Abstract
Background: Metastatic disease of the acetabulum
can be painful and disabling. Operative intervention is indicated
for patients who fail to respond adequately to nonoperative treatment.
We evaluated the functional and oncological outcome of acetabular
reconstruction after curettage for the treatment of refractory symptomatic
acetabular metastases.
Methods: Fifty-five patients with metastatic
disease of the acetabulum were treated with operative acetabular reconstruction
combined with a total hip replacement. The most common primary tumor
was carcinoma of the breast (eighteen patients), followed by carcinoma
of the kidney (seven patients) and carcinoma of the prostate (seven
patients). Forty (73 percent) of the patients presented with multiple
skeletal metastases, and eighteen (33 percent) had associated visceral
metastases. Twenty-eight (51 percent) had severe pain requiring
continuous use of narcotics, twenty-four (44 percent) had moderate
pain requiring periodic use of narcotics, and the remaining three
(5 percent) had mild pain requiring use of non-narcotic analgesics.
Eighteen (33 percent) of the patients could not walk, twenty-three
(42 percent) needed a walker or crutches, twelve (22 percent) used
a single cane, and two (4 percent) walked without assistive devices.
Intralesional curettage of the tumor was performed in all of the
patients. Fifty-four of the hips were reconstructed with a protrusio
cup and one, with a hemipelvis endoprosthesis. Large defects were reinforced
with cement and pin or screw fixation (the modified Harrington technique),
which allowed transmission of weight-bearing forces to the remaining
intact pelvis. Thirty-six acetabular reconstructions were performed
with antegrade pins or cannulated screws; fifteen, with long retrograde screws;
and four, with cement.
Results: The median period of survival was nine
months. Patients with visceral metastases had a median period of
survival of three months compared with twelve months for patients
without visceral metastases (p < 0.001). Patients with breast
cancer presented later in the disease process (p < 0.004) and
lived longer than did those with other carcinomas (p < 0.004).
Forty-five patients were evaluated three months after reconstruction.
Thirty-four (76 percent) of them had relief of pain as determined
by decreased use of narcotics. Nine of the eighteen patients who
could not walk preoperatively regained the ability to walk. Fourteen
of the seventeen patients who originally were able to walk in the
community retained that ability. Thirty-three patients were available
for evaluation at six months. Twenty-five (76 percent) still had
relief of pain, and nineteen (58 percent) were able to walk and
function in the community. Overall, fourteen (25 percent) of the
fifty-five patients had moderate local progression of the disease,
and five of these patients had failure of the fixation. Fourteen
early complications developed in twelve (22 percent) of the patients.
One patient (2 percent) died perioperatively.
Conclusions: Patients who have acetabular metastases
that are refractory to radiation and chemotherapy have a short life
expectancy. The early, gratifying results of reconstruction validate
the role of operative treatment as a short-term palliative procedure. Protrusio
acetabular cups presumably compensate for deficiencies of the medial
wall, while cement and pin fixation can be used effectively to reconstruct
large defects in the acetabular column and dome. The low rate of
fixation failure supports the biomechanical principles of the reconstruction. Generally,
the reconstructions are sufficiently durable to exceed the life
expectancy of the patients.
We evaluated the oncological and functional outcome of operative
treatment of periacetabular metastases. Pelvic and periacetabular
metastases occur frequently in patients with cancer. Some of those
metastatic deposits are associated with disruption and mechanical
instability of the acetabulum, which may cause severe pain that
is increased by walking or motion of the hip. An effective method
of treatment should relieve the pain and restore function for as
long as possible. Correction of mechanical instability can restore
or preserve walking ability and help to maintain the independence
and comfort of the patient during the last months of life. Operative
treatment should enhance the patient's quality of life.
At our institution, many patients present with disruption and
mechanical instability of the acetabulum secondary to metastatic
disease. Nonoperative management with protected weight-bearing and analgesics
is initially recommended. Anti-neoplastic treatment consisting of
radiation therapy, chemotherapy, immunotherapy, or hormonal manipulation
is initiated or continued, as appropriate. Operative treatment is
recommended if (1) acute symptoms do not abate after a combination
of protected weight-bearing, administration of analgesics, and anti-neoplastic
treatment; (2) restoration of satisfactory function with control
of pain is not achieved within one to three months following radiation
therapy; or (3) an ipsilateral fracture or an impending fracture
of the femur requires operative treatment.
