Professional Documents
Culture Documents
Color Atlas of The Anatomy and Pathology of The Epitympanum - (Palva-Ramsay-Northrop) 2001
Color Atlas of The Anatomy and Pathology of The Epitympanum - (Palva-Ramsay-Northrop) 2001
Color Atlas
of the Anatomy and
Pathology of the
Epitympanum
171 figures, 166 in color, 2001
Tauno Palva, MD
Professor Emeritus of Otolaryngology
University of Helsinki, Finland
The Temporal Bone Foundation has generated a
In collaboration with 3D video model of one of the temporal bones
described in Chapter 2. This model is available on
Hans Ramsay, MD a CD, which can be obtained by contacting:
Associate Professor of Otolaryngology The Temporal Bone Foundation, Inc.
University of Helsinki, Finland 9 Brimmer Street
Boston, MA 02108 (USA)
and Phone: +1 617 742 5927, Fax: +1 617 742 4666
Clarinda Northrop, B.A. E-Mail: rindyn@aol.com
Palva, T. (Tauno)
Color atlas of the anatomy and pathology of the epitympanum / Tauno Palva, Hans Ramsay, Clarinda Northrop.
p. ; cm.
Includes bibliographical references.
ISBN 3805572271
1. Middle ear--Anatomy--Atlases. 2. Middle ear--Histology--Atlases. I. Ramsay, Hans. II. Northrop, Clarinda.
III. Title.
[DNLM: 1. Ear, Middle--anatomy & histology--Atlases. 2. Ear, Middle--anatomy & histology--Case Report.
3. Ear, Middle--pathology--Atlases. 4. Ear, Middle--pathology--Case Report. WV 17 P184c 2001]
RF220 .P25 2001
617.8)4--dc21
2001029653
Bibliographic Indices. This publication is listed in bibliographic All rights reserved. No part of this publication may be translated
services, including Current Contents® and Index Medicus. into other languages, reproduced or utilized in any form or by any
means, electronic or mechanical, including photocopying, recording,
Drug Dosage. The authors and the publisher have exerted every microcopying, or by any information storage and retrieval system,
effort to ensure that drug selection and dosage set forth in this text are without permission in writing from the publisher.
in accord with current recommendations and practice at the time of
publication. However, in view of ongoing research, changes in govern- © Copyright 2001 by S. Karger AG,
ment regulations, and the constant flow of information relating to drug P.O. Box, CH–4009 Basel (Switzerland)
therapy and drug reactions, the reader is urged to check the package Printed in Switzerland on acid-free paper by
insert for each drug for any change in indications and dosage and for Reinhardt Druck, Basel
added warnings and precautions. This is particularly important when www.karger.com
the recommended agent is a new and/or infrequently employed drug. ISBN 3–8055–7227–1
OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
Contents
IX Preface
Part 1
Part 2
VI Contents
Part 3
Contents VII
OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
Preface
In many different fields of research progress has been major epitympanic compartments, contained features
made so rapidly that the most advanced knowledge of that I was at a loss to understand. Some sketches of the
today may be old-fashioned in a few year’s time, and the epitympanic folds were so elaborate that they appeared to
established procedures and techniques need constant preclude all effective aeration and drainage.
modernization. This fast progress is also obvious in many I finally went back to the original writing of Prussak
areas of medicine where, for example molecular biology and this experience made me reread other related original
and gene research have opened entirely new horizons for works, for example, by von Tröltsch, Politzer, Wittmaack
both diagnosis and treatment. On the other hand, notably and many others. However, only after a perusal of Ham-
in macroanatomy, previous generations have already mar’s study of the fetal development of epitympanic folds
done the basic studies, and their results can rarely be fun- did I feel close to understanding the epitympanic com-
damentally improved upon. partments and their origin. This was followed by going
The knowledge of human anatomy and pathology, personally back to the dissection laboratory and begin-
obtained at autopsies and complemented by light-micro- ning with the microdissection of temporal bones in an
scopic examination, has remained essentially unchanged effort to verify the folds and basic compartments that
for many decades, even if at a molecular level many new Hammar had described.
discoveries are being made. In modern otology, after the These data obtained by microdissection have during
Second World War, clinicians have been able to take the last 10 years been published in otological periodicals
advantage of the superb magnification of the surgical field and have helped to clear up many earlier misconceptions,
provided by the operation microscope, which has in- the epitympanic anatomy has emerged as a sensible and
creased the knowledge of the anatomy based solely on the logical arrangement. I had enormous help in elaborating
naked eye. The magnified view, nevertheless, has not ren- and deepening of the knowledge I obtained through mi-
dered conventional microscopic anatomy redundant for crodissection by having the continuing cooperation with
those who wish to understand the nature of the structures Clarinda Northrop of the Temporal Bone Foundation in
they are dealing with. In this area the work of the early Boston who permitted me to use their magnificent collec-
anatomists, even without our sophisticated tools for mag- tion of serially sectioned newborn and infant temporal
nification, still contains many valuable observations. Es- bones. Hans Ramsay, my former resident, who has now
pecially in the latter part of the 19th century the anatomy taken over my surgical work in the Department of Otola-
and pathology of the ear was a topic favored by many ryngology in Helsinki, made the material and facilities for
prominent scientists and clinicians. microdissection available to me. After my retirement
During the latter part of my 40 active clinical years, from clinical activity we have continued to enjoy fruitful
ending in the early 1990s, I became more and more dissa- joint research activity, now over a period of 10 years.
tisfied with the articles discussing the anatomy of the epi- During the last decade I have led the department semi-
tympanum and was convinced that some of the described nars of anatomy, pathology and surgery of the ear. It
features did not reflect the facts. The contemporary au- appeared that the initial level of knowledge of the resi-
thoritative descriptions, for example, of the anatomy and dents regarding epitympanic anatomy and pathology was
aeration of Prussak’s space, and of the structure of the not sufficient for a future ear surgeon. It also became clear
IX
that for self-education there were no reliable books which In Part 2 of this atlas we present in compact form the
would cover the entire area without the need to peruse present knowledge of the spread of the amniotic fluid cel-
individual reports from periodicals. Furthermore, it ap- lular content into the different compartments of the mid-
peared that the earlier literature did not contain anatomic dle ear. We also present a great deal of documentation of
documents but showed mostly sketches which often did the intensive foreign body-type tissue reaction caused by
not conform with reality. For these reasons I decided to amniotic fluid cellular content, at times combined with
write the present atlas which summarizes the work of our infection, on the soft tissue structures of neonates and
research groups during the last decade. In order to obtain infants. These sometimes massive reactions were already
maximal clarity it was decided to use only colored photo- well known to the investigators at the end of the 19th cen-
graphs for the description of the anatomy and pathology. tury, and both Aschoff and Wittmaack produced volumi-
Part 1 of this atlas was compiled in the hope that it nous documents, the contents of which are still valid
would, by ample documentation, help both the younger today. We believe that part of the long-standing problem
people in training as well as experienced specialists to of middle ear infection in infancy is initially related to this
have in one volume all data necessary for an understand- forgotten cause.
ing of the mysteries of the epitympanum. We hope that The photographs of serial sections shown in the atlas
reading it would stimulate the otologists to go personally were made by Mauri Laakso, Ari Aalto and Richard Cor-
to the dissection laboratory and practise the microdissec- tese.
tion approaches we have advocated, to become familiar The Ear Research Foundation, Helsinki, the Depart-
with the structures forming the epitympanic compart- ment of Otolaryngology, University of Helsinki, and the
ments. We are convinced that afterwards work in the Temporal Bone Foundation, Boston, Mass. with their
operation theater becomes even more interesting, when generous financial support have made the publication of
structures discovered and understood as a result of mi- this atlas possible.
crodissection are once again encountered. The procedures
for improving aeration and drainage we have outlined in December 2000 Tauno Palva
Part 3 are not difficult and are likely to lead to better func-
tional results once the surgeon becomes accustomed to
them.
X Preface
Part 1
OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
1
Part 1
OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
Introduction and General Review him, from now on will call either epitympanum or attic.
The word aditus has become established as indicating the
Development of the Concept of Epitympanum posterior open portion of the epitympanum connecting it
to the mastoid antrum. Siebenmann’s criticism has been
The division of the middle ear spaces into different accepted in that the word ‘recess’ has been dropped in the
anatomic subcompartments was established during the context of the epitympanum and is presently used in con-
latter part of the 19th century, and initially there was nection with the supratubal recess which indeed usually is
some confusion which name should be used for the por- a blind space as the word implies.
tion superior to the main tympanic cavity. In his concise Siebenmann cited the measurements of Bezold for the
chapter entitled ‘Mittelohr und Labyrinth’ in Karl von dimensions of the epitympanum, which are still valid
Bardeleben’s Handbuch der Anatomie des Menschen, Sie- today: just behind the level of the malleus head, the
benmann [1] still referred to this superior segment as ‘adi- breadth had a mean of 6.6 mm (range 5.3–8.0 mm). The
tus (ad antrum)’ but admitted there was additional usage height from the tip of the short process of the incus to the
in the form of ‘recessus epitympanicus’, ‘epitympanum’ superior wall was 5.7 mm (range 5.0–6.3 mm). The inferi-
and in the German ‘Kuppelraum’. Siebenmann pointed or limit of the attic anteriorly was formed by the tensor
out that most of his contemporary authors still considered tendon, from which the folds originated directed upwards
this area to be part of the tympanic cavity (Paukenhöhle) in form and direction (plicae transversae), attaching supe-
but von Tröltsch, Eysell and Schwartze, all well known riorly to the transverse crest. Siebenmann observed that
researchers, started to use the name ‘aditus’ for this supe- there was often a defect in this fold, varying in structure,
rior region, and that it was Bezold who described its ana- which then connected the epitympanum to the protym-
tomic limits. panic spaces. The area in front of the fold formed a ‘large
Siebenmann was horrified that the Americans referred pneumatic cell’, a portion of the tympanic cavity.
to the recessus epitympanicus by the name of ‘attic’ since The roof of the epitympanum or the floor of the middle
he argued that this term ‘was used in architecture to fossa was medially part of the petrous bone, the lateral
describe a half story adnex on top of the house proper’ and portion being part of the squamous bone, the suture line
did not like the emerging French usage of ‘attique’. He running lateral to the midline making the medial portion
warned his colleagues about adopting ‘this barbaric word, larger. The anterior portion of the tegmental bone, to-
even if latinized into the form of “atticus”, into the Ger- wards the eustachian tube, showed a downwards sloping
man language but one should continue to use the estab- form while the posterior portion remained horizontal.
lished word of “aditus”’. He also criticized the recently The lateral limit of the epitympanum was formed superi-
established nomenclature commission for recommending orly by a very hard bone, lacking air cells, extending in a
the ‘difficult concept of recessus epitympanicus’ because curved form to the notch of Rivinus, the edges allowing an
the word recess implies a blind extension of a certain insertion ring for the lateral inferior portion, Shrapnell’s
space which ‘aditus’ is not. membrane. The medial wall was formed by the labyrinth
Siebenmann then went on to describe exactly the limits capsule and posteriorly the space was open to the mastoid
of the ‘aditus’, which we nevertheless, with due respect to antrum.
