The most common response to this from White is the Studier Attack with 6.Bc4, followed by 0-0, Qe1 and Qh4 and Bh6. Black normally "castles into it" and relies upon counterplay to defuse White's attack, since if Black leaves the king in the centre then White has tricks based on Nf3-e5. However, after 6.Bc4 Bg7 7.0-0 0-0 8.Qe1 Nc6 9.Qh4 Bg4 I have been unable to find a good way for White to keep a strong attack going.
Black threatens to exchange bishop for knight on f3 and then take on d4, going a second pawn ahead and forcing the queens off. White must back-pedal for a move to prevent this, with 10.Ne2 or 10.Be3, and this slows down the white attack. My main problem with this system is that White is relying entirely on piece play to attack Black's king, and with the white king castled to the kingside, White cannot advance the g and h-pawns (the h-pawn hack is often very effective against the fianchetto formation, as seen in openings like the Sicilian Dragon, Yugoslav Attack) without exposing the white king.
Thus, 6.Bf4 is most likely a better bet for White, with the idea of Qd2 and 0-0-0, followed by Bh6 and h2-h4-h5 (it is important to play Bh6 before h4 in many variations to prevent Black from slowing down White's attack with ...h7-h5). The bishop is better-placed on f4 rather than e3 or g5, to gain some control of the e5 and d6-squares.
In the Pirc Defence lines where White plays Be3/f4/g5, Qd2, 0-0-0, Bh6 etc., Black often delays castling and/or goes queenside in order to leave White's kingside offensive firing at nothing, but in this line, White's missing f-pawn helps White to generate tricks down the e and f-files and put pressure on f7. If Black swiftly castles short and goes straight ahead with queenside counterplay, though, White's missing f-pawn is sometimes more of a hindrance than a help, because White cannot play f2-f3 to cover the e4-square and support g2-g4. Thus, Black's most challenging continuation appears to be 6.Bf4 Bg7 7.Qd2 0-0 8.0-0-0 a5!?, which was suggested to me by Mark Nieuweboer, by analogy with Abby Marshall's Chesscafe.com article which suggested this approach against 6.Be3.
Black's idea is simply to cause disruption in the white queenside and encourage White to take time out to defend, thus slowing down the kingside attack. The best response is probably 9.Bh6 a4 10.Bxg7 (White could also consider 10.a3 immediately, with the idea of getting in 11.h4 before exchanging off the dark-squared bishops, e.g. 10...Bf5 11.h4 Nbd7 12.Bxg7 Kxg7 and now the daring 13.h5 is one idea, though Black's control of the g4-square remains a problem.) 10...Kxg7 11.a3 (allowing Black to get in ...a4-a3 would force b2-b3, leaving White's queenside dark squares weak) 11...Bf5 12.Be2. White has long-term kingside attacking chances with the idea of h2-h3 and g2-g4, chasing away Black's bishop on f5, followed by h3-h4-h5, but it is quite slow, so it is debatable whether White has enough for a pawn.
White doesn't have many problems getting enough compensation for the pawn against other lines. I was recently involved in a Chesspublishing.com discussion on the line 8...c5 9.d5 a6, intending ...b7-b5, but instead of Christoph Scheerer's hasty 10.d6 White can play 10.Be2 completing development and if 10...b5 11.Ne5 with good compensation for the pawn.
One popular continuation is 8...Bf5 whereupon the game Seidel-Wolczek, email 2008 continued with the fairly typical line 9.h3 Nc6 10.g4 Ne4 11.Qe3 Nxc3 12.bxc3 Bc8.
Although I am not 100% convinced that White has full compensation for the pawn, I think that from a club player's perspective White's chances are quite good in the 6.Bf4 variation, but the Studier Attack approach following 6.Bc4 needs significant repair work.