Harrington suggested a method of classifying metastatic acetabular
lesions and described an excellent method to reconstruct the hip
in the presence of extensive periacetabular destruction4. Although his results were encouraging,
we are unaware of any subsequent large published series that substantiate
his findings14. Our report examines
these issues in a large cohort of patients, presents an anatomically
based classification, and describes techniques that may make pelvic
reconstruction easier and safer in patients with periacetabular
metastases.
Fifty-five patients who had symptomatic metastatic acetabular
disease with structural insufficiency were treated with operative
reconstruction of the acetabulum in combination with a total hip
replacement from June 1988 to June 1998.
Demographics
The mean age (and standard deviation) of the patients at the
time of presentation was 62 ± 13 years (range,
thirty-four to eighty-two years). There were thirty-one female and
twenty-four male patients. The most common primary tumor was carcinoma
of the breast (eighteen patients) (Table I). The median time from
the diagnosis of the primary cancer to the manifestation of the
acetabular disease was twenty-four months (range, zero to 283 months).
The mean time to presentation of acetabular metastases in patients
with breast carcinoma (ninety-one months; range, three to 283 months)
was longer than that in patients with the other types of carcinoma
(thirty-three months; range, zero to 192 months). This difference
was significant according to the two-tailed, independent Student
t test (p < 0.004).
Preoperative Functional Status
The preoperative condition of the patients was evaluated on the
basis of the Eastern Cooperative Oncology Group (ECOG) performance
status6, hematocrit, serum albumin
level, serum calcium level, technetium-99 methylene diphosphonate scintigraphic
findings, and the presence of visceral metastases. The ECOG performance
status scale is commonly used to quantify the functional status
of patients with cancer, to determine eligibility for clinical trials,
and to predict survival2,7,10,11,13.
The interobserver and intraobserver reliability of the ECOG performance
status has been validated in several studies7,11.
There is also a relationship between duration of survival and the
other parameters mentioned earlier1,3,5,9,15,16.
The ECOG performance status was assessed on the basis of clinical
or admission notes (Table II). Forty (73 percent) of the fifty-five
patients were too sick to have been considered for most investigational
chemotherapy protocols at our institution because they had an ECOG
value of more than 1.
Thirty-five patients were anemic at the time of presentation.
Thirteen of these patients had moderate-to-severe anemia (a hematocrit
of less than 30 percent). The mean serum albumin level was 3.9 ± 0.6 grams per deciliter (39 ± 6
grams per liter). Twenty-five patients had an albumin level of less
than 4.0 grams per deciliter (forty grams per liter). The mean serum
calcium level was 9.8 ± 1.2 milligrams per
deciliter (2.45 ± 0.30 millimoles per liter).
Three patients presented with hypercalcemia. Forty (73 percent)
of the patients had multiple skeletal metastases as seen on bone
scans. Disease was confined to the hemipelvis in the remaining fifteen
patients (27 percent). Eighteen patients (33 percent) had known visceral
metastases, and four of them had multiple visceral metastases. The
lung was the most common site of visceral involvement (fourteen
patients). Other metastatic sites included the liver, the adrenal
gland, the contralateral kidney, and the brain.
Preoperative Pain and Walking Status
All fifty-five patients were incapacitated by hip pain. Twenty-eight
(51 percent) had severe pain requiring continuous use of narcotic
analgesics, twenty-four (44 percent) had moderate pain requiring
daily but periodic use of narcotic analgesics, and the remaining
three (5 percent) had mild pain requiring use of non-narcotic analgesics.
Eighteen (33 percent) of the patients were unable to walk. Twenty
could walk in their home, and seventeen could walk in the community.
Twenty-three (42 percent) used a walker or crutches, twelve (22
percent) used a single cane, and two (4 percent) walked without
assistive devices.
