Sir, Scott et al. (BDJ 2007; 202: 32–33) describe an interesting case of double vision and local cutaneous vasoconstriction in the upper lip, just lateral to the philtrum, following inferior dental and long buccal nerve blocks. As the authors suggest, some of the more obvious anatomical explanations seem unlikely. They report deep anaesthesia with the patient feeling 'odd' after administration of local anaesthetic, which suggests there was direct leakage of anaesthetic into the vascular system, despite aspiration to try and avoid this. A more likely explanation for the diplopia is that anaesthetic entered veins communicating with the pterygoid venous plexus surrounding the pterygoid muscles. This plexus anastomoses freely both with emissary veins that pass through the foramina spinosum and ovale (occasionally also through a foramen of their own, the foramen of Vesalius). All of this valveless system of veins communicates with the cavernous venous sinus intracranially, and through this, with veins of the orbital cavities, forehead and face.

The tough lateral wall of the cavernous sinus, to which the dura mater of the middle cranial fossa is tightly bound, contains within it cranial nerves III, IV, Vi and Vii. However, the VIth nerve runs free within the venous blood of the cavernous sinus on the lateral aspect of the internal carotid artery, which itself also courses through the venous blood accompanied by a plexus of sympathetic nerves destined for the face and forehead – and which are vasomotor and secretomotor to sweat glands in the skin. Local anaesthetic reaching the cavernous sinus is, therefore, immediately in contact with the VIth nerve, which is likely to be affected by it, resulting in either partial or sometimes complete inability to abduct the eye until it clears.

The blanching of the upper lip is harder to explain but the deep facial vein runs in the cheek with the long buccal nerve and artery. The rich anastomosis of arteries around the upper lip would suggest it is unlikely that any one of them would have a discrete cutaneous distribution. Nonetheless, in this patient it remains a possibility that the buccal artery may indeed extend anteriorly through the cheek and into the skin of the upper lip. Anaesthetic and vasoconstrictor in the buccal artery would lead to immediate vasoconstriction. Anaesthetic in the deep facial vein here could again also easily pass back to reach the cavernous sinus. Whether this poor patient experienced two rare bouts of bad luck, that each resulted in a separate sign, or just one associated with the long buccal nerve block, is debatable, but these rare and alarming occurrences remind us of why we all spent so long learning head and neck anatomy. (The best account of these kind of anatomical details is undoubtedly Hollinshead W H. Anatomy for surgeons, volume 1, head and neck. Philadelphia: Harper Row, 1982.)

Editor's note: Regrettably, this letter was printed with a number of inadvertent errors in BDJ 2007; 202: 237, due to typographical misunderstandings. 'Vi' and 'Vii' were given as 'VI' and 'VII'. The version printed here is the correct version. We would like to apologise to Professor Dean, and would also like to thank Miss Sheila Lumley for bringing it to our attention. Notwithstanding the correction of these errors, there are other elements of Professor Dean's letter and the subject in general that Miss Lumley would like to raise (see below).

Sir, in respect of the article by Scott et al. (BDJ 2007; 202: 32–33) and the subsequent response from Professor Dean (BDJ 2007; 202: 237) I believe there are still factual errors.

Firstly, cranial nerve VII (facial) does not enter the cavernous sinus at all and secondly, only nerves III (oculomotor), IV (trochlear), and Va (ophthalmic division of trigeminal) lie in the lateral wall. In addition Vb (maxillary division of trigeminal) enters the cavernous sinus for a short distance before leaving it to pass thorough foramen rotundum. VI (abducent) does not lie in the lateral wall but passes through the centre of the sinus with the internal carotid artery. I understand that these errors arose from an editorial misinterpretation (see above) in which Vi and Vii were confused for VI and VII respectively. In response to this, I would suggest that the designations Va, Vb and Vc are more appropriate, or even better the name of the nerve in full ie the mandibular division of the trigeminal nerve. This would help to prevent any future confusion.

It is also interesting to note that there is in fact an endothelial lining of the cavernous sinus1 meaning VI (abducent) and the internal carotid artery do not 'run free within the venous blood of the cavernous sinus', nor 'course through the venous blood' as was suggested in Professor Dean's original letter.