International Journal of Pediatric Otorhinolaryngology

57 (2001) 55 – 65
www.elsevier.com/locate/ijporl

Aeration and drainage pathways of Prussak’s space

Tauno Palva a,*, Clarinda Northrop b, Hans Ramsay c
a
Department of Otolaryngology, Uni6ersity Central Hospital, Haartmanstr. 4, fin-00290 Helsinki, Finland
b
Temporal Bone Foundation, Boston, MA, USA
c
Uni6ersity of Helsinki, Helsinki, Finland

Received 17 June 2000; received in revised form 10 October 2000; accepted 10 October 2000

Abstract

Objecti6e: the purpose of this study was to document the aeration and drainage pathways of Prussak’s space.
Methods: 55 temporal bones with an age range from neonate to 11 years of age were serially sectioned to 20 microns,
every 10th section was saved and stained by Hematoxylin eosin. Each consecutive section was studied as to the
connections of Prussak’s space to adjacent compartments and measurements of the dimensions were made for both
Prussak’s space and its aeration pathways. Results: the classic aeration pathway via the posterior pouch, as described
by Prussak in 1867, was found in 34 ears (62%). In 19 (36%), aeration occurred superior to the posterior pouch in
the region of the lower lateral attic and the upper mesotympanum. In these ears the posterior pouch had formed, but
its superior limit ended blindly in the tympanic membrane. In two temporal bones the aeration pathway was from the
anterior pouch and in these cases also a blind posterior pouch had developed. In one temporal bone an auxiliary
pathway in addition to the route via the posterior pouch was through the roof (the lateral malleal ligamental fold)
of Prussak’s space to the overlying lateral malleal space. The height of the posterior pouch varied with a range from
0.5 to 2.4 mm, and the greatest width, which was towards the posterior tympanic spine, varied from 1.6 to 3.2 mm.
The superiorly blind posterior pouch was shorter, ranging from 0.4 to 1.2 mm in height. The most frequent route of
the chorda tympani nerve was running from 0.5 to 1 mm medial to the posterior malleal ligament in the anterior half
of the pouch, joining it posteriorly, or it was connected to it by a short fold, 37 ears (67%). In 18 cases (33%) the
chorda was surrounded by its own fold, thus entirely separate from the posterior malleal ligamental fold during its
entire course across the tympanum. Conclusions: we have made a detailed documentation of the aeration and
drainage pathways of Prussak’s space and the findings concur, with important modifications, with Prussak’s original
description. There is no evidence for contemporary claims that Prussak’s space would be aerated superiorly between
the lateral incudal and malleal folds. Prussak’s space and its aeration pathway is an unit of its own, entirely separate
of the major epitympanic compartments which are aerated via the tympanic isthmus. Due to frequent disease
processes in the lower lateral attic and posterior mesotympanum, Prussak’s space and its aeration pathway are likely
to become blocked. This may lead to obliteration of Prussak’s space and to the development of retraction pocket or
papillary ingrowth cholesteatoma. © 2001 Elsevier Science Ireland Ltd. All rights reserved.

Keywords: Epitympanic aeration; Posterior pouch; Prussak’s space

* Corresponding author. Tel. + 358-9-47173019; fax: + 358-9-47175010.

0165-5876/01/$ - see front matter © 2001 Elsevier Science Ireland Ltd. All rights reserved.
PII: S 0 1 6 5 - 5 8 7 6 ( 0 0 ) 0 0 4 4 3 - 2
56 T. Pal6a et al. / Int. J. Pediatr. Otorhinolaryngol. 57 (2001) 55–65