3
The bony limits of the epitympanum have thus been membrane. The conviction of von Tröltsch that especially
established for already well over 100 years. Analyzing the posterior pouch was a duplicate of the tympanic mem-
them again and writing a new atlas of the anatomy of the brane was based on the observation that both of them
epitympanum would only be a repetition of known facts started from the same site in the annular bone and that
and a waste of time. However, the soft tissue structures their fibrotic layers were identical. Von Tröltsch believed
inside the epitympanum, or bordering it, have remained that this pouch membrane, thanks to its fibrotic layer,
obscure to most contemporary writers even if much of increased the elasticity and vibrations of the tympanic
their basics had been worked out satisfactorily at the end membrane. The name of von Tröltsch has since been
of the 19th century. As we did microdissection and histo- attached to these pouches.
pathological studies in the 1990s we also became familiar In his textbook von Tröltsch [2] gave a detailed de-
with the considerable knowledge available in the old writ- scription of the structure forming the posterior pouch. Its
ings. We were astonished by the concise description of the dimensions, measured after removal of the incus, were 3–
anatomic details and can but admire the quality of the 4 mm in height and up to 4 mm in width. It was irregular-
work the pioneers did with simple tools. ly triangular and extended from the annular bone to the
There were other authors who have contributed signifi- handle of the malleus. The fold was superiorly attached to
cantly to the knowledge of the epitympanum as we will see the tympanic membrane but deviated inferiorly from it
later on. Of the many we would like to mention several forming the pouch, open to the posterior mesotympanum.
here: von Tröltsch [2] who established the concepts of the The chorda tympani nerve became connected to the
anterior and posterior pouches as common knowledge, medial surface of the posterior portion of the membrane
Prussak [3] who described the superior pouch, Helmholtz before entering its bony canal. He definitely rejected
[4] who described the malleal ligaments, particularly the Henle’s observation from 1875 that the fold contained
posterior, and Hammar [5] whose contribution to the fetal distinct stiff bundles of ligamental fibers of connective tis-
development of the middle ear remains to be surpassed. sue.
As far as the histology of the epitympanum is concerned, The medial border of the anterior pouch, on the other
Wittmaack’s work [6, 7] has been colossal and much of it hand, was no longer considered [2] as a duplicate tym-
is still valid today. Many other research workers made sig- panic membrane. The pouch formed between the tym-
nificant contributions at the turn of the 19th century and panic membrane and a composite structure, consisting of
we will refer to them during the discussion of specific the long process of the malleus (processus gracilis), the
chapters. anterior malleal ligament, the chorda tympani nerve and
the inferior tympanic artery, enveloped by mucous mem-
brane. We may say even now that this observation has
Early Data of the Soft Tissues in the Epitympanum stood the test of time whereas, as we shall see later, the
concept of the posterior pouch has not. Several of the crit-
The anterior and posterior pouches became known in ical remarks that von Tröltsch made on the work of oth-
the first half of the 19th century. In his textbook (in Ger- ers, referred to above, were unfounded.
man) on anatomy and diseases of the ear from 1881 von The next step forward was made in St. Petersburg by
Tröltsch [2] reviewed the then current knowledge and the Russian otologist Prussak [3] who in 1867 described
found that Wildberg already in 1795 was aware of these the superior pouch of the tympanic membrane. Prussak’s
‘duplicates of the tympanic membrane’ but erroneously study testifies that astute anatomical observations with
considered them as malleal muscles. The dissertation of limited tools can lead to new, lasting discoveries. His
Cornelius from Dorpat in 1825 suggested, according to description of the superior pouch, now known as Prus-
von Tröltsch, the right idea of the folds forming the sak’s space, essentially holds true today. Although Prussak
pouches but he failed to note the connection the folds had accepted the concept of von Tröltsch [2] that the mem-
with the tympanic membrane. Arnold in 1839 had de- brane forming the posterior pouch represented a dupli-
picted the folds correctly both in writing and figures but cate of the fibrotic layer of the tympanic membrane, he
von Tröltsch thought he made a mistake by describing questioned the idea of an anterosuperiorly blind pouch
them as mucosal folds, and not as duplicates of the tym- and started to search for an opening either into the tym-
panic membrane. A further mistake was that Arnold con- panum, or over the malleus neck to the anterior pouch.
sidered the folds not constant and left them out of several His starting point was the observation that Shrapnell’s
figures which showed the medial side of the tympanic membrane was not fixed to the neck of the malleus but
normal pictures (fig. 3, 4), it is the picture representing In 1897 Siebenmann [1] summarized the main epitym-
pathological changes that many of our contemporary panic compartments and presented two alternative pic-
authors have reproduced as the sketch for Prussak’s space. tures of the somewhat varying relationships. Figure 5
This was seen e.g. in the book by Wullstein and Wullstein shows a still accurate representation of the posterior,
[11] who speculated on the important role of these ‘air superior and anterior pouches in relation to the lateral
cushions’, actually sequelae of inflammation in the lateral surface of the tympanic membrane. It is remarkable that
malleal space, in the mechanics of the middle ear. In he also presented here the anterior limit of the epitympa-
1995, Tos [12] no longer gave it a place in his manual, but num, the tensor fold, even if it was drawn in both sketches
several of his other sketches do not conform with Prus- as a vertical fold between the tensor tendon and the trans-
sak’s original description. verse crest. He thus also had a clear representation of the
supratubal recess, open to the mesotympanum and part of the eustachian tube to the inferior part of the tympanic
it. The discontinuous fold he showed near the tubal orifice membrane. As the air sacs expand, more and more of the
appears to represent an inflammatory fold, probably aris- meso- and epitympanic bony cavities open until finally
ing due to a, at that time common, diphtheric infection in most of the embryonal tissue has disappeared and the epi-
this ‘pneumatic cell’. tympanic subcompartments formed. In this context we
will not follow these processes in detail but first limit our-
selves to single out one essential point. When the air sacs
Fetal Development of Epitympanic Folds and expand, the epithelium from the eustachian tube orifice
Compartments expands covering the sac walls but loses its cilia and
becomes thinner, endothelium-like with one cell layer.
The only study recording different phases of the devel- When a sac meets a fixed structure it provides it with
opment of the various soft tissue structures in the epitym- mucosa and this is how the position-fixed structures, e.g.
panum, from a 3-mm embryo to the term fetus, was done ligaments, are coated by epithelium.
in 1902 by Hammar [5] in Uppsala. He sectioned the tem- When an air sac meets a position-fixed structure of
poral bone sagittally or frontally in 26 fetuses and made a another type, e.g. the long process of the incus or the chor-
model of each, recorded the findings in great detail and da tympani nerve, which can be passed on both sides, the
illustrated them by 67 drawings. This study is one of the sac divides into two, one portion continuing over the lat-
classics in otology but a great deal of effort and concentra- eral, the other over the medial side of the obstacle. Having
tion are necessary to read it, and being written in the Ger- passed it the connective tissue surfaces of both sacs merge,
man language it is unfortunately not available to the the epithelium remaining on both sides of the common
English-speaking research community. central layer. This is one of the ways in which the posi-
The discussion important for a clinician starts with the tion-changing duplicate folds arise and during the fetal
190-mm embryo when Hammar described 4 air sacs, the development they initially arise every time a structure is
anterior, medial, superior and posterior, which begin to passed by two sides. The other way is when two air sacs
replace the mesenchyme, starting anteriorly in the upper approach each other from different directions: the meet-
portion of the medial wall of the middle ear. Up to this ing connective tissue sides merge, and the epithelium pre-
stage of fetal development the bony middle ear space is vails on both sides. Large duplicate folds are likely to per-
filled only with the ossicles and mesenchyme, the em- sist even if they sometimes show membrane defects.
bryonic connective tissue, except for a slit running from Small duplicate folds, however, may remain slender and
malleal ligamental fold, the lateral malleal space remains in about 6% of our total material a membrane defect con-
in direct communication with the lower lateral attic. As nected the lateral malleal space to Prussak’s space (fig. 8,
already pointed out by Hammar [5], the interconnections 9). The lateral malleal ligamental fold, the roof of Prus-
of the lateral malleal space and the upper and lower lateral sak’s space, is not a curved half circle-type structure as
attics show considerable variation. In no case, however, is indicated by Proctor [13, 14] but more or less horizontal,
the lateral malleal space entirely closed because it receives as shown already by Helmholtz (fig. 1).
