close

Вход

Забыли?

вход по аккаунту

?

Epidural spinal cord compression from metastatic tumor Results with a new treatment protocol.

код для вставкиСкачать
Epidural Spinal Cord
Compression from Metastatic Tumor:
Results with a New Treatment Protocol
Harry S. Greenberg, MD,* Jae-Ho Kim, MD, and Jerome B. Posner, MD
Eighty-three patients with epidural spinal cord compression from metastatic cancer were treated with high-dose
adr_enocorticosteroids and a new radiation fractionation protocol. Only those patients were included who had complete or almost complete block on myelography and who had not received prior radiation therapy to the area of
compression. Patients were given 100 mg of dexamethasone intravenously at the time of diagnosis and 500 rads of
radiation on each of the first three days. After a four-day rest, radiation therapy was continued at 300 rads to a total
dose of 3,000 rads.
The effects of this new protocol on the patient's motor abilities did not differ from those of previous protocols,
namely, 47 of 83 patients (57%) were ambulatory after treatment, with no rewonses in Datients d
before treatment. However, early administration of high doses of dexamethasone substantially ameliorated pain in
the majority of patients, with relief often coming within hours after the drug was given. On the basis of these data,
we recommend high doses of adrenocorticosteroids combined with ra&,.g&n therapy for acute treatment of spinal
cord compression. The optimum fractionation schedule for radiation therapy is not established.
Greenberg HS, Kim J-H, Posner JB: Epidural spinal cord compression from metastatic tumor: results with a
new treatment protocol. Ann Neurol 8:361-366, 1980
The treatment of epidural spinal cord compression
from metastatic cancer is unsatisfactory. In the best
series reported, fewer than two-thirds of the patients
have maintained or regained the ability to walk [ 4 , 6 ] ,
and in most series the ambulation rate after treatment
is 50% or less [l-3, 5, 71. In large series, the results
have been similar whether the patient underwent decompressive laminectomy followed by radiation therapy (RT) or was treated with RT alone [4, 71. Work
with an experimental model of epidural spinal cord
compression in rats has suggested that high doses
of dexamethasone accompanied by RT delivered
at doses somewhat higher than those conventionally used might be more efficacious than previous protocols [S]. Therefore, we undertook a prospective study of patients with epidural spinal cord
compression from metastatic tumor using high-dose
dexamethasone and a new radiation fractionation
course. This paper reports the results of that treatment in 83 patients compared with the outcome
in a historical control group from the same institution [4].
From the Departments of Neurology and Radiation Therapy,
Memorial Sloan-Kettering Cancer Center, Cornell University
Medical College, New York, NY.
"Present address: Department of Neurology, University of Michi-
Materials and Methods
For purposes of this study, the spinal cord or cauda equina
was considered to be compressed by epidural metastatic
tumor if there was either complete or greater than 80%
obstruction of the subarachnoid space at myelography. From
October, 1976, to July, 1978, 137 patients with compression of t h e spinal cord or cauda equina from metastatic
tumor were treated at Memorial Sloan-Kettering Cancer
Center. Forty-four patients were excluded from the study
either because they had received prior RT to the site of the
lesion or because their radiotherapists selected a fractionation schedule different from the one used in our study. Ten
additional patients were excluded because they underwent
decompressive laminectomy rather than primary RT. (Surgery was performed in these 10 patients because: 5 had
previously been irradiated to the area of spinal cord compression; 3 had no tissue diagnosis of cancer prior to the
cord compression; 1 had pancreatic carcinoma that his
physicians believed would be resistant to RT; and 1 had
worsening clinical symptoms during the course of RT.)
The 83 patients who entered this study were treated by
the protocol illustrated in the Figure: When the clinical
diagnosis was established (either before or immediately
after myelography), the patient received an intravenous
Received Nov 16, 1979, and in revised form Jan 24, 1980. Accepted for publication Feb 2, 1980.
Address reprint requests to Jerome B. Posner, MD, 1275 York
Ave, New York, NY 10021.
gan Medical School, Ann Arbor, MI 48109.