1. The early anatomic concepts of the tympanic membrane, and of its communi-
cation with the posterior pouch was very accurate
The anterior and posterior pouches were estab- and essentially holds good today.
lished during the latter half of the 19th century. In Helmholz [3] in 1868 found strong bundles of
his ‘Textbook of Ear Diseases and Anatomy of fibers in the medial wall of the posterior pouch
the Ear’ (in German) from 1881 V. Tröltsch [1] which led fanwise from the neck of the malleus to
ascribed the first descriptions of the pouches to the posterior tympanic spine and he gave them the
Cornelius in 1825 and to Arnold in 1839. How- name ‘posterior malleal ligament’. Medially it was
ever, he emphasized that only in his own descrip- found to be accompanied by the chorda tympani
tion from 1856 was the structure and meaning of nerve. He considered the ligamental fibers to be
the membrane forming the posterior pouch inter- important, together with the lateral and anterior
preted correctly, and the name of V. Tröltsch is malleal ligaments, in keeping the malleus in place.
since then attached to this structure. V. Tröltsch Touching the tough fibers with a needle caused a
believed that this membrane, thanks to its fibrotic clear movement of the malleus, whereas, touching
layer, increased the elasticity and vibrations of the the soft fold portion along the chorda tympani
tympanic membrane. The observation that he nerve had no effect. Helmholz agreed that the
thought linked the origin of these two structures superior pouch was in open connection with the
together was that, they both started from the posterior pouch and pointed out that a defect in
same site in the annular bone and their fibrotic the lateral malleal ligament would lead to a space
layers were identical. now called as the lateral malleal space [4].
For the posterior pouch, v. Tröltsch described Prussak’s description of a superior pouch was
how the removal of incus revealed a 3 – 4 mm followed by comments from several other authors.
high, and about 4 mm wide, irregularly triangular In 1870, Politzer [5] also observed connections
fold which extended from the annular bone to the from the superior pouch to the anterior pouch,
handle of the malleus. The fold was superiorly and in two temporal bones there was a connection
attached to the tympanic membrane but inferiorly superiorly, to a space above the lateral malleal
deviated from it forming a pouch open to poste- ligament. These observations were later confirmed
rior mesotympanum. The chorda tympani nerve by Hammar [6] by a study done in 1902 of
became connected to the medial surface of the develomental histology. In fetuses the advance-
posterior portion of the membrane before enter- ment of an air sac from the posterior pouch led to
ing its bony canal. The anterior pouch, on the a regression of embryonic tissue in Prussak’s
other hand, was not considered as a duplicate space and the anterior membrane of Prussak’s
tympanic membrane but was a composite struc- space was formed when the advancing front met
ture consisting of the processus gracilis of the air sac advancing from the anterior pouch.
malleus, the anterior malleal ligament, the chorda Only rarely did aeration occur from the anterior
tympani nerve and the inferior tympanic artery, pouch. Politzer later [7] confirmed this by pouring
enveloped by mucous membrane. liquid mercury into Prussak’s space and observed
Prussak [2] accepted the concept that the mem- its coming regularly out from the posterior pouch,
brane forming the medial border of the posterior but occasionally also from the anterior pouch.
pouch represented a duplicate tympanic mem-
brane. However, he questioned the idea of an
anterorosuperiorly blind pouch and searched for a 2. Contemporary concepts
possible opening either in the tympanum or over
the malleus neck to the anterior pouch. His start- We have reviewed the earlier studies at this
ing point was the observation that Shrapnell’s length because the well established aeration and
membrane was not fixed to the neck of malleus drainage pathways of Prussak’s space were largely
but left an air filled space between the two. Prus- forgotten during the latter part of the 20th cen-
sak’s description of this space, the superior pouch tury and the detailed anatomic knowledge thus
 T. Pal6a et al. / Int. J. Pediatr. Otorhinolaryngol. 57 (2001) 55–65 57

