A Quantitative Histologic Comparison of the Thymus in
100 Healthy and Diseased Adults
SUZANNE MELEG SMITH, M.D. AND LUIS JAVIER OSSA-GOMEZ, M.D.
Smith, Suzanne Meleg, and Ossa-Gomez, Luis Javier: A quantitative histologic comparison of the thymus in 100 healthy and
diseased adults. Am J Clin Pathol 76: 657-665, 1981. The
characteristics of the normal adult thymus in both sexes were
determined in 50 cases of accidental death by a simplified quantitative histologic technic. A table of normal values derived from
these findings was used for making comparisons with thymus
from autopsies of 50 additional patients suffering terminal illness. It was observed that thymic changes considered to be part
of physiologic involution occurred earlier and were more advanced at a given age in males than in females. These changes
included disappearance of septae, diminished demarcation between cortex and medulla, and a decreased number of Hassall's
corpuscles with an increase in their size. Changes ascribed to
disease included accelerated involution of the thymus accompanied by loss of septae, smaller lobules, increased adipose
tissue and fusiform cells, a reduced number of lymphocytes and
Hassall's corpuscles and a relative increase in the number of
cystic corpuscles. No lymphoid follicles were observed. Changes
were not identical in all disease conditions. In neoplasia, the
limit between the cortex and medulla was preserved; in immune
diseases, the epithelial nests of the medulla were increased.
(Key words: Table of normal thymic histology; Histometric
thymic characteristics; Adult thymus; Disease-induced thymic
changes; Thymus involution; Age-related thymic changes; Sexrelated thymic changes.)
CONSIDERABLE DISAGREEMENT exists in the
literature regarding normal and disease-induced changes
in the thymus. The difficulties in interpreting the data
have been underlined by Hammar's 8 study of weight
variation in normal thymus and also by Rosai16 in his
description of histologic characteristics.
In extensive studies of the thymus in relation to autoimmune diseases3'4'5'6,710 and various neoplasias, 91519
findings have been controversial. Lymphoid follicles,
considered by some to be normal constituents of the
thymus, 1214 have been reported to be increased in autoimmune diseases. 410 Atrophy of the thymic medulla has
been found in mammary cancer.15 Hassall's corpuscles
have been reported to decrease in number in some neoplasias19 and to increase in renal cell carcinoma. 9
In an attempt to clarify some of these points, data
Received December 19, 1980; received revised manuscript and accepted for publication May 20, 1981.
Presented at the Annual Meeting of the Colombian Society of Pathology, Cali, Colombia, South America, June 1980.
Address reprint requests to Dr. Smith: Apartado Aero 90299, Bogota, Colombia, South America.
Department of Pathology, The Military Hospital and
Department of Pathology, Forensic Medicine Institute,
Bogota, Colombia, South America
are presented from simple quantitative histologic methods which provide objectivity and can be easily adapted
to routine laboratory procedures.
Organizing the data into a table of normal values for
histologic features of the thymus by sex and age, provides a basis for comparison with changes due to disease
and should be of value to the surgical pathologist in
evaluating thymus biopsy specimens.
Materials and Methods
Samples were taken in 100 consecutive autopsies
from the anterior and superior mediastinum of 50 patients considered to be normal (control group) who had
died from traumatic lesions with a survival period of
less than 12 hours and from 50 additional patients with
terminal illness due to infections (27 cases), immunerelated diseases (seven cases), diseases with severe biochemical disbalance (five cases), malignant neoplasia
(seven cases), and various conditions of three days' duration or less (four cases).
One to three slides, stained with hematoxylin and
eosin, were prepared from each case and studied without knowledge of the clinical history. The following
measurements were recorded:
1. Transverse diameter of the largest lobule. The largest lymphoepithelial group was measured (Fig. 1)
when no lobules surrounded by septae were found.
Measurements were made with a micrometer disc
ruled to 0.1 mm in one ocular and calibrated with
a stage micrometer on the microscope.
2. Presence of septae was indicated by + + + and their
absence by 0.
3. Fatty replacement of the lobules was labelled from
0 to + + + and thymic remnants in adipose tissue
were classified + + + + .
