To crash or not to crash? Probability in the management of benign lacrimal gland tumours

Benign tumours within the lacrimal gland fossa are very rare and this makes their clinical management controversial. Major changes in imaging techniques, historic variations in treatment, and the paucity of clinicians with long-term experience of these patients make the best practice rather uncertain. The practical interpretation of a few significant historic series—based, as these are, on retrospective evidence—is further thwarted where the typical reader will lack experience in treating the complications of mismanaged lacrimal gland tumours.

With rare exceptions, benign tumours of the lacrimal gland are pleomorphic adenomas. These tumours probably arise in young adults, and gradually enlarge, expanding the cortical bone of the lacrimal gland fossa over several decades (Figure 1). Typically arising in the orbital lobe of the lacrimal gland, the tumour consists of a very firm mass that leads to compression atrophy of normal gland, displacement of residual lacrimal tissues, and is surrounded by a ‘pseudocapsule’ into which small ‘sprouts’ of adenoma may project. Unless the patient has a short life expectancy, all lacrimal gland pleomorphic adenomas should be excised because untreated or partially excised tumours carry a significant long-term risk of malignant transformation.^{1, 2} Such a change is associated with a high morbidity, and subsequent intervention is very rarely curative.^{3, 4, 5, 6, 7}

Figure 1

Patient presenting with a long history of painless swelling in left lacrimal gland area. Thin slice orbital CT shows (a) proptosis because of a well-defined mass in the left orbital lobe, this being associated with smooth remodelling of bone, (b) flattening of the globe (arrow) by the mass, and (c) smooth expansion of the lacrimal gland fossa, with no evidence of destructive erosion of neighbouring cortical bone (arrows).

A common misconception is that incomplete tumour excision is associated with an increased risk of malignant transformation compared to no intervention, but there is no evidence to support this contention.^{8, 9} Indeed, logic would predict a decreased risk of malignant transformation after subtotal excision of an adenoma. Assuming that malignancy arises from a single cell within the benign tumour, then, for a given tumour mass with, say, a 1% annual risk of malignant transformation, removal of 95% of the mass will logically reduce this risk to 0.05%. Continuing logic ab extremis, intact removal of all benign tumour cells will leave a 0% risk both of malignant transformation and benign recurrence; this deduction is being strongly supported by studies with long-term follow-up.^{10, 11} A truly greater risk of malignancy after partial excision of benign tissues could only arise if each cell within the residual mass carried a significantly raised probability of malignant transformation; the sources for such an increased risk might be the physical injury of preceding biopsy, per se, or a cellular seeding into a new tissue environment, beyond the ‘pseudocapsule,’ which forms an effective physical barrier around the intact benign tumour.

Irrespective of the dreaded risk of malignant transformation, piecemeal removal of lacrimal gland pleomorphic adenomas carries a major risk of benign recurrence.^{3, 12, 13, 14} Clinical characteristics and management of lacrimal and salivary adenomas have been considered identical by many,^{15, 16, 17, 18} but this is completely inappropriate: whereas intact lacrimal adenomas can always be removed intact, salivary gland tumours often require piecemeal excision to preserve the adjacent major neurovascular structures. Furthermore, a major difference in biological behaviour of these tumours is evident in the starkly contrasting rates of recurrence after piecemeal excision—many publications are identifying a major risk for lacrimal gland adenomas,^{1, 13, 14} but a surprisingly low risk in the case of salivary adenomas.^{15, 16, 17} The behaviour of lacrimal adenomas after breach of their ‘pseudocapsule’ provides another compelling reason for intact excision—not only do lacrimal tumours have a high rate of benign recurrence with open biopsy (approximately 10%)¹¹ and an even higher rate after piecemeal excision (20–30%),^{1, 13, 14} but also such recurrence is typically pervasive throughout the orbital tissues. Currently, the only treatment for recurrent benign disease is subtotal orbital exenteration^{10, 19}—a radical approach to a cure when compared with intact tumour excision; having actual experience of managing this devastating complication—usually afflicting patients in their third or fourth decade (Figure 2)—engenders absolute respect for sanctity of the tumour ‘pseudocapsule’ and a realisation that intact excision remains the optimum treatment for what is a completely curable tumour.^{10, 11}

Figure 2

Orbital CT for a 30-year-old woman who had piecemeal excision of a lacrimal gland adenoma 11 years before referral for recurrent disease. An extensive pervasive benign recurrence was present throughout neighbouring orbital tissues, requiring subtotal orbital exenteration for cure. (a) Axial images show a recurrent mass moulding around the globe, with invasion of the temporalis fossa and orbital fat. (b) Coronal images confirm tumour recurrence in the skull base and infiltration of the lateral rectus, and superior rectus/levator complex.