did not remain at the level set by the pioneers. We were infants up to an age of 23 months. There
have discussed many of these publications earlier were no evident congenital ear anomalies except
[4,8] and will return only to the extensively cited the left ear of a case of Goldenhar Syndrome. A
early publication by Proctor [9] in which Prus- few specimens were essentially normal but most
sak’s space was said to be aerated superiorly showed either an influx of amnion fluid cellular
‘‘between the lateral malleolar and the lateral content (AFCC) into the middle ears, or in older
incudal fold’’. However, both these folds are re- infants, a varying degree of inflammation caused
lated to the lateral malleal space, and are superior by AFCC combined by infectious changes. Four
to Prussak’s space, rarely with any contact to it. neonate and three infant temporal bones were
In Proctor’s more recent book [10] several of the made available by the ENT Department in
drawings are not consistent with reality, e.g. the Helsinki, plus one 8 and two 11 years of age.
figure 108 depicts the superior aeration route. Totalling in 55 temporal bones.
This erroneous concept seems to arise in part The preparation of celloidin embedded tempo-
from the fact that the lateral incudomalleal dupli- ral bones for horizontal serial sectioning after
cate fold was considered identical to the lateral fixation in 10% formalin was made in a routine
malleal ligamental fold, which is a separate posi- fashion and has been described earlier [13]. The
tion-fixed fold and forms the roof of Prussak’s sections were cut to 20 micron thickness, every
space. It seems likely that Proctor has seen a 10th section saved, and stained by Hematoxylin
defect in the anterior portion of the lateral incud- eosin. All the sections were studied by microscopy
omalleal fold, which sometimes occurs [11], but it and alterations in various compartments were
leads to the lower lateral attic, not to Prussak’s recorded. Additionally, scanned color images, us-
space. The problem with both these publications ing a Polaroid Sprintscan 35 Plus, were printed
[9,10] is that there is not a single figure of docu- out for an easy side by side recognition of the
mented anatomic evidence of the aeration and magnified compartment boundaries. The height of
drainage pathways from Prussak’s space. various spaces and structures was determined
In our earlier studies [4,8,11,12] we found occa- from the numbered sections, each 20 microns with
sional deviations from the regular aeration path- an interval of 0.2 mm. Horizontal widths were
way via the posterior pouch. Some rare abnormal measured directly from the microscope by using a
connections of Prussak’s pouch to other compart- measure (Graticules Ltd., Tonbridge, Kent), di-
ments were similar to those pointed out already vided in 100 lines, in one of the oculars.
by Hammar [6]) and by Politzer [7], but others
have been different and their frequency has not
been evaluated. The microdissection of temporal
bones [4] was not sufficiently sensitive to do the 4. Results
analysis in a detailed manner which is possible
only in serial sections. In this study, in addition to Microscopic analysis of the sections showed
reporting in detail of the structures forming the that the posterior pouch was present in all exam-
communications from Prussak’s space to adjacent ined ears. It functioned as the aeration and
compartments, we will report about their relative drainage pathway to Prussak’s space in 34 ears
frequency with an analysis of larger number of (62%) and in 21 specimens (38%) it had formed
serially sectioned temporal bones. but ended in a superiorly blind space. In this
latter, smaller group it did not communicate with
Prussak’s space but ended as a slit like structure
3. Material and methods inside the superior portion of the posterior half of
the tympanic membrane. In both the groups it
Forty-five temporal bones of children from the showed a considerable variation as to its height
Temporal Bone Foundation in Boston were avail- and width and to its relation to the chorda tym-
able for study. Of these 30 were neonates and 15 pani nerve.
58 T. Pal6a et al. / Int. J. Pediatr. Otorhinolaryngol. 57 (2001) 55–65

When the posterior pouch functioned as the panic spine. Maximally an open space between
aeration pathway, the changeover occurred in the tympanic membrane and the annular bone
0.2 – 0.4 mm when Prussak’s space moved poste- had a length of 1.5 mm (Fig. 2D).
rior to the malleus to become the posterior pouch. The posterior malleal ligament, starting from
This roundish space widened in a few sections to the posterior surface of the malleus handle near
its full size occupying the area between the the neck, showed in all cases a fanwise spreading
malleus and the posterior tympanic spine (Figs. 1 of the bundles which were attached to the
and 2). The height of the pouch varied in a range posterior tympanic spine and around it. The
of 0.5–2.4 mm, and the greatest width towards posterior portions of the ligament thus
the posterior tympanic spine varied from 1.6 to appeared in several sections while a full, continu-
3.2 mm. The tympanic membrane inserted regu- ous ligament bundle could be seen only in one or
larly to the lateral edge of the annular bone and a few sections. The number of such ligament
the posterior tympanic spine, continuing directly bundles in the fold varied largely, and even the
as the ear canal skin (Fig. 1D). In a few cases the defects in the thin fold portions were observed
tympanic membrane joined the ear canal skin allowing an additional communication to the
posterior to the edge of the annular bone, even if lower lateral attic and posterior tympanum
the pouch membrane ended in the posterior tym- (Fig. 3).