Preoperative Radiation and Chemotherapy
Thirty-one (56 percent) of the patients received radiation to
the involved acetabulum prior to operative intervention. The mean
dose of radiation was 3400 960 centigray. Forty-five patients (82
percent) received systemic chemotherapy, hormonal therapy, or immunotherapy.
Fifty-three patients (96 percent) received either chemotherapy or
radiation therapy preoperatively. All of the patients remained seriously
symptomatic at the time of presentation to the orthopaedic oncology
service.
Anatomical Pathology
The extent of acetabular disease was evaluated preoperatively
with plain radiographs for all fifty-five patients, with computerized
tomography scans for thirty-eight patients, and with magnetic resonance
imaging scans for fourteen patients. Three-dimensional imaging of
the acetabulum was performed when the integrity of the anterior
and posterior columns could not be adequately assessed with plain
radiography. Computerized tomography was the study of choice to
evaluate bone integrity.
The extent of the acetabular defect was evaluated with preoperative
radiographs and on the basis of the intraoperative findings. The
anterior and posterior columns, superior dome, and medial wall were assessed
separately for structural compromise and were then classified as
sufficient or insufficient. Sufficient bone provided adequate support
for the acetabular component, while insufficient bone did not. Fractures
of the anatomical regions as well as segmental and cavitary defects
were classified as insufficient bone. This system was based on both anatomical
and reconstruction considerations. Both the anterior and the posterior
column were insufficient in ten (18 percent) of the patients, whereas
either the anterior or the posterior column was insufficient in
thirty-six (65 percent). Forty-two patients had an insufficient
medial wall and an insufficient column or dome.
Twenty-nine patients also had a radiographically visible lesion
in the ipsilateral femur. Eighteen of them had a lesion in the femoral
head or neck, and the other eleven had a lesion in the femoral shaft.
Forty-seven patients had an acetabular fracture, and the remaining
eight had an acetabular fracture associated with a fracture of the
proximal aspect of the femur.
Embolization
Selective arterial embolization was performed in twelve patients:
four with renal-cell carcinoma, two with thyroid carcinoma, two
with lytic carcinoma of the breast, and one each with multiple myeloma,
leiomyosarcoma, esophageal carcinoma, and pheochromocytoma.
Operative Reconstruction
One patient received a hemipelvis endoprosthesis. Fifty-four
patients received a cemented total hip replacement: three had a
reconstruction with a protrusio acetabular cup, fifteen had a modified
Harrington reconstruction with a protrusio acetabular cup and long
retrograde screws, and thirty-six had a modified Harrington reconstruction
with a protrusio acetabular cup and antegrade pins or cannulated
screws.
Technique of Periacetabular Reconstruction
The patient was placed in a true lateral decubitus position6 to assist with proper orientation
of the acetabular component. Slight flexion of the table at the
patient's waist facilitated subsequent placement of the antegrade
pin or cannulated screw within the iliac crest. The entire extremity,
iliac crest, and flank were prepared and draped. The hip was exposed
through an extensive posterolateral approach, which provided optimal
acetabular exposure. Soft tissues around extensive acetabular lesions
were carefully retracted with a hand-held long, narrow, smooth,
right-angle retractor rather than with typical spiked retractors
that could puncture the tumor pseudocapsule or disrupt the only remaining
intact acetabular column. This avoided uncontrolled bleeding or
soft-tissue injury early in the procedure. All gross tumor was then
removed by curettage. Bleeding was often profuse at this time, but
it subsided after the tumor was removed and the socket was packed
with sponges. Structurally intact bone was preserved. Cortical bone
was preserved to maintain the anatomical landmarks that help to
guide insertion of the cup and also to increase structural support
for the cup.
