Author's response to reviews Title: MRI-negative PET-positive Temporal Lobe Epilepsy (TLE) and Mesial TLE differ with Quantitative MRI and PET: a case control study Authors: Ross P Carne (carnero@svhm.org.au) Terence J O'Brien (obrientj@unimelb.edu.au) Christine J Kilpatrick (Christine.Kilpatrick@mh.org.au) Lachlan R MacGregor (Lachlan.Macgregor@mh.org.au) Lucas Litewka (Lucas.Litewka@svhm.org.au) Rodney J Hicks (Rod.Hicks@petermac.org) Mark J Cook (Mark.COOK@svhm.org.au) Version: 2 Date: 11 April 2007 see over Author's response to reviews:
Sandra Le, Ph.D. Assistant Editor BMC-series journals BioMed Central Ltd Middlesex House 34-42 Cleveland Street London W1T 4LB UK 11/04/07 Dear Dr Le, Re: MS: 1163920501258368 MRI-negative PET-positive Temporal Lobe Epilepsy (TLE) and Mesial TLE differ with Quantitative MRI and PET: a case control study Ross P Carne, Terence J O'Brien, Christine J Kilpatrick, Lachlan R MacGregor, Lucas Litewka, Rodney J Hicks and Mark J Cook Thank you for the reviewer s constructive criticisms and suggestions regarding the above paper. The authors appreciate Reviewer MV s comments and understanding that this work has been laborious, with a clear protocol and followed strictly, and the recommendation that the article should be accepted without revision. Reviewer PD made several comments and suggestions and in response major and minor revisions to the paper have been undertaken. I will comment on the reviewer s comments presented in italics - in turn: In this manuscript, the authors describe the comparison of temporal lobe epilepsy with hippocampal sclerosis (HS+ve) and non lesional temporal lobe epilepsy without hippocampal sclerosis (HS-ve) using quantitative MRI and FDG PET. This is an interesting issue which is already partially described in two previous publications by the same group and using the same patient data. In its current form the added value of this manuscript is limited but it can be improved by some additional analyses. The authors feel that this paper adds significantly to the previously published data. Neither of the two previously published papers (Carne R, et al. MRI-negative PETpositive temporal lobe epilepsy: a distinct surgically remediable syndrome. Brain 2004;127:2276-2285 & Carne RP, et al. MRI negative PET positive temporal lobe epilepsy (TLE): FDG-PET pattern differs from mesial TLE. Molecular Imaging and Biology 2007 9(1):32-42) dealt with quantitative measures of MRI beyond the hippocampal volumetry, which was one of the criteria used to define the two study groups. The quantitative cortical volumes and metabolism data presented in the current manuscript for these two groups of patients are unique in the literature. The primary hypothesis was that the epileptogenic focus in HS-ve involves primarily lateral rather than mesial temporal structures, and that the quantitative structural
and functional changes would reflect this. However, the volumetric analyses using quantitative MRI is too limited. The hippocampus, whole brain, hemispheric and lobar volumes were used but additionally, a voxel based morphometry study would be in place (and if possible including a group of normal matched controls). Indeed, the authors conclude that HS+ve patients showed more hippocampal, but also marginally more ipsilateral cerebral and cerebrocortical atrophy but this is based on small differences in volumes in large parts of the brain. In this respect, the primary hypothesis was not investigated fully. The voxel based morphometry study suggested by Reviewer PD is an interesting project and one that we had been considering performing - we note the Reviewer's own excellent similar study in the HS+ve group (Nelissen et al, 2006). However on balance we feel that, if performed, this should be presented in a separate report. The measurements performed in our studies, particularly of cortical lobar volumes, are laborious, as recognised by the Reviewer MV, and are unlikely to be repeated. The volumes reported, given the clear definition of protocol, could be readily verified and compared by other groups. The use of automated analyses such as voxel based morphometry while appealing, is not without its limitations, and its addition in our study would require further protocol definition, and presentation of pictorial and tabular results, adding to an already sizeable manuscript. Overall, with respect, the authors feel that the volume measurements as presented provide novel and important data, and a voxel based morphometry analysis of our data set, if performed, should constitute a separate paper. Also the use of FDG PET is adding little additional value compared to what is reported in one of the previous publications of the authors. In this manuscript, the authors are discussing the asymmetry (instead of the actual uptake) in FDG metabolism (again in larger structures) and given the fact that morphological changes are reported, the differences in metabolism might be due to partial volume effects (see also a recent study by our group; Nelissen et al Neuroimage 32 (2006), 684-695). Our previous publication on FDG PET reported a voxel based SPM analysis, and offered data complementary to the ROI based FDG-PET asymmetry indices presented in the current manuscript. The former type of analysis has the advantage of not being hypothesis driven with the potential to identify changes in the whole brain; however because of the multiple testing performed and the use of population controls this method has a lower sensitivity to detect changes in specific regions. This is the advantage of hypothesis driven ROI analysis as presented in this current paper. The authors also believe that it is unlikely that partial volume effects have played a significant role in the reported results, given the marked metabolic (FDG PET) asymmetries seen in structures shown to be volumetrically much less asymmetrical: for instance the ipsilateral cortical hypometabolism which is more marked in the HSve group, despite the fact that these is less asymmetry or ipsilateral cortical atrophy in this group. This is the opposite of what would be expected if partial volume effect were to explain the results. The arguments we have put forward in the manuscript (pp 13-14) regarding the lack of statistical correlation between neuronal loss and degree of metabolic asymmetry could equally be used to refute the argument that the observed metabolic asymmetries relate to partial voluming. The discussion in the manuscript has been clarified to read (p21):