Fig. 1. Case 1, A 92-168, aeration pathway from the posterior pouch. An infant aged 8 months, right ear. (A) Section level 173.
Prussak’s space (p), 0.9 mm high, measures 1.7 × 0.9 mm, the chorda tympani nerve (c) is still medial to the malleus (m). (B) Section
0.9 mm inferior to A. The pouch (pp) posterior to malleus (m) measures 1.5×0.5 mm, the tympanic membrane (t) is of even
thickness and the ligamental fiber bundles (vertical arrow) appear only in the fold’s posterior portion. The chorda (c) is posterior
to the malleus and well distanced from the pouch. (C) Section 0.2 mm inferior to B. The pouch measures 2 × 0.6 mm, the curved
ligamental portion (oblique arrow) in the lateral surface of the fold is continuos from malleus to the posterior tympanic spine (s).
Separation of the surface layer of the tympanic membrane (vertical arrow) is artefactual. The chorda tympani nerve (c) remains
separate and surrounded by inflammatory tissue. (D) Section 0.4 mm inferior to C. The pouch measures 2.8 × 0.9 mm and its
posterior 2/3 (oblique arrow) is open to the mesotympanum full of granulation tissue. The remaining fold (vertical arrow) shows no
ligamental fibers. The tympanic membrane (t) fastens to the anterior tympanic spine (s), the chorda tympani nerve (c) is enveloped
by granulation tissue and runs 0.8 mm medial to the tympanic membrane. The total height of the pouch is 1.2 mm. This regular
aeration pathway contains secretion full of round cells, and there were signs of incipient organinization of the secretion. Original
magnification × 20.
 T. Pal6a et al. / Int. J. Pediatr. Otorhinolaryngol. 57 (2001) 55–65 59

Fig. 2. Case 2, A91-01, aeration pathway from an unusually large posterior pouch. A child aged 23 months, left ear. (A) Section
level 262. Prussak’s space (p), 0.9 mm high, measures 1.1 × 0.9 mm, the anterior membrane separates it from the anterior pouch (ap).
A thin membrane (vertical arrow) appears towards the upper portion of the posterior pouch (pp). The posterior malleal ligamental
fold, 1.8 mm long, connects malleus (m) to the posterior tympanic spine, an oblique arrow points to a ligament bundle. The chorda
tympani nerve (c) is medial to the malleus. (B) Section 0.9 mm inferior to A. The pouch (pp) measures 2.2 × 1.3 mm, the fold inserts
to the posterior tympanic spine (s) and extends slightly past it. The main portion of the fold is thin (oblique arrow) without
ligamental fibers, the chorda tympani nerve (c) is about to join the fold. (C) Section 1 mm inferior to B. The pouch becomes longer
and narrower, measures 2.8 ×0.7 mm and extends for 1.3 mm on the annular bone (b). The discontinuous ligamental bundles
(vertical arrow) became continuous in the next three sections and the anterior portion of the fold united with the tympanic
membrane halfway to the spine. Chorda tympani nerve (c) is united with the fold. (D) Section 1 mm inferior to C. The posterior
pouch is open to the mesotympanum (oblique arrow). Due to the remaining fold components the tympanic membrane anterior to
the vertical arrow is three times thicker than the posterior portion. i =incus. The total height of the pouch is 2.4 mm. The regular
aeration pathway is filled throughout with mainly acellular secretion. Original magnification ×20.

The other major aeration pathway from Prus- from the malleus to the spine, for a distance of
sak’s space led directly to the lower lateral attic or generally 2–3 mm (Figs. 4 and 5).
mesotympanum in 19 (34%) specimens. This path- The historically well known aeration pathway
way became effective in 0.2 – 0.4 mm after Prus- from Prussak’s space to the anterior pouch (5,6)
sak’s space had come to lie posterior to the was rare, appearing bilaterally in 2 (4%) of these
malleus, thus being much shorter than the more temporal bones (Fig. 6). The opening was imme-
frequent pathway via the posterior pouch. The diate, Prussak’s space still extended posteriorly
blind superior end of the posterior pouch was past the malleus when the anterior portions al-
0.2 – 0.6 mm inferior to Prussak’s space’s new ready had a large communication to the anterior
pathway to the lower lateral attic or mesotympa- pouch (Fig. 6B). A blind extension from the lower
num, and the pouch was only 0.4 – 1.2 mm in lateral attic towards Prussak’s space was seen in
height. Weak bundles of the posterior malleal both ears, an attempt towards the development of
ligament connected the medioposterior portion of a normal aeration pathway. Also here a small
the malleus to the inner layer of the tympanic posterior pouch had developed and the ligamental
membrane (Fig. 4) or to the superior portion of fibers spread to the posterior tympanic spine
the posterior tympanic spine (Fig. 5). Another set partly along the pouch wall and partly along the
of stronger ligamental bundles run in this fold medial surface of the tympanic membrane. This
60 T. Pal6a et al. / Int. J. Pediatr. Otorhinolaryngol. 57 (2001) 55–65