4. The percentage in each subgroup of cases with presence of clear boundary between cortex and medulla
was calculated.
5. Thickness of the cortex and medulla was measured
0002-9173/81/2011/0657 $00.95 © American Society of Clinical Pathologists
657
658
SMITH AND OSSA-GOMEZ
A.J.C.P. • November 1981
FIG. 1 (upper). In this thymic remnant, cut perpendicular to the vertical
axis, the delimitation between cortex
and medulla is clearly defined and permits easy measurements of the cortex
(185 microns) and the medulla (300
microns) Hematoxylin and eosin X 40.
FIG. 2 (lower). Hassall's corpuscle
in the center of thefieldcan be distinguished from epithelial nests by the
perinuclear vacuoles. It measures 106
microns Hematoxylin and eosin X 400.
in the largest lobule that was cut in a plane perpendicular to the vertical axis which showed roughly
the same cortical thicknuss on both sides of the
medulla (Fig. 1).
6. Lymphocytic infiltration was indicated by + (from
1 to 3 in medulla and from 1 to 4 in cortex).
7. The diameter of the largest Hassall's corpuscle was
measured in microns.
8. The number of Hassall's corpuscles per ten contiguous medium power fields (X100) was counted,
starting in the area where they were more frequent.
Solid (epithelial) Hassall's corpuscles (Fig. 2) were
distinguished from epithelial nests on the basis of
their having more than three concentrically layered
cells with a perinuclear vacuole.19 Cystic corpuscles
were defined as showing a central space containing
keratin or amorphous proteinaceous material, 16 surrounded by one or more layers of cells.6
9. The percentage of Hassall's corpuscles that was
cystic or that contained calcium was determined.
10. The presence of detritus in Hassall's corpuscles was
indicated by + + + .
11. The presence of epithelial cells (lying freely without
nest formation) was also indicated by + + .
ADULT THYMUS
Vol. 76 • No. 5
FIG. 3 (upper). Epithelial nests surrounded by limited number of lymphocytes taken from the thymus of a
case of bacterial endocarditis Hematoxylin and eosin X 100.
FIG. 4 (lower). Thymus from a female in the fourth decade. There is a
sharp delimitation between cortex and
medulla and three Hassall's corpuscles
can be seen Hematoxylin and eosin
X 100. Compare with Fig. 5.
*#»*>
\*
12. Fusiform cells were studied separately from the
previous group.
13. Epithelial nests were identified as cell aggregates
with a tubular or glomeruloid disposition (Fig. 3)
without corpuscle characteristics. 6 Their presence
was indicated by + + .
The quantitative histologic data were tabulated, for
both the normal and the diseased group, according to
sex and age by decade. Morphologic variations in the
various disease groups were compared with those in the
"Control Table" (Table of normal thymic values Table
1) developed from the data derived from the group of
50 normal individuals.
Results
The thymus of young people is easily distinguished
from the adipose tissue that surrounds it. An average
weight of 20 gm (minimum 12 and maximum 26 gm)
was found between the ages of 16 and 39 years. After
40 years of age the thymus is heavily infiltrated by
adipose tissue and the net weight is difficult to obtain
even though the parenchyma can be easily palpated. At
660
Table I. Quantitative Hi stologic Values Obtained from Thymus of Normal Controls (Control Table)
Decades of Age
2nd
3rd
5
3
II
3
Number of cases
M
F
Lobular size in microns
Mean ± SD (range)
M
4th
5th
6th
7th
575 ± 285
(167-968)
726 ± 218
(572-880)
860 ± 411
(528-1320)
616 ± 62
(572-660)
470 ± 309
(176-792)
Septac
M
F
+ 4—
4-4—
4-44-4-4-
4-4++
Fat
M
F
4-4+
4-4-44-4-
4-4-44-4-
4-4-4-44-4-4-4-
