Predictive value -- Cancer Trials
Pepmoron:
(talking about MYG and TLN) "As for outcome for both of them. Impossible to predict."
Another humilation... LOL.
Truly moronic to imply that in vivo, in vitro trials have no value at all.
In fact, accoring this article, 70 ot 90% correlation:
https://pediatricca.asco.org/portal/site/ASCO/menuitem.34d60f5624ba07fd506fe310ee37a01d/?vgnextoid=76f8201eb61a7010VgnVCM100000ed730ad1RCRD&vmview=abst_detail_view&confID=16&index=y&abstractID=360
This study examined the clinical predictive value of threepre-clinical laboratory cancer models. A group of 31 cancer drugs withknown Phase II clinical trial and pre-clinical performance in breast,non-small-cell lung (NSCLC), ovary and colon cancers was selected. Theliterature (Medline and CancerLit) was searched for Phase II trial,human xenograft and mouse allograft publications. For each drug,clinical activity was scored as the mean response rate across all PhaseII trials of the same tumour type and pre-clinical activity as the meanT/C% across all xenografts/allografts of the same disease site.Publicly available (Internet) "mean graph" data from the NCI HumanTumor Cell Line Screen were used to calculate two measures of drug in vitropre-clinical activity for each cancer type: the mean LogGI50 and ameasure we termed the "Activity Fraction". Phase II activity vs.pre-clinical activity correlation analysis was conducted for eachmodel, by tumor type (disease-oriented approach), from the perspectiveof one tumour type for the other three (compound-oriented approach) andfor all four tumour types together. The results showed that, when themean LogGI50 measure of pre-clinical activity was used, the in vitrocell line model was predictive of the Phase II performance of cytotoxiccancer drugs in all cases except colon cancer in the disease-orientedcontext. In contrast, the mouse allograft model was not predictive. Thehuman xenograft model was generally predictive for NSCLC and ovarian,but not for breast or colon cancers. Interestingly, this modelpredicted clinical performance much better when panels of xenograftsrather than a single xenograft were used. Among different drugs, theNSCLC and ovarian human xenograft panels that predicted Phase IIperformance had similar sub-histology frequency constitutions for eachdisease site. These results suggest that, under the right framework,the in vitro cell line and human xenograft models can be useful inpredicting the Phase II clinical trial performance of cytotoxic drugs.
Powerpoint pres (full data)
https://media.asco.org/asco/meetings_education/2002posters/Abstract360/poster.htm?XZ=1&mediaURL=/media&ServerName=media.asco.org