"In our study, the finding of significant mesial hypometabolism in the MRI negative (HS-ve) group, despite volumetrically normal hippocampi, is evidence against the contention that volume loss and hypometabolism are significantly correlated. It also argues against the contention that partial volume effect may have significantly affected the results..." and the study by Nelissen et al cited (p21): "Quantitative FDG-PET uptake studies have shown the most marked deficits when comparing HS+ve TLE patients to controls bilaterally in frontal and parietal lobes, in comparison to the more marked abnormalities detected in temporal lobes with asymmetry indices [39]." Table 2: add the standard deviation. This has now been added as suggested by the reviewer. I would also add a graph indicating all individual ratios for the three areas since I find it difficult to understand that an average chance of e.g. 2% is highly significant (cerebrum, ratio 0.98, p <0.001) given the variability within humans. This graph has now been added as suggested by the reviewer (figure 3). The p-value quoted relates to side-to-side comparisons within individuals across the group, then compared to the age and sex matched control, rather than being a population comparison of ipsilateral to contralateral structures. Use of the contralateral structure as the internal control allows comparisons with more statistical power than would be seen with a population comparison. Age and sex account for most of the variance seen within the population (see our previous study: Carne RP, Vogrin S, Litewka L, Cook MJ: Cerebral cortex: an MRI based study of volume and variance with age and sex. J Clin Neurosci 2006, 13: 60-72). This is now mentioned in the Key to Table 2: Key Table 2. The p-value quoted relates to side-to-side comparisons within individuals across the group. The between group comparison relates to comparison of the case to the age and sex matched control, performed across the group, rather than being a population comparison of ipsilateral to contralateral mean volumes. These data are presented graphically in Figure 3. Figure 3 gives a graphical representation of the data: Figure 3. MRI volumetric comparisons: HS-ve versus HS+ve. Case-control pairs are presented in order of ascending age. Ratios presented relate to epileptogenic/contralateral MRI volumes: a. hippocampus, b. cerebrum and c. cerebral cortex. Mean and standard deviation figures are given in Table 2.
Table 5: I don't understand why for the HS+ve only the data of 24 instead of 27 cases are reported and similarly for the 27 instead of 30 HS-ve cases. I guess that the three cases (and their case-control counterparts in the HS+ve group) in which visually no lateralised hypometabolism was found are left out. If this is the case, the exclusion of these three PET negative cases should be made in the beginning (and these should also not be included for the MRI analysis since the title suggests that the patient population is 'MRI negative PET positive'. This has now been addressed in the text (p 10) "[18F] FDG-PET scans were available for analysis in 30 HS-ve patients and 27 HS+ve patients. Three of the 27 FDG PET scans in the HS+ve group were technically suboptimal, and did not allow accurate coregistration to be performed. These three HS+ve FDG PET scans all showed lateralised ipsilateral temporal hypometabolism, and had been included in the previously published study reporting visual analyses (Carne et al. 2004). Coregistered FDG-PET analyses of these HS+ve controls and of the FDG PET scans of their matched MRI negative PET positive counterpart cases were not included in this analysis, allowing 27 HS-ve and 24 HS+ve patients to be analysed." Page 9: The PET acquisition was 30-40 minutes but nowhere it is mentioned at what time post injection the acquisition started. Also the details of the reconstruction should be mentioned as well as the type of corrections performed. The methods section has been adjusted to read: (p9) The dose of FDG administered was 37-111 MBq (1-3 mci). One bed position was used. All scans were commenced 45 minutes after injection of FDG with the patient lying comfortably in a darkened room during the uptake period. The scan acquisition time was 30-40 minutes to achieve comparable statistical quality between scans. The duration of acquisition was based on count rates measured at the commencement of imaging, achieving total counts of >40 million. An empiric attenuation correction (ellipse) was applied. The data were processed using a Wiener prefilter (scaling value = 0.5) and ordered-subsets expectation maximization (OSEM) iterative reconstruction. OSEM provides significant resolution recovery and minimizes partial volume effects. Wiener filtering attempts to reduce blurring of an object by restoring the amplitude of the object's power spectrum in certain ranges. In particular, the filter identifies frequencies that define the resolving power of the spectrum Page 9: how was corrected for the FDG dose injected? Correction for the dose injected was not done. This was not needed as absolute quantification of FDG hypometabolism was not performed. All PET analyses were of relative asymmetry in structures of interest between sides, and therefore normalisation for the overall intensity of the images was not necessary. Page 10: the conversion of MRI and FDG PET to binary volumes was performed in order to coregister the two data sets. If this step is done properly, the coregistration is working well but the authors should give more details about this binarisation itself.