0364-5134/8O/l00361-06$01.25 @ 1979 by Jerome B. Posner 361
- 50G x 3 Radiation
therapy
-
8
16
ti
8
Dose
Irads)
1
z
3
4
96
96
96
48
Oav
Steroids
- 300 x 5
D e - m b n e
8
1 lQ"l"sllnll
IWQO
iwmpma~
tv n i v i '
5
48
6
32
7
32
9
16
Sferord dose incredred rl SynrDlomr
10
8
12
4
13
14
2 0
4
WcJSen
+
~
0)Prtrn~rqdose) 500 l a d s x 3 = 1500 rads
(21 Rest period 4 days
(31F r d c t l O i a l i O n 3w lads x 5 = 1 9 0 rads
Total DOSP 3000 rad5 In 12 days
Eou~vnlenl01 4 000 r i d s in 4 weeks 1200 rads x 2CI
Adrer2ocorticosteroid dose schedule and radiation therapy fracrionution schedule used in this study.
bolus of 100 mg of dexamethasone. The patient was then
continued on a dose of 96 mg of dexamethasone in four
divided doses (usually orally) for three days, then tapered
according to the schedule in the Figure. If the patient's
clinical condition worsened at any time during the tapering,
the dosage of dexamethasone was raised to the next higher
level and maintained there for another 48 hours before the
tapering schedule was begun again. RT was started within 2
hours of myelography in all but 1 patient, at a dose of 500
rads to a posterior port centered on the site of the block
and encompassing two vertebral bodies above and below it.
(The site of the lesion was determined by lumbar myelography. When a complete block was present, cisternal or lateral C2 myelography defined the upper margin of the lesion. The contrast material was not removed after the
procedure.) A daily dose of 500 rads was given for three
consecutive days, followed by a four-day rest, then 300
rads was given daily over five consecutive days. Repeat
fluoroscopic myelography was performed at the end of
treatment and again 2 months following treatment.
The patients were divided into four grades, based on
motor function. Grade I patients were ambulatory without
assistance, although they might be weak or ataxic. Grade I1
patients were not ambulatory but were able to move their
legs against ghvity in bed. Grade 111 patients were unable to
move their legs against gravity but could contract one or
several leg muscles. Grade IV patients were paraplegic.
Attention was also paid to other aspects of the neurological
examination, in particular to the degree of pain o n admission and its resolution with treatment; sphincter function
and sensory loss were also evaluated. Successful treatment
was defined as the patient's ability to walk when discharged
from the hospital two to four weeks following therapy,
whether o r not there was improvement in sphincter function or sensory loss.
Along with analysis of these patients, we reviewed 19
previously reported patients who were ambulatory 1 year
after successful therapy for spinal cord compression [ 4 ] to
define the long-term results of successful treatment.
Results
Table 1 lists the primary tumor and location of spinal
cord compression in the 83 patients. Carcinoma of
the breast was the most frequent primary neoplasm,
followed by prostatic and lung carcinomas. There
were 39 male and 44 female patients, with a mean age
of 56 years (range, 13 to 84 years). The interval between diagnosis of cancer and development of spinal
cord compression varied from one week to 23 years.
The 23-year interval was in a patient with epidermoid
carcinoma of the bladder. One patient with carcinoma of the breast developed spinal cord compression 19 years after diagnosis of the primary tumor.
The site of compression was the cervical spinal
cord in 12 patients (14%), the thoracic cord in 59
Table I . Type m d Location of Primary Tumor in 83 Patients with Epidrtral Spinal Cord Compression
Site of Compression
Primary Tumor
Breast
Lung
Prostate
Kidney
Lymphoma
Myeloma
Melanoma
Sarcoma
Head and neck
Gastrointestinal
Embryonal
Neuroblastoma
Female
reproductive
Miscellaneous
(thyroid, thymus, bladder)
Unknown
Total
No. of
Patients
Cervical
21
11
12
2
3
3
2
1
5
4
5
6
2
1
0
1
1
1
0
0
Thoracic
16
8
9
3
Lumbosacral
3
0
No. of Patients
Ambulatory at
Completion of
Treatment
3
0
0
0
1
1
2
1
0
0
1
12 (57%)
3 (27%)
8 (67%)
2 (67%)
2 (100%)
1(100%)
3 (60%)
2 (50%)
3 (60%)
3 (50%)
1(50%)
0 (0%)
2 (50%)
1
0
4
3
2
3
2
1
2
4
1
2
0
0
0
3
1
1
1
2 (67%)
3
0
12
3
59
0
83
12
2 (67%)
47 (57%)
362 Annals of Neurology Vol 8 No 4 October 1980
(71oJo), and the lumbar cord or cauda equina in 12
(14%) (see Table 1). Because our previous studies
showed no difference in the outcome of treatment
between cauda equina compression and spinal cord
compression, the two entities are considered together as epidural spinal cord compression.
