Page 39 - New epidemiological and PSMA-expression based paradigms in salivary gland tumors

Salivary gland pleomorphic adenoma in the Netherlands
For this present research project, we excluded malignant transformations of primary SGPA, diagnosed as carcinoma ex pleomorphic adenoma at first presentation without a history of SGPA. In earlier research, however, we found 34 cases of salivary gland carcinoma ex pleomorphic adenoma in the same period of investigation. [11]. Four occurred in recurrent SGPA and were added to our database, leaving 30 cases to account for a 1.1% risk of de novo malignant transformation of primary SGPA (30 in 2 749). This is a similar percentage as the 1.7% that could be calculated from the population in Denmark[4]. Earlier publications reported a mean 6.2% risk, but their figures relate to single-center data and may reflect a referral bias [10,35].
The 0.15% secondary malignant transformation rate we found (carcinoma ex pleomorphic adenoma in recurrent SGPA in our SGPA cohort; 3.2% of all recurrences) is in the lower range of earlier findings [6]. These numbers are also lower than in Denmark, reported at 0.35% and 12.6% respectively. To some extent, the differences may be explained by different inclusion criteria, but more importantly, compared to smaller studies, we ruled out referral bias by compiling a nationwide cohort, rather than using single-center data.
Our results confirm that at a population level, complete surgical removal of SGPA can be difficult, leading to a 4.6% first-recurrence rate and a 16% second-recurrence rate (median times to recurrence 7 and 2 years, respectively). Recurrences are often multinodular, with a mean number of 26 nodules (range 1-266) found in the primary resection bed [36]. These figures provide a strong argument for MRI follow-up after all first recurrences, to avoid a need for more extensive surgery at some later point in time.
Risk factors for recurrence: We found margin status to be the primary risk factor for recurrence. However, our margin data were based on microscopy, whereas in practice, margin status is often determined macroscopically by the surgeon. In many resections, sufficient margins cannot be taken because of adjacent facial nerve branches, and the pathologist will only have a very thin capsule to examine. This problem may raise doubt as to the reliability of microscopy data for multivariate analysis. Still, if margins are positive or uncertain, it is highly plausible to expect higher recurrence, since positive microscopic margins are accepted as a primary cause for tumors to recur, as are rupture and spillage[5,6].
2
37