plenty of literature favouring both approaches :
http://www.ncbi.nlm.nih.gov/pubmed/18771811
http://www.ncbi.nlm.nih.gov/pubmed/19792965
http://www.scribd.com/full/46996310?access_key=key-1vz285e1aqipo48homh2 ( MRI vs PET , not USPIO)
both are aggressively advocated , sensitivity and specificity cannot be compared objectively across different patient databases
members' comments and any head on comparisons please
To be honest, I am not sure about any head on comparison. I too have been thinking of this but primarily it would need two major issues to be debated out:
1) The impact of any new imaging modality, how it would lead to upstaging or understaging compared to pathological gold standard.
2) Ensuring complete coverage of the the nodes and followed by para-aortic irradiation and hence leading to better overall survival outcomes with minimal morbidity.
However, what is definitely known is that with increasing depth of invasion, there is incremental increase in the pelvic lymph nodal rate and about 15% patients get upstaged because of para-aortic involvement. Should this cohort be treated prophylactically with extended field radiation?
However, even the BIGGER issue is that of negative nodes because of necrosis. Are necrosed nodes viable or unviable? Should they be targeted in absence of FDG uptake? Hence the differentiating feature (or recommendation of one technique over other ) should come from how well (and accurately) it is able to exclude negative nodes as free of disease rather than showing positivity.
This is my opinion though.
Reality is merely a persistent illusion
thanks ayan for the tedious head to head comparison paper.
I agree with Abhishek's concerns, but in wat %age of Ca Cervix patients in our routine clinical practice do we go in for pathological gold standard staging, n hence we do hav to rely on such non invasive staging procedures.
Similar issues propped up in TMH PET-CT vs CT/MRI trial in Early stage Carcinoma Cervix.
A considerable number of patients had positive nodes on PET CT even in early stage Ca Cervix in the trial; does anyone has any further information on the trial?
Considering higher incidence and variety of infections in our population, can we
1) actually rely on the existing sensitivity and specificity values?
2) justify modifying the existing staging w/up and treatment protocols?
3) blindfold ourselves to the upcoming developments till they r declared standard of care?
| VARIABLE | PET CT | MRI |
| Sensitivity | 72% | 82% |
| Specificity | 97% | 93% |
| PPV | 92% | 69% |
| NPV | 89% | 96% |
There is another big concern for cervix patients specially in India : what about imaging for brachy. If there is a single imaging modality that gives all the info for brachy its MRI. As you all know the MRI is required before EBRT and before each brachy session — so how many imaging sessions a person goes through?
very valid points raised by Santam , but just for the sake of academic discussion , let us leave cost out of it for a minute—anyways cost has to be weighed against under-treating an under-staged patient considering the cost paid by the patient is measured in terms of poorer treatment outcome .
And how true , here is an ongoing(?) study:
http://www.pharmaceuticalonline.com/article.mvc/Cervical-Cancer-Treatment-PETCT-USPIO-MRI-0002
could not find the results though, so i presume the study is still recruiting , or closed for poor accrual
any others ?
Positive nodes showing uptake does not always mean malignancy and negative nodes does not exclude malignancy!!!
This is the biggest fallacy of PET-CT and I have faced this situation often; I think it has been mentioned somewhere in RSNA protocol that PET positive nodes NEED TO BE proven as "malignant" before the change of treatment plan is contemplated.
Another issue:
How do you prove it is malignant? By using directed FNAC? The chances of "positive yield" is still around 30-40% (again dependent on the skill). Do you make the patient undergo staging laprotomy? In absence of overtly enlarged nodes, it may still be prone for a major sampling error unless the vessels are "skeletonised".
Still then, the major issue and stumble is at the grossing. A resident not attuned to the idea of looking for the nodes would give up the "fat" and then you need thin sections (I think around 10 microns or so) to be able to identify any presence of disease. Although there are investigators who unless use a fancy PCR array to detect malignancy and we are back to square one.
The big question. Is it worth it?
Hence before it can be recommended as part of the staging work up, I feel that it's pathological correlation is mandatory.
@Santam, I would argue against the study (in Science Direct). I don't have access to the complete paper but prostate can be dicey. One major issue is the risk of skip metastasis (increases with Gleason's score) and the fact that nomograms grossly understage the disease. I would not recommend a surgery upfront (in a prostate specially when we have state of the art radiation); but if the patient is undergoing surgery, he needs a thorough PLND with at least a median of 25-30 nodes identified for complete staging.
Reality is merely a persistent illusion
Hi a very intensive debate.
Find the full text here
The question asked in the prostate study was a simple and fundamental one - how well can MRL (MR Lymphangiography) pick up nodes - in other words can it do it so well that we dont need to do a dissection in order to confirm it. Even considering the variability of natural history of prostate and cervix this is the fundamental question we need to ask before putting any investigation as a staging investigation. EL prior to treatment is not possible in each patient and when we are devising investigations for staging even FIGO the so called surgeon dominated body understands that. That is why even till date its a clinical staging mainly.
The answer from this study is important the MRL has a very high negative predictive factor. The post test probablity of a false negative result is only 4%. How well does MDCT score ? - 12% Which means that MDCT actually will accurately predict that you dont need to do the surgery (ipso facto any irradiation either) in 88% of the patients and MRL in 96% if the test results come as negative. In cervix things are easier, you will be treating almost all patients for the whole pelvis starting form stage I and paraortic nodes if positive can be irradiated as EFRT. Will we see a shift in the pattern of irradiation from WPRT to local RT alone based on negative MRL results for pelvic nodes - time will tell.
The biggest advantage the MRL has over PET is that it is not affected by inflammation as inflammed nodes filled with macrophages will come as black (negative) on MRL but can be picked up as positive in PETCT. MRL has no additional Radiation dose exposure and can be used along with other sequences for brachy related imaging also. PET has no such advantages.
Given all this where do we stand? All imaging modalities MDCT/PET/MRL/MRI have size of the node as an important criteria. For nodes above 1 cm the decision is simple - its involved unless otherwise proven for the Grigsby protocol or the MRL protocol. For nodes less than that CT fails, PET works by picking up activity if there is enough, MRL picks up disturbances in nodal architecture. A planning CT can therefore tell you what the nodal involvement is in more than 80 - 88% of the patients reliably based on the size only if negative. If you have a node positive target them but dont upstage them - FIGO doesnot give you any recommendations on treatment. Giving a higher dose to the node is possible today easily with IMRT. And even extended fields are tolerated reasonably well. For an early stage cervix go ahead get a surgical colleague involved and do a dissection if required.
Here are the recent ESMO recommendations http://annonc.oxfordjournals.org/content/21/suppl_5/v37.full
If any member wants the ACRIN protocol send an blank email to contact.isocentre at gmail.com with the words ACRIN in the subject line. Please make sure you type ACRIN as ACRIN not acrin or Acrin.
