SEARCH

SEARCH BY CITATION

Keywords:

  • immune thrombocytopenia;
  • splenectomy;
  • predictive factors;
  • age;
  • IVGG

Abstract

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

Sixty-one consecutive patients undergoing splenectomy for chronic immune thrombocytopenia were retrospectively evaluated. Platelet response was considered as complete (CR) when platelet count rose to > 100 × 109/l, partial (PR) when 30–100 × 109/l or absent (NR) if otherwise. Follow-up (mean time 7·6 years) was possible in 54 patients. Forty-eight patients (88%) had an immediate response to splenectomy (39 CR, 9 PR) whereas six (12%) were NR. Thirty-six responders (67%) had sustained remission (31 CR; 5 PR) without further treatment; thrombocytopenia recurred in 12 patients (33%). The probability curve of continued remission showed a constant relapse-rate during the first 36 months; a further step of relapse was observed beginning 70 months after surgery. The only positive predictive factor for the long-term response to splenectomy was age < 40 (P < 0·005). Neither duration of thrombocytopenia nor previous response to medical treatment (steroids and/or intravenous immunoglobulins) were related to splenectomy response.

Thrombocytopenia is commonly encountered in clinical practice with an estimated incidence of 4·5 per 100 000 in men and 7·4 in women. About 50% of cases are the result of autoantibodies against platelet and megakaryocyte surface antigens (McMillan & Imbach, 1994).

The acute form of immune thrombocytopenic purpura (ITP) occurs more frequently in childhood and is self-limiting in more than 70% of cases, whereas chronic ITP occurs mainly in adults ( McMillan & Imbach, 1994; George et al, 1996).

The therapeutic strategies in ITP include: (a) treatment of acute bleeding episodes; (b) long-term improvement in platelet count; and (c) potentially definitive cure of the disease. Because few chronic ITP patients achieve remission with medical therapy, splenectomy is the corner-stone treatment (George et al, 1996). However, there is no definitive evidence concerning the optimal timing for splenectomy, nor about the related prognostic factors of response or long-term outcome. We retrospectively analysed the clinical response to splenectomy in 61 patients with chronic ITP belonging to a larger cohort of 200 consecutive patients treated at our clinics.

Patients and methods

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

Patients. The diagnostic criteria for ITP were the following: (a) platelet count less than 100 × 109/l for at least 6 months; (b) normal or increased number of megakaryocytes at bone marrow examination; and (c) exclusion of (i) recent administration of drugs known to possibly induce thrombocytopenia, (ii) increased splenic sequestration, (iii) sepsis, (iv) microangiopathic disorders and (v) familial forms of inherited thrombocytopenia.

Haemorrhage on presentation was defined as severe mucosal bleeding, melena, haematemesis, haematuria, haemophtysis or metrorrhagia requiring red cell transfusion; none of the patients had cerebral bleeding.

All patients were given oral prednisone 1 mg/kg until response or for a maximum of 4 weeks, then tapered until the minimal effective dose required to maintain platelets > 30 × 10 9/l was reached. Several different drugs were also administered concurrently with prednisone: 24 patients were given i.v. gamma globulins (IVGG) 0·4 mg/kg over 3 or 5 consecutive d; 10 patients were given danazol 200–400 mg/d for at least 30 consecutive d; eight patients were given i.v. vincristine 1–2 mg/weekly, for a maximum of three consecutive cycles; Azathioprine was administered to seven patients at a dose of 100 mg/d for at least 5 weeks and than tapered to 50 mg/d; three to five courses of plasma exchange followed by one course of IVGG were performed in four patients. All patients underwent medical treatment for at least 6 months before splenectomy.

Clinical outcome and follow-up. Response to either medical or surgical treatment was considered as complete (CR) if platelet count rose to > 100 × 109/l, partial (PR) if platelets were between 30 and 100 × 109/l or absent (NR) if otherwise. Remission was defined as partial or complete response persisting after all treatments were discontinued.

Follow-up was obtained in 54 patients. The immediate response to splenectomy was considered within 6 months of surgery and the long-term response was considered from 6 months to 34 years later. All records were updated in December 1999 by interviewing either the patients or their family physicians.

Statistical analysis. The Kaplan–Meier method was used to estimate the probability of overall thrombocytopenia-free survival as a function of time from splenectomy for subgroups of patients. The Wilcoxon model was used to assess the significance of the difference between subgroups with 95% confidence intervals. The chi-square test with a continuity correction was used to compare proportions; P < 0·05 were considered significant. The Cox-proportional-hazards model was used for multivariate analysis of possible predictors of response to splenectomy.

Results

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

Response to splenectomy

The response to splenectomy is reported in Table I. Forty-eight of 54 evaluable patients (88%) had immediate complete or partial remission, whereas thrombocytopenia persisted in six (12%). The probability curve of persistent remission showed a constant relapse rate during the first 36 months of observation; a further step-down relapse was observed starting 70 months after surgery (Fig 1).

Table I.  Characteristics of patients related to response to splenectomy.
 All patientsRespondersNon-respondersP-values
  • *

     Mean ± standard deviation [range].   

