy as a predictor of response to surgery. In a review of 202 adult and pediatric patients undergoing splenectomy for ITP, Coon et al demonstrated that the likelihood for a positive response to splenectomy could be suggested by the degree of response to steroid treatment where 24% of splenectomy failures and 72% of splenectomy successes had had a platelet response of greater than 50,000 platelets/mm3 to steroid therapy.’ Brennan et al similarly found in an adult population that all 15 steroid responders had an excellent response to splenectomy while only 8 of 1.5 steroid failures responded to splenectomy.” In a study of 36 children undergoing splenectomy for ITP, Weinblatt found a 100% response rate to surgery among 20 steroid responders with a failure rate of 85% among children not responding to steroids.14 In contrast, studies by Wilde et al,” Block et al,” and Akwari et al” failed to find a predictive relationship between steroid responsiveness and surgical outcome. The present results support the findings of Weinblatt and others that steroid responsiveness is a reasonable predictor of surgical response in pediatric patients with a 100% response rate among steroid responders. IV IgG has been used in recent years in the treatment of ITP. It has in general been used as a means to boost platelet counts in the acute situation and preoperatively.‘2.‘9 Its mechanism of action has been postulated to be one of competitive inhibition of Fc receptor sites on the macrophages in the RE Evidence suggests that the macrosystem (Fig 4).6,20.21 phage can not distinguish between bound and free IgG. Other possibilities include an antidiotypic suppression of autoantibodies or a passive antibody-
Fig 4. Hypothesis proposed by Fehr and others of the mechanisms of IV IgG treatment of ITP. Exogenous IgG saturates the Fe receptors on the macrophage by competitive inhibition decreasing the binding and subsequent phagocytosis of IgG-coated platelets. A second proposed mechanism is the entiidiotypic binding of exogenous gammaglobulin to platelet-associated immunoglobulin (PAIgG).
DAVIS, WILLIAMS, AND SHAMBERGER
induced suppression of the humoral immune response.22-24Its utility has been limited primarily by its high cost and transient efficacy. We found a suggestive relationship between response to IV IgG and response to subsequent splenectomy, but the numbers were too small to be statistically significant. A review of a larger group of IV IgG treatment patients undergoing splenectomy may be helpful to clarify this issue. No major complications from surgery and only minimal perioperative morbidity were observed. The population of 40 is small, but the findings are consistent with the observation that elective splenectomy in ITP is a relatively safe procedure. Overwhelming postsplenectomy sepsis is a significant concern with a reported incidence of 5% to 7%.25,26No documented episodes of this entity in any of the patients reviewed was observed. In a review of 78 patients with uncomplicated ITP, Eraklis et alz5also found no incidence of postsplenectomy sepsis. With the current use of routine preoperative immunization as well as antibiotics postsplenectomy, the risk of overwhelming sepsis after splenectomy for ITP will continue to be small as has been reported in an early series of patients postsplenectomy for Hodgkin’s disease.27 In this review, one patient developed recurrent thrombocytopenia due to the presence of an accessory spleen and subsequently responded to its removal. Block et al, in a review of 67 patients, found an incidence of 19% accessory spleens at the time of primary splenectomy.” An incidence of 18% accessory spleens identified at the time of primary splenectomy is noted by Akwari et al in a review of 100 patients.” Accessory spleens were also found in 12 patients following splenectomy by radionuclide scan performed for recurrent (5) or persistent (7) thrombocytopenia. Of nine patients electing to undergo repeat splenectomy, six (66%) had successful outcomes. Wallace et al reported on three patients with recurrent thrombocytopenia that developed up to 11 years after primary splenectomy.w All responded to removal of the accessory spleen. Verheyden et al reported response to accessory splenectomy in only one of six patients.29 Search for accessory splenic tissue should be pursued at primary operation in all patients and by radionuclide scan in any patient who demonstrates recurrence of thrombocytopenia. The results of Akwari et al** suggest that all primary splenectomy failures should also be scanned. In recent years we have noted a trend in the preoperative preparation of patients undergoing splenectomy. There has been a clear effort to avoid transfusion of blood products perioperatively in recog-
411
ITP: SURGICAL THERAPY AND RESPONSE
nition of the risk of transmission of hepatitis and particularly human immunodeficiency virus (HIV) infection. Hence, treatment with steroids, IV IgG, or a combination of both has developed as a means to increase the platelet counts immediately prior to elective splenectomy for ITP. Wilde et al, in a review of 43 patients undergoing splenectomy for ITP, recommended routine steroid therapy perioperatively to “improve capillary fragility and diminish troublesome oozing.“” In addition, they recommended routine platelet transfusions. Steroid therapy alone has clearly been adequate without the need for additional platelet transfusions. IV IgG has also proven useful as a preparatory treatment for splenectomy. Akwari et al reports six patients thus treated with IV IgG with preoperative platelet counts greater than
100,000/mm3.‘* Schmidt et al also recommended IV IgG as a preoperative therapy.” The response to IV IgG is generally as efficacious as steroids and it has the added advantage of having less associated side effects.5 Most adverse responses to IV IgG are minimal, such as fever and local reactions at infusion sites. Major complications such as anaphylaxis and transmission of diseases (HIV, hepatitis) are not reported with current preparations.30.3’ Corticosteroid therapy, particularly high-dose perioperative administration, has many well known side effects of concern to the surgeon, including poor wound healing, increased risk of infection, and gastrointestinal bleeding.32’3 Although serious perioperative complications relating to steroid therapy were not encountered, we feel that IV IgG as a preoperative strategy is warranted.