aeration from one of the adjoining spaces. Prussak’s Space. Part of the walls of Prussak’s space are
The mucosa of the lateral malleal space, while descend- bone, like the neck and the short process of the malleus,
ing towards the floor, becomes often thicker and undulat- forming medial and inferior walls, respectively. The soft
ing and does not seem to cover the structures tightly as is walls show some variability in size and shape, even if the
more the case at the superior portions. This apparently is structures are basically similar. The roof, discussed alrea-
related to the necessity of the mucosa to move freely, dy above, shows a similar unevenness as the floor of the
together with the nearly continuous movements of the lateral malleal space, so that a certain portion of the roof
malleus. The floor is not even but shows local areas that extends in a dome-like form more superiorly than the
extend deeper towards Prussak’s space than the adjacent main roof.
portions (fig. 7). When the roof of Prussak’s space is opened from
The floor of the lateral malleal space, the roof of Prus- above, or if a membrane defect allows a view of the inside
sak’s space, is subject to great variations in thickness. In of the space, the first structure seen is the round horizon-
all dissected bones the anterior portion has always been tal portion of the neck of the malleus (fig. 9). By viewing
the thickest due to the thick bundles of the lateral malleal anterosuperiorly past the neck of the malleus, and turning
ligament. More posteriorly the ligament is much weaker the anterior end of the specimen slightly medially, a por-
and individual separate bundles may be seen as transpar- tion of the thin anterior membrane of Prussak’s space is
ent whitish structures through the covering epithelium. revealed (fig. 10). However, this approach does not allow
The floor often shows a weak area posteriorly (fig. 7) and a full view of its inferior portion. Viewing along the neck
Serial Sections
Floor of the Lateral Malleal Space (Roof of Prussak’s
Space). In a section across the superior portion of the lat-
eral malleal space well above its floor, the anterior wall
consisting of the upper portion of the anterior malleal liga-
mental fold is relatively thin but thicker than the upper
portion of the posterior wall which consists only of
the downward turning thin lateral incudomalleal fold
(fig. 15a). Sections further down show stronger bundles of
the anterior malleal ligamental fold while the posterior
Fig. 14. Series A, adult case 7, left ear. Lateral anterior approach. The
wall remains thin or may, as in the ear shown, end before
tympanic membrane (TM) has been detached from the anterior sul-
reaching the posterior malleal ligamental fold (fig. 15b). cus and turned posteriorly together with Shrapnell’s membrane.
Thus in this case the lower lateral attic and the lateral mal- Prussak’s space (horizontal arrow) appears superior to the short pro-
leal space became united. Further sections showed the cess of the malleus (M) and extends superiorly past the rim of the
weaker bundles of the lateral malleal ligament, adjacent to lateral attic bone (B). This lateral anterior approach provides a view
of the low portion of the tensor fold (curved arrow).
the strong anterior malleal ligament, to spread from the
malleus neck to the opposite lateral attic wall, thus form-
ing the resistant anterior half of the floor. When the dome
of Prussak’s space made its appearance the lateral malleal parts of Prussak’s space are situated above the bony rim of
ligamental bundles were still present anteriorly (fig. 15c). the attic bone where the membrane has its insertion. At
Shrapnell’s Membrane. We limit ourselves in this con- times it is large, covering a sizeable Prussak’s space under
nection only to the observation that Shrapnell’s mem- the bony rim. Both these situations were well illustrated
brane (fig. 15d) is also subject to a considerable variation also by Politzer [10] in 1908 (fig. 3, 4). Even normally the
in size and in structure. It may be very small when most covering epidermis may show distinct changes, the papil-
lae at times being short and inactive, sometimes long and tendon. Being a duplicate fold it is thin, consisting of a
active, looking capable of deeper invasion towards the tiny connective tissue core with the surface constituted by
basal cell layer. a one-cell layer epithelium on both sides. Sometimes it
Anterior Membrane of Prussak’s Space. This frontal extends past the neck of the malleus to the tensor tendon
wall, limiting Prussak’s space anteriorly, arises as a dupli- (fig. 15d, 16) and in such cases a few collagen fibers from
cate fold when the anterior air sac, advancing posteriorly, the tensor tendon may extend even up to the tympanic
meets the expanding projection of either the medial or membrane. It should be stressed once more that this ante-
superior air sac, advancing anteriorly and forming Prus- rior membrane is a separate structure of its own and in no
sak’s space. The membrane fastens laterally to the tym- way part of the lateral malleal ligamental fold as indicated
panic membrane and medially regularly to the anterior in Proctor’s sketches [13, 14].
portions of the malleus neck, or anterior to it, to the tensor
book from 1881. It might be that all of them happened to tween the ‘lateral incudal and lateral malleolar folds’
be those in which the pouch ended blindly and von advocated by Proctor [13, 14] has not been documented
Tröltsch thus had no chance of finding the connection. and is not realistic.
The documents shown of the normal aeration and
drainage pathways of Prussak’s space by both microdis-
section and by serial sections leave no place for specula- Pathology of Prussak’s Space and the Lateral Malleal
tion. The two pathways, one via the posterior pouch and Space
another directly to the lower lateral attic and mesotym-
panum, are the norm and slight deviations along these Inflammation of the middle ear, be it caused by
routes may occur. The routes to the anterior pouch, or viruses, bacteria or foreign body-type noninfectious
through the lateral malleal ligamental fold are rare, facts agents, spreads to all spaces interconnected by the aera-
already noted by the early investigators [4, 5, 10]. The tion pathways. There is no anatomic obstacle between the
misconception that Prussak’s space is aerated from be- upper mesotympanum, the lower lateral attic and the orif-
ice of the posterior pouch, and thus the route is open via data of the frequency of these changes which occasionally
these regular aeration pathways to Prussak’s space. Simi- are noted either as indrawing or bulging of the membrane.
larly, if Prussak’s space is aerated from the anterior Perforation of Shrapnell’s membrane due to severe infec-
pouch, the route for the spread of infection is both short tion in Prussak’s space is very rare as a result of the escape
and wide. If there is an open communication between route for pus via the posterior pouch or even easier via the
Prussak’s space and the lateral malleal space, inflammato- lower lateral attic. Even if these routes were blocked, the
ry agents can spread in either direction. In acute otitis thin limiting anterior membrane, or the weak point in the
media the congested mucosa may temporarily occlude lateral malleal ligamental fold, would be much more likely
these communications, leading to absorption of air from to break than the relatively thick Shrapnell’s membrane,
Prussak’s space. conducting pus from Prussak’s space into either the ante-
In acute infection, changes in the pars tensa are so rior pouch or the lateral malleal space.
marked that they often mask those occurring in the small
Shrapnell’s membrane. There are no generally applicable
Fig. 21. Series G, adult case 12, right ear, superior view after removal Fig. 22. Series G, adult case 16, left ear, superior view after removal
of the incus. The roof of Prussak’s space lateral to the head of the of the incus. Prussak’s space (horizontal arrow) is filled with secre-
malleus (M) has been opened, revealing slimy secretion with a few tion and granulation tissue. The posterior malleal ligament bundles
strands of granulation tissue (horizontal arrow). The sucker touches (asterisks) show a fanwise spread, starting from the malleus (M) and
the anterosuperior edge of the posterior malleal ligamental fold inserting into the annular bone (B) and the posterior tympanic spine.
which shows both ligament bundles and thin membranous areas The membrane between the bundles is thin, the chorda tympani
(oblique arrow) and is united with the chorda tympani nerve (C). A nerve (C) runs in the inferomedial edge of the ligamental fold. A large
small portion of the anterior membrane of Prussak’s space (curved inflammatory web partly obstructing the tympanic isthmus extends
arrow) appears in front of the neck of the malleus. S = Stapes. to various directions from the head of the stapes (S). A = Anterior
malleal ligamental fold; T = tensor tendon.
of the lateral malleal space. If the floor contains a larger sak’s space has occurred, Shrapnell’s membrane being
weak area, this thin portion is in such cases retracted united with the neck of the malleus, this can be ascer-
towards the neck of the malleus into Prussak’s space tained by microdissection, but the information is meagre
(fig. 20). When the entire epitympanum shows granula- as compared to serial sections.
tion tissue the lateral malleal space shows similar involve-
ment. Serial Sections
In chronic infections, as soon as the roof of Prussak’s In our entire temporal bone material with chronic
space is opened, a mucoid fluid is frequently seen (fig. 21). inflammatory pathology there has been one constant find-
In the secretion there may be strands of granulation tissue ing both in adults and children, that is the accumulation
crossing and partially filling the space (fig. 22). These of inflammation products in the lower lateral attic. The
changes are observed similarly in the lateral approach via latter together with the superior and posterior mesotym-
the posterior tympanotomy. If a full obliteration of Prus- panum attracts both amniotic fluid cellular content
(AFCC) in the neonates [31] as well as sticky mucus with way via the posterior pouch can be seen in figure 23. Prus-
inflammatory cells in older children and adults [8, 31– sak’s space contains much secretion with round cells
33]. The early authors, including Prussak [3] and Politzer (fig. 23a) which continue to be present when the space has
[9] already noted that Prussak’s space may be filled with moved behind the malleus (fig. 23b). More inferiorly, the
the same thick mucus that was seen in the lower lateral pouch has become longer and narrower, contains secre-
attic and in the posterior pouch. A permanent block of tion, and the chorda tympani nerve has joined the medial
aeration to Prussak’s space results in a retraction of pouch wall (fig. 23c). Finally, when the pouch opened to
Shrapnell’s membrane and organization of the secretion the mesotympanum, it continued for 1.5 mm on the later-
starts, fibrocytes and capillaries entering the secretion via al side of the annular bone (fig. 23d).
portals developing in the areas of epithelial breaks. Even if the short and generally wide aeration pathway
We start this section by showing a documentation of from the lower lateral attic seems likely to provide a gen-
the different modes of aeration and drainage pathways in erous pathway for drainage of secretions (fig. 18), exten-
the presence of inflammation. The regular aeration path- sive inflammatory processes in the lower lateral attic may
also spread to Prussak’s space. In this type of a process (fig. 25b). A blind extension from the lower lateral attic
Prussak’s space can be invaded by granulation tissue aris- towards Prussak’s space testified of a disturbed attempt
ing from the lower lateral attic and advancing to Prussak’s towards the development of a normal aeration pathway
space (fig. 24a, b). The more inferiorly situating blind pos- (fig. 25b, c). Here a small posterior pouch had also devel-
terior pouch was relatively free of pathology even if there oped and the ligamental fibers spread to the posterior tym-
was much inflammatory tissue around the chorda tym- panic spine both along the pouch wall and partly in the
pani nerve and its fold (fig. 24c, d). medial surface of the tympanic membrane. This made the
In serial sections we have seen the aeration pathway to tympanic membrane much thicker than the regular tym-
the anterior pouch in both ears of one case. There was a panic membranes, until the pouch opened (fig. 25d).