made the tympanic membrane in some sections The measurements on the dimensions of Prus-
(Fig. 6C) much thicker than regular tympanic sak’s space and the posterior pouch did not show
membranes. age related differences between the neonate and
Only in one ear in this series there was a infant bones. Prussak’s space in both groups
communication through the roof of Prussak’s could contain embryonal connective tissue. In a
space to the lateral malleal space (Fig. 7). This 8-year-old child Prussak’s space was obliterated
must be considered as an auxiliary aeration route fully but the posterior pouch was open. In the
as the regular aeration pathway was also present, 11-year-old children Prussak’s spaces were shal-
similar to that seen in Fig. 1, from Prussak’s space low and the posterior pouches of the same size as
via the posterior pouch to the mesotympanum. in the major portion of the material.
The chorda tympani nerve was in most in-
stances 0.5–1 mm medial to the ligamental fold in
the anterior half but posteriorly joined it, or was 5. Discussion
connected to it with a short fold, in 37 (67%) of
the cases (Figs. 2–4). The nerve entered its bony The concept of V. Tröltsch [1] of the nonvary-
canal adjacent to the insertion of the fold. In 18 ing constant structure of the membrane forming
cases the chorda tympani, surrounded by its own the posterior pouch, and its intimate kinship to
fold, was separate from the posterior malleal liga- the tympanic membrane, was partly erroneous
mental fold, the medial wall of the posterior and very black and white. He criticized Arnold,
pouch, during its entire course across the tympa- who had 17 years prior to him described the
num (Figs. 1, 5 and 6). membrane and the pouch correctly, of not under-

Fig. 3. Case 3, A88-28, A combined aeration pathway from the posterior pouch, the lower lateral attic and posterior mesotympa-
num. A newborn, left ear. (A) Section level 171. The medial wall of the posterior pouch (p) shows a short and broad ligament which
connects malleus (m) with the posterior spine (s). Chorda tympani nerve (c) is still medial to the malleus. (B) Section 0.6 mm inferior
to A. Only a tiny posterior tip of the ligament remains and a defect in the thin portion of the fold connects the pouch to the lower
lateral attic and posterior mesotympanum (oblique arrow). The chorda tympani nerve with its fold (c) blocks the strictly central
connection from the pouch to the lower lateral attic. (C) Section 0.4 mm inferior to B. The fold defect is passed and the medial wall
of the pouch (p) contains again ligamental bundles and the chorda tympani nerve (c) is united with it. (D) Section 0.6 mm inferior
to C. The pouch (p) is fully open to mesotympanum. The chorda tympani nerve (c) is close to the posterior tympanic spine (s). Total
height of the pouch is 1.6 mm. Despite the presence of AFCC in the mesotympanum and the lower lateral attic the aeration pathway
has remained free. Original magnification × 15.
 T. Pal6a et al. / Int. J. Pediatr. Otorhinolaryngol. 57 (2001) 55–65 61

Fig. 4. Case 4, A89-295, aeration pathway from the lower lateral attic and mesotympanum. A child 15 months of age, left ear. (A)
Section level 211. The pathway from a 0.8 mm high Prussak’s space is posterior to the malleus (m) where the space measures
0.8×1.1 mm and is open inferiorly to the lower lateral attic (ll) and mesotympanum. The chorda tympani nerve (c) is in connection
with the fold (oblique arrow) which no longer contains ligamental fibers. (B) Section 0.3 mm inferior to A. The pathway is also
medially open to the mesotympanum (vertical arrow), the chordal fold (horizontal arrow) remains connected to the tympanic
membrane (t) which posterior to the malleus (m) is much thicker than anterior to it. Total height of the aeration pathway is only
0.4 mm and it is free of disease. (C) A section 0.8 mm inferior to B. A separate slit-like posterior pouch (oblique arrow) appears
inside the tympanic membrane and measures 1.6 × 0.1 mm, the chorda tympani nerve (c) is connected to it with a broad pedicle.
(D) Section 0.1 mm inferior to C. Anterior portion of the pouch is open (vertical arrow) to the mesotympanum. The chorda tympani
nerve (c) remains separate from the fold but is connected to it. Total height of the superiorly blind pouch is 0.5 mm. i = incus,
s = posterior tympanic spine. Original magnification × 20.