If an antegrade modified Harrington reconstruction was indicated,
then the anterior or posterior portion of the iliac crest was exposed
through a separate incision measuring approximately six to eight
centimeters in length. The incision was made parallel to the iliac
crest, and the crest was then subperiosteally exposed. The insertion
sites for the iliac crest pins were prepared prior to curettage
of the tumor in patients with vascular tumors to minimize the time
during which tumor bleeding occurred. Use of a triangulation guide
assisted proper antegrade pin or cannulated screw placement within
the iliac wing (Fig. 1). An entry site near the inner one-third
of the crest was used to prevent pins from cutting through the outer
pelvic table. Smooth or threaded one-quarter-inch (0.64-centimeter)
Steinmann pins or cannulated 6.5-millimeter screws were used. The
goal of pin or screw placement was to reinforce deficiencies in
the dome, medial wall, or columns (Table III).
Pins or screws in the anterior iliac crest were aimed posteriorly
toward the ischium to reinforce deficiencies of the posterior column,
whereas those in the posterior iliac crest were aimed anteriorly
toward the pubis to reconstruct insufficiencies of the anterior
column. Pins were inserted to the desired depth, cut to the appropriate
length, and impacted within the iliac crest. Impaction helps to prevent
protrusion of the pins, which causes local pain and discomfort.
Pins or screws were placed at least one centimeter apart along the
iliac crest. Typically, two pins or screws in the anterior iliac crest
and two in the posterior iliac crest were easily placed when necessary.
Defects of the superior dome were reconstructed with antegrade
or retrograde pins or screws. Incisions in both the anterior and
the posterior iliac crest were used to place antegrade pins or screws in
the superior dome. Broad pin or screw placement in the superior
dome created a pedestal that provided additional cup support. A
pedestal located superomedial to the acetabular component transmitted
weight-bearing forces across the acetabulum to healthier areas.
Large defects of the superior dome were further supported by long
(sixty to 100-millimeter) 6.5-millimeter-diameter screws inserted
retrograde through the cup into the exposed innominate bone or embedded
in cement.
Insufficiencies of the medial acetabular wall were reconstructed
with a flanged acetabular cup (Biomet, Warsaw, Indiana) in thirty-five
(83 percent) of the forty-two patients with medial wall deficiencies.
The flange was placed over the most intact part of the remaining
rim of the acetabulum to improve anchorage of the cup. The superolateral part
of the rim was usually chosen. This position lateralized the cup
so that it did not impinge on the acetabular pins or screws. Accurate
orientation of the cup was often difficult when there was loss of cortical
and acetabular rim bone. Cups were inserted in 40 to 55 degrees
of horizontal inclination. Notching the acetabular rim allowed the
flange of the protrusio cup to fit better, thus avoiding excessive
vertical placement of the cup. Supplemental fixation screws were
placed through the dome and the flange of the cup to secure the
component better and to integrate it with the cement within the defect
in the superior dome. Dome screws were placed before flange screws
were inserted, to prevent the cup from cantilevering on the acetabular rim
and dislodging the component (Figs. 2-A, 2-B, and 2-C).
The flanged cup was used in a total of forty-six patients. Insufficiency
was noted in both the medial wall and the superior dome in thirty
patients, in the superior dome only in nine, and in the medial wall only
in six. One patient had sufficient bone in both the superior dome
and the medial wall. The flange on the cup prevents protrusio, and
the multiple holes within the cup add stability to the construct by
allowing cement interdigitation or screw placement into the dome
or flange for additional fixation of the cup and reinforcement of
the superior dome.
Cementing on the acetabular side was performed in two stages
in patients with large defects and a deficient inner table. During
the first stage, the defect was filled entirely with cement and
the pins or screws were incorporated into the cement. An acetabular
cup was then cemented over this construct. Attempts to simultaneously
fill the defect and cement the cup led to incomplete filling of large
defects. Additionally, pressurization in the presence of a medial
defect can dangerously force cement into the pelvis. A one-stage
cementing technique was used for smaller defects with an intact
inner table.
The femoral side was reconstructed with a standard-stem-length
prosthesis in nineteen patients and with a long-stem prosthesis
(220 to 300 millimeters) in thirty-six patients. All long-stem prostheses
were cemented after the femur was vented to prevent fat, tumor,
and cement embolization8.
