Page 62 - Reconstructive Techniques in Musculoskeletal Tumor Surgery

Chapter 3
One or more complications occurred in 15 patients. There were seven mechanical and 11 non-mechanical complications. We were unable to identify any risk factor which was signi cantly associated with the occurrence of complications.
Type I complications22 (dislocation) occurred in ve patients, after a mean of 18.5 months (17 days to 8.5 years). Three patients experienced recurrent dislocations. Of these, two had type 2/3 resections and two had MUTARS attachment tubes in place. Two patients required open reduction, one of whom subsequently underwent revision of the acetabular component.
Type II complications22 (aseptic loosening) were diagnosed in three patients, after a mean of 19 months (16 to 24). The construct was reinforced by percutaneous bone cement injection in one patient. No attempt was made to reinforce or revise the other constructs, either because of a poor prognosis or because of a lack of remaining bone stock.
Type III complications22 (structural) occurred in four patients; they had undisplaced crack fractures of the remaining ilium during implantation of the stem. All healed uneventfully.
Type IV complications22 (infection) occurred in nine patients, six of whom required removal of the pedestal cup. The nal outcomes of these patients included revision to a LUMiC prosthesis (implantcast) in two, a type BII rotationplasty26 in one and a hindquarter amputation in one. In the remaining patients, no attempts were made to reconstruct the defect, either because of a lack of remaining bone stock or poor prognosis.
Type V complications22 (tumor progression) occurred in four patients: local recurrence and lung metastases were each diagnosed three times. Two local recurrences resulted in hindquarter amputation.
The prosthesis was removed in eight patients after a mean of 19 months (29 days to 4.2 years). None failed for mechanical (type I/II/III) reasons. For all reasons, the estimated two- and ve-year survival rates were 72% and 50%. For non-oncological reasons, these were 78% and 61%, respectively ( gure 5). Survival of the prosthesis was signi cantly worse for patients with an infection (log rank, p = 0.008).
The median postoperative hospital stay was 13 days (IQR 11 days to 6.6 weeks); all patients were able to walk postoperatively. A total of 13 patients had one or more further operations: the total number of secondary procedures was 85. In all, 59 reoperations (69%) were performed in the rst postoperative year, 69 (82%) for infection or wound problems. Four patients, all with a deep infection, underwent ten or more reoperations and accounted for 59 (69%) of all reoperations.
60