Published ahead of print on April 21, 2011
doi: 10.3174/ajnr.A2602

American Journal of Neuroradiology 32:E124, June-July 2011
© 2011 American Society of Neuroradiology

K. Lin^a
aDepartment of Radiology
New York Hospital Medical Center of Queens
Flushing, New York

I read with great interest the retrospective study by Hom etal,¹ in which a model combining clinical metrics with blood-brainbarrier (BBB) permeability data from dynamic CT was presentedto predict symptomatic hemorrhagic transformation (HT) and malignantedema in acute ischemic stroke. The authors stated that HT wasclassified according to the convention of the European CooperativeAcute Stroke Study (ECASS II), of which there are 4 subtypes²: hemorrhagic infarction types 1 and 2 (HI1 and HI2) and parenchymalhematoma types 1 and 2 (PH1 and PH2). It was shown in ECASSII that PH2, defined as a space-occupying hematoma of >30%of the infarct zone with substantial mass effect attributableto the hematoma, is the only subtype of HT that portends poorprognosis.³ HI’s, on the other hand, are defined as petechialhemorrhages without mass effect and are likely epiphenomenal,reflecting reperfusion into the infarct. (For examples of the4 HT subtypes by ECASS criteria, see Berger et al.³)

The authors report that 3 of the 32 patients in their cohort(9.4%) had HT that was classified as PH2, which is unusuallyhigh, especially given that not all patients received tissueplasminogen activator (tPA) (though the exact number of patientswho did was not reported). PH2 is an uncommon event, occurringonly in 1.7% of the 6444 intravenous tPA-treated patients inthe Safe Implementation of Thrombolysis in Stroke-MonitoringStudy⁴ and in 1.9% of the 418 intravenous tPA-treated patientsin ECASS III.⁵ Presumably, the most illustrative of the 3 PH2swas used by the authors for publication in Fig 1 (http://www.ajnr.org/cgi/content/full/32/1/41/F1).However, looking at the follow-up CT at 23 hours postictus,the HT is petechial rather than a space-occupying hematoma,let alone one that would be categorized as PH2. While thereis some mass effect and midline shift, these are to be expected,given the edema from the large infarct, and should not be attributedto the small focus of hemorrhage (red arrow in Fig 1). I believethe more appropriate classification of the HT shown in Fig 1is HI2. Without strict adherence to the established ECASS classificationscheme, the reader is left uncertain as to what the true incidenceof PH2 was in this cohort and whether the technique of CT permeabilityimaging as described can actually predict the only sub-typeof HT that is clinically relevant and most feared when administeringthrombolytic or endovascular therapy.

Since our original report describing quantitative measurementsof BBB permeability from dynamic CT perfusion imaging,⁶severaladditional studies have appeared in the literature.^7–10Imaging protocols (ie, first-pass^6,7,10 versus delayed acquisition^8,9) and methods of modeling the dynamic data (Patlak model,⁸ modified delay-corrected Patlak model,^6,10distributed parametermodel,⁷ adiabatic Johnson-Wilson model⁹) vary considerably.A detailed comparison is beyond the space and scope of thisbrief letter, but a major concern that should be raised is theparadoxical depiction of nonzeropermeability values in normalbrain by Patlak modeling of delayed-acquisition data used inthis study (again see Fig 1, as well as figures and tables fromprevious reports^8,10).

Clearly all of these methodologies require further validation.Despite the optimistic conclusions offered by Hom et al,¹ Ido not believe there are yet any compelling data that demonstratethe ability to predict PH2 unambiguously from other HT subtypesbased on thresholds of permeability values derived from CT.As such, the use of dynamic CT permeability imaging for stroketriage is not warranted at this time, and the results of thisarticle should not alter current clinical practice guidelines.