REFERENCES 1. Cheung NV, Hilgartner MW, Schulman I, et al: Plateletassociated IgG in childhood idiopathic thrombocytopenic purpura. J Pediatr 102:366-370, 1983 2. Lusher JM, Zuelzer WW: Idiopathic thrombocytopenia purpura in childhood. J Pediatr 68:971-979.1966 3. Walker JH, Walker W: Idiopathic thrombocytopenic purpura in childhood. Arch Dis Child 36:649-657, 1961 4. Imbach P, Barandum S, Hirt A, et al: Intravenous immunoglobulin for idiopathic thrombocytopenic purpura (ITP) in childhood. Am J Pediatr Hematol Oncol6:171-174,1984 5. Imbach P, Berchtold W, Hirt A, et al: Intravenous immunoglobulin versus oral corticosteroids in acute immune thrombocytopenic purpura in childhood. Lancet 2:464-468, 1985 6. Bussel JB, Schulman I, Hilgartner MW, et al: Intravenous use of gammaglobulin in the treatment of chronic immune thrombocytopenic purpura as a means to defer splenectomy. J Pediatr 103:651-654, 1983 7. Newton WA, Zwelzer WW: Idiopathic thrombocytopenia purpura in childhood. N Engl J Med 245:879-885,195l 8. Benham ES, Taft LI: Idiopathic thrombocytopenic purpura in children: Results of steroid therapy and splenectomy. Aust Paediatr J 8:311-317, 1972 9. Coon WW: Splenectomy for idiopathic thrombocytopenic purpura. Surg Gynecol Obstet 164:225-229,1987 10. Block GE, Evans R, Zajtchuk R: Splenectomy for idiopathic thrombocytopenic purpura. Arch Surg 92:484-489,1966 11. Wilde RC, Ellis LD, Cooper WM: Splenectomy for chronic idiopathic thrombocytopenic purpura. Arch Surg 95344-350, 1967 12. Akwari OE, Itani KMF, Coleman RE, et al: Splenectomy for primary and recurrent immune thrombocytopenic purpura (ITP). Ann Surg 206:529-541.1987 13. Schwartz SI, Hoepp LM, Sachs S: Splenectomy for thrombocytopenia. Surgery 88:497-506, 1980 14. Weinblatt ME, Ortega JA: Steroid responsiveness: A predictor of the outcome of splenectomy in children with chronic immune thrombocytopenic purpura. Am J Dis Child 136:1064-1066, 1982 15. Cola B. Tonielli E, Sacco S, et al: Surgical treatment of chronic idiopathic thrombocytopenic purpura: Results in 107 cases. Int Surg 71:195-198, 1986 16. Gernsheimer T, Stratton J, Ballem PJ, et al: Mechanisms of response to treatment in autoimmune thrombocytopenic purpura. N Engl J Med 320:974-980,1989
17. Najean Y, Ardaillou N: The sequestration site of platelets in idiopathic thrombocytopenic purpura: Its correlation with the results of splenectomy. Br J Haematol21:153-164, 1971 18. Brennan MF, Rappeport JM, Moloney WC, et al: Correlation between response to corticosteroids and splenectomy for adult idiopathic thrombocytopenic purpura. Am J Surg 129:490-492, 1975 19. Schmidt RE, Budde U, Broschen-Zywietz C, et al: High dose gammaglobulin therapy in adults with idiopathic thrombocytopenic purpura (ITP) clinical effects. Blut 48:19-25,1984 20. Fehr J, Hofmann V, Kappeler U: Transient reversal of thrombocytopenia in idiopathic thrombocytopenic purpura by high-dose intravenous gamma globulin. N Engl J Med 306:12541258,1982 21. Templeton JG, Cocker JE, Crawford RJ, et al: Fey-receptor blocking antibodies in hyperimmune and normal pooled gammaglobulin. Lancet 1:1337. 1985 22. Stuart MJ, Kelton JG: The platelet: Quantitative and qualitative abnormalities, in Nathan DG, Oski FA (eds): Hematology of Infancy and Childhood. Philadelphia, PA, Saunders, 1987, pp 1348-1373 23. Sultan Y, Maisonneuve P, Kazatchkine MD, et al: Antiidiotypic suppression of autoantibodies to factor VIII (antihaemophilic factor) by high-dose intravenous gammaglobulin. Lancet 2:765-768,1984 24. Gadd SJ, Brogan MP, Ashman LK: Specificity of the passive antibody-induced suppression of the humoral immune response of mice to surface antigens on human cells. Immunology 54:223-231, 1985 25. Eraklis AJ, Kevy SV, Diamond LK, et al: Hazard of overwhelming infection after splenectomy in childhood. N Engl J Med 276:1225-1229,1967 26. Posey DL, Marks C: Overwhelming postsplenectomy sepsis in childhood. Am J Surg 145:318-321, 1983 27. Muraji T, Hays DM, Siegel SE, et al: Evaluation of the surgical aspects of staging laparotomy for Hodgkin’s disease in children. J Pediatr Surg 17:843-848, 1982 28. Wallace D, Fromm D, Thomas D: Accessory splenectomy for idiopathic thrombocytopenic purpura. Surgery 91:134-136, 1982 29. Verheyden CN, Beart RW, Clifton MD, et al: Accessory
412
DAVIS, WILLIAMS, AND SHAMBERGER
splenectomy in management of recurrent idiopathic thrombocytopenic purpura. Mayo Clin Proc 53:442-446,1978 30. Stiehm ER, Ashida E, Kim KS, et al: UCLA conference: Intravenous immunoglobulins as therapeutic agents. Ann Intern Med 107:367-382,1987 31. Centers for Disease Control: Safety of therapeutic immune globulin preparations with respect to transmission of human
