close

Вход

Забыли?

вход по аккаунту

?

472

код для вставкиСкачать
1161
Necrosis as a Prognostic Factor in
Glioblastoma Multiforme
Fred G. Barker II, M.D.'"
Richard 1. Davis, M.D?
Susan M. Chang, M.D?
Michael 0. Prados, M.D?
' Neurosurgical Service, Massachusetts General
Hospital, Boston, Massachusetts.
Neuro-Oncology Service. Brain Tumor Research Center, Department of Neurological Surgery, University of California, San Francisco,
California.
Neuropathology Unit, Department of Pathology, University of California, San Francisco, California.
BACKGROUND. Many pathologists require the presence of tumor necrosis within
an astrocytic neoplasm to establish the diagnosis of glioblastoma multiforme (GM).
Two new grading systems for astrocytic neoplasms allow a tumor to be diagnosed
at the highest level of anaplasia without requiring the presence of tumor necrosis.
METHODS. To determine whether GMs without necrosis had biologic behavior most
compatible with a diagnosis of GM or of anaplastic astrocytoma (AA), we examined
the survival of 299 patients whose tumors were diagnosed as GM because they
contained endothelial proliferation and who were treated according to prospective
clinical protocols. Multivariate proportional-hazards survival analysis was used to
assess the importance of tumor necrosis after adjustment for other prognostic
factors.
RESULTS. Of 275 patients with GMs containing endothelial proliferation, 88% had
tumor necrosis. Absence of necrosis was associated with younger age and with
less extensive surgical resection. The absence of necrosis predicted longer survival
in univariate analysis ( P = 0.02) and after adjustment for age, Karnofsky performance score, and extent of resection in a multivariate analysis ( P = 0.04). However,
the magnitude of the survival difference was not clinically important: patients
without tumor necrosis had a median survival of 12.5 months, and those with
tumor necrosis had a median survival of 10.9 months. Kaplan-Meier survival rates
two years after diagnosis were 13.0% for patients with necrosis and 27.1% for
patients without necrosis; the difference in 2-year survival was not statistically
significant (difference in survival rates = 14.1%, 95% confidence interval -2.2% to
30.4%).
CONCLUSIONS. The survival of patients with astrocytic neoplasms containing endothelial proliferation and no necrosis conforms best to the pattern expected for
patients with GM rather than AA. This supports the classification of these tumors
as GM in the revised World Health Organization grading system and as grade 4
astrocytomas in the St. Anne-Mayo grading system. Cancer 1996; 721161-6.
0 1996 American Cancer Sociev.
Supported by National Institutes of Health
Grants CA 13525 and CA 09291 to the Brain
Tumor Research Center, Department of Neurological Surgery, University of California, San
Francisco.
Address for reprints: Fred G. Barker, M.D., Neurosurgical Service, Warren 905. Massachusetts
General Hospital, Fruit Street, Boston, MA
02114.
Received July 20, 1995; revision received October 25, 1995: accepted November 20, 1995.
8 1996 American Cancer Society
KEYWORDS: giioblastoma multibnne, anaplastic astrocytoma, necrosis, survival
analysis, prognostic factors.
V
arious grading systems for astrocytic neoplasms have been proposed
since the pioneering studies of Tooth' and Bailey and Gushing.' Most
grading systems in current or recent use divide astrocytic neoplasms into
t h ~ e e or
~ -fouP-"
~
levels of anaplasia. Correlation between the level of
anaplasia and survival after diagnosis is the method used to validate
grading systems for astrocytic neoplasm^.^^^^^^'*-'^
The presence or absence of necrosis has been used as a key diagnostic
criterion for the most anaplastic form of astrocytic neoplasm, the glioblastoma multiforme (GM).5,6,16817
However, two recently proposed grading
systems allow tumors that do not contain foci of necrosis to be classified
1162
CANCER March 15, 1996 I Volume 77 I Number 6
as GM (in the revised World Health Organization system4
[WHO-21) or Grade 4 astrocytoma (in the St. Anne-Mayo
as proposed by Daumas-Duport et al.). This is
possible in the WHO-2 system when the tumor contains
endothelial proliferation and in the St. Anne-Mayo system when the tumor contains three criteria: nuclear
atypia, mitoses, and endothelial proliferation. In the grading system used at the University of California, San Francisco (UCSF) since 1980, endothelial proliferation has
been a required criterion for the diagnosis of GM, without
a requirement for necrosis. Because the work of Burger
et al.5.18 and of OtherSb.tl.12.13,17
has established the importance of tumor necrosis in predicting poor survival, we
studied a cohort of patients with tumors diagnosed as
GM by UCSF criteria to compare survival in relation to
the presence of tumor necrosis. Specifically, we examined
whether the biologic behavior of tumors containing endothelial proliferation and no necrosis best conformed to
the pattern expected for GM or to that expected for anaplastic astrocytoma (AA).
