Pfizer (NYSE: PFE) has had quite a few pipeline failures lately (see exhibits one, two, and three for instance). Individually, it's easy to blow them off. No one drug is going to make or break a drugmaker the size of Pfizer.

But collectively they can be a problem. Perhaps today more than at any time in its past, Pfizer needs big hits from its pipeline. Megablockbuster Lipitor is on the decline and will hit free fall in a little more than a year. It's doubtful that any single drug will replace Lipitor completely, but a few solid new drugs could help soften the blow.

At its last update, Pfizer had 133 programs in its pipeline, with 34 in the most critical phase 3 stage testing 24 different molecules. Here's a highlight of a few that investors should be keeping an eye on.

I heart this drug
Pfizer's apixaban, which is partnered with Bristol-Myers Squibb (NYSE: BMY), has the potential to knock off long-used, but long-hated generic warfarin. Doctors don't like the latter drug because the dose range that works for patients is narrow and foods that patients eat can change that window.

The duo has already shown that apixaban works better than aspirin at reducing strokes and blood clots, but the true test of the potential blockbuster is a head-to-head trial against warfarin. That trial is scheduled to be completed next year.

A tiny number times a really big one can sometimes work out
Pfizer has a couple of drugs in its pipeline that won't be used on very many people but could still bring in a decent amount of revenue because they'll be priced extremely high.

Taliglucerase alfa, which Pfizer licensed from Protalix BioTherapeutics, treats a rare disease called Gaucher disease. Genzyme's (Nasdaq: GENZ) floundering in its manufacturing of Cerezyme, which also treats the disease, has opened the door for the duo. The FDA already has the marketing application in its hands and is expected to hand down a decision in February.

Pfizer also has a lung cancer drug, crizotinib, in the pipeline that only treats a subset of patients with a certain mutation. Fortunately, crizotinib seems to treat them very well and should have an easy time gaining FDA approval for the hard-to-treat population.

Don't forget about these
The Alzheimer's disease market is wide open with the current offering doing little to help slow down the debilitating disease. Any drug that can show a reasonable efficacy is sure to be an instant blockbuster.

Pfizer has two shots at making its mark. Actually, maybe it's more like one and a half.

Dimebon, which is partnered with Medivation (Nasdaq: MDVN), already flunked one of its phase 3 trials. The duo is testing Dimebon in combination with Eisai and Pfizer's Aricept and Forest Labs' (NYSE: FRX) Namenda, which might show an effect, but the first trial shouldn't give investors a whole lot of confidence.

Pfizer's other shot at conquering Alzheimer's disease comes from bapineuzumab. Pfizer owns half of the drug with the remaining split between Johnson & Johnson (NYSE: JNJ) and Elan (NYSE: ELN). The phase 2 results were confusing, to say the least, making bapinezumab's success very hard to handicap. (There are a few others, but they're in earlier phases of development.)

Opportunity to grow
Of the drugs I've highlighted, all but one came to the company through a licensing deal. Pfizer needs those kinds of deals to grow its pipeline further.

While Pfizer made a fairly large acquisition last year that I wasn't all that fond of, the drugmaker was smart enough to fund much of Wyeth's purchase with shares and loans. At the end of the first quarter, Pfizer still had more than $17 billion in the bank to acquire additional drugs.

As long as it uses the money to continue licensing drugs and doesn't go off and buy a medium-sized drugmaker with drugs already on the market, the company's pipeline should do just fine.

Editor's note: An earlier version of this article noted that Pfizer had more than $17 million in the bank. While correct, that number drastically understated the position, and it has been updated. The Fool regrets the slipped decimal point.