http://www.sciencedirect.com/science/article/pii/S0040403904012833

http://pubs.acs.org/doi/abs/10.1021/jo00335a058

It can be done but typically needs heating to drive the reaction to completion. alternative is hydrolyse the ester to acid, and use borane or NaBH4/I2 (to generate the borane). 

Basically it always depends on your substrate: here's an extract from a paper That you would find helpful (SYNTHETIC COMMUNICATIONS, 2003, 33, 1733–1750): The reduction of aliphatic esters with sodium borohydride in protic solvents is extremely slow. In aprotic solvents, such as dichloromethane, the reduction of ethyl laurate with tetrabutylammonium borohydride is only 25% complete after four days at 25 oC. However, if an electron withdrawing atom or functional group is adjacent to the carbonyl group, the reduction can be carried out much easily. This has been demonstrated when the electron withdrawing group is an epoxy, chloro, or cyano function. Some saturated acids and esters can be reduced to alcohols by combining NaBH4 and AlCl3 at room temperature. Unsaturated acids or esters formed complex compounds under these conditions. Esters have been reduced by NaBH4 when catalyzed by LiCl (or LiI) in THF. Other metallic ions, such as TiCl4  and CaCl2 were used to enhance the reduction efficiency of NaBH4 in various reductions. Reductions of aromatic carboxylic acids and aromatic esters with NaBH4 in diglyme at 162 oC have been actively investigated under a variety of conditions, both with and without LiCl present. Ethyl benzoate was readily reduced to benzyl alcohol at high temperature in diglyme. Aromatic amides and nitriles were also reduced under these conditions. High temperature borohydride reductions also can dechlorinate aromatic chloro compounds such as 4-chlorobiphenol and PCBs.

I recommend you for an old book. Organicum 449p. 

Thank you very much Eugene Yang and Fady Nahra. There are no any electron withdrawing groups at alpha or beta position of the ester. I'm gonna consider using the lewis acid in the reaction. However when I tried to use LiAlH4, the reaction is done so fast at below 0 C (ice + sat. KCl) but the lactam ring also be reduced and got the mixture product. Is it OK to use LiAlH4 but in very low temperature such as -78 C. Is the LiAlH4 active at that temperature?

Of course you can try to reduce ester with LAH  under low temperature (such as -30 or -78 degree). Hope it works for your substrate.

You could also try using LiBH4 or 0.5 eq LiALH4 (or lower)

Sodium borohydride is used to reduce a lot of ester but need some harsh condition ( RT to reflux)while lithium borohydride and LAH are used to reduced ester at low temerature.

Sodium borohydride is a good reducing for of carbonyl compounds in protic solvents like alcohols, water etc.  However since the ester carbonyls are less electrophilic than the aldehyde / keto carbonyls, are difficult to reduce with it. The aromatic esters (in which resonance of aryl group further decreases the electrophilic character of the ester carbonyls)  are rarely reduced with NaBH4 even in protic solvents. However aliphatic esters can be reduced with excess of reagent and on prolonged reaction time depending on the substrate. An electron with drawing group enhances the reaction rate as it makes the ester carbonyl a better electrophile. In the reduction with NaBH4 the Na+ has practically no role in activating the electrophilic character of the carbonyl group. Hence a protic solvent like EtOH, MeOH or water is necessary and the solvent has a distinct role here. Here the proton activate the carbonyl group. However  in case of LiBH4 an aprotic solvent like THF can be used. Here LI+ being a better coordinating ion can activate the carbonyl functionality by coordination with C=O making the hydride transfer to C=O much easier. Thus here Li+ has a distinct role (also LiBH4 is more soluble than NaBH4 in THF). In approtic solvents other Lewis acids like anhyd. MgCl2, anhyd ZnCl2 can also catalyze the reaction on the same ground that it activate the C=O by coordination. 

I preferably use NaBH4/CaCl2 combination for ester reductions. The reaction works selectively in most cases and under mild conditions (RT to 60 C). Ester (1 eq) in THF/alcohol (1:2, 0.2 M) is treated with CaCl2 (1.5 eq) and NaBH4 (3 eq). Stirred at rt to 60 C. After completion the reaction is quenched with acetone and isolated the usual way.

If you use 2.5 to 3.0 equivalents of sodium borohydride and seal the flask well. It will generate enough hydrogen pressure that could potentially reduce ester. That happens with me while reducing ketoester and I ended up with diol. You may not need any harsh condition. Diethyl ether : Methanol (10:2) is more preferable. THF can be used, if you want to avoid diethyl ether.. 

If above trick doesn't work, you may replace sodium borohydride with Lithium borohydride which is known for reducing esters to alcohols.

Borohydride anion is not so reactive, than alumohydride, but it remains good reducing agent.