Occlusion for an implant supported prosthesis?

Thats not a short answer by any means - you will probably find this turning into a general occlusion chat centred on recording CR more so than anything to do with implants! Ans thats a prosthos opinion!

In general forces that are directed in line with the long axis of the implant will be best resisted. Canine guidance on a free standing implant would not be a good thing in most cases. Group function would be better than canine guidance. Implants are not teeth and the occlusion for implants is not the same as it is for natural teeth. Multiple implants splinted will be more resistant to lateral forces than multiple non spited implants. Implant retained overdentures that are supported by the implants with little soft tissue support are very different as far as forces applied to the implants than a tissue supported overdenture where the implants mostly resist displacement of the denture rather than resisting occusal forces. There is no absolute rule here and analysis of the forces that the implants will be subject to has to be taken into consideration in planning the occlusion.

What factors go into establishing patient risk factors for overload failure; in this case dental implant supported prosthesis ?

Just speaking about single posterior implants. It is good to have a narrower occlusal table to reduce load. Centric occlusion with no excursive contacts.

For posterior fixed partial dentures it would be best to have only MI (centric) contacts and no excursive contacts. This is assuming that there is already canine rise on a natural tooth in excursive movements. If the FPD is in the anterior there should be contact on all four incisors natural or implant supported in protrusive. Arguably the contact on the implant supported crowns might be a little lighter than the natural teeth. There should only be contact on the canines(natural teeth) or possibly first Bi in right and left lateral movement. As previously mentioned if the implant is a canine then I would suggest group function

I am a general dental practitioner placing implants in many and varied situations over the last 30 years and restoring my own implant placements. I have recently been reviewing the literature generally on occlusion and find that there is indeed no evidence based formula for any occlusal system for any of the situations you want answers for. There are a number of items to consider however. Bone likes a little action, not too much and not too little. Don't put your implant crown in heavy load but make it function under stress. It seems single teeth on implants need occlusion in centric such that the occlusal table is narrow with central fossa contact as much in line with the implant and a relatively flat cusp plane, such that the restored tooth is not in heavy function through the lateral excursion but can come into function (shim contact) under heavy load. Pay attention to the natural CR and lateral excursions and any parafunction. It seems that most natural canine guided occlusion changes to group function with age. Watch out for Class 2 and bruxists. It also seems that any form of occlusal system for full arch replacements with implants can be acceptable, but you need to take into account the quality and strength of the restoration as there is insignificant proprioception to the neuromusculoskeletal input for implants. So possibly group function with freedom and balance is the way to go for older patients. I can't yet determine if age and type of system is an issue with full implant supported bridges. I have also determined that the masticatory system and teeth guidance of any form is very adaptable in all situations, but don't overload it unnecessarily. Possibly the least overloaded design is canine guidance but can the prosthesis withstand the unnatural forces. We know the bone in the golden triangle (Canine area) is very strong but up until now where we now have monolithic materials it wasn't uncommon for porcelain shearing and delamination of acrylic etc. Experience is valuable! This is a huge topic and I have only touched the surface. Keep on keeping on as the literature is extremely short on good formulae.