References

1. Hom J, Dankbaar JW, Soares BP, et al. Blood-brain barrier permeability assessed by perfusion CT predicts symptomatic hemorrhagic transformation and malignant edema in acute ischemic stroke. AJNR Am J Neuroradiol 2011;32:41–48[Abstract/Free Full Text]
2. Larrue V, von Kummer RR, Muller A, et al. Risk factors for severe hemorrhagic transformation in ischemic stroke patients treated with recombinant tissue plasminogen activator: a secondary analysis of the European-Australasian Acute Stroke Study (ECASS II). Stroke 2001;32:438–41[Abstract/Free Full Text]
3. Berger C, Fiorelli M, Steiner T, et al. Hemorrhagic transformation of ischemic brain tissue: asymptomatic or symptomatic? Stroke 2001;32:1330–35[Abstract/Free Full Text]
4. Wahlgren N, Ahmed N, Dávalos A, et al. Thrombolysis with alteplase for acute ischaemic stroke in the Safe Implementation of Thrombolysis in Stroke-Monitoring Study (SITS-MOST): an observational study. Lancet2007;369:275–82[CrossRef][Medline]
5. Hacke W, Kaste M, Bluhmki E, et al. Thrombolysis with alteplase 3 to 4.5 hours after acute ischemic stroke. N Engl J Med 2008;359:1317–29[CrossRef][Medline]
6. Lin K, Kazmi KS, Law M, et al. Measuring elevated microvascular permeability and predicting hemorrhagic transformation in acute ischemic stroke using first-pass dynamic perfusion CT imaging. AJNR Am J Neuroradiol2007;28:1292–98[Abstract/Free Full Text]
7. Bisdas S, Hartel M, Cheong LH, et al. Prediction of subsequent hemorrhage in acute ischemic stroke using permeability CT imaging and a distributed parameter tracer kinetic model. J Neuroradiol 2007;34:101–08[Medline]
8. Dankbaar JW, Hom J, Schneider T, et al. Dynamic perfusion CT assessment of the blood-brain barrier permeability: first pass versus delayed acquisition. AJNR Am J Neuroradiol 2008;29:1671–76[Abstract/Free Full Text]
9. Aviv RI, d’Esterre CD, Murphy BD, et al. Hemorrhagic transformation of ischemic stroke: prediction with CT perfusion. Radiology 2009;250:867–77[Abstract/Free Full Text]
10. Schneider T, Hom J, Bredno J, et al. Delay correction for the assessment of blood-brain barrier permeability using first-pass dynamic perfusion CT. AJNR Am J Neuroradiol 2010 Nov 24. [Epub ahead of print]

Reply

Published ahead of print on April 21, 2011
doi: 10.3174/ajnr.A2629

American Journal of Neuroradiology 32:E125, June-July 2011
© 2011 American Society of Neuroradiology

J. Hom^a
aStanford University
Department of Internal Medicine
Stanford, California

M. Wintermark^b
bUniversity of Virginia
Department of Radiology, Neuroradiology Division
Charlottesville, Virginia

We sincerely appreciate the feedback from Dr Lin and his interestingcomments regarding our article entitled “Blood-Brain BarrierPermeability Assessed by Perfusion CT Predicts Symptomatic HemorrhagicTransformation and Malignant Edema in Acute Ischemic Stroke”published in the January issue of the American Journal of Neuroradiology.¹ In this article, we reported the predictive value of blood-brainbarrier permeability measurements from CT perfusion scanningfor complications observed in patients with acute ischemic stroke.We found that permeability measurements were 100% sensitiveand 79% specific in predicting symptomatic hemorrhagic transformationand malignant edema.

As mentioned above, our primary outcome included symptomatichemorrhagic transformation (not parenchymal hematoma type 2[PH2]) and malignant edema, both of which were defined accordingto the European Cooperative Acute Stroke Study (ECASS III) criteria.² Some of our patients, such as the one shown on the figurein the article discussed here, had a combination of hemorrhagictransformation and edema, both of which contributed to theirclinical deterioration. Of note, symptomatic hemorrhagic transformationas defined by ECASS III criteria does not necessarily coincidewith PH2, even if often these 2 entities overlap.