standing its origin and function as an inner tym- in the 1881 textbook of the anatomy of the ear [1].
panic membrane, and appearing as a constant Today we can conclude that Helmholz, Rüdinger
structure. Similarly he did not accept Henle’s view and Henle were right in considering the mem-
from 1875 of the membrane as a structure which brane as a ligamental fold whose primary function
contained strong parallel fibers serving for hold- is to assist in holding of malleus in its proper
ing the malleus in position. Rüdinger’s view from position by posterior fixation.
1873 was similarly described as erroneous, be- The present results show that the dimensions of
cause he regarded the membrane as an indepen- the posterior pouch vary individually but usually
dent ligamental band of the malleus. it is bilaterally symmetrical. It formed the regular
Such was the influence of V. Tröltsch that only aeration pathway to Prussak’s space in 62% of the
his name has prevailed to the contemporary gen- specimens. In 36% the aeration and drainage
eration of specialists as the otologist who discov- pathways of Prussak’s space led directly to the
ered and described the pouches, comparable to lower lateral attic and mesotympanum thus
Prussak [2]. While he was the one who described forming another main aeration pathway. The lat-
them, his conclusion of the origin and function of ter route is much shorter than the regular path-
the pouch membrane was obviously at fault. As way, and could thus be more efficient. However,
Helmholtz [3] had described the posterior malleal the granulation tissue was seen to invade
ligaments already in 1868 V. Tröltsch should have Prussak’s space also from the lower lateral attic
been able to rewrite his early 1856 communication (Fig. 5).
62 T. Pal6a et al. / Int. J. Pediatr. Otorhinolaryngol. 57 (2001) 55–65

In all cases studied the posterior pouch was the lateral malleal space was found only in one
present, but in the specimens with the short aera- bone (2%) or slightly less than the 6% we saw in
tion pathway it was reduced in height and started 120 microdissections. This communication might
inferior to the drainage route of Prussak’s space. be missed in serial sections when they are exam-
The fiber bundles were present also in these blind ined at 0.2 mm distance. However, in the present
pouches testifying of its original function, provid- series the partition between the lateral malleal
ing support for posterior fixation of the malleus. space and Prussak’s space appeared fully intact at
A fact that needs emphasizing is that both main least in one and frequently in two or more sec-
aeration pathways start at levels inferior to Prus- tions leaving no doubt of its integrity.
sak’s space, from the mesotympanum, not at the
level superior to Prussak’s space (8, 9), 5.1. Clinical implications
epitympanum.
As already established by Politzer [5,7] and Both in microdissection and in serial sections
Hammar [6], aeration of Prussak’s space can oc- mucoid secretion is frequently seen to collect in
cur from the anterior pouch. We found this to the posterior mesotympanum and in the lower
happen earlier in 2% (4) and in the present series lateral attic. It adheres to the surfaces which are
it was seen in similar frequency (3.6%). In both close to each other, creates epithelial damage and
these ears the posterior pouch ended blindly in the gives rise to portals for organization. V. Tröltsch
tympanic membrane. A superior connection to [1] already noted that as a result of inflammatory

Fig. 5. Case 5, A89-120, aeration pathway from the lower lateral attic. An infant 5 months of age, right ear. (A) Section level 171.
Prussak’s space, 1 mm high, measures 1 × 1.1 mm and contains secretion and a portion of an epithelialized polyp (horizontal arrow).
The empty anterior pouch (ap) shows a tip of a mass of nonepithelialized granulation tissue (oblique arrow). Weak ligamental
bundles connected malleus (m) to the posterior tympanic spine. (B) Section 0.4 mm inferior to A. The pathway to low portion of
Prussak’s space (p) from the lower lateral attic (ll) is open only in this section, the epithelialized strand of granulation tissue
(horizontal arrow) blocked the superior sections. (C) Section 1.3 mm inferior to B. The separate posterior pouch (pp), starting 0.7
mm superior to this section, measures 2.8 × 0.5 mm. The ligament bundles from the malleus (m) to the posterior tympanic spine (s)
are continuous (vertical arrow). Chorda tympani nerve (c) is separate from the fold, and its own fold is heavily inflamed. (D) Section
0.4 mm inferior to C. The superiorly blind pouch is open to mesotympanum (oblique arrow) and had a total height of 1.1 mm. A
remnant of the ligament (vertical arrow) appears near the chorda tympani nerve which is in contact with the fold. Polyps have
formed (horizontal arrow) medially to the tympanic membrane (t). i = incus. Original magnification × 20.
 T. Pal6a et al. / Int. J. Pediatr. Otorhinolaryngol. 57 (2001) 55–65 63