Postoperative Treatment
Eighteen patients received postoperative radiotherapy. The mean
dose was 2900 620 centigray. Forty-one patients (75 percent) received
postoperative chemotherapy. Overall, fifty-four patients received
chemotherapy or radiation therapy either preoperatively or postoperatively.
Forty-seven patients received chemotherapy and radiation therapy,
five received chemotherapy only, and two received radiation therapy
only.
Below-the-knee compression boots were used intraoperatively and
until the patient could walk. Prophylactic chemical anticoagulation
was not used routinely. The patients began walking one to four days
postoperatively depending on their general condition. They were
instructed to bear weight as tolerated.
Statistical Analysis
Differences in the means of continuous data were tested with
the Student t test. Survival data was analyzed with the Kaplan-Meier
method of estimating survival distribution. The log-rank test was used
to compare survival distributions. The Fisher exact test was used
to compare survival and functional status of patients in the present
study with those of patients in other studies4,14.
Formal Kaplan-Meier survival distributions and comparisons could
not be performed between studies with the data available. A p value
of less than 0.05 was considered significant. All statistical analyses
were performed on a personal computer with the statistical package
SPSS for Windows (version 10.0; SPSS, Chicago, Illinois).
Operative Time, Estimated Blood Loss,
and Duration of Hospital Stay
The mean operative time was 290 ± 72 minutes.
The mean estimated blood loss was 2200 ± 1100
milliliters. Two of the patients had excessive intraoperative bleeding
(14,000 and 8000 milliliters). One of these patients had renal-cell
carcinoma, and the other had a lytic carcinoma of the breast. The
mean postoperative hospital stay was 24 ± 11
days.
Postoperative Complications
There were fourteen early postoperative complications in twelve
patients (22 percent). One patient died perioperatively (less than
one month after the procedure) from seizure activity secondary to brain
metastases and uncontrolled hyponatremia. The other complications
included five deep-vein thromboses, three superficial wound infections, one
mild disseminated intravascular coagulation-like coagulopathy, one
sacral decubitus ulcer, one wound hematoma, one colonic pseudo-obstruction (Ogilvie
syndrome), and one hip subluxation requiring revision of the acetabular
component. None of the patients had a deep infection.
There were five late fixation failures due to progression of
disease and one prominent pin in the iliac wing in a thin patient.
The mean time to failure was 12 ± 7 months
(median, eleven months). Four of these patients had to have one
or more of the pins removed from the iliac wing. One patient underwent a
revision to a bipolar hemiarthroplasty because of progressive protrusio
acetabuli. One patient remained asymptomatic and thus did not require
further intervention.
Survival and Progression of the Disease
Forty-four patients (80 percent) had died by the time of this
writing. The cause of death of the majority of the patients was
systemic progression of the disease. Twenty-three of these patients
had survived for more than twelve months postoperatively. At the
time of this writing, six patients were alive with less than two
years of follow-up (mean, 16 ± 2.8 months;
median, sixteen months). All of these patients survived longer than
the overall median period of survival of nine months as determined
by Kaplan-Meier analysis (95 percent confidence interval, 5.9 to
twelve months; mean, eighteen months). Overall, ten patients were
alive at the time of this writing (mean period of survival, twenty-eight
months; median, nineteen months; range, twelve to seventy-three
months) and one was censored as alive at the time of the most recent follow-up.
In the entire series, only seven patients survived for more than
two years: four were still alive at the time of this writing, two
had died, and one was lost to follow-up.
The median period of survival of the patients with visceral metastases
was three months (95 percent confidence interval, 0.9 to 5.1 months;
mean, 5.7 months), whereas that of the patients without visceral
metastases was twelve months (95 percent confidence interval, 7.3
to seventeen months; mean, twenty-five months) (Fig. 3). This difference
was significant according to log-rank analysis (p < 0.001).
Patients with breast cancer had a median period of survival of nineteen
months (95 percent confidence interval, 4.6 to thirty-three months;
mean, twenty-two months) compared with a median period of 6.3 months
(95 percent confidence interval, 1.4 to eleven months; mean, 9.4
months) for patients with other carcinomas (p < 0.004) (Fig.
4).