The rates of symptomatic hemorrhage in patients from the SafeImplementation of Thrombolysis in Stroke-Monitoring Study³ andECASS III² who received alteplase were 2.4% and 1.7%, respectively.In the pooled analysis of the National Institute of NeurologicalDisorders and Stroke (NINDS), Altephase Thrombolysis for AcuteNoninterventional Therapy in Ischemic Stroke (ATLANTIS) andECASS trials, the rate of PH2 was 5.9%.⁴ In our study, 8 patientsof 32 showed a significant clinical deterioration (a NationalInstitutes of Health Stroke Scale [NIHSS] score increase of>4 or death). Three of these 8 (9.3% of our 32 patients)had symptomatic hemorrhagic transformation (of these, 2 receivedintravenous tPA while 1 received intra-arterial tPA plus treatmentwith the Merci retriever [Concentric Medical, Mountain View,California]). Another 3 patients (or 9.3% of our 32 patients)had malignant edema (2 received intravenous tissue plasminogenactivator[tPA], while 1 received intra-arterial tPA plus treatmentwith the Merci retriever). Finally, there were 2 patients inour study who had significant clinical deterioration due toa non-neurologic cause—aspiration pneumonia and septicemia—withoutsymptomatic hemorrhagic transformation or malignant edema. The6 patients with a significant clinical deterioration (an NIHSSscore increase of >4 or death) and with symptomatic hemorrhagictransformation or malignant edema were included in those withpositive outcomes. Our 9.2% is just above the superior limitof the confidence interval for the rate observed in the pooledanalysis of the NINDS, ATLANTIS, and ECASS trials.⁴

Regarding the nonzero permeability values in the normal brainparenchyma, one has to remember a very important element. Ourmodel of blood-brain barrier permeability, as all models ofbrain perfusion and all models in general, is not the truth,but only a simplification of the truth. Indeed, the permeabilityof the blood-brain barrier is a complex phenomenon that involvesselective passage of certain molecules in a specific amountand nonpassage or limited passage of other molecules. The conceptof modeling the blood-brain barrier permeability is meant tosimplify it to a numeric value that can be used to make clinicaldecisions. This simplification is justified as long as 1) themodel is an accurate simplification, and 2) the model is clinicallyrelevant. From this perspective, a recent study that comparesour model of blood-brain barrier to histologic findings of permeabilityin rats (the criterion standard) addressed issue number 1 andindicates that our model accurately depicts global changes ofthe blood-brain barrier permeability in the setting of brainischemia.⁵ Our current study addresses issue number 2 and suggeststhat our model and approach are clinically relevant.

We completely agree with Dr Lin that further confirmation ofour findings in a large prospective study is required beforeany change to the current clinical practice guidelines is considered.

References

1. Hom J, Dankbaar JW, Soares BP, et al. Blood-brain barrier permeability assessed by perfusion CT predicts symptomatic hemorrhagic transformation and malignant edema in acute ischemic stroke. AJNR Am J Neuroradiol 2011;32:41–48[Abstract/Free Full Text]
2. Hacke W, Kaste M, Bluhmki E, et al. Thrombolysis with alteplase 3 to 4.5 hours after acute ischemic stroke. N Engl J Med 2008;359:1317–29[CrossRef][Medline]
3. Wahlgren N, Ahmed N, Dávalos A, et al. Thrombolysis with alteplase for acute ischaemic stroke in the Safe Implementation of Thrombolysis in Stroke-Monitoring Study (SITS-MOST): an observational study. Lancet2007;369:275–82[CrossRef][Medline]
4. Hacke W, Donnan G, Fieschi C, et al for the ATLANTIS Trials Investigators; ECASS Trials Investigators; NINDS rt-PA Study Group Investigators. Association of outcome with early stroke treatment: pooled analysis of ATLANTIS, ECASS, and NINDS rt-PA stroke trials. Lancet 2004;363:768–74[CrossRef][Medline]
5. Hoffmann A, Bredno J, Wendland MF, et al. Validation of in vivo MRI blood-brain barrier permeability measurements by comparison with gold standard histology. Stroke. In press