Fig. 6. Case 6, A80-126, aeration pathway from the anterior pouch. An infant 5 months of age, left ear. (A) Section level 121.
Prussak’s space (p), 1 mm high, measures 1.5 × 0.9 mm, is anteriorly limited by the anterior and lateral malleal ligaments (oblique
arrow). (B) Section 0.4 mm inferior to A. Prussak’s space is aerated from and drains to the anterior pouch (ap) with a large pathway.
A vertical arrow points to a blind extension of the lower lateral attic towards Prussak’s space. (C) Section 0.8 mm inferior to B. The
blind pouch has united fully with the lower lateral attic (ll), had a height of 0.9 mm and is also medially open to mesotympanum
(vertical arrow). The chorda tympani nerve (c) and its fold, 0.7 mm medial to the tympanic membrane, is connected both to the
incus (i) and to the fold by inflammatory bridges. The highest point of the posterior pouch (vertical arrow), measuring 0.5 ×0.1 mm,
appears in the thick tympanic membrane. (D) Section 0.2 mm inferior to C. The fully formed posterior pouch measures 1.6 ×0.3
mm. Two polyps emerge from the fold, the larger uniting with the chorda (c). Ligamental fibers fasten to the posterior tympanic
spine (s), the tympanic membrane continues directly as the skin lateral to the spine. i =incus. The superiorly blind posterior pouch
had a total height of 0.8 mm, opened widely to the mesotympanum and the chorda tympani nerve entered bone separately from the
fold. Original magnification ×20.

processes the inner epithelial surfaces of the poste- pouch was not obliterated in these infant ears but
rior pouch sometimes adhered together so that the an ongoing organization process (Fig. 1D) sug-
space became narrowed, or fully obliterated. Fill- gested a likely future blockage or full obliteration
ing of Prussak’s space and the posterior pouch by of the aeration pathway. Even the short pathway
mucus was noted also by Prussak in his original directly from the upper portion of the lower lat-
publication [2]. The beginning organization of the eral attic did not protect Prussak’s space from the
secretion in Prussak’s space in a child led Politzer invasion of massive granulation tissue (Fig. 5B).
[5] to erroneously interprete these inflammatory The processes at these sites initially continue
strands as forming a genuine network of spaces intact with the basic disease processes. However,
between Shrapnell’s membrane and the neck of it should be pointed out that Prussak’s space and
the malleus. its aeration pathways, either via the posterior
The present material includes many temporal pouch or the lower lateral attic, form a small
bones with inflammation of the middle ear, compartment of its own, independent of the large
mostly resulting from an influx of AFCC. One of epitympanic compartments aerated via the tym-
the spaces nearly always affected was the lower panic isthmus. Disease processes in Prussak’s
lateral attic and in these ears there was generally space thus as such do not contribute to the persis-
also disease in Prussak’s space and its aeration tence of otitis media. The use of tympanostomy
pathways (Figs. 1, 2, 5 and 6). The posterior tubes do not necessarily guarantee healing of
64 T. Pal6a et al. / Int. J. Pediatr. Otorhinolaryngol. 57 (2001) 55–65

Fig. 7. Case 1, A91-01. An auxiliary aeration pathway from the lateral malleal space to Prussak’s space. A child aged 23 months,
right ear, section level 223. A membrane defect (p) connects the two spaces and measures 0.8 ×1 mm. Sections both superior and
inferior to this level showed increasing dimensions. Many bundles of the lateral malleal ligament are seen anterior to the membrane
defect, the one closest to the anterior limit of the defect is strong (oblique arrow). One weak bundle (horizontal arrow) outlines the
posterior defect margin. Secretion with small clusters of round cells fills both the defect area and the lower lateral attic (ll).
m= malleus, i = incus, a = anterior malleal ligament c = chorda tympani nerve, s =anterior tympanic spine, b = lateral attic bone.
Original magnification ×40.

