Late Relapses in Patients Successfully Treated for Thrombotic Thrombocytopenic Purpura

Methods

The details of the Canadian Apheresis Group trial have been reported elsewhere [6]. Briefly, all patients with a diagnosis of thrombotic thrombocytopenic purpura who presented at any of 16 participating medical centers in Canada were considered for entry into the Canadian Apheresis Group randomized trial. This trial compared plasma exchange and plasma infusion as treatments for thrombotic thrombocytopenic purpura. The diagnostic criteria for thrombotic thrombocytopenic purpura were thrombocytopenia (platelet count <100 × 10⁹/L); microangiopathic hemolytic anemia (erythrocyte fragmentation in a peripheral blood film); and no identifiable cause for the thrombocytopenia or microangiopathic hemolytic anemia.

Eligible patients were randomly assigned to receive either 1) plasma exchange using fresh frozen plasma [a minimum of seven procedures during the first 9 days in the hospital; treatment on the first 3 days was essential]; or 2) daily plasma infusion until the end of a treatment cycle. A treatment cycle was defined as the period from randomization to one of the following events: death; clinical deterioration to the extent that the patient was taken off the protocol (for patients receiving plasma infusion, this usually meant crossing over to plasma exchange); early response; or the completion of 7 days of treatment. Patients receiving plasma exchange were treated with exchanges of 1.5 times their predicted plasma volume for the first three procedures and with exchanges equal to their plasma volume thereafter (average total exchange ±SD, 21.5 ±7.8 L). Patients receiving plasma infusion received 30 mL/kg body weight over the first 24 hours and 15 mL/kg daily thereafter (average total infusion ±SD, 6.7 ±3.3 L). Thus, patients receiving plasma exchange received approximately 3 times as much plasma as patients receiving plasma infusion.

Patients with a history of thrombotic thrombocytopenic purpura were not excluded from the study. All patients received dipyridamole (400 mg/d) and aspirin (325 mg/d) orally for a minimum of 2 weeks after entry.

Outcome was assessed at the end of the first treatment cycle (day 9 after randomization) and again 6 months after randomization. The clinical trial ended with this 6-month assessment, after which there was no prescribed follow-up protocol. Instead, patients were seen at intervals at the discretion of their own physicians or sooner if they developed symptoms. Similarly, the laboratory tests required as part of the clinical trial concluded with the tests done at the 6-month assessment. Thereafter, although all patients had complete blood counts done at each follow-up visit, any other laboratory tests done were as ordered by the individual attending physicians. The Canadian Apheresis Group attempted to collect clinical follow-up information on the 72 surviving patients; the attending physicians were able to provide this information for 63 of the 72 patients. Particular attention was paid to documenting any relapses of thrombotic thrombocytopenic purpura, which were defined as the patient again meeting the diagnostic criteria for thrombotic thrombocytopenic purpura that were used in the original trial, and the treatment and outcome of these relapses.

Relapse-free survival over the 10 years of observation was analyzed using the Kaplan-Meier method [8], and the approximate 95% CIs were computed according to the method of Simon [9].

Results

Seventy-two of the 102 patients in the original trial survived 6 months or more after entry. Nine of these patients were lost to follow-up. Thirty-seven of the remaining 63 patients are alive and well and have not had recurrence of thrombotic thrombocytopenic purpura during follow-up, which has ranged from 3 to 10 years.

Six other patients have not had recurrence of thrombotic thrombocytopenic purpura but have developed other medical problems (each of the six has had one of the following: systemic lupus erythematosus, non–insulin-dependent diabetes mellitus, chronic hepatitis, lymphoma, rheumatoid arthritis, and a fatal illness related to the acquired immunodeficiency syndrome [AIDS] 4 years after study entry). Three other patients who have not had recurrence of thrombotic thrombocytopenic purpura had neurologic deficits from the original episode (Table 1).

Seventeen patients (11 of whom were originally treated with plasma exchange and 6 of whom were originally treated with plasma infusion) have had recurrences of thrombotic thrombocytopenic purpura (Table 2). Clinically, the recurrences were indistinguishable from the initial presenting episodes. Only 3 of the 17 patients had relapses within 1 year. Nine of the 17 had relapse only once, 7 months to 7 years after entry into the study. The other 8 patients had recurring relapses as shown in Figure 1; the first relapse in 1 of these patients did not occur until 8 years after the patient was entered into the study. In all, there have been 30 relapses, and 2 patients have died during relapse. Relapses were seen in 13 of the 42 women and in 4 of the 21 men for whom follow-up had been obtained after the original trial. One patient had four relapses at 3, 8, 8.5, and 9 years but is currently surviving 10 years after her original presentation. As shown in Figure 1, the time at which the relapses occurred in patients with recurring episodes of relapse ranged from 8 months to 9 years.

View larger version:

• In this window
• In a new window

Figure 1. Time at which relapse occurred in eight patients with recurring episodes of thrombotic thrombocytopenic purpura.

Both patients who died during relapse failed to respond to plasma exchange and other treatments, including splenectomy in one patient. Both were among the few (11 of 102) patients who had a previous history of thrombotic thrombocytopenic purpura when they were entered into the original trial. Ten of these 11 patients had had only a single previous episode of thrombotic thrombocytopenic purpura and 1 had had two previous episodes.