  •  Median (percentage).   

  • ‡ 

    Response before splenectomy.   

  • §

     Five thyroid diseases, one systemic lupus erythematosus, two Evans, one Down's syndrome, four HBV positive, two HCV and three HBV + HCV positive.

  • PR, CR: partial or complete response.

Immediate response5448 (88) 6 (12) 
Relapses1212 (25) 
Long-term response5436 (67)18 (33)
Female/male34/2024//1210/80·60
Children 4 (7·4) 3 (8·3) 1 (5·5)0·85
Mean age (years)38 ± 17*[11–70] 3336 ± 19 [11–70] 3143 ± 17 [13–66] 470·19
Platelet count (x109/l)23 ± 21 [2–84] 1525 ± 25 [2–84] 1518 ± 13 [3–45] 200·26
Years of follow-up 7·6 ± 7·0 [0·4–33·7] 6 7·7 ± 7·4 [0·4–33·7] 6 7·3 ± 6·3 [0·6–21·7] 60·84
Years of thrombocytopenia 3·6 ± 4·0 [0·5–18·5]1·52·9 ± 3·1 [0·5–13·7] 1·5 4·3 ± 4 [0·6–11·8] 30·17
Associated diseases§18 (33)11 (30) 7 (39)0·76
Major haemorrhages13/47 (28) 3/30 (10)10/17 (60)0·001
Splenectomy by 12 months15 (28) 9 (25) 4 (22)0·92
PR + CR to therapy
 Steroids30/54 (55)23/36 (64)  7/18 (39)0·27
 Steroids + Danazol 6/9 (66) 1/2 (50) 5/7 (71)0·77
 Vincristine 2/8 (25) 1/1 (100) 1/7 (14)
 IVGG13/24 (54)11/19 (58) 2/5 (40)0·65
 Azathioprine 0/4 (0) 0/2 (0) 0/2 (0)
 Plasmapheresis + IVGG 2/4 (50) 1/2 (50) 1/2 (50)
image

Figure 1. Thrombocytopenia-free survival (platelet count > 30 × 109/l) among the cohort of patients (Kaplan–Meier estimate) and the subgroups defined according to age. The estimated curve of all patients shows a constant relapse-rate during the first 36 months, a plateau at 80% of remission rate and further relapse beginning 70 months after surgery. The 28 patients aged more than 40 years had a significantly lower probability of remission than that of 33 patients less than 40 years.

Download figure to PowerPoint

Thrombocytopenia recurred in 12 of 48 responders (25%); three patients had accessory spleen, one relapsed during pregnancy and eight had no apparent causative or precipitating cause of recurrence. Three patients relapsed within 1 year of splenectomy, whereas thrombocytopenia recurred more than 5 years later in the other cases. Therefore, 36 patients (67% overall) had complete or partial long-term remission. Of the remaining 18 patients, seven had partial response to low-dose prednisone; five were partially responsive to steroids plus Danazol; two were responsive to IVGG courses; and four, with platelet count < 30 × 109/l, were maintained without any therapy.

Predictive factors of response

The percentage of major haemorrhages at presentation was significantly higher in non-responders (Table I). The probability curve of persistent remission showed that patients younger than age 40 (median age 23·5 years) had a higher probability of remission than the older group (median age 58 years) (P < 0·005) (Fig 1). The cut-off age of 40 was selected after analysing the frequency distribution of age versus response to splenectomy.

Sex, age, platelet count, years of thrombocytopenia, associated disease, time of splenectomy and the response to medical therapy were also analysed using a multivariate model to assess their association with splenectomy response. Age over 40 was identified as the only significant risk factor for response to splenectomy using the Cox multivariate analysis (hazard ratio 6·51, 95% confidence interval 1·9–21·9, P = 0·0025), whereas previous haemorrhages lost significance after correction for age.

Discussion

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

Chronic ITP in elderly patients usually has an insidious onset and a prolonged benign course with only minor bleeding manifestations such as purpura, petechiae, epistaxis and gingival oozing. Therefore, in the absence of surgery or risk factors, patients should be treated only when platelet count is lower than 30 × 109/l or in the presence of major haemorrhage (George et al, 1996).

Steroids increase platelet count in more than 50% of patients, but complete remission of thrombocytopenia is obtained only in a few (McMillan & Imbach, 1994; George et al, 1996;) who would probably have done well spontaneously. Patients who are refractory, or responsive only to high-dose steroids or complain of severe side-effects of medical therapy, are candidates for splenectomy. However, there is no evidence-based consensus about the optimal timing for surgery. Our patients underwent late splenectomy in agreement to the presently prevailing opinion among American haematologists (George, 2000). Accordingly, the rationale for late surgery was good tolerance to low platelet count, the possibility of spontaneous remission and the concern of life-threatening infections after splenectomy. The results from several studies published in recent years indicate no difference in outcome between early (< 12 months from onset) (Mazzucconi et al, 1999) and late surgery (Fenaux et al, 1989; Stasi et al, 1995).

Splenectomy has been successful in about 80% of patients, consistent with the fact that the spleen is the predominant site for both autoantibody production and platelet destruction (Gernsheimer et al, 1989).

