The purpose of this multicenter study was to analyze the results of shoulde
r arthroplasty for the treatment of the sequence of proximal humerus fractu
res and establish an updated classification system and treatment guidelines
for these complex situations. Seventy-one sequelae of proximal humerus fra
ctures were treated with shoulder replacement with the use of the some nonc
onstrained, modular, and adoptable prosthesis: the Aequalis prosthesis (Tor
nier Inc, St Ismier, France). The average time between initial fracture and
shoulder arthroplasty was 5 years and 5 months. On the basis of anatomic c
lassification schemes, sequelae were divided into 4 types: type 1, humeral
head collapse or necrosis with minimal tuberosity malunion (40 cases); type
2, lacked dislocations or fracture-dislocations (9 cases); type 3, nonunio
ns of the surgical neck (6 cases); and type 4, severe malunions of the tube
rosities (76 cases). The mean postoperative follow-up was 79 months (range,
12 to 48 months). Overall, the postoperative Constant score was excellent
in 11 cases (16%), good in 19 cases (16%.), fair in 18 cases (25%), and poo
r in 23 cases (33%). There were 18 complications (27%). Fifty-nine of 70 pa
tients (81%) stated that they were satisfied with the result. The most sign
ificant factor affecting functional outcome was greater tuberosity osteotom
y (P < .005). Regarding both surgical treatment and postoperative prognosis
, we identify 2 categories of proximal humerus fracture sequelae: category
1, intracapsular/impacted fractures sequelae (associated with both cephalic
collapse or necrosis [type 1] and chronic dislocation or fracture-dislocat
ion [type 2]), in which an articulating joint can be reconstructed without
a greater tuberosity osteotomy; and category 2, extrocapsular/disimpacted f
ractures sequelae (associated with both surgical neck nonunions [Pipe 3] an
d severe tuberosity malunions [type 4]) where the proximal humerus cannot b
e reconstructed without a greater tuberosity osteotomy All of the excellent
and good postoperative Constant scores were obtained in type 1 and 2, in w
hich osteotomy of the greater tuberosity was not required. All patients in
type 3 and 4, who underwent a greater tuberosity osteotomy, had either fair
or poor results and did not regain active elevation above 90 degrees. We c
onclude that a greater tuberosity osteotomy is the most likely reason for p
oor and unpredictable results after shoulder replacement arthroplasty for t
he treatment of the complex sequelae of proximal humerus fractures. Shoulde
r arthroplasty for the treatment of the sequelae of fractures of the proxim
al humerus should be performed without an osteotomy of the greater tuberosi
ty when possible. if prosthetic replacement is possible without an osteotom
y, surgeons should accept the distorted anatomy of the proximal humerus and
adopt the prosthesis and their technique to the modified anatomy. A modula
r and adaptable prosthesis with both adjustable offsets and inclination may
allow surgeons to adopt to a large number of malunions and may help to avo
id the troublesome greater tuberosity osteotomy in a higher proportion of c
ases.