Frederic G. Kenyon is noted for observing that the "last foundation for any doubt that the Scriptures have come down to us substantially as they were written has now been removed."
Kenyon's statement seems to be correct: Scripture has substantially come down to us as it was written. Despite the attempts to change, alter or corrupt the written Word of God, the Most High Deity Jehovah has protected or guarded His Word, as He duly promised in Holy Writ.
For example, in ancient times, the famed Johannine Comma was added to the first Epistle of John (1 John 5:7). However, many noted scholars (themselves Trinitarians) have produced mounds of evidence showing that the Comma Johanneum should not be included in the Bible. Granted, some may still contend that the Comma belongs in Scripture, but the preponderance of evidence is certainly not in their favor.
Somewhat more innocuously, we find examples of textual variants in the book of Philippians. Philippians 3:21 (in the Majority Text) reads:
EIS TO GENESQAI AUTO SUMMORFON.
But this reading does not appear earlier than the seventh century CE.
The Majority Text also adds XRISTWi in Phil 4:13.
Moises Silva points out nonetheless that while this variant is
attested as early as the fourth century CE, "the omission
has much wider and stronger attestation" (Philippians,
236). All the same, I don't find either reading
particularly problematic for my faith.
Finally, we have 1 Timothy 3:16, which has QEOS in some
texts rather than hOS. But If one seriously examines this
matter, he/she will find that the reading
QEOS appears no earlier than the eighth or ninth
century CE. Metzger demonstrates the superior status of
the lectio hOS over QEOS in 1 Tim 3:16, something even Isaac Newton discerned years prior to Metzger and
other textual critics.
In closing, I'd also like to refer you to a book
written by Neil Lightfoot entitled "How We Got the
Bible." Lightfoot's work is accessible but scholarly.
He convincingly demonstrates that we can put trust in
the 66 holy books that we have in our
possession today. He also shows how claims about
textual variations have often been exaggerated. For
instance, let us suppose that we have 1000 copies of a
work and 500 copies contain one clausal variant. That
is, there is one clause in 500 copies of the book that
differs from the other 500 books that have the same identical
clause in each of them. In this kind of situation,
how many variant readings might one be justified in
saying there are in the 500 books? Would there be one clausal
variant that occurs 500 times or 500 clausal variants altogether?
Some who (wittingly or unwittingly) desire to magnify
biblical variants would probably--in fact, many times
are--be inclined to say that there were 500 different
readings in the scenario presented above. But why
could we not say that there is simply one variant that
occurs 500 times? My point is that sometimes disparate
readings are counted where there is really no need to
do so. Lightfoot discusses this point and provides a
lucid account of why it is wrong-headed to tabulate
repeats of a distinct reading as additional
occurrences of the said variant.