The clinical symptoms are detailed in Table 2. The
symptoms of spinal cord compression were present
from a half day to 2 years prior to diagnosis, and pain
was the presenting symptom in 92% of the cases,
preceding other symptoms by a median of four
weeks. Although the percentages varied somewhat
from our previous studies, the clinical symptoms
were qualitatively similar to those found in both our
own and other previous studies. In 9 patients (1l%),
all with thoracic lesions, pain was the only symptom
or sign despite complete block on myelography in 4.
Plain radiographs of the spine were helpful in
diagnosis in that of 7 2 patients, 66 (91%) had metastases to the vertebral body at the level of compression. Complete block was shown at myelography in
5 1 patients and partial but greater than 80% block in
32.
Patient grading and the effects of treatment on
motor function are outlined in Table 3. The most
noteworthy difference in results between the new
protocol and our previous study was not in motor
function but in relief of pain. In 61 of the 83 patients,
the degree of pain relief was documented by both the
patient’s own report and nursing records of analgesic
intake. Of those 61 evaluable patients, 39 (64%) had
substantial reduction of pain the first day following
therapy. Six patients had complete relief of pain after
the 100 mg intravenous dose of dexamethasone, before the first dose of RT was administered. Such a
degree of pain relief is entirely different from our
experience using conventional doses of steroids or
R T without steroids. Many patients were able to
sleep lying supine for the first time in weeks. The
patient’s report of pain relief was verified by a substantial reduction in narcotic requirements on the
Table 2 . Signs and Symptoms of Epidural Spinal Cord
Compression in 83 Patients
Symptom
or Sign
Pain
(tenderness)
Weakness
Autonomic
dysfunction
Sensory loss
Ataxia
No. with
Initial
SvmDtoms
No. with
Symptoms
Present at
Diagnosis
No. with
Signs
Present at
Diagnosis
76
82
52
5
58
64
0
37
37
1
1
46
4
55
6
Table 3. Weakness by Grade, Incidence, and PatientJ
Ambularory at Completion of Treatment
~~
Weakness Grade
at Diagnosisa
No. of
Patients
No. Ambulatory
after Treatment
I (ambulatory)
38 (46%)
24 (29%)
13 (16%)
8 (10%)
83
34 (89%)
11 (nonambulatory)
111 (nonambulatory)
IV (paraplegic)
Total
‘Grades 111 and IV
=
10 (42%)
3 (23%)
0
47 ( 5 7 % )
Grade 111 of Gilbert and Posner [4].
first day following steroid treatment. A few patients
became tremulous and showed signs of narcotic
withdrawal after voluntary and abrupt discontinuation of their pain medication. Pain relief was also
achieved in 11 of the remaining 22 patients by the
end of RT, yielding 82% pain relief overall. The
other 11 patients either required narcotic drugs for
relief of pain (6 patients) or complained of pain but
did not require narcotics ( 5 patients). There was no
correlation between immediate relief of pain and
successful treatment outcome: 23 (59%) of the 39
patients with immediate pain relief eventually became ambulatory, similar to the outcome in the series
as a whole.
Serious complications of high-dose steroids occurred in only 1patient, who developed acute abdominal pain from a ruptured duodenal ulcer on the
fourth day following treatment but recovered following laparotomy. A few patients experienced hallucinations, which were mild and not disturbing.
Many patients complained of a severe but transient ( 5
minutes) burning sensation, either limited to the
genitalia or generalized, with the intravenous bolus
of dexamethasone. Some patients experienced insomnia and tremulousness (not due to narcotic withdrawal), and several became euphoric.
It was not possible to assess the effects of steroids
alone on motor function since RT was begun immediately after steroid therapy. In 1 patient, however,
an inadvertent overnight delay of R T allowed us to
discover that there was not only relief of pain but
considerable improvement in motor function within
Grade 11. However, the patient became able to walk
only after RT was completed. There were no other
short-term complications of RT.
