Interesting question!
The paper you read should have inserted an appropriate citation. Moreover, "little input" is a subjective statement and doesn't make much sense in the context, as it can mean that there are few efferent axons (which may still have large effects on retinal functioning), or that the efferent input is relatively unimportant when compared to the afferent visual projections (which may indeed be so). In fact, I think what the authors actually try to say is that there is not much data available on efferent retinal input.
To answer your question: The efferent projections from the brain back to the retina, referred to as centrifugal projections, are indeed not extensively studied. The studies done were mainly performed in avian species. However, centrifugal projections have been shown to exist in monkeys and humans. In the macaque, they originate in the brain in the hypothalamus, and constitute histaminergic neurons which are thought to be involved in regulating blood flow in the inner plexiform layer of the retina. They may also be involved in regulating retinal dopaminergic amacrine cells, at least in mice and other species. Specifically, dopaminergic amacrine cells in the retina may be receiving centrifugal input from the suprachiasmatic nucleus (SCN, situated in the hypothalamus) where circadian rythms are initiated from melanopsin containing ganglion cells. The dopaminergic amacrine cells in the retina target photoreceptors, horizontal cells and amacrine cells and thereby regulate the retina depending on the state of the circadian clock. Most studies are histological in nature and functional data remains scant.
Reference:
Webvision - Feedback Loops (by Helga Kolb)