Kaplan-Meier analysis of the data Figure 2 done 3 years after the last patient was entered into the trial and 10 years after the first patient was entered into the trial shows that 18% (95% CI, 9% to 27%) of patients can be expected to have had one or more relapses by 4 years after initial treatment (study entry). By 10 years after initial treatment, 36% (CI, 23% to 59%) of the patients can be expected to have had one or more relapses. When analysis was done only on those patients without a previous history of thrombotic thrombocytopenic purpura, 14% (CI, 4% to 28%) were found to have had a relapse by 4 years after initial treatment and 27% (CI, 7% to 61%) were found to have had a relapse by 10 years after initial treatment. No particular feature of the disease (including platelet count) or the treatment appeared to be a predictor for relapse, but our sample size was too small for meaningful statistical analysis of this possibility.

View larger version:

• In this window
• In a new window

Figure 2. Kaplan-Meier analysis of the probability of relapse-free survival in patients who survive an acute episode of thrombotic thrombocytopenic purpura.

Several treatments (plasma exchange, plasma infusion, steroids, and splenectomy) were used for the 30 episodes of relapse. When plasma exchange was used, a mean of 13.4 exchanges (range, 1 to 37 exchanges) was given. This Figure compares with a mean of 15.8 exchanges (range, 3 to 36 exchanges) in the original trial.

Fourteen of the 63 patients for whom follow-up was obtained had had a splenectomy at some time during their clinical course. One patient had had a splenectomy for acute thrombotic thrombocytopenic purpura 2 years before having relapse and being entered into the original clinical trial. This patient had another relapse 2 years later. Eight patients had had splenectomy during the episode of acute thrombotic thrombocytopenic purpura that had prompted their entry into the original clinical trial. One of these patients had relapse 3 years after splenectomy, but the other 7 have not had relapse. Five other patients had splenectomy at the time of relapse. One of these patients died after splenectomy; the other 4 have not had relapse.

Discussion

Patients with an acute episode of thrombotic thrombocytopenic purpura should be treated with plasma exchange as soon as possible [6], probably using plasma that has been depleted of cryoprecipitate and thus of the largest multimers of von Willebrand factor [10]. The use of cryosupernatant plasma as the fluid of choice for first-line therapy for thrombotic thrombocytopenic purpura is suggested by the results of a recent nonrandomized study in which patients who received exchanges of cryosupernatant plasma had better outcomes than patients who received exchanges of fresh frozen plasma [11]. Steroids may also be beneficial [5] but this has never been proved. The essential and most important treatment is immediate plasma exchange when the diagnosis of thrombotic thrombocytopenic purpura is made. The roles of other forms of treatment, including aspirin, dipyridamole, splenectomy, and vincristine sulfate are less clear.

Most patients who have acute thrombotic thrombocytopenic purpura now survive with these treatments. However, the prognosis for patients who have recovered from acute thrombotic thrombocytopenic purpura has been unclear. Recurrence of acute thrombotic thrombocytopenic purpura shortly after the apparent induction of remission is common [5]. After the disease has truly remitted—that is, when the patient is stable after cessation of plasma exchange—the reported relapse rate varies from 7.5% in a retrospective review of 25 patients [4] to 37% in a retrospective review of 38 patients [7]. There have been no reports on the incidence of relapse in a large series of patients followed prospectively for many years after an acute episode of thrombotic thrombocytopenic purpura had subsided.

In our study, it is possible that some episodes of mild, self-limited relapse were not detected by the attending physicians because the patients did not report symptoms. One or more episodes of relapse indistinguishable from the initial acute episodes were documented in 17 of the 63 patients who had survived an acute episode of thrombotic thrombocytopenic purpura. Some of the 63 patients are at only the 3-year follow-up point and, given that we have seen a first relapse as late as 8 years after the initial episode of thrombotic thrombocytopenic purpura, it is likely that more patients will have relapse during additional follow-up time. Kaplan-Meier analysis of the data indicates that the rate of relapse when all patients have been followed for 10 years will be 36% (CI, 23% to 59%).

The role of splenectomy in preventing relapse is unclear; our results show that late relapse of thrombotic thrombocytopenic purpura can occur in patients who have had splenectomy. However, our study does not allow conclusions about whether relapse of thrombotic thrombocytopenic purpura is more or less likely to occur in patients who have had splenectomy.

Patients who survive an initial episode of thrombotic thrombocytopenic purpura should be followed indefinitely for recurrence of this disease. Relapse may occur on several occasions and at variable intervals, in some instances after a long period of remission. Plasma exchange is usually effective in treating a recurrence of thrombotic thrombocytopenic purpura but, just as in patients with an initial episode, some patients who have such a recurrence are unresponsive to plasma exchange and are at risk for death.

The members of the Canadian Apheresis Group are Grenfell Adams, MD; Barrett Benny, MD; Noel A. Buskard, MD; Stephen Caplan, MD; Robert Card, MD; William F. Clark, MD; Peter Ford, MD; John Freedman, MD; Phil Gordon, MD; Max Gorelick, MD; John Kelton, MD; John Klassen, MD; Pierre Leblond, MD; Mariette Lepine-Martin, MD; Jack McBride, MD; Marc Monte, MD; Rama C. Nair, PhD; Harry Rayner, MD; Gail A. Rock, PhD, MD; Tsiporah Shore, MD; Kenneth H. Shumak, MD; Marion Sternbach, MD; David M.C. Sutton, MD; and John Verrier Jones, MD.