Patients with a poor performance status (an ECOG status of more
than 1) had a shorter period of survival (median, seven months;
95 percent confidence interval, 4.1 to 9.9 months; mean, sixteen months)
compared with patients with a good performance status (an ECOG status
of 0 or 1) (median, fifteen months; 95 percent confidence interval, twelve
to eighteen months; mean, twenty-one months.)
Fourteen patients (25 percent) had moderate local progression
of disease. Five of these patients had fixation failure with progressive
migration of the cup and subsequent pin protrusion. Thus, in nine patients,
fixation of the implant was maintained despite progressive local
disease (Figs. 5-A, 5-B, and 5-C). Remarkably, forty of fifty-four
patients who survived more than one month had radiographic evidence
of local control of the tumor despite progression of the disease
systemically.
Functional Results
At Three Months
Functional evaluation was performed during postoperative clinic
visits. Pain, walking, and the use of assistive devices were assessed.
Forty-five patients were available for follow-up at three months
(eight had died, and two were lost to follow-up). Five patients
(11 percent) continued to have severe hip pain requiring continuous
use of narcotic medication, eleven (24 percent) had moderate pain
requiring oral narcotic analgesics, eighteen (40 percent) had mild
pain requiring occasional non-narcotic analgesics, and eleven had
no pain. Overall, thirty-four patients (76 percent) had less pain
compared with preoperative levels as determined by decreased use
of narcotics. Eight patients had the same level of pain, whereas three
had worse pain.
Six patients were unable to walk, fourteen could walk only in
their home, and twenty-five could walk and function in the community.
Thirty of the thirty-two patients who could walk preoperatively retained
that ability postoperatively. Eleven of them had decreased use of
assistive devices, fourteen maintained the same requirements, and
five had increased use of assistive devices. Of the thirteen patients
who could not walk preoperatively, nine regained the ability postoperatively
(two without the use of assistive devices, two with a cane, and
five with a walker) and the other four were still unable to walk.
At Six Months and Beyond
Thirty-three patients were available for follow-up (nineteen
had died, and three were lost to follow-up) at six months. Twenty-five
(76 percent) still had less pain than they had had preoperatively, three
(9 percent) had the same level of pain as they had had preoperatively,
and five (15 percent) had increased pain. Nineteen (58 percent)
maintained the ability to walk and function in the community. Of
the nine patients who had regained the ability to walk after acetabular
reconstruction, all remained alive and six maintained the ability
to either walk in their home or to walk and function in the community.
Twenty-one patients had more than twelve months of follow-up
(thirty-two had died, and two were lost to follow-up). Fourteen
(67 percent) still had decreased pain, and twelve (57 percent) maintained
the ability to walk and function in the community.
Of the fifty-five patients, seven had more than two years of
follow-up (forty-one had died, one was lost to follow-up, and six
were alive with less than two years of follow-up). Six maintained
pain relief, while five maintained the ability to walk and function
in the community.
Acetabular reconstruction in patients with metastatic disease
is an oncological and reconstructive challenge. The overall outcome
in the current series of patients was satisfactory. The present
study differs from earlier reports on survival and late functional
outcome after acetabular reconstruction in patients with metastatic
disease4,14.
Harrington reported that thirty (52 percent) of fifty-eight patients
survived for two years after operative treatment4.
In contrast, only seven (13 percent) of fifty-five patients survived
for two years in the present study. This difference in survival
is significant according to chi-square analysis (p < 0.003),
even after assuming the worst possible two-year survival rate (thirty
alive and twenty-eight dead) in the Harrington study and the best
possible two-year survival rate (fourteen alive and forty-one dead)
in the present study.
The difference in survival may be due to differences in the patient
populations and the indications for operative treatment. Harrington
treated patients on the basis of the presence of metastasis and
the patient's life expectancy rather than using the currently reported
approach, which was based on the degree of acetabular involvement
by the metastases and the amount of pain produced by walking. The
process used by Harrington resulted in the selection of patients
whose life expectancy was longer than that of the patients in the
current study.
In our study, any patient with both symptomatic acetabular metastases
that were refractory to nonoperative intervention and a life expectancy
of more than three months was a candidate for operative treatment.