normal Prussak’s space which contained a small amount of The superior pathway through the roof of Prussak’s
secretion with round cells (fig. 25a). The communication space is shown in figure 26 and we have observed it only
with a large pathway opened early to the anterior pouch once in serial sections. It appeared to be an auxiliary aera-
when Prussak’s space still was present lateral to the malleus tion route as the regular aeration pathway went normally
Fig. 27. Series P, case 9, a child aged 18 months, right ear. Prussak’s
space (P) contains only a short stretch of flat epithelium laterally;
there is connective tissue ingrowth from all directions. A vertical
arrow points to one of the several organizing tissue strands with capil-
laries. The cells in Prussak’s space are mostly fibrocytes with some
round cells. A tissue strand (oblique arrow) has separated a small
air-filled corner off the main space. Shrapnell’s membrane (S) is sepa-
rated from Prussak’s space by a thick connective tissue layer infil-
trated by round cells. A change of this kind leads to obliteration of
Prussak’s space. AP = Anterior pouch; LL = lower lateral attic. M =
malleus. Hematoxylin eosin stain. Magnification !54.
from Prussak’s space via the posterior pouch to the meso- severe retraction of Shrapnell’s membrane (fig. 30). Res-
tympanum. One might speculate that because the sections toration of an air filled Prussak’s space is no longer possi-
are 0.2 mm apart, the intervening sections could in many ble [33].
other cases easily have shown a similar defect. However, This retraction can result in two other different end
experience in microdissection, during which a membrane stages. One is the invagination of Shrapnell’s membrane
defect can always be seen, testifies that the frequency of forming so deep a retraction that the horizontal migration
such an occurrence is low. may soon be incapable of removing the keratin layer out
The filling of Prussak’s space with cell-rich secretion of the pocket (fig. 31). This is a nonreversible stage and
leads to the initial organization from portals which devel- will lead to the development of a retraction pocket choles-
op through epithelial breaks in the mucous membrane. teatoma. Another possibility is the filling of Prussak’s
Organization may occur in Prussak’s space itself (fig. 27) space with abundant granulation tissue which finally,
or anywhere along the aeration pathways (fig. 28). If the after maturing, fills the entire space allowing even consid-
pathway is sufficiently large an infection need not block erable retraction but no pocket formation, the keratin
the routes fully even if some organization of secretion transport being unimpaired (fig. 32). Such ears, even if
appears through the entire pathway (fig. 29a–d). Some of Shrapnell’s membrane is retracted and nonmobile, are
these processes, when intense, totally block the posterior stable and the papillae do not advance to the dense con-
pouch, causing absorption of air from Prussak’s space and nective tissue.
leading to obliteration of Prussak’s space, without or with
Cholesteatoma in Prussak’s Space advantage is that it can be diagnosed and treated already
in the early stages.
As appears above, cholesteatoma in Prussak’s space The second form is the development of cholesteatoma
develops in two basically different forms. Retraction from Shrapnell’s membrane by papillary ingrowth, the
pocket cholesteatoma presupposes absorption of air from start of which was already demonstrated by Wittmaack
Prussak’s space due to blocked aeration pathways, the [6, 7]. This form, in addition to a destruction of the lining
posterior pouch and the lower lateral attic. This is fol- epithelium of Prussak’s space, also presupposes a break in
lowed by retraction of the posterior portion of the pars the basement membrane of Shrapnell’s membrane. If the
tensa and of Shrapnell’s membrane, and by gradual reten- inflammation continues the papillae become longer and
tion of keratin and bone resorption at the posterior retrac- thicker (fig. 29). An ingrowing papilla passes through an
tion edge. The disease progresses relentlessly but the area of mucosal damage and gradually its advancing tip
allows aeration and drainage to occur predominantly on air space without any limiting borders. Above the superi-
its medial side and superiorly. The connection between or surface of the ossicles the upper lateral attic, together
the spaces formed by the upper lateral attic and the lateral with the medial attic, forms the superior attic, which nor-
malleal space to the anterior epitympanum is narrow and mally extends from the lateral to the medial attic bony
may be fully closed by the suspensory malleal ligament wall, separated only by the superior malleal ligament and
and the anterior malleal ligament with their folds. The by its generally short anteriorly trailing, mostly absent
medial portion of the posterior attic has normally a large fold [18] . The superior attic changes into the medial attic
communication with the anterior attic from the tensor without any borders at the level of the ossicles, and com-
tendon to the attic roof. municates with the mesotympanum via the large tym-
panic isthmus.
ment is around 6 mm. In the middle of it, but 1–2 mm have unquestionably been seen in serial sections. They
inferior to the level of the isthmus, is the incudostapedial usually start from the incus, do not always reach the oppo-
articulation (fig. 33, 34). Even if it serves as an area pro- site medial attic bone and it is likely that they atrophy
viding insertion points for inflammatory folds which may with the growing of the child, hence their general absence
seriously restrict the passage to the isthmus itself, the isth- in older children and adults. There are no grounds for
mus in normal ears is clearly one single unit without ascribing them a prominent role by normally obstructing
obstacles for aeration. the posterior portion of the tympanic isthmus as was done
In our entire series of microdissections of normal ears, in the undocumented sketches of Proctor [13, 14]. Nev-
the tympanic isthmus has been fully open and free of ertheless, as we will see later in connection with the
folds. This applies to neonates as well as to children and AFCC-induced sterile otitis media of the neonate, their
adults and conforms with Hammar’s [5] finding of the remnants may have a definite role in that they provide
atrophy of the composite fold affecting the medial attic fixation points for AFCC in the tympanic isthmus.
during the fetal stages. As will be remembered, this is the
fold deriving from passing of the air sacs over both sides Normal Anatomy by Serial Sections
of the incus long process and of its continuation posterior- In the absence of inflammatory changes, microdissec-
ly as the three stapedial folds. They may continue also tion can instantly provide an idea of the compartments
anteriorly for a limited distance, and we have earlier compared to serial sections (fig. 44), necessitating exami-
called them the incus intercrural fold or the medial mal- nation of a large number of slides. The lateral incudomal-
leal fold [18]. To make matters more simple, we have leal fold, being largely horizontal and thin, less than
recently [32] suggested that this structure be called the 0.2 mm in thickness, generally can be seen only partially
medial ossicular fold which would allow its presence as a in one single serial section in adults (fig. 45). In neonates
long fold or folds limited to short stretches along the and infants larger areas and thicker portions are seen
medial side of the entire ossicular chain. because the formation of the fold has not yet been final-
In some ears of neonates and infants in the material of ized, embryonal tissue remaining between the upper and
the Temporal Bone Foundation shorter or longer lower lateral attics. The anterior portion of the fold, on the
stretches of this vertically oriented medial ossicular fold other hand, because as a rule it becomes directed inferi-
Fig. 47. Series G, adult case 2, right ear. a A superior view of the dal fossa (open arrow). M = Malleus; I = incus. b Removal of the
epitympanum discloses a few postinflammatory tissue strands in the incus discloses a large thin postinflammatory web which contains
anterior portion of the tympanic isthmus (horizontal arrow), the pos- many thicker fibrotic strands and only small defects (oblique arrow)
terior portion (oblique arrow) appears free. The lateral incudomalleal allow for aeration. The medial wall of the posterior pouch, the poste-
fold is intact but indrawn and contains a few thicker areas (curved rior malleal ligamental fold (between horizontal arrows), is distinct.
arrow). The lateral malleal space (vertical arrow) is normal and a pos- The floor of the lateral malleal space is intact (curved arrow). T =
terior incudal fold closes the auxiliary aeration pathway via the incu- tensor tendon.
orly towards the posterior malleal ligamental fold Pathology as Seen in Microdissection
(fig. 15a) is well visible at thin cross sections. The majori- The mild forms of inflammatory pathology causing
ty of the sections further inferiorly show the air-filled large soft tissue sequelae appear in the form of mature tissue
compartments of the medial attic down to the tympanic strands and narrow or broader webs crossing the tym-
isthmus on one side, and either the upper or the lower panic isthmus (fig. 47a). In such cases microdissection
lateral attic on the other side of the incus (fig. 46). In the may make an instantaneous evaluation still possible, by
area of the incudal fossa the serial sections show the rela- allowing a look between the strands to the changes deep
tively spacious area underneath the incus tip, bordered by down in the mesotympanum. However, sometimes quite
the air cells, while the larger posterior incudal aeration unexpectedly the removal of the incus may reveal in the
pathway route appears behind the tip. inferior portion of the isthmus postinflammatory thin
webs which may greatly obstruct the air flow through it
(fig. 47b). One of the very frequent, almost routine find-
mus and a direct view to the tympanic isthmus becomes the granulation tissue may be so extensive that only a cast
obscured due to a full blockage (fig. 53a). In microdissec- of the incus remains after its removal (fig. 53b). A biopsy
tion layer after layer of the webs and granulation tissue of such a mass may show various stages of maturation in
must be removed to arrive at the deeper spaces. It may be the granulation tissue (fig. 54), the final stage of which
necessary to remove the entire incus to get an idea of the appears years later when the tissue mass has become
deepest layers at the low border of the tympanic isthmus dense scar tissue with full ossicular fixation and oblitera-
and around the incudostapedial articulation. Even then tion of the middle ear spaces.
