The importance of clinical judgment for the diagnosis of thrombotic thrombocytopenic purpura

In this issue of TRANSFUSION, Kim and colleagues1 analyze the cost-effectiveness of different strategies for the initial management of patients with suspected thrombotic thrombocytopenic purpura (TTP) using previously published data. They analyzed four hypothetical scenarios: two different rates of reporting ADAMTS13 activity measurements, each with and without the prior use of a clinical prediction score (the PLASMIC score2). Scenarios 1 and 2 evaluated the rates of reporting ADAMTS13 activity: 3 days after the blood was drawn (the test is performed at a distant site, “send-out”) or on the same day that blood was drawn (the test is performed “in-hospital”). Scenarios 3 and 4 were the two rates of reporting ADAMTS13 activity combined with the PLASMIC score; the ADAMTS13 test was not performed when the PLASMIC score was less than 4 (maximum score, 7), indicating a low probability of ADAMTS13 activity of less than 10%.2 Kim and colleagues1 conclude that an in-hospital ADAMTS13 test combined with the PLASMIC score provides the most cost-effective management. Even though there is cost required for the implementation of an in-hospital ADAMTS13 test, cost saving would result from avoidance of plasma exchange treatment (TPE) for patients with a low probability of TTP (PLASMIC score of 4 or less ADAMTS13 activity ≥ 10%). Avoidance of TPE would also mean fewer TPE-associated major complications and their cost of extended hospitalization and care.3, 4 The question for readers of TRANSFUSION may be: Why are these data important for clinicians requesting TPE or transfusion medicine specialists providing TPE on a consultative basis? For clinical practice, these data would be applied to the evaluation of patients who have anemia with features of microangiopathic hemolysis and thrombocytopenia, the signs that create initial suspicion for the diagnosis of TTP. Evaluation of these patients leads to one critical decision. Should TPE be requested immediately or is it better to hesitate and to intensify the search for an alternative diagnosis? How do the data of Kim and colleagues1 help this decision? Their conclusion that a rapid report of ADAMTS13 activity together with a PLASMIC score provides the most cost-effective practice may not be the same as providing the most clinically effective practice. However, my interpretation is that these two outcomes, cost-effective and clinically effective, are aligned. My interpretation is that a rapid report of ADAMTS13 activity of less than 10% together with a PLASMIC score of more than 4 should support a decision for urgent initiation of TPE treatment. Alternatively, ADAMTS13 activity of 10% or more or a PLASMIC score of 4 or less may (or may not) support a decision to hesitate requesting TPE and to intensify the search for an alternative diagnosis. These data of Kim and colleagues,1 together with the data that supported the derivation of the PLASMIC score,2 are important for clinicians whenever they evaluate a patient for the possible diagnosis of TTP. The decision to begin TPE, or to hesitate, is urgent because of the risk of death if TTP is not treated5 and the effectiveness of TPE for treatment of TTP.6 The decision to begin TPE is related to the physician's impression that the risks from hesitation are greater than the risks from TPE. This dilemma is evident from our experience. Among 363 patients who have been enrolled in the Oklahoma Registry during the past 20 years because TPE was requested (almost always for suspected TTP), only 78 (21%) had ADAMTS13 activity of less than 10% and also a clinical diagnosis of TTP.4 These data document that physicians in Oklahoma have had a low threshold for ordering TPE when TTP was suspected. Most patients did not have TTP. I assume that physicians in Oklahoma are not different from physicians everywhere. Our experience may be similar to data reporting the frequency of TPE for TTP in Canada7 during the clinical trial comparing plasma infusion with TPE (1982-1988) and after its completion and publication in 1991.6 Even when this clinical trial was still enrolling patients, TPE use for suspected TTP increased, presumably because of increasing awareness of the effectiveness of TPE. After publication of this trial, TPE use for suspected TTP doubled.7 With the availability of an effective treatment for TTP (and without the availability of the yet to be discovered ADAMTS13 activity measurements to support the diagnosis), the feeling of urgency for beginning TPE for suspected TTP would have been great. The feeling of urgency for beginning TPE for suspected TTP may become even greater with the future approval of caplacizumab for treatment of TTP.8 Caplacizumab effectively blocks formation of platelet–von Willebrand factor (VWF) thrombi and therefore halts the pathogenesis of TTP. The recommendation for caplacizumab will be that it should be given to patients with TTP as soon as possible. And it will certainly be very expensive. Therefore, the prompt availability of ADAMTS13 activity is clearly an important advantage for clinicians, as it is important for cost-effectiveness. TTP is defined by a severe deficiency of ADAMTS13 activity and ADAMTS13 deficiency is responsible for the clinical features of TTP. A recent consensus panel stated that ADAMTS13 activity of less than 10% confirms the diagnosis of TTP.9 Therefore, it is appropriate to begin TPE in a patient with suspected TTP if the ADAMTS13 activity is less than 10%. But the diagnosis of TTP and beginning treatment with TPE cannot be based on the ADAMTS13 activity alone. The clinician's judgment is also critical for the diagnosis of TTP. This additional diagnostic help is the contribution of the PLASMIC score and other prediction models developed from clinical features10, 11 to assist the evaluation of patients with suspected TTP. These scores combine clinical features to provide a numerical value. They quantify clinical judgment. They convert clinical judgment into data. Then these clinical data can be combined with the ADAMTS13 activity data. They can be synergistic. Thrombotic thrombocytopenic purpura provides a perfect example for the essential value of clinical judgment. I accept that ADAMTS13 deficiency is the definition of TTP and that ADAMTS13 deficiency is responsible for the clinical features of TTP. These concepts have been documented for many years.12-14 But my experience documents that measurements of ADAMTS13 activity alone are insufficient for the clinical decision to initiate TPE or to hesitate. Measurements of ADAMTS13 activity may be inconsistent. Patients