4-4-4-44-4-4-4-
4-4-4-4-
Delimitation of cortex and medulla
M
F
100%
100%
70%
100%
14%
100%
40%
50%
33%
0
33%
220
(220-;.20)
106
(44-167)
294
(176-484)
159
(132-185)
88
35;:
(308-MO)
607
(246-968)
472
(352-704)
278
(282-273)
403
(246-572)
308
4-4-4-44-4-4-4-
4-4-4-44-4-4-4-
4-4-4-44-4-
4-4-4-44-4-
4-4-4-4-
4-4-+
4-4-
4-4-44-4-4-
4-4-44-4-4-
4-4-44-4-
4-4-44-4-
4-4-4-
402 ± 172
(229-i',16)
185 ± 118
(53-282)
240 ± 98
(132-440)
185 ± 9
(176-194)
282 ± 161
(79-458)
205 ± 90
(106-325)
239 ± 68
(132-308)
132 ± 124
(44-220)
207 ± 168
(159-440)
127 ± 56
(88-167)
132+44
(88-176)
22 ± 18
(8-50)
33 ± 10
(25-45)
11 ± 5
(7-21)
12 ± 5
(5-19)
17 ± 8
(6-25)
8±3
(5-13)
13 ± 9
(5-18)
6±3
(3-8)
8 ±4
(5-10)
2±3
(0-6)
F
Cortex size in microns (range)
M
F
Medulla size in microns (range)
M
F
Lymphocytes (cortex)
M
F
Lymphocytes (medulla)
M
F
Largest Hassall's
corpuscle in microns
Mean ± SD (range)
M
Number of Hassall's
corpuscles*
Mean ± SD (range)
M
4- +
1018 ± 535
(352-1760)
1012 ± 440
(528-1584)
4- 4-
884 + 286
(264-1584)
792 ± 305
(616-1144)
4- +
4- +
933 ± 253
(616-1320)
965 ± 582
(343-1496)
20 ± 5
(15-24)
231
(88-440)
0
0
0
0
0
264
132
352
528
ADULT THYMUS
Vol. 76 • No. 5
661
Table 1. (Continued)
Decades of Age
2nd
3rd
4th
5th
6th
7th
Percentage of cystic
corpuscles
M
F
68%
60%
58%
65%
59%
55%
72%
45%
73%
50%
95%
Detritus in corpuscles
M
F
++
+++
+
0
0
+
0
++
0
0
0
Epithelial cells
M
F
++
++
++
+
+
+
+
+
++
++
+
Fusiform cells
M
F
++
0
+
0
0
0
0
Epithelial nests
M
F
+
0
0
++
• Per 10 medium power (XI00)
microscopic fields.
autopsy, it is unusual to recognize the thymus after the
sixth decade of age but remnants will generally be found
on histologic examination.
Microscopic cysts that contained colloidal material
surrounded by a single layer of cuboidal or respiratory
epithelium were found in 10% of specimens. In no case
was the thymus infiltrated by plasma cells nor was there
lymphoid follicle formation. No ectopic parathyroid
glands were found; there was no relation between thyroid pathology and changes in the thymus.
Control Group
The slides did not show thymus in two of the 50
specimens in this group, representing a recovery rate
failure of 4%. The Control Table (Table 1) shows the
findings for each decade of age in males and females.
Lobular size was similar for both sexes in the second
decade of age, measuring 933 microns in males and 965
in females. In both sexes the lobular size was greatest
in the fourth decade of age and later diminished with
the exception of males in the sixth decade. In both sexes
septae were clearly defined structures in the second
decade of age and disappeared in the fifth decade. In
the intervening years they were found more frequently
in females than males. Fatty tissue increased at the
same rate as septae disappeared, with a greater quantity
of fat present in males. In females, the boundary between cortex and medulla was sharp in all the cases up
0
0
+
0
++
0
0
0
SD Standard deviation =^ /sum of squared differences from average
0
0
++
to the fourth decade of age (Fig. 4); in the fifth it was
less frequent and by the sixth it could not be found. In
males it was less frequent in all the decades (Fig. 5)
except for the second. In some cases, despite the presence of a separation between cortex and medulla, these
components could not be measured because they lacked
the other required characteristics. Size of the cortex
fluctuated in both sexes between 88 and 294 microns
in different decades and the medulla between 278 and
607 microns (Table 1). Lymphocytes in cortex as well
as in medulla were numerous in males of any age; in
females they diminished in the fifth and sixth decades.