Pathology as Seen in Serial Sections lium becomes edematous and filled with large capillaries,
While the serial sections in normal ears were not as and cell-rich secretions appear on all sides of the posterior
quickly informative as microdissection, in the presence of incudal ligamental fold (fig. 55d). We have found that the
pathology they are invaluable in giving information about incudal fossa itself is free of granulation tissue and some
the exact location and histological nature and phase of the aeration of the mastoid cells may occur in children suffer-
disease. The presence of secretion, either serous or muci- ing from secretory otitis media even if the main pneuma-
nous together with its cellular content, and the appearance tization process remains halted.
of the epitympanic mucosa make it possible to draw con- Extensive development of granulation tissue in the
clusions of the duration and activity of the process. When posterior epitympanum occurs, even during the era of
organization phenomena are ongoing, the maturity of the antibiotics and disappearance of florid mastoid infec-
granulation tissue, epithelialization of its surface and the tions, since the process of organization involves the non-
formation of pseudocysts are important markers of the removed masses of mucoid secretion filling the attic. This
severity and duration of the process. may be seen in association with an undiagnosed AFCC-
Chronic inflammation, even if mild, causes many linked foreign body otitis media of neonates and infants,
changes some of which appear in figure 55. The duplicate especially in cases with superimposed low grade bacterial
folds become thicker with a distinct increase of capillaries infections of longer duration As both these processes often
combined with a slight infiltration by the inflammatory go hand in hand we will discuss this type of pathology seen
cells (fig. 55a). The air cells in the attic walls may be filled in serial sections in Part 2 of this atlas in connection with
with secretion containing round cells, some cells are obli- AFCC-related pathology.
terated by connective tissue and some by cholesterol crys-
tals. Round cells containing secretion also appear in the Anterior Epitympanum
major air compartments (fig. 55b), the amount depending The anterior attic is one of the major epitympanic
upon the available open drainage pathways. The tym- compartments which has been subject to a considerable
panic isthmus may become partially blocked by inflam- confusion. The early scientists like Siebenmann [1] at the
matory webs or by polypoid tissue at times covered by end of the 19th century already had an accurate idea of
advancing squamous epithelium from the mesotympa- this space, defining it as a forward continuation of the
num (fig. 55c). These changes become also well apparent medial attic past the head of malleus. Its inferior limit was
around the incudal fossa where the ligamental fold epithe- formed by the tensor tendon and anteriorly and upwards
it was closed by the vertically rising tensor fold, the inser- concepts [19, 20] of the anterior epitympanum concern-
tion of which was into the transverse crest in the attic teg- ing the superior insertion of the tensor fold. As to the
men (fig. 5). The writers of that period defined the space pneumatic cell limited inferiorly by an incomplete fold we
anterior to the fold as a ‘pneumatic cell’ which was partial- have proposed tentatively that the fold was simply an
ly limited at its low portion by a free fold-like structure inflammatory web arising during the long-standing
arising from the tegmen tubae. Siebenmann noted that chronic diphtheric disease.
this ‘pneumatic cell’ harbored trapped disease, but he spe- We have previously discussed the current chaotic and
cifically pointed out that it is part of the mesotympanum, confused concepts of using many different names for the
not the epitympanum. His description differs from our anterior epitympanum, also regarding it as the pneumatic
cell of Siebenmann, and confusing it with the supratubal tinctly shorter than the medial portion which extends
space in front of it [19, 20, 24], and will not repeat them anteriorly until the tensor fold (fig. 56a).
here. Suffice it to say that even in the year 2000, as an The bony walls of the anterior epitympanum generally
answer to our letter [34] to the editor of the American showed good pneumatization, sometimes it was extensive
Journal of Otology concerning the ideas of supratubal and extended a long distance anteriorly above the tubal
recess and pathways to Prussak’s space in epitympanic tegmen (fig. 56b). In a few cases a major sinus-like forma-
endoscopy [35], the authors still sought to backup their tion appeared superior to the insertion of the tensor fold,
argument by referring to the erroneous writings of Proctor extending for more than 10 mm anteriorly (fig. 57, 58).
[13, 14]. Among other unfounded claims they maintained Transverse Crest. The position and extent of this struc-
that the anterior epitympanum was the supratubal recess ture have been misunderstood by most of the contempo-
and continued that: ‘By definition, supratubal recesses rary writers even if Siebenmann already in 1897 pub-
and intact plicae tensoris are mutually exclusive’! The lished an accurate drawing depicting its dimensions. The
truth of course is the opposite: when the tensor fold is crest starts from the anterior tympanic spine and crosses
intact, the anterior epitympanum and the supratubal the tegmen transversely 1–2 mm in front of the malleus
recess form two separate compartments. head (fig. 5), while its medial leg may continue until the
cochleariform process. This portion, which may be inde-
Normal Anatomy by Microdissection pendent and without any continuity with the transverse
The medial attic continued in all dissected bones unob- portion, was called a ‘cog’ by Sheehy [36] as it looks like a
structed, superior to the tensor tendon and medial to the cog of a wheel. As a rule the crest is low, around 0.5 mm,
head of the malleus, and became the anterior epitympa- and only in very rare cases, in our series 2%, has it
num. Superiorly the connecting space is narrower and exceeded a height of 1 mm. Its lower edge is sharp general-
somewhat obstructed by the superior malleal ligament ly with an uneven curve with many bony ridges and tags.
and its variable short fold (fig. 43). Connections from the Siebenmann believed that the tensor fold attached to
lateral portion of the anterior attic with the posterior com- the transverse crest and therefore the tensor fold drawn by
partments, the upper lateral attic and the lateral malleal him was vertical and the anterior epitympanum rather
space are often narrow and may be closed by a short fold small in size. This insertion, however, is not the regular
between the malleal ligaments and the attic roof. The lat- one but appeared only in 10% of our microdissections.
eral portion of the anterior epitympanum is as a rule dis- The tensor fold insertion is regularly anterior to the trans-
verse crest into a thinner or thicker ring of a composite niques have been proposed for removing this ‘bony ob-
tissue, fixed with a broad base to the tegmen. One major struction’ to provide epitympanic aeration directly from
misconception is, as the transverse crest is situated in the the supratubal recess. The error in this reasoning can be
tegmental roof, that it forms the frontal border of the ante- shown very simply by the removal of the tensor fold: a
rior epitympanum and the space in front of it is the supra- large communication between the anterior epitympanum
tubal recess. However, the crest only divides the anterior and the supratubal recess appears between the transverse
epitympanum into two portions, anterior and posterior, crest and the tensor tendon (fig. 59a, b). In our series of
both of which merge inferior to the crest and form one 125 microdissections the crest has only twice taken up the
common space. upper half of the separating wall while the thin tensor fold
Another, even more major misconception by contem- formed the lower half. In all others the crest played an
porary writers is that the transverse crest would form a insignificant role in the separation of these two major
bony separation between the anterior epitympanum and compartments.
the supratubal recess. This has been suggested especially Tensor Fold. As already emphasized, this duplicate
by Japanese workers [37, 38] and even new surgical tech- fold has a highly important strategic position as it normal-
ly prevents communication between the supratubal re- that does not break even if subjected to vigorous move-
cess, part of the mesotympanum, and the anterior epitym- ment after deliberate filling of the entire middle ear with
panum (fig. 56, 57, 60). This means that the only aeration water and sucking it out with force.
and drainage pathway in the majority of ears is through Removal of bone from the epitympanic and supratubal
the tympanic isthmus. We have found a membrane defect space tegmen reveals the soft tissue insertion ring and a
in this fold, allowing for direct attic aeration from the relatively thin mucosa on either side. Cutting through the
supratubal recess, in around 25% of microdissections soft tissues superiorly shows the thick insertion ring which
(fig. 61). Siebenmann’s finding [1] of a defect in the even by gross examination is seen to contain varying
majority of cases probably depended upon the corrosive amounts of fat (fig. 62, 63). These views, and fig. 61, dem-
methods used at that time in the handling of the speci- onstrate well how the removal of the fold connects the two
mens, which were likely to destroy the thin fold portions big compartments, the anterior epitympanum and the
looking like a pseudomembrane. Hammar [5] did not supratubal recess, with a large aeration and drainage route
observe it in his fetal series but considered Siebenmann’s continuing unobstructed to the eustachian tube.
finding possible because of a secondary fold atrophy. The insertion of the tensor fold anterosuperiorly can
Aimi’s [22, 23] figure for tensor fold membrane defects well be studied in separate biopsies obtained during
was 10%. microdissection. The thin insertion rings consisted of
The tensor fold starts from the tensor tendon upwards dense collagenic fibrous tissue with small bone fragments
in a slightly concave fashion and may have a greatly vary- but without adipose tissue (fig. 64). The thicker insertion
ing further course. If the insertion is to the transverse rings showed a pronounced broader base with many bone
crest, then the direction is nearly vertical, when the inser- fragments, with collagenic fibrous tissue and many islands
tion is to the tubal tegmen, the fold is nearly horizontal. In of adipose cells (fig. 65). The fat-specific Sudan black
most cases the angle is around 45° and the insertion ring staining in fresh specimens was always positive for the
situated halfway between the downsloping combined an- areas which in the normally processed specimens with the
terior epitympanic-supratubal tegmen. The fold appears hematoxylin-eosin stain showed only the cell outlines.
generally thicker close to its borders and may have cross- Maximally the insertion ring, measured from successive
ing strands, while the central portion is thin like a pseu- sections, was found to extend even to 4 mm in the inferi-
domembrane (fig. 60). Nevertheless it is a tough structure or-superior direction in the tegmental bone.