METHODS AND CLINICAL MATERIAL
Patient Population
The present report includes a consecutive series of all
supratentorial GM patients enrolled in UCSF nonbrachytherapy protocols for newly diagnosed GM from January
1988 through November 1993. All patients had newly diagnosed, histologically proven supratentorial GM and
were enrolled prospectively in one of two consecutive
clinical protocols administered by the UCSF Neuro-Oncology Service. One hundred forty-seven patients were
treated on the first protocol and 154 patients were treated
on the second protocol. Patients included in this report
constitute a subset of total enrollment for these protocols:
the first protocol also enrolled patients with AAs and contrast-enhancing, low grade astrocytomas, and enrollment
for the second protocol was still in progress at the time
this analysis was performed. In addition, patients with
high grade oligodendroglial tumors were eligible for enrollment in both protocols, but are not included in this
report. Results for both protocols will be reported separately. Informed consent was obtained from each patient
or a responsible relative, and the protocols were approved
by the UCSF Human Experimentation Committee.
Patients entered the protocols after maximal tumor
resection and central pathologic confirmation of GM. In
the first protocol, patients received 60 Gy focal external
beam radiation therapy in standard daily fractions to the
tumor and a 2- to 3-cm margin. Hydroxyurea (300 mg/
mz orally every 6 hours on Monday, Wednesday, and Friday) was administered during radiation therapy. Adjuvant
treatment with 6-thioguanine (100 mglm’ orally every 6
hours; 12 doses) and carmustine (210 mg/m’ intravenously 4 to 6 hours after the last 6-thioguanine dose) was
begun after completion of radiation treatment. Six cycles
at 6-week intervals were planned.
In the second protocol, patients were randomized to
one of four groups in a two-by-two design. All patients
received focal external beam radiation therapy to the tumor and a 2- to 3-cm margin. Half of the patients received
60 Gy in standard fractions, and the other half received
70.4 Gy in twice-daily fractions of 160 cGy each (accelerated hyperfractionation). Half of the patients received
eflornithine (1.8 g/m2 orally every 8 hours; provided by
the National Cancer Institute) as a radiosensitizer and
half did not. Adjuvant chemotherapy was not used.
Patients whose tumors progressed after treatment on
protocol were offered treatment for recurrent GM, including reoperation, gamma-knife stereotactic radiosurgery,
interstitial brachytherapy (sometimes with iododeoxyuridine or bromodeoxyuridine as a radiosensitizer), and chemotherapy with standard or experimental agents.
Patient age and Karnofsky performance score (KPS)
were recorded at the time of protocol enrollment. Extent
of resection was defined as “biopsy” (less than 10% resected), “subtotal resection” (10-90% resected), and
“gross total resection” (more than 90% resected), based
on the surgeon’s intraoperative impression. Resections
classified by the surgeon as “gross total,” in which the
postoperative scan was felt to show less than 90% resection, were classified as “subtotal.” Survival was defined
as the interval from the date of diagnosis of GM until
death.
Pathologic Criteria
GM was diagnosed according to criteria used at UCSF.”
GM is defined as a neoplasm that is at least focally highly
cellular and shows nuclear pleomorphism, cytoplasmic
pleomorphism, and vascular endothelial hyperplasia
(cushions of apparent endothelial cells protruding into
the vascular lumen).” Tumor necrosis was not considered essential for the diagnosis of GM. Tumors meeting
these criteria would also be classified as GM according
to WHO-2 criteria.‘
Statistical Methods
All P values reported are two-tailed. The Mann-Whitney
U test and Spearman rank correlation were used for ordinal variables.” Survival curves were calculated with the
Kaplan-Meier product limit estimation. For KaplanMeier survival rate confidence intervals, Greenwood’s
standard error estimate was used; confidence intervals
for median survival were derived using the “reflection”
meth~ d. ‘ ~The
. ‘ ~ log rank test was used for univariate survival analysis; Cox proportional hazards models were
used for multivariate survival analy~is.‘~-‘~
Survival analyses were stratified by protocol enrollment.