Hassall's corpuscles were always smaller in females
(Fig. 6); they shrank with age from 402 microns (for
males) and 185 microns (for females) in the second
decade to 207 and 127 microns, respectively, in the sixth
decade. Corpuscles were less numerous in males (Fig.
7) (22 per ten medium power fields in the second decade) than females (33 in the same circumstances) and
in both sexes they became sparse with age, reaching
counts of six and eight in the sixth decade. Hassall's
corpuscles are considered diagnostic of thymus; but, in
their absence, the organ can be recognized by the perivascular disposition of lymphocytes. The percentage of
cystic corpuscles increased with age in males; the opposite occurred in females. Calcium was present in 1020% of the corpuscles of females of all ages. In males
in the first five decades, calcium (Fig. 7) was present
in 10-20% of the corpuscles; there was an increase in
Vol. 76 • No. 5
ADULT THYMUS
663
FIG. 5 (upper, left). Thymus from a male in the fourth decade; there is no clear delimitation between cortex and medulla Hematoxylin
and eosin X 100.
FIG. 6 (upper, right). Hassall's corpuscles in a female in the second decade. The largest measures 190 microns; the smallest is solid and
measures 105 microns Hematoxylin and eosin X 100. Compare with Fig. 7.
FIG. 7 (lower). Hassall's corpuscles in this male in the second decade measure 273 microns; they are cystic and contain calcium
Hematoxylin and eosin X 100.
the sixth decade and, by the seventh, calcium was observed in 40% of the corpuscles. Cellular detritus was
present in cystic corpuscles in some decades (Table 1).
The amount of epithelial cells was the same in both
sexes, being more frequent at the two extremes of life.
Fusiform cells were never found in females (Table 2);
in males they were observed in some decades (second,
third, and fifth). Epithelial nests were only found at
certain ages: males in the second, sixth, and seventh
decades of age and females in the third.
Disease Group
Thymic tissue could not be identified microscopically
in 11 of the 50 specimens of hospital deaths, representing a recovery rate failure of 22%. Morphologic
characteristics for both sexes of the disease group are
shown in Table 3 and include diminished size of lobules
and reduced number of Hassall's corpuscles when compared to the Control Table (Table 1). In disease, septae
were less frequent (Table 4) and adipose tissue was
increased, while lymphocytes diminished and fusiform
cells increased. The percentage of cystic corpuscles was
increased in illness.
Cases with a disease of at least one week's duration
were divided into several groups, according to the most
important diagnosis.
a. Infectious diseases, 27 cases.
b. Malignant neoplasia, seven cases with diverse primary tumors.
c. Immune-related diseases: seven cases with various
diagnoses such as disseminated lupus erythematosus,
rheumatic fever, hepatic cirrhosis, and chronic renal
failure.
d. Diseases with severe biochemical disbalance (five
cases) such as diabetes mellitus, pancreatitis, and
chronic arsenic poisoning.
e. A small group of four patients was studied separately
as their disease had lasted less than three days, which
is not a long enough period for changes to become
apparent in the thymus. 1
Differences between these groups included persistence
of the boundary between cortex and medulla in neoplasia (Table 5). Epithelial nests were numerous in immune-related diseases, less so in infectious diseases and
absent in neoplasia.
Discussion
The simple quantitative, histologic method for study
of the thymus described here can be applied in routine
autopsy work or surgical pathology. As the study progressed recognizing the thymus at autopsy on palpation
of the mediastinum was increasingly simple, but (except
for younger patients) it was very difficult to decide what
to weigh because parenchyma and fat were intermingled. This could be a possible explanation for the great
variations that have been found in the weight of the
normal thymus. 8 Several microscopic characteristics
that represent physiologic (age) involution of the thymusi.6,i6.i7,i8,i9 w e r e c o n f i r m e d j n b 0 th sexes, such as loss
of septae and fatty infiltration. With increasing age, the
limit between cortex and medulla disappeared and it
was found that Hassall's corpuscles became smaller and
less numerous. It has been suggested that the size of
the corpuscle depends on its immunologic activity, as
it may play a major role in the destruction of forbidden
clones."
