Where v or ƀ appeared before u in Quenya it vanished. It is unclear whether this loss occurred before or after the shift of ꝑ, ƀ > f, v, but given the (probable) lateness of the latter change, I suspect the vanishing was first. This sound change is only mentioned in Outline of Phonetic Development (OP1) from the 1930s in a deleted note:

ƀ was lost before u (cf. phū̆ > hū̆) (OP1: PE19/32 note #15).

However, evidence of this sound change can be seen in the future tense forms of basic verbs ending in v:

• √NDAB [> *navuva] > Q. nauva “will try” (PE22/151).
• ᴹ√KAV [ᴹ√KAB] > kavuva > ᴹQ. kauva “shall be able to” (PE22/105, 121).

Another (limited) description of this sound change appears in notes on Quenya verb suffixes from the late 1960s:

the sequence uvu was not favoured and unless altered became ú (PE22/155).

In this document Tolkien explored two grammatical variations resulting from this phonetic development. First, suffixal -uite was altered to -oite when following v to avoid the loss of the consonant. Compare Q. yuluite “drinking (as a habit)” vs. Q. kuvoite “hiding, secretive” (rather than **kuvuite > **kúite). Second, the future tenses of tuv- “to find” were described as:

the futures such as tuvuva “will find” which became túva or (since this was the same as the pres. continuous) tuvua (PE22/155).

Here it seems the normal phonetic development was tuvuva > túva, but this was modified to tuvua (why tuvuva wasn’t maintained or restored isn’t clear). Another document on verbal suffixes from the 1940s provided a different example of uv survival or restoration: ᴹQ. kúvula “flexible, pliant” (PE22/111) rather than **kúla. Thus it seems grammatical and morphological considerations (the need to preserve the stem and/or grammatical suffix) could sometimes inhibit or (partially) revert this sound change.

Conceptual Development: The combination vu appears in the Early Qenya of the 1910s, so this sound change was not part of Tolkien’s earliest conception of the languages.