Before high-resolution orbital imaging, the diagnosis of pleomorphic adenoma was based on a relatively long history of a painless, firm mass in the lacrimal gland fossa,² with the correct diagnosis being more likely where there was radiographic evidence of a smooth expansion of the lacrimal gland fossa with intact cortical bone. Given these early limitations in preoperative diagnosis, however, there were inevitable cases where a short history or inflammatory features prompted incisional biopsy of the gland, and the advocated management in this unfortunate scenario was intact excision of the gland and the biopsy tract.¹⁰ Although such surgery not infrequently caused troublesome lagophthalmos, because of adhesions between the excised lid skin (biopsy site) and orbital rim, early results suggested that this treatment was curative.¹⁰ Data from a study with longer follow-up have, unfortunately, identified late orbital recurrence in about 10% of such cases,¹¹ further enforcing the folly of open biopsy where pleomorphic adenoma of the lacrimal gland is suspected.

Together with a logical clinical approach to the management of lacrimal gland masses, the advent of thin-slice orbital CT has revolutionised the treatment of these lesions: a well-defined mass, compressing—rather than moulding around—the globe (Figure 1b), with preservation of orbital rim in front of a smoothly expanded lateral orbital wall (Figure 1a) or roof (Figure 1c) is sufficient evidence to confirm the likely presence of a pleomorphic adenoma. The author currently explores such a mass through an incision based in the lateral two-thirds of the upper lid skin crease; the mass being completely mobilised (without the use of any sharp instrumentation to handle the gland) on a sheet of orbital periosteum and, if it is clearly a hard and abnormal gland, the lacrimal neurovascular bundle is cauterised and divided at the posterior pole of the gland. Should there be any threat to an intact excision of the gland—because of difficulty in reaching the posterior pole of a giant tumour or because of a (very rare) soft and cystic pleomorphic adenoma—then lateral osteotomy is readily carried out after extending the same incision 1 cm laterally, with the lateral wall being swung outwards on a small temporalis flap, which preserves the blood supply to the bony wall (Figure 3). With experience, bone-swinging lateral orbitotomy (required in a minority of cases) is very safe and not associated with the high morbidity suggested by some authors.¹⁸ Since adopting this approach, the writer has not had a single case of inadvertent biopsy among 76 tumour-operated patients, over the last 15 years, with a histological diagnosis of lacrimal pleomorphic adenoma. Inevitably, such a policy may lead to the occasional excision of lacrimal glands with other pathologies that often being suspected preoperatively. Such masquerading conditions have included lymphoma, sclerosing dacryoadenitis, carcinoma in pleomorphic adenoma, or extreme rarities, such as intralacrimal neurilemmoma or haemangiopericytoma (solitary fibrous tumour). Indeed, many of these conditions benefit from intact excision, as the gland had been completely replaced by abnormal tissue. Thus, although the risk of tumour recurrence after fine-needle biopsy of adenomas may be very low, the performance of such a biopsy actually adds nothing useful to the management algorithm; there was no functioning glandular tissue in any of the author's cases where the mass (after intact excision) proved to be other than pleomorphic adenoma.

Figure 3

With experience, intact excision of the orbital lobe of the lacrimal gland is generally possible through an upper lid skin-crease incision. Where safe access to the posterior pole of the gland is not possible, however, the anterior orbitotomy is readily converted to a lateral approach by a 1–2-cm lateral extension of the incision into the rhytids. (a) A large window is available between the upper (light arrow) and lower (dark arrow) osteotomies, with the bone swung laterally on a temporalis muscle pedicle. (b) The extended upper lid skin-crease incision heals very well, as shown in this patient with only mild lateral redness evident at 4 weeks after removal of a large right tumour.

The use of appropriate imaging criteria to detect possible pleomorphic adenoma appears to provide a technique with complete sensitivity and high specificity; this being the case, for which lesions should incisional biopsy be reserved? Especially where inflammation is present, the author will carry out a biopsy of all persistent lacrimal gland masses that show moulding around the globe, ill-defined extension outside the lacrimal gland (generally either alongside lateral rectus, towards the superior orbital fissure or across the orbital roof), or that show destruction of the cortical bone of the gland fossa (such lesions representing probable invasive disease for which surgery is rarely curative). Although fine-needle biopsy will often provide a correct diagnosis, open biopsy is readily carried out, preserves macroscopic tissue architecture, and permits high-quality histological analysis.

In summary, because of the high risk of widespread and pervasive benign recurrence (this in addition to the risk of malignant transformation), it is indefensible to carry out piecemeal excision of any lacrimal gland mass that, on high-quality imaging, might be a pleomorphic adenoma. Longer-term follow-up suggests that incisional biopsy also carries a significant risk of recurrent benign disease—even where the lacrimal mass is later excised intact with the associated biopsy track.¹¹ If high-quality imaging suggests pleomorphic adenoma, a strict non-biopsy policy^{1, 2, 10, 11, 12} continues to provide the only chance of complete cure of these tumours. To crash, or not to crash: can we really—either through misguided belief or perhaps for convenience—advise a patient to take an aeroplane with a 10% chance of crashing,¹⁸ or should we advise the one with a proven long-term record of safety?¹¹