The effects of the new treatment protocol on
motor function are analyzed in Tables 3 and 4 and did
not differ substantially from our previous findings
or those reported in the literature. After treatment,
47 (57%) of the patients were ambulatory: 34 of 38
patients (89%) ambulatory at the time of diagnosis
remained ambulatory (see Table 3). Among the patients who were ambulatory initially, the best results
Greenberg et al: Spinal Cord Compression by Tumor
363
were in those with cord compression from lymphoreticular disorders (3 patients) and breast carcinoma (8 patients), all of whom remained ambulatory; in prostate carcinoma, 7 of 9 patients (78%)
remained ambulatory. Lung carcinoma had the worst
outcome, with only 3 of 5 patients (60%) remaining
ambulatory (Table 4).
Thirteen of 37 patients (35%) improved from
nonambulatory to ambulatory with treatment, 10
from Grade I1 and 3 from Grade 111. Four patients
with breast carcinoma, 4 with miscellaneous tumors,
and 1 each with melanoma and head and neck cancer
improved from Grade I1 to ambulation. One patient
each with melanoma, prostate cancer, and miscellaneous tumors improved from Grade I11 to Grade I.
No patient with lung carcinoma and cord compression improved from nonambulatory to ambulatory
status with treatment. No patient who was paraplegic
(Grade IV) attained ambulation or even the ability
to move the legs against gravity (Grade 11).
Table 4. Outcome by Grade for Individual Tumor Types
Tumor TvDe
Breast
No. of
Patients
8
(N = 21)
Prostate
( N = 12)
5
4
4
9
2
1
0
Lung
( N = 11)
5
3
1
L
Melanoma
(N
=
5)
1
3
1
Head and neck
(N = 5 )
0
2
2
0
Lyrnphoproliferative
1
3
0
(N = 3)
0
0
Miscellaneous
( N = 26)
10
9
6
1
Total
364
83
Annals of Neurology
Entry
Grade
I
I1
I11
IV
I
I1
111
IV
I
I1
111
IV
I
I1
I11
IV
I
I1
I11
IV
I
Grade When
Discharged
I
I1
I11
IV
8
4
7
1
1
1
3
1
1
1
1
3
1
3
1
3
1
1
2
1
1
1
2
2
1
During treatment by RT, 6 patients became suddenly worse, with motor function declining one
grade. Three of these 6 had a sudden increase in pain;
2 had had no decrease in pain, and 1 developed a
confusional state. Three of the 6 improved with an
increase in steroid dosage.
Sphincter involvement was associated with a poor
prognosis: 37 of the 83 patients (4596)had sphincter
involvement and only 10, or 27%, regained ambulation. Return of sphincter function paralleled and
often preceded the return of motor function. In patients in Grade I1 o r I11 after treatment, sphincter
function often returned during their RT. In patients
who remained paraplegic, sphincter control did not
return.
When the tumors are divided into two groupsthose considered highly sensitive to R T and those
considered less sensitive [4]-the ambulation rate
was 62.5% for the patients with sensitive tumors and
55% for those with less radiosensitive tumors. Surprisingly, patients with prostatic and renal tumors,
generally not considered very radiosensitive, had the
highest rate of ambulation. Patients with lung carcinoma did poorly, with only 27% having a successful
outcome. The data for lung carcinoma are consistent
with those previously reported in the literature [7].
Changes on myelography generally correlated with
clinical improvement (Table 5 ) : 35 patients were ambulatory at the completion of treatment, and 21 of
them showed resolution of the block. In only 5 of the
14 nonambulatory patients did the block resolve. In
2 patients ambulatory after treatment, the block was
still present at the completion of R T but had resolved
when a myelogram was repeated 2 to 3 months following RT. In several others the block was still present weeks or months after RT, although the patient
remained ambulatory. Two other patients did not regain ambulation until after discharge, although their
block had improved while they were in the hospital.
Because this study has closed so recently, longterm follow-up is not possible in these patients.
However, we were able to analyze long-term
follow-up from the series reported by Gilbert et a1
1
1
3
Table 5 , Correlation of Repeat Fluoromye!ograms at Completion
of' Treatment with Ambulation in 49 Putients
I1
Patient State
No. with
Myelogram
Improved
(block
open)
No. with
Myelogram
Unimproved
(block
unchanged)
Ambulatory (N = 35)
Nonarnbulatory (N = 14)
Total 49
21 (60%)
5 (36%)
26 (53%)
14 (40%)
9 (64%)
23 (47%)
111
IV
I
I1
111
IV
10
4
1
1
2
1
2
3
1
1
47
13
11
12
Vol 8 No 4 October 1980
141, which included 6 patients treated with surgery
followed by R T and 13 treated by R T alone (using
another protocol) who were ambulatory at 1 year.
