For early stage NSCLC to be treated with SBRT, is there any difference in the technology being used such as robotic v. linac based VMAT?

In my opinion, as long as the SBRT team can follow the principles and safety measures, there shouldn't be any difference in treatment outcomes and toxicities among different treatment devices for early stage NSCLC. In other words, the expertise of the team is much more important than the treatment device used. In our center, we have both LINAC-based and robotic SBRT. Having said that, it is very important to note that when one is planning robotic-based SBRT for thoracic tumors, Monte Carlo calculation should be used instead of Ray Tracing algorithm (still very widely used) as the latter will overestimate the dose around the PTV resulting in underdosing. There have been a few papers published addressing that very issue:

Sharma et al. J Applied Clin Med Phys 2010

Ove et al. J Radiation Oncology 2015

For very posteriorly located lung tumors, LINAC-based SBRT may be preferred from the dosimetric standpoint because the robotic radiosurgery system cannot direct beams from the back (all beams from the front and the sides), resulting in more low to intermediate dose spillage to the ipsilateral lung.

For LINAC-based SBRT, use RTOG approved algorithms (rpc.mdanderson.org). Stay away from PBC.

All-in-all, no matter what machine you have, you can still deliver excellent SBRT.

I respectfully disagree with @Dr. First Last.  In our experience, we can acheive much tighter plans using Cyberknife than with Linac based radiosurgery, probably due to the much higher number of beam angles being used and the use of real-time fiducial tracking rather than a 4-D integrated tumor volume.  The lower lung doses may not be clinically relevant in a majority of cases, but the dosimetry is clearly superior. 

I agree that there are certain aspects that robotic SBRT is better compared to LINAC-based SBRT for lung cancer. One drawback of the robotic radiosurgery system is the lack of volumetric verification prior to treatment. Occasionally, there is tumor growth seen in merely 2 weeks between CT simulation and start of SBRT. We came across cases where the fiducial marker positions changed during the set-up and repeat CT showed tumor growth in 2 weeks. When reviewing the medical literature on SBRT for stage I NSCLC, the results (LC and toxicities) are comparable between robotic and LINAC-based SBRT for NSCLC. Ultimately, these are the most important endpoints. We have both (used to have TomoTherapy as well) and haven't noticed any difference.

Because of our experience with tumor growth between CT simulation and start of SBRT, if we use a robotic system, we always do a CT chest on the same day as treatment to ascertain there is no tumor growth (most of the time, there isn't but we have seen a few cases of progression).

And we use Monte Carlo for all SBRT cases if the tumor is in the chest or located close to any air cavities (such as paranasal sinus). We have proved that if Ray Tracing is used for thoracic spinal tumors, we will underdose the PTV and may overdose the spinal cord (Okoye et al. TCRT 2015).