Additionally, eighteen patients (33 percent) in the current study
had visceral metastases, which were associated with a reduced median
period of survival (p < 0.001) and may have contributed to the
difference in survival between studies. Poor preoperative performance
status was also associated with a reduced median period of survival.
The large number of patients (forty of fifty-five) with a poor preoperative
performance status in the current study may also explain the difference
in survival between the two studies.
Pain relief, ability to walk, and use of an assistive device
were analyzed to evaluate functional outcome. Seventy-six percent
of our patients (twenty-five of the thirty-three who were alive
at six months) had satisfactory pain relief. These results are similar
to the pain relief reported by several authors who performed reconstructive
procedures in patients with metastatic acetabular disease4,12,14. On the other hand, only nineteen
of thirty-three patients in our study maintained the ability to
walk and function in the community six months after the reconstruction,
whereas Harrington reported that thirty-nine of fifty-one patients
maintained these abilities (p < 0.037). Patient selection may
have contributed to this difference. The current study included
patients with shorter life expectancies and more patients with combined
medial wall and superior dome or column defects (forty-two of fifty-five)
compared with Harrington's study (twenty-five of fifty-eight) (p < 0.001).
The optimal time to evaluate the outcome of this procedure is debatable,
particularly because of the presence of other sites of disease and
the short life expectancy of the patients. It is notable that there
were similar rates of pain relief and maintenance of function among surviving
patients at each of the time-points chosen (three, six, twelve,
and twenty-four months).
Although the oncological and functional results of the present
study differ from those of previous reports, we believe that acetabular
reconstruction provides worthwhile palliative care for patients with
disabling metastatic acetabular disease. The magnitude of the operation,
however, requires judicious use of this procedure. Most patients
respond to nonoperative measures. Only a few patients have major
structural deficiencies that require operative treatment. Preselection
of patients with minimal disease may lead to longer periods of survival
after reconstruction but will subject many patients to the risk,
expense, and inconvenience of operative treatment that is perhaps
unnecessary. On the other hand, not treating patients with disabling
metastatic acetabular disease denies many of them a better quality
of life during their remaining few months.
We suggest modifying the Harrington system of evaluation of metastatic
disease of the acetabulum4. Although
the Harrington classification identifies overall structural compromise,
the system described in our report defines structural deficiencies of
the acetabulum anatomically. Column and superior dome deficiencies
were effectively reconstructed with use of the remaining intact
bone to anchor the implants and to transmit forces to stronger bone.
Medial wall deficiencies were successfully treated with reconstruction
with a protrusio cup.
The Harrington operative technique was modified to aid placement
of pins or cannulated screws within the remaining intact bone. A
triangulation guide facilitated antegrade placement of pins or cannulated
screws. They could be placed accurately, rapidly, and safely, optimizing
the reconstruction. Another modification was the use of retrograde
long-screw fixation in an attempt to anchor the cement and cup to
intact bone.
The results of the current series support Harrington's findings4, and they endorse the concept of
reconstructing pathological fractures of the acetabulum by transmitting
forces to stronger proximal pelvic bone. Although fourteen patients
(25 percent) had local disease progression, only five of them had
fixation failure (Figs. 5-A, 5-B, and 5-C). They had bone resorption
and loss of fixation of the devices to the underlying bone. The
disease progressed in patients in whom the tumors did not respond
to chemotherapy, radiation, or intralesional excision. Proliferation
of residual cancer can weaken the construct, contribute to implant
migration, and result in deterioration of the patient's functional
status. The reconstruction can be no better than the quality of
the underlying bone. It is unknown whether patients who do not respond
to local radiation or systemic chemotherapy or hormone therapy would
benefit from a tumor excision that is more extensive than the incisional
curettage used in our series. However, few patients with acetabular
disease are candidates for wide excision because of their short
life expectancy and widespread metastatic disease. The intralesional excision
of the tumor and the reconstruction of the acetabulum described
in this report helped most patients.