Necrosis in Glloblastorna MuMfonne/Barker et al.
-Necrosis
TABLE 1
Baseline Characteristics of the Study Group
Age (median; range)
Male sex [no., percent)
KPS [no.,percent)
60
70
80
90
100
Unrecorded
Extent of resection (no., percent]
Biopsy
Subtotal resection
Gross total resection
1163
----
No necrosis
55 p (16-79 p)
184 (62%)
14 (5%)
28 (9%)
63 (21%)
113 (38%)
66 (22%)
15 (5%)
48 (16%)
219 (73%)
32 (11%)
Y)
5
0.4
P
e
a
K P S Karnofsky performance score.
Oe2
1
:
0.0
0.0
RESULTS
Three hundred and one patients with supratentorial GM
entered the 2 protocols during the study period. Of these,
two patients were excluded from analysis because their
initial KPS (50) was too low for protocol enrollment. Baseline characteristics of the study population are shown in
Table 1. As of January 1995, 275 of the 299 patients had
died. Median survival for the entire cohort was 11.3
months.
Pathologic specimens from 289 (97%)of the 299 patients were reviewed at UCSF; pathologic reports from
referring institutions were examined for the other 10 patients (3%).Presence or absence of necrosis was recorded
for 275 of the 299 patients (92%).There was no significant
difference in age, KPS, extent of resection, or survival in
the subgroup of patients for whom the presence or absence of tumor necrosis was not recorded. Necrosis was
present in 243 of 275 tumors (88%) and absent in 32
tumors (12%).
Necrosis in relation to other prognostic variables
In a prior multivariate proportional hazards analysis of
prognostic factors in this patient cohort, age, KPS, and
extent of resection were all found to be significant?' Necrosis was positively correlated with age, with older patients more likely to have tumors containing necrosis
(Spearman P = 0.01). There was no correlation between
necrosis and KPS ( P= 0.5).A trend ( P = 0.05) was noted
toward more frequent presence of necrosis in specimens
from gross total resections (30 of 31 patients; 97%) compared with subtotal resections (175 of 198 patients; 88%)
or biopsies (38 of 46 patients; 83%).
Necrosis and survival
The absence of necrosis predicted longer survival in univariate analysis ( P = 0.02). Patients without tumor necro-
I
I
I
1
0.5
1.0
1.5
2.0
Years after diagnosis
FIGURE 1. Kaplan-Meier curves showing survival of patients with glioblastomas multiforme stratified by presence of tumor necrosis. Solid
line: patients with tumor necrosis; dashed line: patients without tumor
necrosis. The difference in survival is statistically significant ( P = 0.02).
sis had a median survival of 12.5 months (95%confidence
interval, 10.2 to 22 months), and those with tumor necrosis had a median survival of 10.9 months (95%confidence
interval, 9.6 to 12.5 months) (Fig. 1). Kaplan-Meier survival rates 2 years after diagnosis were 13% for patients
with necrosis and 27.1% for patients without necrosis;
the difference in the 2-year survival was not statistically
significant (difference in survival rates 14.1%;95% confidence interval, -2.2% to 30.4%).
In multivariate analysis, after adjustment for age,
KPS, and extent of resection, absence of necrosis was still
a statistically significant predictor of longer survival ( P=
0.04). The difference in predicted survival (increment in
instantaneous mortality) between patients without and
with tumor necrosis was approximately equivalent to that
predicted for patients aged 50 years versus patients aged
65 years; patients with KPS of 90 versus those with a KPS
of 80; and patients who had gross total resection versus
those who had subtotal resection.
DISCUSSION
Since the work of Burgefs" and other^^*^.".'^*'^ demonstrated the importance of tumor necrosis in predicting
shorter survival for patients with astrocytic tumors, necrosis has been a critical diagnostic criterion for many
pathologists in diagnosing GM. Because two recently proposed grading systems for astrocytic tumors allow diagnosis of GM or Grade 4 astrocytoma in the absence of
1164
CANCER March 15,1996 / Volume 77 / Number 6
TABLE 2
Frequency of Grade 4 Astroeytoma without Necrosis in Previously
Reported Series
Authors
No. of tumors
Daumas-Dupofl et aln
Kim et al."