with TTP may have reported ADAMTS13 activity of 10% or more. Patients who do not have TTP may have reported ADAMTS13 activity of less than 10%. Patients with TTP may have reported ADAMTS13 activity of 10% or more with their initial episode and then ADAMTS13 activity of less than 10% with subsequent episodes. Although these inconsistencies represent only a small minority of patients with TTP, our responsibility as clinicians is to protect minorities. Here are the data for the apparent discrepancies between ADAMTS13 activity and the clinical diagnosis of TTP. For patients in the Oklahoma Registry, ADAMTS13 activity is measured by two methods15: a fluorogenic assay using a VWF732 peptide (FRET method, the common commercial method) and quantitative immunoblotting (IB method, the original method for measuring ADAMTS13 activity12). We report the lower of the two results. Among the 78 patients whom we have diagnosed as having TTP by both ADAMTS13 activity of less than 10% and also clinical features, ADAMTS13 activity was less than 10% by both FRET and IB assays in 60 patients. In 15 patients, the FRET method measured ADAMTS13 activity as less than 10% while the IB method measured ADAMTS13 activity as 10% to 68%. Among these 15 patients, the three with the highest ADAMTS13 activities by IB measurements (35%-68%) have all relapsed, confirming their diagnosis of TTP. At the time of their relapse, the IB measurement was less than 10% in two patients and 18% in the third. In three other patients, the IB method measured ADAMTS13 activity as less than 10% while the FRET assay measured ADAMTS13 activity as 11% to 23%.4 Therefore, it should be considered that patients may still have TTP even when neither the FRET nor the IB measurement of ADAMTS13 activity is less than 10%. When we reviewed the presenting features and clinical course of patients whose lower measurement of ADAMTS13 activity was 10% to 20%, we determined that four of 22 patients had characteristic clinical features supporting a diagnosis of TTP.16 One additional Registry patient had characteristic clinical features of TTP but his initial ADAMTS13 activities were 53% by the FRET measurement and 60% by the IB measurement; he subsequently had five relapses. With his last three episodes, both ADAMTS13 measurements reported activities of less than 10%.17 The reverse situation has also occurred. Five patients who were treated with TPE for suspected TTP had ADAMTS13 activity of less than 10% but their subsequent clinical diagnoses were severe systemic infections or malignancy; four died, and two had autopsies with no evidence of TTP. Two of these patients had high serum bilirubin levels (24 and 64 mg/dL), which can interfere with the FRET measurement, causing falsely low ADAMTS13 activity.18 These observations emphasize that the interactions among ADAMTS13, anti-ADAMTS13 autoantibodies, and VWF are complex and measurements of ADAMTS13 activity may be susceptible for falsely high values.19 These observations document the essential role for clinical judgment in the evaluation and management of patients with suspected TTP. Although an individual physician's clinical judgment is essential, an individual physician's experience with patients who have suspected TTP may be limited. Enhancement of an individual physician's experience and clinical judgment is the goal of the quantitative scores, such as the PLASMIC score2 and its predecessors.10, 11 The derivation and validation of the PLASMIC score was an impressive effort.2 Although some of the seven components of the PLASMIC score may not appear to be common diagnostic variables of TTP, they were selected and validated objectively. The value of these scores is that they provide a consistent, quantitative measure. They supplement, but they do not replace, our individual clinical judgment. For example, in my own experience the occurrence of transient focal neurologic abnormalities has a strong association with TTP. Older age (≥70 years) and a long duration of illness or hospitalization preceding the suspicion for TTP appear to be evidence against the diagnosis of TTP.16 These clinical features are not part of the previously published scores for evaluating patients with suspected TTP.2, 10, 11 Individual physician experience can enhance the ability of a standardized clinical score, but the reproducible qualities of the clinical scores provide a quantitative measure that our individual experience cannot match. Now, back to the report by Kim and coworkers.1 They used data from many published sources (see their Table S1) and they used multiple methods of analysis. Their conclusions are clear. The rapid report of ADAMTS13 activity improves cost-effectiveness and the addition of the PLASMIC score provides additional value. Their principal focus is for hospital administration and economic policy, to present data on the value of introducing a new laboratory test that will be used infrequently but may help to control health care cost. This issue is important for population health, but issues of population health may override the needs of exceptional patients. The needs of exceptional TTP patients may be reflected in another result reported by Kim and coworkers.1 If effectiveness is measured only by the number of averted deaths, rather than the cost, then the send-out ADAMTS13 test is the most effective. The interpretation of this result is that TPE for 3 days before the report of ADAMTS13 activity prevents some deaths. This result is consistent with our experience that some patients with TTP have ADAMTS13 activity of 10% or greater and therefore would not have received TPE if the ADAMTS13 value was known immediately. My conclusion is that three things are important for the initial evaluation of a patient with suspected TTP: 1) prompt availability of ADAMTS13 activity; 2) the PLASMIC score or a comparable, quantitative clinical score for predicting the presence of ADAMTS13 activity of less than 10%; and 3) perhaps most importantly, the primary physician's individual clinical judgment, reviewed and supported by experienced transfusion medicine specialists. All three components contribute to the critical, initial decision: Should TPE begin urgently, or should we hesitate while we search for an alternative etiology for the microangiopathic hemolytic anemia and thrombocytopenia? The author has disclosed no conflicts of interest. James N. George, MD e-mail: [email protected] Department of Medicine, College of Medicine Department of Biostatistics and Epidemiology, College of Public Health University of Oklahoma Health Sciences Center Oklahoma City, OK

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.