Prussak’s space which may become obliterated extensive surgery. Especially in retraction pro-
even if the disease in the major compartments cesses, the most frequent disease form, early inter-
may appear quiet [8,14]. vention is both simple and allows a good chance
This process may have a variety of different for hearing preservation [15].
outcome. In some patients the obliteration may
lead to the development of a retraction pocket
cholesteatoma if Prussak’s space has been volumi- References
nous and Shrapnell’s membrane consequently has
suffered a deep retraction. Another mode is the [1] Von Tröltsch A Lehrbuch der Ohrenheilkunde mit Ein-
ingrowing of long papillae from Shrapnell’s mem- schluss der Anatomie des Ohres. 7.Aufl. Leipzig: Verlag
von FCW Vogel, 1881.
brane through its inflamed inner lining to the
[2] A. Prussak, Zur Anatomie des Menschlichen Trom-
secretion undergoing organization inside Prus- melfells, Arch. Ohrenheilkd 3 (1867) 255 – 278.
sak’s space. A third possibility is filling of Prus- [3] H. Helmholtz, Die Mechanik der Gehörknöchelchen und
sak’s space with a broad layer of connective tissue des Trommelfells, Pfüger’s Arch. Ges. Physiol. 1 (1868)
making Shrapnell’s membrane immobile but pre- 16 – 25.
[4] T. Palva, H. Ramsay, T. Böhling, Prussak’s space revis-
venting deep retractions. Such ears are likely to ited, Am. J. Otol. 17 (1996) 512 – 520.
remain stabile for life [8,14]. [5] A. Politzer, Uber die Höhlensysteme zwischen Trom-
Because of the two earlier mentioned possibili- melfell und Hammerhals, Med. Wochenschr. (Wien) 20
ties we recommend that children with recurring (1870) 252 – 253.
otitis media and immobile Shrapnell’s membranes [6] J.A. Hammar, Studien über die Entwicklung des Vorder-
darms und einiger angrenzenden Organe. 1,. Abt: allge-
be examined by otomicroscopy once a year until meine Morphologie der schlundspalten beim menschen.
adult age. Early discovery of a limited disease Entwicklung des Mittelohrraumes und des äusseren
process in Prussak’s space can be cured by less Gehörganges, Arch. Mikrosk Anat. 59 (1902) 471 – 628.
 T. Pal6a et al. / Int. J. Pediatr. Otorhinolaryngol. 57 (2001) 55–65 65

[7] A. Politzer. Lehrbuch der Ohrenheilkunde. 5 Aufl Stutt- cellular content to attic aeration pathways. Histological
gart, Verlag von Ferdinand Enke, 1908, pp 24–28. observations of infants aged 2 – 4 months, Am. J. Otol. 21
[8] T. Palva, L.-G. Johnsson, The epitympanic compartments, (2000) 62 – 71.
surgical considerations: a re-evaluation based on findings [13] C.C. Northrop, J. Piza, R. Eavey, Histological observa-
in a pair of temporal bones and a literature review, Am. tions of amniotic fluid cellular content in the ear of neonates
J. Otol. 16 (1995) 505–513. and infants, Int. J. Ped. Otolaryngol. 11 (1986) 113 – 127.
[9] B. Proctor, The development of the middle ear spaces and [14] T. Palva, L.-G. Johnson, H. Ramsay, Attic aeration in
their surgical significance, J. Laryngol Otol. 78 (1964) temporal bones from children with recurring otitis media.
631 – 648. Tympanostomy tubes did not cure disease in Prussak’s
[10] B. Proctor, Surgical Anatomy of Ear and Temporal Bone, space, Am. J. Otol. 21 (2000) 485 – 493.
Thieme, New York, 1989. [15] T. Palva, H. Ramsay, Chronic inflammatory ear disease
[11] T. Palva, H. Ramsay, Incudal folds and epitympanic and cholesteatoma creation of auxiliary attic aeration
aeration, Am. J. Otol. 17 (1996) 700–708. pathways by microdissection, Am. J. Otol. 20 (1999)
[12] T. Palva, C. Northrop, H. Ramsay, Effect of amnion fluid 145 – 151.