epilogue on the
lung ENI...
Less is more in radiotherapy target volume planning: lessons from the PET-plan trial
...target volume reduction based on 18F-FDG PET alone was not inferior to traditional large volume radiotherapy with elective nodal irradiation with regard to locoregional control, with a suggestion of improvement in locoregional control; allowed further escalation of the radiation dose; and was not associated with a substantial reduction or increase in toxicity... In view of the result of this study, target volume reduction based on 18F-FDG PET alone could be considered standard of care. International guidelines already recommend the omission of elective nodal irradiation guided by 18F-FDG PET/CT in locally advanced non-small-cell lung cancer, making this study practice confirming.
The PET-plan trial is an interesting trial, but I do have some ... issues with its interpretation.
First of all, let's have a look at one of its results:
What we have here is a almost 20% higher risk of locoregional progression in patients who received ENI (conventional target group).
Now this may have happened because:
a) the patients in the non-ENI (FDG-PET-based) group received higher doses?
Perhaps. But the dose difference was merely 2 Gy
"The mean escalated total radiotherapy reference dose was significantly higher in the ¹⁸F-FDG PET-based target group (67·3 Gy [SD 5·2]) than in the conventional target group (65·3 Gy [5·3]; appendix p 7)."
2 Gy more of RT won't result in a 20% absolute reduction of locoregional failures.
b) adherence to treatment may have been better in the non-ENI group?
Perhaps, but doesn't seem so.
"Eight (8%) of 99 patients in the conventional target group and ten (10%) of 105 patients in the ¹⁸F-FDG PET-based target group had relevant dose reductions in radiotherapy (seven vs five) or chemotherapy (three vs eight). Two patients (both died) in the conventional target group and four (one died, two had oesophagitis, and one had pneumonitis) patients in the ¹⁸F-FDG PET-based target group discontinued treatment for toxicity. 20 deaths potentially related to study treatment were reported by treating physicians (seven in the conventional target group and 13 in the ¹⁸F-FDG PET-based target group)"
c) some immuno-related event happening because the non-affected nodes were not irradiated in the non-ENI group.
Perhaps, but none of them got adjuvant immunotherapy.
What do I think?
There is a (suspicious) discrepancy in the reported values of GTV/PTV
Patients in the non-ENI group seem to have had bigger tumors in terms of volume (GTV). Although these patients did not receive ENI, the resulting PTV was as big as in the ENI group. This makes little sense... 15ml of "extra" GTV do not justify a PTV that's as big as if you were giving ENI. And in fact, this brings up the question, why patients in the non-ENI group were dose escalated more often than in the ENI-group, when their PTVs were actually as big (or slightly bigger)?
"The mean escalated total radiotherapy reference dose was significantly higher in the ¹⁸F-FDG PET-based target group (67·3 Gy [SD 5·2]) than in the conventional target group (65·3 Gy [5·3]; appendix p 7)."
What is even more striking is the fact that in the non-ENI group the GTV of the primary was drawn based on FDG-PET, while in the ENI group the GTV of the primary included atelectasis too. Yet, in the non-ENI groupt the GTVs of the primaries were bigger... Did FDG-PET-based drawing of the primary result in bigger GTVs or was the randomization so flawed?
It simply makes no sense...
What I suspect:
Physicians were not blinded (at least I couldn't find this being described in the publication/protocol) as to which arm the patients were in, when they contoured the volumes and accepted plans. This could have been different. It would have meant they would have to draw two sets of CTV/PTV for each patient and calculate two plans according protocol but then only one set would be used for the actual treatment according to randomization. Extra burden, but manageable IMHO and would eliminate all bias.
This shortcoming may have led to:
a) tighter margins around CTV in the ENI-group and perhaps "less than actually per protocol supposed" elective nodes and tissue around primary tumor contoured (--> smaller than supposed CTVs in the ENI group). This would explain, why the PTVs remained small, although ENI was used.
b) less likely escalation in the non-ENI group decided by the physicians, knowing that the patients where in the ENI group out of fear of toxicities.
The discrepancy becomes even bigger, if you look at the per-protocol analysis...
The authors did describe an extensive quality assurance project, but not many details were revealed on what came out of it.
Thoughts on PET-Plan?
P.S. ENI is dead, don't get me wrong.