The data are detailed in Table 6. Of the 6 surgical
patients, 3 remained ambulatory 20, 27, and 44
months, respectively, after treatment. Three have
died, and 2 were ambulatory just prior to death at 23
and 36 Months; the third, a patient with chondrosarcoma, was ambulating 76 out of 84 months.
Of the 13 patients treated with RT without surgical
decompression, 5 remained ambulatory. Three patients suffering from lymphoreticular disorders were
still ambulatory an average of 42 months after treatment. O n e of them developed a second compression
at a site above the first, which was successfully
irradiated. Two patients, 1each with carcinoma of the
prostate and breast, were ambulatory 49 and 28
months, respectively, after RT. Another patient, with
prostatic carcinoma, was still ambulatory when lost to
follow-up 28 months after RT.
Seven patients treated by R T alone have died. Four
were ambulatory until just prior to death: 2 with
multiple myeloma were ambulatory an average of 6
years after treatment; 2 with breast carcinoma were
still ambulatory until death 12 and 26 months, respectively, after treatment. Three patients died
nonambulatory, each walking until the last month of
life, 16, 22, and 44 months after treatment. A patient
with breast carcinoma suffered recompression of her
spinal cord at the original site and underwent successful laminectomy, walking for 3 months until a
third compression occurred for which she was reirradiated. She then walked for 5 more months but
became paraplegic 1 month prior to death. One patient with thymoma died of brain abscess 17 months
after treatment.
Discussion
Two elements in this new protocol for the treatment
of spinal cord compression differ from those used
previously. The first is the high dose of dexamethasone, delivered first by bolus and then orally over the
first several days, and the second is the radiation
fractionation course. Both were chosen on the basis
of animal studies done by Ushio et a1 [ S ] , which
showed that high-dose dexamethasone improved neu-
Table 6. Long-term Outcome in Previously Reported Patients Treated by Radiation and Sargery Who Were Ambulatory cst 1 Year
Level
Treatment
Duration of
Ambulation
(mo)
I
I
I
I
I
I1
T5-7
C7-T3
T1
C8-Tl
Surg + RT
Surg RT
Surg + RT
Surg + RT
Surg RT
Surg + RT
20+
27 +
44 f
23
36
76
Lymphoma
Myeloma
Hodgkin's
I
I
I1
TI0
L2-4
T8-11
RT
RT
RT
24 +
46 +
54+
Breast
Prostate
Prostate
Myeloma
Myeloma
Breast
Breast
Th ymoma
I
I
I
I
I
I
I
11
T4-6
T4
L3
L5
T7
T2
T8-10
T8-12
RT
RT
RT
RT
RT
RT
RT
RT
28+
49 +
28?
Lung
I
TI1
RT
22
Breast
I
T7
RT
44
Tumor
Weakness
Grade
Chondrosarcoma
Myeloma
Breast
Myeloma
Lymphoma
Chondrosarcoma
c5
T9- 12
+
+
60
90
12
26
16
Outcome
Alive and well
Alive and well
Alive and well
Died (ambulatory)
Died (ambulatory)
Died (ambulatory 76 out of 84
m0)
Alive and ambulatory
Alive and ambulatory
Alive and ambulatory; recompression of T6 at 48 mo, reradiated
successfully
Alive and ambulatory
Alive and ambulatory
Ambulatory when last seen
Died (ambulatory)
Died (ambulatory)
Died (ambulatory)
Died (ambulatory)
Died (ambulatory 16 out of 17
mo, multiple recompressions
treated by R T + steroids)
Died (ambulatory 22 o u t of 23
mo)
Died (ambulatory 44 out of 45
ma; recompression at 36 mo,
laminectomy ; recompression at
39 mo. reradiated)
Greenberg et al: Spinal Cord Compression by Tumor
365
rological function in the first day after treatment and
maintained improvement for three to four days, even
if the animals were not irradiated. Improvement was
not sustained permanently, however, because the
animals all relapsed unless RT was delivered. The
second element of that study indicated that ambulation improved earlier in those animals given higher
initial doses of RT. Thus, 1,000 rads in a single dose
improved ambulation to a greater degree than did
three doses of 500 rad each or eight 200-rad doses.
