Can antimatter be used as fuel for nuclear reactors? I completely understand the difficulties of making and storing antimatter, so I am not talking about the mechanism or the way of doing it here, I am just talking about the concept.
As far as I know, nuclear power plants use the heat from the nuclear fission reaction to heat water and use the steam through turbines and generators to generate electricity. So, if we could somehow use the annihilation of matter-antimatter inside a reactor, would it still be a viable way of generating heat and thus electricity ? or is there something special for nuclear fission that is not available for matter-antimatter annihilation ?
 A: Short answer: Yes, it can. Although for near-future application the utility of antimatter would be not as a fuel per se but as a catalyst of nuclear reaction.
The energy density of proton antiproton annihilation is $1.8\times 10^{14}\text{J}/\text{g}$ of antiproton is hundreds times that of fusion or fission reactions.
One field where antimatter could be of use is space where enormous cost of its production is offset by the small mass of the product and relative small size of devices utilizing this energy. Therefore most of concepts for utilizing such energy (at least in the context of near future technology) is for propulsion purposes.
The reaction $\bar{p} + p$ produces mainly $\pi^{+}$, $\pi^{-}$, $\pi^{0}$ mesons, so about 2/3 of reaction energy is available as charged light energetic particles which could rapidly heat up matter or/and initiate other nuclear reactions (both fission and fusion). This would allow to derive most of the energy from such reaction thus reducing antimatter requirements and at the same time maintaining small size (usually much smaller than full scale conventional fusion or fission reactors). Some of the concepts mentioned in this review:
Antimatter-Catalyzed Micro-Fission/Fusion (ACMF): 

Here, a pellet of D-T and U-238 is compressed with particle beams and irradiated with a 
  low intensity beam of antiprotons. The antiprotons are readily absorbed by the U-238 and 
  initiate a hyper-neutronic fission process that rapidly heats and ignites the D-T core. 
  The heated fission and fusion products expand to produce thrust ... Gaidos et al. 7 have 
  shown that the interaction between the antiproton beam and target exhibits extremely 
  high-gain yielding a ratio of fusion energy to antimatter rest mass energy $\beta$ of 
  $ 1.6 \times 10^7$ ... Assuming a 3-order of magnitude improvement in the efficiency of 
  producing antiprotons  over current values, the net energy gain is 640.

Antimatter-Initiated Microfusion (AIM) 

Here, an antiproton plasma within a special Penning trap is repetitively compressed via 
  combined electric and magnetic fields. Droplets containing D-T or D-He3 mixed with a 
  small concentration of a metal, such as Pb-208 or U-238, are synchronously injected into 
  the plasma. The main mechanism for heating the liquid droplet is antimatter-induced 
  fission fragments which have a range of 45 microns ($\mu$m) in the droplet. The power 
  density released by the fission fragments into the D-T or D-He3 is about 
  $5 \times 10^{13}$ W/cm$^3$, which is enough to completely ionize and heat the fuel atoms
  to fusion ignition. The heated products are directed out magnetic field lines to produce 
  thrust. The $I_{sp}$ and energy efficiency for this concept are higher than ACMF 
  ($I_{sp} \sim 67,000$ sec and $\eta_e \sim 84\% $ with D-He3, and $I_{sp} \sim 61,000$ 
  sec and $\eta_e \sim 69\% $ with D-T). The gains $\beta$ are $10^5$ for D-He3 and 
  $2.2 \times 10^4$ for D-T.  Again assuming a 3-order of magnitude improvement in
  antiproton production efficiency, these gains are near breakeven in terms of net
  energy flow.

The requirements of antimatter is thus dramatically reduced and, for instance, ACMF propulsion for manned flight to Jupiter (100 tonnes payload) would require only 10$\mu$g of antiprotons
(see here (pdf))
A: If you could I would not call it "fuel", and it would definitely not be an energy resource.
You are brushing over the bit where you make the antimatter, and you just can't do that. This is not a technical issue (yes, it's hard to actually make antimatter, but you can brush over the technical difficulties), it's an issue with the energy balance of it. Antimatter, as a fuel you could get electrical energy from, requires an input of at least that much energy to make it.
This means that unless you can "mine" antimatter directly from nature (i.e. find some antimatter asteroid and get it back here), the antimatter in your hypothetical reactor is not an energy resource but an energy transmission medium much like electricity is.
Let me make an analogy with hydrogen fuel cells for cars. These are fantastic in that they allow cars to run without burning petrol, but they do need hydrogen to run. Where do you get the hydrogen? From electricity. And if you got that electricity from burning coal, then your car is still responsible for CO$_2$ emissions. Hydrogen fuel cells are no greener than the electricity used to make them.
Similarly for your antimatter: it would simply be a (very expensive and inefficient) way to transport energy from your huge accelerator antimatter factory to your hypothetical antimatter reactor. Not a lot to gain there.
A: My particle physicist's answer would be No, not for fuel, not in existing reactor designs. The necessary technical methods would be such that a small part of the energy available would end up as heat and useful energy as in the fission type reactors, it will not be cost effective.
Consider that the antiprotons will have to exist in a plasma with positrons suspended in a magnetic field. Plasma is not very dense, orders of magnitude less dense than uranium. It has to impinge on ordinary matter in a controlled manner and the products of the annihilation will have to travel some distance. The products are on the average 5 pions, 1/3 of which are pi0 which decay practically immediately into two high energy gammas.
The charged pions decay in ~10^-8 seconds into a muon and a neutrino, neutrinos will leave without interacting and muons will interact minimally electromagnetically,without depositing much energy. The energy that can be trapped is the electromagnetic ionization that the pions and muons leave in whatever material is used for the trapping of the energy. One would need detailed calculations to establish what type of material would trap most of the energy , but it will have to be a new design, from scratch. The difference lies in the kinetic energies of the products of the interaction: in fission they are low and the products can be trapped in ordinary matter because further decays are improbable. In an antiproton on matter reactor the kinetic energies are large and the charged products have to be trapped before the decay, while a new design is needed for the gammas of the pi0.
clarification after comment: Fission reactors have developed specific methods of turning the kinetic energy of the fission products into heat, this means specific materials suitable for this purpose. Fission works with energies of particles of the order of MeV, and a small part of that is kinetic energy. These heavy slow moving fragments can be stopped by ordinary materials, their kinetic energy absorbed by them.
The energy  of the antimatter matter annihilation is almost 2 GeV with the average multiplicity of 5 pions it means that each pion gets on average 400 MeV. This energy is mainly kinetic, they are very light particles moving fast and different materials  and methods would have to be used than the ones in a fission reaction.
In this bubble chamber picture we see lots of charged  pions ( the chamber is transparent to the gammas of pi0) from the annihilation ,and the decay to a muon and an appropriate  neutrino for one of them. The medium , probably liquid hydrogen, is not dense enough to contain the energy of the interaction. The objective for a reactor would be to absorb all the kinetic energy of the pions before their decay in 10^-8 seconds. Maybe magnetic fields would be necessary to trap them in suitable dense material so that their kinetic energy turns into ionization energy and finally heat. New technology is needed, is all I am saying, and maybe so expensive that the output would not be cost effective.