Histologic characteristics of the thymus are not identical in the two sexes. Septae and delimitation between
cortex and medulla (Fig. 4) were present into later decades of age in females. Other differences included a
wider cortex and a greater amount of adipose tissue in
males; more numerous Hassall's corpuscles (Fig. 6) in
females; larger corpuscles in males and the presence of
fusiform cells in males but never in females. The percentage of cystic corpuscles and the presence of calcium
(Fig. 7) in them increased with age in males. In females
the number of cystic corpuscles diminshed with age and
the presence of calcium was variable. It becomes ob-
Table 2. Differences in Histologic Characteristics
Between the Sexes in the Control Group
Septae
Fat
Delimitation of cortex and medulla
Thickness of cortex
Number of lymphocytes
Size of Hassall's corpuscles
Number of Hassall's corpuscles
Fusiform cells
Epithelial nests
Males
(34 cases)
Females
(16 cases)
±
++
+
++
++
++
+
++
++
++
+
++
+
+
±
++
-0+
SMITH AND OSSA-GOMEZ
664
A.J.C.P. . November 1981
Table 3. Quantitative Histologic Characteristics of the Thymus in Disease
Decades of Age
2rd
3rd
4th
5th
6th
7th
8.
i.
3
4
3
2
3
2
5
5
13
686 ± 388
(290-1320)
836 ± 311
(616-1056)
836 ± 311
(616-1056)
642 ± 205
(458-880)
850 ± 254
(704-1144)
603 ± 106
(528-678)
692 ± 261
(449-962)
343 ± 137
(246-440)
409 ± 106
(334-484)
499 ± 134
(352-616)
211 ± 122
(97-440
—
289 -: 177
(132-616)
594 :: 31
(572-616)
127 ± 81
(70-185)
154 ± 35
(106-185)
167 ± 102
(70-273)
440 ± 373
(176-704)
167 ± 92
(106-273)
154 ± 44
(123-185)
369 ± 447
(53-686)
147 ± 102
(88-264)
106 ± 118
(26-420)
—
19 :: 10
(5-32)
22 :: 14
(12-32)
8 ±6
(4-12)
12 ± 11
(1-26)
9+8
(4-18)
12 ± 4
(10-15)
8 ±6
(5-14)
12 ± 2
(11-14)
5±0
(5-5)
7±2
(5-8)
3±3
(1-10)
—
Number of patients
M
F
Size of lobules in
microns
Mean ± SD (range)
M
F
Largest Hassall's
corpuscle in microns
Mean ± SD (range)
M
F
Number of corpuscles*
Mean ± SD (range)
M
F
fields.
• Per 10 medium power (XIOmicroscopic
O)
SD Standard deviation
vious that any table of normal values (control table)
must account for the differences in histologic characteristics due to sex19 and age. The dissimilarities described between males and females might be relevant
to the variations in incidence of lymphoblastic lymphoma, a presumed thymic malignancy. 13
Histologic characteristics in the two sexes did not
follow a linear pattern through the decades; they
showed an abrupt change in the fourth decade when
compared with contiguous aj;e groups. As seen in Table
1, there was an increase in the size of the lobules and
of the corpuscles in the fourth decade for both sexes.
In the second and third decades the cortex was larger
in males than in females; in the fourth decade it became
smaller in males. Changes occurring in this age group
could reflect variations in the hormonal status of the
individual18 and suggest a need for further investigation
of the immunological status during middle age.