Fig. 64. Thin soft tissue insertion ring and portion of the tensor fold
(vertical arrow) removed during microdissection. A thin slice of bone
(B) remains at the insertion surface of the attic roof, otherwise the
block consists of connective tissue with only little fat. Hematoxylin-
eosin stain. Magnification !21.
filled the entire anterior epitympanum (fig. 69a, b). When ious levels of the sections (fig. 71, 72) and inflammatory
the entire tegmen was removed, revealing the soft tissue webs with trapped secretion going across the space
insertion ring, the tensor fold was seen to effectively limit (fig. 73). Also, the air cells around the anterior attic space
the disease to the anterior epitympanum (fig. 70). In one showed signs of mucosal edema, round cell secretion or
case, however, we have seen the granulation tissue pro- total obliteration of their lumen. Some air cells contained
gress by bone destruction to the supratubal recess. cholesterol crystals. Further examples associated with ac-
tive inflammation and development of granulation tissue
Pathology as Seen in Serial Sections in the anterior epitympanum are shown in Part 2 dealing
Inflammation in the anterior attic appeared as the with the AFCC-associated pathology.
presence of cell-rich secretion and mucosal edema at var-
Fig. 72. Case A92-168, an infant aged 8 months, left ear, Temporal Fig. 73. Case A89-195, an infant aged 15 months, left ear, Temporal
Bone Foundation. Thick mucoid secretion appears in the anterior Bone Foundation. The anterior attic (AA) forms a compartment with
attic (A) and in the supratubal recess (R). The tensor fold (horizontal trapped secretion and is divided into smaller spaces by inflammatory
arrow) is slightly edematous and polypoid structures appear in the folds (vertical arrows). The anterior malleal ligament, the superior
mucosa (vertical arrows). M = Malleus; I = incus; P = Prussak’s space. portion of which is present (A), forms the border towards the empty
Hematoxylin-eosin stain. Magnification !25. lateral malleal space (L). An inflammatory web (horizontal arrow)
closes the route to the medial attic (MA). M = Malleus; I = incus.
Hematoxylin-eosin stain. Magnification !15.
Supratubal Recess (Space) fore, developed the anterior approach for microdissection
which allowed evaluation of the entire supratubal recess
The supratubal recess is a mediosuperior extension of without disturbing any of the normal or pathological
the protympanum and anterior mesotympanum, a space in structures in the region [20]. It also makes possible the
front of the tensor fold. The size of the recess is mainly evaluation of a large portion of the mesotympanum and of
related to the degree of the angle of the tensor fold being the anterior portion of the tympanic isthmus.
largest in ears with more vertically oriented folds and smal-
lest in more horizontally oriented folds. We have observed Normal Anatomy by Microdissection
that the shape and size of the supratubal recess existed in a The anterior approach gives initially a very limited
nearly similar form in both temporal bones in a newborn view, depending upon how near the saw cut is to the ante-
[39] as well as in normal adult ears [20].The formation of rior sulcus of the tympanic membrane. Usually a good
the supratubal recess during the fetal period was already deal of enlargement must be done with a drill to give ade-
described by Hammar [5]. As reviewed above (pages 6, 7), quate light for documentation of the structures deep
the early authors [1, 5] often called this area the ‘pneumatic inside. Once near the sulcus, a full view opens to the
cell’ and clearly understood it to be part of the mesotym- mesotympanum (fig. 74, 75) of which even the regions
panum, outside the borders of the epitympanum. past the stapes can be evaluated. The anterior portion of
In the conventional superior microdissection removal the tympanic isthmus can also be visualized well and the
of the tensor fold made a reasonable evaluation of a nor- area examined for the presence of blocking folds between
mal supratubal recess possible (fig. 59). It also allowed the incudostapedial articulation and the tensor tendon. If
evaluation of the soft tissue insertion ring after removal of the saw cut is made too much posteriorly and across the
the tegmental bone and incising the ring (fig. 62, 63). anterior portion of the tympanic membrane, the tensor
Pathological changes, however, were found difficult to fold becomes partly destroyed but at the same time the
evaluate properly, because removal of the tensor fold excellent communication to the anterior epitympanum
destroyed much of the anatomy of the inflammatory tis- can instantly be seen (fig. 76).
sues, fixed to the anterior surface of the fold. We, there-
Looking more superiorly along the malleus handle, in fibers seen in the thin portions of the lateral malleal liga-
addition to the superior mesotympanic structures, the mental fold [16]. The membrane appeared resistant to
view includes the anterior pouch, the anterior malleal movement but even a slight touching with the suction tip
ligamental fold and the lateral lower portion of the supra- produced an opening, testifying to the fold’s tension. In
tubal recess (fig. 77). The anterior membrane of Prussak’s the entire series of microdissection we have only once
space, its anterior wall, appears as a truly delicate dupli- noted an anomaly when neither the tensor tendon nor the
cate membrane, inferior to the lateral malleal ligamental fold had developed, even though the tensor muscle was in
fold (fig. 78). The reason why Proctor [13, 14] described its canal (fig. 81, 82). The anterior approach, similarly to
the anterior wall as being a curved portion of the lateral the superior approach, usually reveals the insertion of the
malleal ligamental fold is apparently due to the fact that tensor fold into the soft tissue ring, while in a few ears it is
he never saw the membrane, neither from Prussak’s space inserted into the transverse crest.
nor from the supratubal recess side. With a mediosuperior Membrane defects in the tensor fold, as seen already
view the size and shape of the entire supratubal recess and from the superior approach (fig. 61), demonstrated the
the tensor fold (fig. 79, 80) can be evaluated in minute large size of this new aeration route. Seen from the recess
detail. side (fig. 83) it appears convincing that a full removal of
Similarly to a view from the anterior epitympanum in the tensor fold will provide an adequate auxiliary path-
the superior approach, a look from the supratubal recess way for aeration and drainage of the epitympanum in
side shows varying textures in the tensor fold. Intact folds cases where the tympanic isthmus posterior to the tensor
had at least portions of thin and smooth surfaces and in tendon does not function at full capacity.
some specimens the surfaces were undulating with cross- The supratubal recess has a rounded dome anterior
ing thicker strands. The fold margins generally appeared and slightly superior to the insertion of the tensor fold
thicker and could contain whitish strands resembling (fig. 79, 80, 83). Its anterior wall, the recess tegmen,
measured from the malleus handle to the medial leg of the Pathology as Seen in Microdissection
transverse crest varied between 2.5 and 3.5 mm. The anterior approach also allows a good view of the
pathological changes in the anterior portion of the tym-
Normal Anatomy by Serial Sections panic isthmus. These ranged from mature tiny tissue
The supratubal recess extends to levels superior to the strands from the stapes towards the long process of the
highest point of the anterior pouch, hence a large portion incus and the tensor tendon (fig. 85) to webs completely
of the anterior wall of the recess is seen only in serial sec- closing the anterior portion of the isthmus, from the
tions. Along the tegmen the ongoing composite tissue stapes to the tensor tendon (fig. 86). In some cases these
insertion ring appears clearly in each section (fig. 44, 46). webs also extended to both sides of the stapes and its ten-
The largest insertion ring observed measured 4 mm in don, the medial ones partly blocking the posterior portion
height on the tegmental bone, the narrow ones were of the isthmus (fig. 87). Removal of the entire tegmental
around 1 mm. At lower levels the anterior pouch and the bone together with the tensor fold gives a particularly
recess are separated by the anterior malleal ligamental good view of the webs closing the anterior portion of the
fold (fig. 55b) until its lowest mucosal surface is passed tympanic isthmus (fig. 88). If the entire supratubal and
when the pouch and the recess merge to one large single epitympanic bony tegmen is removed and the mucosa
compartment (fig. 55c). At the level of the tensor tendon incised, microdissection gives a full general view of the
the sections usually show also the eustachian tube lumen, eustachian tube lumen, the supratubal recess, the soft tis-
laterally there is the anterior sulcus of the tympanic mem- sue insertion ring, the head of the malleus and epitympan-
brane, medially the lowest portion of the supratubal recess um. Pathological changes in different compartments be-
with the tensor tympani muscle (fig. 84). The detailed come obvious and can be studied in detail.
structure of these tissues is apparent only in serial sections Pathological changes in the supratubal recess itself
but as with the superior microdissection in normal ears, a were clearly less frequent than those seen in the epitym-
single observation in microdissection from the tubal orif- panum. Chronic changes in the anterior pouch appeared
ice mediosuperiorly reveals the entire normal gross ap- as mature inflammatory strands and webs attached to the
pearance of the supratubal recess. frontal membrane of Prussak’s space and to its margins
Fig. 93. Same ear as in figure 91. The central Fig. 94. Case A89-13, right ear of a neonate, Temporal Bone Founda-
block of the fibrotic mass has been removed, tion. AFCC is concentrated in the medial attic (MA), a lesser amount
disclosing a thin tensor fold membrane (TF) appears in the anterior attic (AA) and in Prussak’s space (P), already
superior to the tensor tendon (T). On both posterior to the malleus (M). The section goes through the superior
sides (horizontal arrows) portions of the fi- portion of the supratubal recess (R) which shows three separate
brotic mass remain. The soft tissue insertion spaces, all containing acellular fluid. The surface of the tensor fold on
ring of the tensor fold is between oblique the recess side (vertical arrow) is covered by secretion containing
arrows. Removal of the granulation tissue round cells. The anterior pouch (AP) shows mainly acellular fluid
has opened the route to the anterior epitym- and one cell cluster. A = Inferior mucosal surface of the anterior mal-
panum (curved arrow). M = Head of the mal- leal ligamental fold. In more inferior sections the two spaces merged.
leus; TC = transverse crest. Hematoxylin-eosin stain. Magnification !12.
53
Part 2
OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
Introduction and Short Review of the Literature thick mucus mixed with pus and that an amniotic fluid
on Amniotic Fluid Cellular Content cellular content (AFCC) is present in a great percentage of
the specimens.