Wang and Ho''
Revesz et aIP3
188
181
I8
80
No. of tumors
without necrosis
Percent
37
14
2
27
20%
7.7%
11%
34%
tumor necrosis, we examined the survival of a large cohort of GM patients, some of whose tumors did not contain necrosis.
Necrosis proved to be a prognostic factor predicting
shorter survival in our study group. However, the magnitude of the survival difference between patients with and
without tumor necrosis was small: median survival was
10.9 months and 12.5 months, respectively. Necrosis was
correlated with older age, but remained a statistically significant prognostic factor after multivariate adjustment
for age, KPS, and extent of surgical resection.
Relative prevalence of GM without necrosis
If the diagnosis of GM without necrosis were rarely made,
it would be less interesting as a potential prognostic factor, as well as more difficult to study. Several groups have
reported the frequency of occurrence of the various diagnostic criteria of the St. Anne-Mayo system in their clinical material (Table 2). The proportion of Grade 4 astrocytomas lacking tumor necrosis ranged from 7.7% of 181
Grade 4 tumors12 to 34% of 80 Grade 4 tumors.13 The
latter series included only patients in whom diagnoses
were made by stereotactic biopsy. Because of the histologic heterogeneity of malignant gliomas, sampling bias
due to the small size of stereotactic biopsy specimens can
cause GMs to be misclassified as AAs, particularly when
a single diagnostic criterion, such as necrosis, is required
for the diagnosis of GM. In addition, surgeons usually
select areas of tumor that appear viable on imaging studies as targets for stereotactic biopsy, decreasing the likelihood of tissue necrosis in the samples obtained. We observed that necrosis was less frequent in biopsy specimens, as others have previously rep~rted.~".'~
In our series, containing only tumors diagnosed as
GM because of the presence of endothelial hyperplasia,
the frequency of GM without necrosis was 12% of 275
tumors. This value is a slight overestimate of the true
frequency of WHO-2 GM without necrosis in our population, because a few tumors examined during the study
period were diagnosed as highly anaplastic astrocytomas
by UCSF criteria despite the presence of tumor necrosis.
If these tumors had been included in the study group,
the frequency of GM without necrosis would have been
slightly lower. However, it is clear that the diagnosis of
GM or Grade 4 astrocytoma without necrosis is not rare,
occurring in 8-34% of patients with either WHO-2 GMs
or St. Anne-Mayo Grade 4 tumors.
Survival in patients with GM without necrosis
Although necrosis was a statistically significant prognostic factor in our study group, the predicted difference in
survival was not sufficiently large to be helpful to clinicians in predicting survival for individual patients with
GM. Other groups have examined the survival of patients
with St. Anne-Mayo Grade 4 astrocytoma whose tumors
did or did not contain necrosis. Daumas-Duport et a1.8
reported no statistically significant difference in survival
between patients whose tumors contained three or four
diagnostic criteria, a conclusion that has been confirmed
by other^,".'^ but these groups did not specifically examine the patient subgroup with endothelial proliferation
and no necrosis. Revesz et
did compare survival of
patients with and without tumor necrosis within the St.
Anne-Mayo Grade 4 category and found no difference in
patient age or survival. However, the relatively small size
of the study population in this report (80 Grade 4 tumors,
with 79 deaths) could have caused a significant survival
difference to be overlooked. In addition, misclassification
of tumors actually containing necrosis may have been
frequent in this series because all diagnoses were made
by stereotactic biopsy. No study to date has compared
survival of patients with and without tumor necrosis
within the WHO-2 diagnostic category of GM.
Our survival analysis may have been slightly biased
because some patients whose tumors contained necrosis
but not endothelial proliferation were not included in
our study group. These patients could have had relatively
longer survival than the patients with both tumor necrosis
and endothelial proliferation. Because these patients
were not eligible for inclusion in the second of the two
protocols examined here and instead were eligible for a
different treatment protocol, it is not possible to eliminate
this bias by recapturing these few patients. Because survival of patients with either three or four of the St. AnneMayo diagnostic criteria has been reported to be identical,'.I4 and because our results show slightly longer survival for those patients with GM without necrosis, it is
unlikely that patients with St. Anne-Mayo Grade 4 tumors
and no endothelial proliferation have substantially longer
survival than other Grade 4 patients.