T-lymphotropic views type III/lymphadenopathy-associated infection. MMWR 35:231-233, 1986 32. Glenn F, Grafe WR: Surgical complications steroid therapy. Ann Surg 165:1023-1034,1967
virus
of adrenal
33. Vogel HG: Tensile strength of skin wounds in rats after treatment with corticosteroids. Acta Endocrinol64:295-303,197O
Discussion R T. Schaller (Seattle, WA): ITP is characterized by antiplatelet antibodies that mediate the rapid destruction of these cells by the RE system. The spleen is involved in the pathogenesis of ITP in two respects: it is the major site of both autoantibody production and phagocytosis of opsonized platelets. In a climate of evolving pretreatment regimens over the past 10 years, the authors have convincingly demonstrated the predictive value of response to steroids and/or IgG. Twenty-nine of 30 patients who responded to medical management did well after splenectomy whereas only 3 of 10 patients who failed medical management responded to splenectomy. The reason for this difference must be the production of significant amounts of autoantibody against platelets elsewhere in the immune system. However, low serum titers of autoantibodies and the polyclonal nature of human serum make it very difficult to identify platelet target antigens with plasma antibodies. Work is now being done to isolate human monoclonal antiplatelet antibodies from patients with ITP so we can identify which platelet protein is carrying the immunogenic epitope. This may have implications for the future treatment of patients with ITP. Once a blocking antibody is developed against the antiplatelet autoantibody, surgeons may no longer be needed to treat this disease or, more realistically, an immune solution to the splenectomy treatment failures may be developed. The importance of autoantibodies against platelets must be implied by the therapeutic efficacy of removing accessory spleens that have been identified in some relapsed patients. Another factor of prognostic significance described recently has been splenic follicle size. In a recent review, patients with hyperplasia of splenic follicles greater than 500 pm were more likely to relapse or to develop additional autoimmune disorders than patients without hyperplasia of splenic follicles. In fact, if the follicle size was normal in the resected specimens, there was 100% response to splenectomy whereas one of four patients with follicle hyperplasia failed splenectomy. I congratulate the authors on the lack of complications in the series and the clear description of the progression of their
preparative regimen during the past 10 years. I have several practical questions for them. What was the long-term outcome of those patients who did not respond to splenectomy and did splenectomy make their medical management easier? How often did you identify accessory spleens at operation and do you have any special techniques for locating them? And finally, what is your current 1990 preference for preoperative preparation with steroids and/or IgG? P. W. Davis (response): As far as the specific pathology of the spleens removed, that was not looked at in the study. In terms of our recent preference for preoperative treatment, we do prefer IV IgG for its transient effect and the limitation of potential perioperative complications from steroid use. Regarding the outcome for patients who had failed splenectomy, who typically would have undergone continued medical courses with steroids, some with danozal, there is one patient who received courses of vincristine. The course of thrombocytopenia in several of these patients was somewhat attenuated. RM. Filler (Toronto, Ontario): Could you clarify one point about the 17 patients who received no treatment before surgery. Do you mean they never received treatment or just on that admission? Had they received treatment at a prior time? P. W. Davis (response): They did not receive any immediate preoperative treatment. All patients had received various types of treatment during their illness. D.M. Hays (Los Angeles, CA): Do you have any idea how many of the patients with an original diagnosis of ITP in Boston turned out to have lupus? P. W. Davis (response): I don’t know in terms of all patients with ITP. We had one patient in the series who had lupus and responded well to splenectomy. R. Soninno (Cleveland, OH): Aside from the fact that I always thought that the “I” in ITP stood for idiopathic, do you have any experience with preoperative nuclear platelet trapping studies and, if so, do you think they help in any way to predict outcome? P. W. Davis (response): In review of the literature for this paper, we did identify several reports that used
