Vivek,

below a couple of references that may answer your question.

A low power OTA with adaptive biasing and common-mode compensation
Zhiliang Hong; Yan Wang;
Solid-State and Integrated Circuit Technology, 1995 4th International Conference on
24-28 Oct. 1995 Page(s):188 - 192

Power-efficient super class AB OTAs
Lopez-Martin, A.; Baswa, S.; Ramirez-Angulo, J.; Carvajal, R.G.;
Solid-State Circuits Conference, 2004. ESSCIRC 2004. Proceeding of the 30th European
21-23 Sept. 2004 Page(s):231 - 234

Low-voltage low-power adaptive biased high-efficiency integrated amplifiers
Ferri, G.;
Electronics, Circuits and Systems, 2001. ICECS 2001. The 8th IEEE International Conference on
Volume 3,  2-5 Sept. 2001 Page(s):1529 - 1532 vol.3

A 900-mV low-power ΔΣ A/D converter with 77-dB dynamic range
Peluso, V.; Vancorenland, P.; Marques, A.M.; Steyaert, M.S.J.; Sansen, W.;
Solid-State Circuits, IEEE Journal of
Volume 33,  Issue 12,  Dec. 1998 Page(s):1887 - 1897

900 mV differential class AB OTA for switched opamp applications
Peluso, V.; Vancorenland, P.; Steyaert, M.; Sansen, W.;
Electronics Letters
Volume 33,  Issue 17,  14 Aug. 1997 Page(s):1455 - 1456

Be careful with the adaptivity ratio and related bias stability issues.

Good luck,

Paul

In the case of a two stage amplifier where the first stage is a folded cascode, you can try tying the miller compensation capacitor to the source of the common-gate devices in the folded cascode instead of tying the cap to the high impedance node of the folded-cascode. This will split the poles better and allow you to reduce the capacitive load at the high impedance node of the folded cascode which sets the slew-rate. This architecture has some gotchas; run large signal sims to be certain you have no large signal oscillations.  good luck.