In keeping with such data, an initial response to splenectomy of 87% was observed in our series. However, 25% of our patients relapsed, a slightly higher percentage than reported elsewhere (Pizzuto & Ambritz, 1984; Fenaux et al, 1989; Stasi et al, 1995). Such a discrepancy can be explained by the threshold of 30 × 109/l platelets to define response and the longer follow-up in our series, suggesting a time relationship for response rate. In fact, the probability of remission at 120 months was similar to that reported in another long-term study (Stasi et al, 1995), however we observed more late relapses.

Several factors have been considered as positive predictors for response to splenectomy, such as younger age of patients (Fenaux et al, 1989; Julia et al, 1990), previous response to steroids (Pizzuto & Ambritz, 1984) or IVGG (Law et al, 1997), intensity of bleeding (Julia et al, 1990) and predominantly splenic sequestration (Fenaux et al, 1989).

We analysed the prognostic value of age, sex, severe haemorrhages and platelet count at diagnosis, duration of thrombocytopenia and the response to medical therapy. Our data support the role of positive predictive factor for long-term response to splenectomy for young age and severe bleeding only. Major haemorrhages were also confirmed to be age related, as reported by Cortellazzo et al (1991). The relationship between IVGG and splenectomy response was not statistically significant because patients refractory to IVGG still had a good chance of responding to splenectomy (Fabris et al, 1997).

In conclusion, splenectomy is an effective treatment for chronic ITP but the response rate decreases with time. Choosing splenectomy for ITP treatment requires individual case analysis because only young age seems to be a predictive factor for a good response.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

This study was supported by grants from CNR Rome (grant 9502257CT04).The authors are indebted to Dr F. Noventa for aid with the statistical analysis of the data.

References

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References
  • Cortellazzo, S., Finazzi, G., Buelli, M., Molteni, A., Viero, P. & Barbui, T. (1991) High risk of severe bleeding in aged patients with chronic idiopathic thrombocytopenic purpura. Blood, 77, 3133.
  • Fabris, F., Cordiano, I. & Girolami, A. (1997) High dose intravenous immune globulin and the response to splenectomy in patients with idiopathic thrombocytopenic purpura. New England Journal of Medicine, 337, 10881089.
  • Fenaux, P., Cauller, M.T., Hirshauer, M.C., Beauscart, R., Goudeman, J. & Bauters, F. (1989) Reevaluation of the prognostic factors for splenectomy in chronic idiopathic thrombocytopenic purpura. European Journal of Haematology, 42, 259264.
  • George, J.N. (2000) Treatment options for chronic idiopathic (immune) thrombocytopenic purpura. Seminars in Hematology, 37, 3134.
  • George, J., Woolf, S.H., Raskob, G.E., Wsser, J.S., Aledort, L.M., Ballem, P.J., Blanchette, V.S., Bussel, J.B., Cines, D.B., Kelton, J.G., Lichtin, A.E., McMillan, R., Okerbloom, A., Regan, D.H. & Warrier, I. (1996) Idiopathic thrombocytopenic purpura: a practice guideline developed by explicit methods for the American Society of Hematology. Blood, 88, 340.
  • Gernsheimer, T., Stratton, J., Ballem, P.J. & Slichter, S.J. (1989) Mechanisms of response to treatment in autoimmune thrombocytopenic purpura. New England Journal of Medicine, 320, 974980.
  • Julia A., Araguas C., Rossello J., Bueno J., Domenech P., Olona M., Petit J. & Flores A. (1990) Lack of useful clinical predictors of response to splenectomy in patients with chronic idiopathic thrombocytopenic purpura. British Journal of Haematology, 76, 250255.
  • Law, C., Marcaccio, M., Tam, P., Heddle, N. & Kelton, J.G. (1997) High dose intravenous immune globulin and the response to splenectomy in patients with idiopathic thrombocytopenic purpura. New England Journal of Medicine, 336, 14941498.
  • Mazzucconi, M.G., Arista, M.C., Peraino, M., Chistolini, A., Felici, C., Francavilla, V., Macale, E., Conti, L. & Gandolfo, G.M. (1999) Long-term follow-up of autoimmune thrombocytopenic purpura (ATP) patients submitted to splenectomy. European Journal of Haematology, 62, 219222.
  • McMillan, R. & Imbach, P.A. (1994) Immune thrombocytopenic purpura. In: Thrombosis and Hemorrhage (ed. by J.LoScalzo & A.I.Schafer), pp. 575595. Blackwell Science, Oxford.
  • Pizzuto, J. & Ambritz, R. (1984) Therapeutic experience on 934 adults with idiopathic thrombocytopenic purpura: multicentric trial of the Cooperative Latin American Group on Hemostasis and Thrombosis. Blood, 64, 11791183.
  • Stasi, R., Stipa, E., Masi, M., Cecconi, M., Scimò, M.T., Oliva, F., Sciarra, A., Alessio, P., Adomo, G., Amadori, S. & Papa, G. (1995) Long-term observations of 208 adults with chronic idiopathic thrombocytopenic purpura. American Journal of Medicine, 98, 436442.