Coleman, R. E., and Rubens, R. D.: The clinical course of bone metastases from breast cancer. British J. Cancer,55: 61-66, 1987.5561
1987
Conill, C.; Verger, E.; and Salamero, M.: Performance status assessment in cancer patients. Cancer,65: 1864-1866, 1990.651864
1990
[PubMed]
Dunphy, F. R.; Spitzer, G.; Fornoff, J. E.; Yau, J. C.; Huan, S. D.; Dicke, K. A.; Buzdar, A. U.; and Hortobagyi, G. N.: Factors predicting long-term survival for metastatic breast
cancer patients treated with high-dose chemotherapy and bone marrow
support. Cancer,73: 2157-2167, 1994.732157
1994
[PubMed]
Harrington, K. D.: The management of acetabular insufficiency secondary to
metastatic malignant disease. J Bone Joint Surg,63-A: 653-664, April 1981..63-A653
1981.
Matzkin, H.; Perito, P. E.; and Soloway, M. S.: Prognostic factors in metastatic prostate cancer. Cancer,72: 3788-3792, 1993.723788
1993
[PubMed]
Nercessian, O. A., and Joshi, R. P.:
General principles of surgical technique. In The Hip, pp. 951-958.
Edited by J. J. Callaghan, A. G. Rosenberg, and H. E. Rubash. Philadelphia,
Lippincott-Raven, 1998.
Orr, S. T., and Aisner, S.: Performance status assessment among oncology patients:
a review. Cancer Treat. Rep.,70: 1423-1429, 1986.701423
1986
[PubMed]
Patterson, B. M.; Healey, J. H.; Cornell, C. N.; and Sharrock, N. E.: Cardiac arrest during hip arthroplasty with a cemented
long-stem component. A report of seven cases.. J Bone Joint Surg,73-A: 271-277, Feb. 1991.73-A271
1991
Perez, J. E.; Machiavelli, M.; Leone, B. A.; Romero, A.; Rabinovich, M. G.; Vallejo, C. T.; Bianco, A.; Rodriguez, R.; Cuevas, M. A.; and Alvarez, L. A.: Bone-only versus visceral-only metastatic pattern in breast
cancer: analysis of 150 patients. A GOCS study. Am. J. Clin. Oncol.,13: 294-298, 1990.13294
1990
[PubMed]
Roila, F.; Lupattelli, M.; Sassi, M.; Basurto, C.; Bracarda, S.; Picciafuoco, M.; Boschetti, E.; Milella, G.; Ballatori, E.; Tonato, M.; and Del Favero, A.: Intra and interobserver variability in cancer patients'
performance status assessed according to Karnofsky and ECOG scales.. Ann. Oncol.,2: 437-439, 1991.2437
1991
[PubMed]
Sorensen, J. B.; Klee, M.; Palshof, T.; and Hansen, H. H.: Performance status assessment in cancer patients. An inter-observer
variability study. . British J. Cancer,67: 773-775, 1993.67773
1993
Vena, V. E.; Hsu, J.; Rosier, R. N.; and O'Keefe, R. J.: Pelvic reconstruction for severe periacetabular metastatic
disease.. Clin. Orthop.,362: 171-180, 1999.362171
1999
[PubMed]
Verger, E.; Salamero, M.; and Conill, C.: Can Karnofsky performance status be transformed to the
Eastern Cooperative Oncology Group scoring scale and vice versa?. European J. Cancer,28A: 1328-1330, 1992.28A1328
1992
Walker, R. H.: Pelvic reconstruction/total hip arthroplasty for metastatic
acetabular insufficiency. Clin. Orthop.,294: 170-175, 1993.294170
1993
[PubMed]
Yamashita, K.; Koyama, H.; and Inaji, H.: Prognostic significance of bone metastasis from breast
cancer. Clin. Orthop.,312: 89-94, 1995.31289
1995
[PubMed]
Zinser, J. W.; Hortobagyi, G. N.; Buzdar, A. U.; Smith, T. L.; and Fraschini, G.: Clinical course of breast cancer patients with liver metastases. J. Clin. Oncol.,5: 773-782, 1987.5773
1987
[PubMed]