The following has been added (p10): Coregistration was performed using a surface matching technique, involving conversion of MRI and FDG-PET to binary volumes and subsequent matching of surface points, using the AnalyzeAVW software package (Mayo Clinic). Binary images representing the cerebral surface for each scan were obtained by extracting the brain from extracerebral structures on MRI, as described previously [12], and thresholding out the extracerebral uptake on FDG-PET. Interior holes were then deleted, to give two binary single objects representing brain surface as measured using each modality. Page 11: the authors are using statistical tests without correction for multiple comparisons. In the discussion section (p19) they give as argument that in the situation of an a priori hypothesis it is appropriate not to correct for the multiple comparisons. I agree with this argument if you test only the a priori hypothesis but the authors look at more than this hypothesis only and therefore, a combination is the best. Correction for multiple comparison should be done except when directly testing the hypothesis. While there is some speculation in the discussion regarding the possible meaning of the results that strays beyond the primary hypothesis -that the epileptogenic focus in HS-ve involves primarily lateral rather than mesial temporal structures, and that the quantitative structural and functional changes would reflect this-, all of the data presented relate either to structural or metabolic asymmetry and deal directly with this primary hypothesis. This has now been stated in the discussion (p 19): In this situation it is appropriate, however, to interpret data based on the primary hypothesis, in this case that the epileptogenic focus in HS-ve patients involves primarily lateral rather than mesial temporal structures, and that the quantitative structural and functional changes would reflect this: that ipsilateral atrophy and hypometabolism would be present, and that the extent of hypometabolism would be greater in HS-ve, where the focus was posited to be lateral rather than mesial. The data presented relate either to structural or metabolic asymmetry and deal directly with this primary hypothesis... Figure 1: the caption should be extended, labels should be given to each panel and colour scale representing the FDG uptake should be included This has been done as suggested by the reviewer. The new figure key reads: Figure 1. Volumetric MRI and coregistered FDG-PET: (a) MRI Extracted whole brain is segmented into (b) MRI right and left cerebral hemispheres, in turn thresholded into (c) MRI right and left cerebral cortices. Coregistration of the (d) FDG-PET whole brain allows accurate anatomical definition on the metabolic studies of, for example, (e) FDG-PET right and left cerebral hemispheres and (f) FDG-PET right and left cerebral cortices. Figure 2: can be omitted This has been omitted as suggested by the reviewer.
Figure 3: can be omitted This has been omitted as suggested by the reviewer. Figure 4: I have the impression that there is a mismatch between the hippocampal region and the segmented GM. How will this affect the results? It would be helpful to indicate the orientation of the figure for readers less familiar with this type of analysis/figures. There is a minor mismatch between the superimposition of hippocampus as defined by drawing on the T1 volumetric MRI, the segmented grey matter and the coregistered FDG PET. The image was included to show how well these various measures correlate. Any bias introduced should be systematic. The orientation has now been indicated as left temporal coronal section. Table 3: I notice that the lobar volume of the contralateral temporal and insular cortex is larger in HS+ve patients compared to HS-ve patients. Are the authors willing to speculate that there might be a some sort of compensatory mechanism taking place because of HS? This is an interesting suggestion of the reviewer, but at this stage we believe that this speculation would be extending the interpretation of the data too far. A line clarifying that patient consent was obtained has been added as requested (p 8): The study was approved by the institutional ethics committees of The Royal Melbourne Hospital, St. Vincent s Hospital and The Peter MacCallum Cancer Institute. All patients consented to MRI and PET scans as part of the study. I thank both reviewers again for their thoughtful and considered responses. We feel that this paper is an article of importance in Neurology and Epileptology, and that the data presented have been carefully obtained and analysed. I look forward to your response. Yours Sincerely, Dr Ross Carne for the Authors.