However, maintenance of improvement was superior
when the animals were given three 500-rad doses
rather than one of the other two fractionation protocols. For that reason, we selected a protocol in
which a high dose of RT is delivered fairly rapidly
(over three days) and a consolidation course is given
over five days after a four-day rest period. The
radiobiological rationale is that high doses are more
efficient in inducing rapid tumor cell cytolysis and in
improving reoxygenation because of tumor shrinkage. Reoxygenation theoretically helps eliminate
the hypoxic fraction of a solid tumor, making that
tumor more susceptible to the RT which follows after
the rest period.
The results of steroid treatment were encouraging.
Most patients had rapid and complete amelioration of
pain with minimal morbidity, thus allowing them to
be considerably more comfortable during the days
they were being treated with RT. Side effects were
mild and probably no more frequent than in patients
treated with more conventional doses of steroids.
Thus, the use of high doses of steroids when the
diagnosis is first established seems warranted, at least
for relief of pain. The effects of steroids on motor
function were difficult to assess, and one cannot draw
any conclusions about them from this study.
The results with alteration of the RT protocol were
disappointing. There was a slight but not significant
increase in ambulation rate in each of the groups after
treatment, compared with the study of Gilbert et a1
[4].Two factors hint at the possibility that the higher
dose fractionation schedule may be slightly more
efficacious than our previous fractionation schedule.
The first is that there was no significant difference in
outcome in this series between patients whose
tumors were highly radiosensitive and those whose
366 Annals of Neurology Vol 8 No 4 October 1980
tumors were less so. The reason appears to be an increase in successful outcome among patients whose
tumors were less radiosensitive (from 45 to 55%
compared with our previous series), with no change
in the outcome of those tumors highly sensitive to
radiation. The second hint is that in the present
series, 1 of 3 patients with prostate carcinoma and 2
of 4 with melanoma who were unable to walk prior to
treatment regained ambulation, whereas in our previous series only 1 of 8 patients with prostate carcinoma, and none of 5 with melanoma, regained ambulation after having been paraparetic at the onset of
treatment. The numbers are too small for us to be
certain of the meaning of these results.
The salient finding in this study, as in previous
studies on therapy for epidural spinal cord compression, is that early and vigorous treatment is highly
effective, no matter what the radiation fractionation
protocol or even whether the patient undergoes decompressive laminectomy followed by RT as opposed to RT alone. Patients in whom diagnosis and
treatment are begun while they are still ambulatory
are highly likely to remain ambulatory. Patients in
whom diagnosis and treatment are deferred until
they are severely paraparetic or paraplegic are much
less likely to have a favorable outcome.
References
1. Cobb CA, Leavens ME, Eckles N: Indications for nonoperative
treatment of spinal cord compression due ro breast cancer. J
Neurosurg 47:653-658, 1977
2. Giannotta SL, Kindt GW: Metastatic spinal cord tumors. Clin
Neurosurg 25:495-503, 1978
3. Gilbert H, Apuzzo M, Marshall L, et al: Neoplastic epidural
spinal cord compression. A current perspective. JAMA
240:2771-2773, I978
4. Gilbert RW, Kim J-H, Posner JB: Epidural spinal cord compression from metastatic tumor: diagnosis and treatment. Ann
Neurol 3:40-51, 1978
5. Gorter K, Results of laminectomy in spinal cord compression
due to tumours. Acta Neurochir 42:177-187, 1978
6. Livingston KE, Perrin RG: The neurosurgical management of
spinal metastases causing cord and cauda equina compression. J
Neurosurg 49:839-843, 1978
7. Marshall LF, Langftt TW: Combined therapy for metastatic
extradural tumors of the spine. Cancer 40:2067-2070, 1977
8. Ushio Y , Posner R, Kim J-H, et al: Experimental spinal cord
compression by epidural neoplasms: treatment. J Neurosurg
47:380-390, 1977
Документ
Категория
Без категории
Просмотров
2
Размер файла
552 Кб
Теги
spina, treatment, metastatic, epidural, cord, protocol, compression, results, new, tumors
1/--страниц
Пожаловаться на содержимое документа