ITP: SURGICAL THERAPY AND RESPONSE
platelet trapping studies trying to identify patients who may benefit from splenectomy. Patients were found that had a predominence of either splenic trapping or liver sequestration and in these studies, the patients who had predominantly splenic sequestration seem to do better with splenectomy. However, we have not had any direct experience with these methods. A. Hailer (Baltimore, MD): I am confused by your conclusions in the sense that we continue to use steroids as the primary mode of treatment for ITP and operate on those patients who do not have a response. Your conclusions are that those who have a good response also respond to splenectomy. But what about those patients who do not respond and after they have had their splenectomy have not responded? What’s wrong with them? How are they different? Have you attempted to pick up those patients before so that we can look at another mode of treatment for them? And I guess my question is, are you really suggesting that a patient who responds to steroids still needs to have a splenectomy if they have gotten a good response to it? P. W Davis (response): The question of response to any medical therapy is not implied to be a curative response. Patients remit with or without medical therapies, the medical courses are primarily being used in terms of reversing episodes of thrombocytopenia. So the response we are looking at is in terms of a reversal of the thrombocytopenia that may be temporary. Some of the severe cases have reversal of thrombocytopenia, but after completion of those courses of medical therapy had rapid recurrence of their thrombocytopenia to very low levels and they ended up receiving splenectomy because they required repeated courses of steroids or other medical regimens. As far as those patients who have had no response to medical therapy, we have considered investigating the use of platelet radionuclide scan to try and identify those patients who may have predominantly liver sequestration as a reason for their failure, thereby trying to manage them potentially differently than splenectomy. W.L. Donnellan (Lansing, MI): I just wanted to clear up a point with mathematics. I am sure it is clear in the final paper, but you treated 23 patients preoperatively and 22 got a good response and were
413
operated on and the other 17 didn’t get any prior treatment at all. How do you reach the conclusion that response to preoperative treatment will condition what happens after the operation? P. W. Davis (response): Those numbers specifically are looking at those patients who received an immediate preoperative treatment in an effort to boost their platelet counts to lessen the likelihood of platelet transfusions. So those numbers are specifically looking at that group of 23 patients who received preoperative treatment only one required platelet transfusions, so we are looking at two sort of different categorizations of patients. The overall response rate relative to a response to any type of treatment during the course of disease is what is predictive of a response to splenectomy. J.L. Hill (Baltimore, MD): Did you look at the difference in ages with respect to the responsiveness of either medical treatment or surgical treatment to lump the children that are 1 or 2 years of age together with a teenager? In other problems, you can have quite a difference. For instance, giving ATG to a child versus an adult at a comparative dose, you can get a remarkable thrombocytosis in the child that far exceeds that in a teenager. Second, when did you use your preoperative steroid treatment? There is a test actually for bone marrow function where steroids are given, the response is assessed, and it peaks between 1 and 2 hours and then in fact the platelet count will drop below the baseline. And did you do this in a matter of hours or was it a matter of days in your preoperative management? P. W. Davk (response): Regarding your first question, that is not something that we looked at specifically. The timing of the preoperative treatments were basically within 2 weeks of planned surgery so they were not immediately preoperative. D. Tapper (Seattle, WA): Do you have an idea of the total number of patients with ITP during that lo-year period? In addition to that in the previous years, what’s happening to the incidence of the patients that we are seeing is it less? P. W. Davis (response): I’m afraid I really can’t comment on that. We didn’t look at ITPs that would include those patients with a complete response to steroids or spontaneous recovery who were never considered for splenectomy.