We feel that the magnitude of the survival advantage
for patients with GM without necrosis is not large enough
to justify reclassification of these tumors as AAs. Patients
with AAs have substantially longer survival than patients
with GMs. In prior reports from our center, median survival for patients with anaplastic astrocytomas was 3.3
Necrosis in Glioblastoma MultiformelBarker et al.
years’” to 4 years,31substantially longer than the upper
confidence limit for survival in our cohort of patients
without necrosis. Others have reported similar res u l t ~ . ~ .Patients
” - ~ ~ whose tumors were classified as AA
with endothelial proliferation have been reported to have
median survivals of between 8.4 months (19 patients with
marked widespread endothelial proliferation)3s and 3.5
years (12 patients).34 Burger et aLS noted that patients
with AAs containing endothelial proliferation had shorter
survival ( P < O.OOOl), but the magnitude of the survival
difference was not specified. In the present report, the
median survival of 32 patients whose tumors contained
endothelial proliferation and no necrosis was 12.5
months, which is most consistent with classification of
these tumors as GMs rather than as AAs for purposes of
patient counseling and protocol design. Other investigators may prefer to regard these tumors as occupying an
intermediate position between GM and AA. The small
difference in survival demonstrated here may be important in analyzing the results of large randomized controlled trials of treatment for patients with GM.
In summary, 12% of tumors diagnosed as GM by
UCSF criteria lacked tumor necrosis. Necrosis was more
likely to be absent in tumors from younger patients and
in samples from less extensive surgical resections. The
absence of necrosis was a statistically significant prognostic factor predicting longer survival in both univariate and
multivariate analysis. However, the magnitude of the difference in survival was not large enough to justify reclassification of these tumors as WHO-2 anaplastic (Grade 3)
astrocytomas. These data support the WHO-2 classification of astrocytic tumors with endothelial proliferation
and no necrosis as GMs.
9.
Tooth HH. Some observations on the growth and survivalperiod of intracranial tumours. based on the records of 500
cases, with special reference to the pathology of the gliomata. Brain 1912;35:61-108.
2. Bailey P, Cushing H. A classification of the tumors of the
glioma group on a histogenetic basis with a correlated study
of prognosis. Philade1phia:J.B. Lippincott Co., 1926.
3. Ringertz N. “Grading” of gliomas. Acca Pathol Microbiol
Scarid 1950;29:51-64.
4. Kleihues P, Burger PC, Scheithauer BW. Histologic typing
of tumours of the central nervous system. Ber1in:SpringerVerlag, 1993.
5. Burger PC, Vogel FS, Green SB, Strike TA. Glioblastoma multiforme and anaplastic astrocytoma. Pathologic criteria and
prognostic implications. Cancer 1985;56:1106~11.
6. Nelson JS,Tsukada Y. Schoenfeld D. Fulling K, Lamarche J ,
Peress N. Necrosis as a prognostic criterion in malignant
supratentorial, astrocytic gliomas. Cancer 1983;52:550-4.
7. Zulch KJ. Histologic typing of tumours of the central nervous
system. Geneva:World Health Organization, 1979.
8. Daumas-Duport C, Scheithauer B, O’Fallon I, Kelly P. Grading of astrocytomas. A simple and reproducible method.
Cancer 1988; 622 152-65.
1.
Daumas-Duport C, Szikla G. Delimitation et configuration
spatiale des gliomes ckrebraux. Donnees histologiques,
incidences therapeutiques. Neurochirurgie 1981;27:273-
84.
Sayre GP. Tumors of the central nervous sys10. Kernohan JW,
tem. Washington, DC:Armed Forces Institute of Pathology,
1952.
11. Svien HJ, Mabon RF, Kernohan
12.
13.
14.
15.
16.
JW,
Adson AW. Astrocytomas. Proc StaflMeet Mayo Clin 1949;24:55-64.
Kim TS, Halliday AL, Hedley-White ET, Convery K. Correlates of survival and the Daumas-Duport grading system for
astrocytomas. I Neurosurg 1991;74:27-37.
Revesz T, Scaravilli F, Coutinho L, Cockburn H, Sacares P,
Thomas DG. Reliability of histological diagnosis including
grading in gliomas biopsied by image-guided stereotactic
technique. Brain 1993; 116:781-93.
Wang H-C, Ho Y-S. Clinicopathological evaluation of 78
astrocytomas in Taiwan with emphasis on a simple grading
system. J Neurooncol 1992; 13265-76.