Towards the end of the 19th century the scientific liter- Aschoff’s next observation was that of the 33 ears
ature contained numerous publications on the ‘otitis me- which showed a diseased middle ear with mucoid secre-
dia neonatorum’, a condition that had repeatedly been tion and inflammatory cells, 19 showed a large amount of
seen in the autopsies of neonates. Secretion, or frank pus AFCC in the middle ear whereas of 39 ears with a healthy
had been found in the middle ear compartments and a middle ear only 2 showed some AFCC. Based on these
lively controversy arose as to whether or not this finding data he concluded that it is the AFCC that is responsible
should be considered as a physiological phenomenon or for the presence of the leukocytes. In additional studies of
whether it was due to infection. The great authorities of 9 fetuses of 11–27 cm in length, 2 were found to show
the period took part in the debate and the advances made squamous epithelial cells in the ear fluid. In one case,
by the development of bacteriology were initially thought delivered to him with an intact amnion sac after the
to resolve the question. However, the discovery of sapro- fourth gestational month, he found exactly similar cells in
phytic bacteria in the cultures, associated with the delay the amnion fluid as were present in the middle ear. He felt
in the autopsies, kept the battle lines intact. the evidence was conclusive in showing that the inflam-
In 1897 Aschoff [1] published an article which can be matory cell response in the middle ear secretion was due
regarded as a cornerstone in the discussion of the newborn to the presence of AFCC. The inflammatory changes in
middle ear dilemma. His neonate autopsy material was the neonate were not physiological but a response to the
transported from the obstetrical hospital in Hanover to aspirated or swallowed AFCC. The ‘otitis media neonato-
the Institute of Pathology in Göttingen, a delay that made rum’ thus was a typical reaction against this foreign mate-
the bacteriological investigations futile. Instead, he de- rial.
cided to concentrate on studying the cellular composition The histological studies were in accordance with the
of the middle ear secretion by microscopy and on evaluat- findings based on the above conclusions from the cellular
ing the degree of inflammation of the middle ear mucosa analysis of the secretions. The mucosa showed infiltration
in the sections. During a 2-year period he managed to of the subepithelial layer with inflammatory cells, increas-
evaluate 85 stillborns and neonates, presenting the essen- ing with increasing amounts of AFCC present. In the ears
tial individual data in a detailed table. He concluded that with no or only a small amount of AFCC the mucosa was
the gross nature of the secretion varied considerably but normal. Even in the presence of large contamination,
that the mucoid element was generally prominent. In however, the leukocytic infiltration of the subepithelial
microscopy the findings included the presence of round tissue never reached the amount generally seen in severe
cells, polymorphonuclear leukocytes, nuclei-containing or cases of otitis media of infectious origin.
nonnucleated cells of the squamous epithelium, hairs, The data and conclusions of Aschoff and of other
larger particles of meconium, and cholesterol crystals. His prominent scientists were elaborated on by Wittmaack [2,
first conclusion was that in all children the middle ear 3], an outstanding authority in ear pathology. In 1925 [3]
contains at birth secretion which varies from clear fluid to he considered the battle between the two camps to be over
55
and it was universally accepted that the filling of the neo- Aschoff [1], namely that the results do not correlate with
natal middle ear with a mucoid secretion is definitely the findings of pathology. Nevertheless, even if she specu-
pathological. The process is sterile and develops because lated that the pathological process could be due to a for-
of the presence of AFCC in the middle ear causing the eign body-type reaction, she was more inclined to think
foreign body-type inflammation. He went an important that the inflammatory condition in the ears resulted from
step further by developing the notion that the AFCC is an aspiration of infected amniotic fluid. Similar reasoning
responsible for the infant otitis media, seen most often at can also be found in a study by McLellan et al. [5] who
the age of 4–5 months, but often continuing much longer. found an inflammatory reaction in 19 of 28 middle ears of
It is a condition marked by a strong hyperplastic reaction premature infants. Buch and Jörgensen [6] reviewed serial
of the mucosa that, together with secretion, may fill all sections of 135 temporal bones of neonates and found
middle ear spaces. AFCC in 77% in ample or moderate amounts. They felt
Wittmaack’s histological figures demonstrated the that conclusions of the etiology of the infant otitis media
continuing process in the infant otitis media, starting would be possible only after bacteriological investiga-
from the clusters of AFCC inside a granulation tissue and tions.
leading through the organization process of the secretions In 1973 de Sa [7] reported a series of 130 cases, 36
to large hyperplastic tissue masses in the middle ear com- stillbirths and 94 neonatal deaths. In 56 cases the middle
partments. These processes were particularly severe in the ear cavities were considered as normal, in 55 AFCC was
epitympanum while the mucosal polyps were more present, and in 17 cases otitis media was found to be
marked in the meso- and hypotympanum. Wittmaack present. In 2 cases there was mucoid debris but AFCC
considered that nature could remove the thick tissue-fixed could not be found. AFCC was seen in the stillbirths in
secretion only by resorption or by the process of organiza- infants of over 28 weeks of gestation who had suffered
tion. The open space lost to granulation tissue could be intrapartum asphyxia and also had AFCC in the pulmo-
partly regained by its maturation and subsequent shrink- nary airways.
age. While the second group with AFCC in the middle ears
However, Wittmaack observed that there was a varia- conforms well with the findings of Aschoff [1] and Witt-
tion in this future development of the organizing tissues. maack [2, 3], the third group of 17 ears with frank otitis
If the organization process had completely eliminated the media deserves attention. These children lived from 12 h
foreign bodies, the process calmed down, the granulation to 27 days after birth, 15 over 4 days. 11 of them had
tissue matured and formed postinflammatory scar tissue AFCC present but all had evidence of infection elsewhere,
and webs crossing from one compartment wall to the oth- particularly with septicemia, pneumonia and meningitis,
er. If the child suffered, for example, from periods of vom- the principal invaders being Escherichia coli and Pseudo-
iting, new foreign bodies could be introduced into the ears monas. They thus can be set apart from the group of pure
and the process would start anew. During the course of infant otitis media but represent ears of generalized infec-
months it could be accompanied by secondary, low grade tion independent of AFCC.
bacterial invaders and a latent fight between the mucosa In a subsequent study de Sa [8] reported on histological
and the invader would continue, in some cases over many findings in 72 infants of varying periods of gestation and
years. During certain periods the invaders may become ages, all of whom died after receiving ventilatory support
more virulent and an exudative phase with secretion and and oxygen for longer than 14 days. Only 5 of these
signs of inflammation developed. An infection with viru- infants showed normal middle ear histology. In the re-
lent bacteria would cause symptoms of an acute process. mainder a wide range of changes were observed, from
When the cases were of longer standing the process would glandular metaplasia, retained squamous debris, squa-
prevent pneumatization of the mastoid bone. mous polyps, otitis media signs to destruction of ossicles.
During the following decades the presence of AFCC in None of the middle ear problems was diagnosed before
neonate and infant ears was confirmed by several investi- autopsy, all were associated with pneumonia. De Sa drew
gators. Benner [4] in 1940 in a study of 70 neonates not the important conclusion that similar changes, possibly of
only examined the temporal bones but also the paranasal lesser severity, could be present in the survivors in whom
sinuses and lungs and found 44 cases to be essentially nor- otitis media with conductive hearing losses could be
mal while 26 (37%) showed an inflammatory reaction. It expected.
is interesting to note that in her extensive bacteriological During the last two decades the Boston-Costa Rican
portion of the study she came to the same conclusion as research group has reinvigorated the concept of the
Fig. 103. Case 4, right ear. The tensor fold (oblique arrow) is thin and Fig. 104. Case 1, left ear, section 111. The lower lateral attic (LL) is
has an unusually posterior location. A small amount of secretion, filled with fluid with an AFCC cluster anteriorly. A few cells appear
containing round cells, appears in the anterior attic (AA) and a larger in the fluid lining the medial surface of the incus (I) and the posterior
cluster is seen in the supratubal recess (R). Fetal mesenchyme surface of the tensor fold (TF) in the anterior attic (AA). The supratu-
appears both medially and laterally. A vertical medial ossicular fold bal recess (R) is filled with acellular fluid. The roof (P) of Prussak’s
of a height of 9 mm (horizontal arrow) divides the medial attic partly space lateral to the malleus (M) contains vacuolated fetal tissue. C =
into two portions. M = Malleus; I = incus. Magnification !12. Chorda tympani nerve; A = anterior malleal ligament; L = lateral
malleal ligament; MA = medial attic. Magnification !12.
Fig. 110. Case 2, left ear, section 371. A magnified view of the con- Fig. 111. Case 2, section 391, left ear. The entire mesotympanum
tinuing AFCC cluster in the medial attic, 0.8 mm more inferior to around the long process of the incus (I) contains fluid with AFCC
that shown in figure 108. A large number of squamous epithelial cells clusters. Fluid with few cells fills the anterior mesotympanum (AM).
and some lanugo hair appear surrounded by round cells and multinu- The footplate area is free (vertical arrow), the posterior pouch (obli-
cleated cells. Magnification !200. que arrow) contains scant fluid. E = Ear canal; F = facial nerve; M =
malleus. Magnification !12.
of the cell clusters was particularly noticeable around the was normal in all temporal bones, but there was some
incus long process (fig. 111) and the stapes (fig. 112) and AFCC posterior to it together with prominent capillaries
there were many areas with small epithelial breaks form- and polypous mucosa in case 2. In all ears the tympanic
ing future portals from the subepithelial space to the sinus contained fluid with a varying amount of AFCC and
mucus. In case 3 there were several multinucleated giant distinct areas of contact with the epithelium for the devel-
cells, some of which had ingested fragments of squamous opment of portals for organization.
epithelial cells (fig. 113). The round window membrane
Eustachian Tube cells were filled with AFCC. The inferior portions of the
In all temporal bones the lateral portion of the protym- antrum, separated already by thick bone from the middle
panum and the bony eustachian tube contained moderate ear spaces, were surrounded by smooth-walled red mar-
numbers of amniotic cells and fluid, not attaching to the row bone.
walls.