Because the diseased group consisted of consecutive
cases from a general hospital, the variety of pathology
resulted in subgroups too small to permit the drawing
of definite conclusions regarding the thymus. In agreement with other workers,16,19 it was observed that involutional changes increased in disease, with fusiform
cells becoming more numerous. However epithelial
nests were not decreased in infectious diseases as has
Table 4. Comparison of the Histologic Characteristics
of Thymus in Health and Disease
Size of lobules
Fat
Septac
Lymphocytes
Number of Hassall' s corpuscles
Percentage of cystic corpuscles
Fusiform cells
Health
(50 cases)
Disease
(50 cases)
++
+
++
++
++
+
+
+
++
+
±
++
++
Table 5. Comparison of Histologic Characteristics of
Thymus in Various Diseases
Number of cases
Boundary between
cortex and medulla
Epithelial nests
Infectious
Diseases
Neoplasia
Immune
Diseases
27
7
7
±
++
+++
±
+
+++
Vol. 76 • No. 5
ADULT THYMUS
665
3. Burnet FM, Mackay 1R: Lymphoepithelial structures and autoimmune disease. Lancet 2:1030-1033, 1962
4. Corridan M: The thymus in hepatic cirrhosis. J Clin Pathol
16:445-447, 1963
5. Goldstein G, Mackay IR: Contrasting abnormalities in the thymus in systemic lupus erythematosus and myasthenia gravis:
a quantitative histological study. Aust J Exp Biol Med Sci
43:381-390, 1965
6. Goldstein G, Mackay IR: The thymus in systemic lupus erythematosus: a quantitative histopathological analysis and comparison with stress involution. Brit Med J 2:475-478, 1967
7. Gun A, Michie W: Biopsy of the thymus. Brit J Surg 52:957962, 1965
8. Hammar JA: Die Menschen Thymus in Gesundheit und Krankheit. Z Mikrosk Anal Forsch (Suppl) 6:107-208, 1926
9. Hanash KA, Souadjian JV, Utz DC, et al: Quantitative morphologic studies of the thymus in hypernephroma. J Urol
105:488-490, 1971
10. Kamegaya K, Tsuchiya M, Sambe K: Thymic abnormalities and
the chronicity of liver disease. Keio J Med 29:77-90, 1971
11. Kouvalainen K: Significance of Hassall's corpuscles in the light
of their morphological and histochemical appearance. Ann
Med Exp Fenn 42:177-184, 1964
12. Middleton G: The incidence of follicular structures in the human
thymus at autopsy. Aust J Exp Biol Med Sci 45:189-199, 1967
13. Nathwani BN, Kim H, Rappaport H: Malignant lymphoma, lymphoblastic. Cancer 38:964-983, 1976 '
14. Okabe H: Thymic lymph follicles: a histopathological study of
1356 autopsies. Acta Path Jap 16:109-130, 1966
15. Papaioannou AN, Tsakraklides V, Critselis AN, et al: The thymus in breast cancer. Cancer 41:790-795, 1978
Acknowledgment. Dr. Charles H. Mullinax, D.V.M. provided ed- 16. Rosai
J, Levine GD: Tumors of the thymus. Fascicle 13, Atlas
itorial assistance.
of Tumor Pathology. Washington: Armed Forces Institute of
Pathology, 1976, pp 1-33
References
17. Shier KJ: The syncytial and medullary duct epithelium of the
1. Boyd E: The weight of the thymus gland in health and in disease.
thymus and thymomas. Am J Clin Pathol 71-614, 1979
Am J Dis Child 43:1162-1214, 1932
18.
Simpson
JG, Gray ES, Beck JS: Age involution in the normal
2. Burnet FM, Holmes MC: Thymic changes in the mouse strain
human adult thymus. Clin Exp Immunol 19:261-265, 1975
NZB in relation to the auto-immune state. J Pathol 88:22919. Souadjian JV, Silverstein MN, Titus JL: Morphologic studies of
241, 1964
the thymus in human neoplasia. Cancer 23:619-625, 1969
been previously described.19 This may be explained on
the basis that epithelial nests are condensations of epithelial cells and may result either from loss of lymphocytes or hyperplasia as a reaction to stress.6 The
group of immune-related diseases showed an interesting
gross finding: it was possible to weigh the thymus in
most cases. However, this bore no relation to the microscopic picture, which showed the usual atrophy 6 with
loss of delimitation between cortex and medulla. In this
group there was an increase in the number of epithelial
nests, a finding that has been described in other studies. 510 Epithelial hyperplasia may be due to auto-immune reactions against the thymus. 5 The presence of
lymphoid follicles in the thymus has been the object of
intensive research 3 ever since they were described in
rats with auto-immune anemia. 2 Many authors have
found an increase in lymphoid follicles in immune diseases in humans 4,510 while others 12 describe them in 51%
of autopsies of accidental deaths. In this study lymphoid
follicles were not found in any of the cases, either control or with immune-related diseases. This is not entirely
surprising if one considers that in a series of 1.365 autopsies14 lymphoid follicles were found in only 1.3% of
the thymus glands.