Rougier A, Dartigues IF, Cohadon F. Supratentorial gliomas:
multivariate analysis of prognostic factors. In: Chatel M,
Darcel F, Pecker J, editors. Brain oncology. Dordrecht:Martinus Nijhoff, 1987:249-51.
Cohadon F, Aouad N, Rougier A, Vital C, Rive1 I, Dartigues
IF. Histologic and non-histologic factors correlated with survival time in supratentorial astrocytic tumors. J Neurooncol
1985;3:105-11.
17. Alvord EC Jr. Is necrosis helpful in the grading of gliornas?
Editorial opinion. 1 Neuropath Exp Neurol 1992;51:12732.
18. Burger PC, Green SB. Patient age, histologic features, and
19.
20.
21.
REFERENCES
1165
22.
23.
length of survival in patients with glioblastoma multiforme.
Cancer 1987; 59:1617-25.
Davis RL, Liu HC, Vestnys P, Resser KJ, Levin VA. Correlation
of survival and diagnosis in supratentorial malignant gliomas [abstract].1 Neurooncol 1984;2:267.
Russell DS, Rubinstein LJ. Pathology of tumors of the central
nervous system. 5th edition. Baltimore:Williams and Wilkins. 1989.
Sprent P. Applied nonparametric statistical methods. London:Chapman & Hall, 1993.
Simon R. Confidence intervals for reporting results of clinical trials. Ann Intern Med 1986; 105:429-35.
Machin D, Gardner MJ. Calculating confidence intervals for
survival time analyses. In: Gardner MJ, Altman DG, editors.
Statistics with confidence. London:British Medical Journal,
1989:64-70.
24. Cox DR. Regression models and life tables. 1 R Stat Soc [Ser
B] 1972;34: 187-220.
25. Harris EK. Albert A. Survivorship analysis for clinical studies.
New York Marcel Dekker, 1991.
26. Lee ET. Statistical methods for survival data analysis. New
YorkJohn Wiley & Sons, 1992.
27. Barker FG, Prados MD, Chang SM, Gutin PH, Lamborn KR,
Larson DA, et al. Radiation response and survival in patients
with glioblastoma multiforme. J Neurosurg. In press.
28. Glantz MJ, Burger PC, Herndon JE 2d, Friedman AH, Cairncross JG, Vick NA, et al. Influence of the type of surgery on
the histologic diagnosis in patients with anaplastic gliomas.
Neicrology 1991; 4 1:1741-44.
29. Chandrasoma PT, Smith MM, Apuzzo ML. Stereotactic biopsy in the diagnosis of brain masses: comparison of results
of biopsy and resected surgical specimen. Neurosurgery
1989; 24: 160 -5.
1166
CANCER March 15,1996 / Volume 77 / Number 6
30. Prados MD, Gutin PH, Phillips TL,Wara WM,Larson DA,
Sneed PK, et al. Highly anaplastic astrocytoma: a review of
357 patients treated between 1977 and 1989. Int / Rudiat
Oncol Biol Phys 1992;23:3-8.
31. Levin VA, Prados MD, Wara WM, Davis RL, Gutin PH, Phillips TL, et al. Radiation therapy and bromodeoxyuridine
chemotherapy followed by procarbazine, lomustine, and
vincristine for the treatment of anaplastic gliomas. IntJRadiat Oncol Biol Phys 1995;32:75-83.
32. Nelson DF, Nelson IS, Davis DR, Chang CH, Griffin TW,
Pajak TF. Survival and prognosis of patients with astrocytoma with atypical or anaplastic features. / Neurooncol
1985;3:99-103.
33. Curran WJJr, Scott CB, Horton J, Nelson JS, Weinstein
AS, Nelson DF, et al. Does extent of surgery influence
outcome for astrocytoma with atypical or anaplastic
foci (AAF)? A report from three Radiation Therapy Oncology Group (RTOG) trials. Neurooncol 1992;122 1927.
34. Fulling KH, Garcia DM. Anaplastic astrocytoma of the adult
cerebrum: prognostic value of histologic features. Cancer
1985;55:928-31.
35. Schiffer D, Chi6 A, Giordana MT, Leone M, Soffietti R. Prognostic value of histologic factors in adult cerebral astrocytoma. Cancer 1988;61:1386-93.
~
Документ
Категория
Без категории
Просмотров
8
Размер файла
509 Кб
Теги
472
1/--страниц
Пожаловаться на содержимое документа