Comment
Mastoid Pneumatization A varying amount of embryonic tissue remained in all
The mastoid antrum was well formed in all 8 speci- these neonatal ears. It made the incudomalleal fold areas
mens, open for the most part except in one ear which thick between the lateral attics. In 3 bones it was present
showed a large antrum still filled with mesenchyme The in large amounts in the lateral malleal and in Prussak’s
petrous bone, making its thick posteromedial wall, was spaces and only in case 2 was Prussak’s space fully open
typically red marrow bone containing hemopoietic tissue and comparable to the adult. In case 4 in the left ear the
throughout. It even showed smooth contours with occa- large attic compartments were also only partially open.
sional short trabeculae projecting to the antral lumen and The fetal mesenchymal tissue in the lower lateral attic and
was lined by mucosa which contained some embryologi- around the posterior pouch clearly protected Prussak’s
cal mesenchyme. The squamous bone, the smaller antero- space from AFCC contamination.
lateral portion of the mastoid bone, was much thinner and The anterior membrane of Prussak’s space, the anteri-
showed evidence of a marked preformation extending or malleal ligament and an intact tensor fold form a dead
close to the lateral periosteum. Numerous short and long end for a superior air flow. This is in accordance with the
trabeculae projected towards the antral cavity with em- presence of generally only little or no AFCC in the supra-
bryonal mesenchyme, generally rich in capillaries, be- tubal recess, apart from the main flow of air to the middle
tween them (fig. 98). The marrow spaces close to the per- ear inferior to the tensor tympani tendon. An exception
iosteum contained no hemopoietic tissue, only connective was made in the right ear in case 4 where the recess was
tissue and a few capillaries. The bone preformation con- posteriorly unusually deep (fig. 103). Due to the direct
tinued from the attic roof until the fossa incudis. In sever- aeration route from the eustachian tube to the incudosta-
al specimens there was a marked lateral formation of air pedial articulation and to the lower lateral attic, these
cells with spaces lined by epithelium, the subepithelial tis- areas constantly showed varying amounts of AFCC. From
sue still containing thick mesenchyme. Many such air the upper mesotympanum and the lower lateral attic
Fig. 117. Case 1, right ear, section 210. The pseudocystic granulation Fig. 118. Case 1, left ear, section 132. A long secretion mass fills half
tissue mass (horizontal arrow) adheres to the margins of the tensor of the tympanic isthmus (TI) and continues to the anterior attic (AA)
fold membrane defect (oblique arrows) and shows posteriorly large where part of it drains into the dome of the supratubal recess (R) via a
connecting strands (vertical arrows) to the malleus (M) and incus. defect (oblique arrow) in the superior portion of the tensor fold. The
A = Anterior malleal ligament; R = supratubal recess. Magnification lateral malleal space (L) and the lower lateral attic (LL) contain secre-
!25. tion with round cells. M = Malleus; I = incus. Magnification !12.
Tensor Fold and Supratubal Recess fold was intact but thick and invaded by inflammatory
Case 1. Organized granulation tissue and strands of cells. The supratubal recess contained cell rich secretion
secretion appeared in the anterior attic and around the (fig. 120).
tensor fold which had a 0.4-mm-high membrane defect in
the superior portion of the deformed fold (fig. 116). In the Lateral Malleal Space and the Lateral Attics
following section the defect became patched with the Case 1. In the right ear the lateral malleal space and the
increasing size of granulation tissue (fig. 117) and further lower lateral attic contained cellular mucus and in the lat-
sections showed an intact fold drawn posteriorly by con- ter there was also epithelialized granulation tissue, but the
tracting tissue bridges. Excluding its upper portion the downturning anterior edge of the lateral incudomalleal
supratubal recess was free of pathological changes. In the fold remained thin (fig. 114). Further inferiorly the lower
left ear the membrane defect in the superior portion of the lateral attic contained increasing amounts of pseudocystic
tensor fold was smaller but still provided a tiny passage granulation tissue (fig. 116). In the left ear the upper and
for secretion from the anterior attic to the supratubal lower lateral attics contained thick mucoid secretion with
recess (fig. 118). round cells and many giant cells.
Case 2. The tensor fold was intact in the right ear and Case 2. The lateral incudomalleal fold was thick in the
the supratubal recess contained acellular secretion. right ear, rich in capillaries, and slightly infiltrated by
Case 3. In the left ear the tensor fold showed a defect round cells. Several polypoid formations and crossing
0.2 ! 0.5 mm in width and 0.3 mm high in its medial strands appeared in the lower lateral attic together with
portion (fig. 119), the fold was thick and infiltrated by secretion (fig. 121).
round cells. In some sections mucus was seen to go Case 3. The lower lateral attic, the lateral malleal space
through the defect. The supratubal recess contained large and the anterior attic contained in both ears secretion
polypous epithelialized tissue strands which divided it with round cells and giant cells with a few mucosal polyps
into several compartments, the mucosa being severely (fig. 122).
infiltrated. In the right ear near the recess dome the tensor
Fig. 126. Case 2, right ear, section 221. The posterior tympanum Fig. 127. Case 3, left ear, section 260. The posterior pouch (oblique
shows a large epithelium-covered granulation tissue mass around the arrow) contains dense clusters of round cells. The fold around the
long process of the incus (I). Secretion contains both round cells and chorda tympani nerve (C) is infiltrated by round cells and the mucosa
giant cells in all areas, also in the posterior pouch (horizontal arrow). in the posterior tympanum is polypous throughout. Epithelialized
The chorda tympani nerve (C) is surrounded by granulation tissue. granulation tissue (horizontal arrows) adheres to the tensor tendon
One continuous bundle (vertical arrow) of the posterior malleal liga- (T) and the anterior mesotympanum contains secretion with round
ment appears. S = Posterior tympanic spine; M = malleus. Magnifica- cells (vertical arrow). M = Malleus; I = long process of incus. Magnifi-
tion !12. cation !12.
was free. The stapes was surrounded by heavily infiltrated contained scant secretion and there was slight infiltration
thick, polypous mucosa. In the right ear the whole posteri- of the subepithelial tissue with round cells. In the left ear
or tympanum and facial recess presented polypous muco- of case 3 inferior and anterior to the level of umbo the
sa and there was a long organized pseudocystic tissue mesotympanic spaces were nearly empty. It was observed
sheet starting from the malleus via the long process of the that some mucus had been transported to the intact eusta-
incus to the stapes (fig. 129). chian tube.
Tympanic Sinus and Round Window Niche Elements Specific to Amniotic Fluid Cellular Content
Case 1. In the right ear there was a small amount of In the right ear of case 1 fragments of hair were seen in
granulation tissue in the round window niche but the the granulation tissue surrounding the incus body and the
membrane itself was normal. stapes, and in the mastoid antrum. In the left ear similar
Case 2. In the right ear the round window niche fragments appeared in the anterior attic. In the right ear of
showed both secretion and mucosal polyps. The mem- case 2 fragments of hair were frequently seen the granula-
brane itself was normal and not affected by inflammatory tion tissue, exceptionally also in the free space, in all com-
changes. partments apart from Prussak’s space and posterior
Case 3. In the left ear the sinus tympani was full of pouch. Macrophages or giant cells were a frequent finding
mucus and the mucosa was infiltrated by round cells. An surrounding portions of hair, around one end of a hair
organized pseudocystic tissue mass filled the round win- (fig. 133, 134) or engulfing a fragment. Some areas were
dow niche and the membrane was firmly connected to the surrounded by clusters of activated lymphocytes. In case 3
mucosal surfaces of the niche (fig. 132). In the right ear only a few small fragments of hair were found.
the changes were slightly less intense and the window
membrane itself was still free of the organizing tissue Mastoid Pneumatization
mass present in the niche. The petrous bone in 4 specimens showed similar char-
acteristics as in the neonate series. Distinct embryonal
Eustachian Tube mesenchyme was still present subepithelially, and the
In case 1 both eustachian tubes had normal mucosa bone surface facing the antrum was even or showed only
and practically no secretion. In case 2 the eustachian tube short ridges (fig. 114, 115). At inferior portions of the
antrum the marrow bone surrounded the entire cavity fragments appeared less frequently in case 1, aged 84
and there was no sign of pneumatization. The squamous days, and only rarely in case 3, aged 113 days. It is obvious
bone laterally varied in thickness and close to the perios- that the process of organization gradually eliminates
teum contained marrow bone without hemopoietic tissue. AFCC by enzymatic breaking of keratin and phagocytosis
Towards the antral cavity there was the preformed trabe- by the activated macrophages. The presence in all studied
cular arrangement and distinct air cells had formed be- ears of the still very cellular granulation tissue covered by
tween the bone ridges medially, lined with a thick, mesen- epithelium made it possible to determine that the begin-
chyme-containing mucosa (fig. 115). Only in the right ear ning of the process occurred during the period of birth.
of case 3, showing smaller inflammatory changes, was It must be remembered as pointed out in Ashoff’s
there more advanced air cell formation especially in the study [1] that aspiration during delivery is not the only
squamous bone, less in the petrous bone. Several air cells cause for the presence of AFCC in the middle ear cleft. In
were filled with mucus and showed signs of organization. fact, he found that due to swallowing AFCC may appear
in the middle ears from the 4th fetal month on. Thus
Comment asphyxia and aspiration must not be regarded as the only
The hospital records of these 3 children contained nei- criteria indicating that a neonate can have AFCC present
ther data of the presence of meconium in the amniotic in the middle ear. In the present ears there were also asso-
fluid nor information of a possible aspiration during ciated factors which could have contributed to an inflam-
delivery. It is apparent that there had been no massive matory middle ear disease. In case 1 the nasogastric tube
pulmonary aspiration because the asphyxic state would could have aided the development of signs of the secretory
have been obvious clinically. The focal pulmonary infil- otitis media in the left ear, containing large amounts of
trates in case 1 might indicate some aspiration even if not yet organizing mucoid secretion. In case 3 the associ-
diarrhea was the cardinal symptom. The initial presence ated vomiting might have forced contents of the mouth
of AFCC in the middle ears could, nevertheless, be ascer- into the middle ears even if squamous, mouth-derived
tained by the identification of specific histology. Case 2, epithelial cells were not seen.
67 days of age, had fragments of hair at several sites inside The areas most affected both by collection of mucoid
the granulation tissue, a finding in line with the earlier secretion and fresh granulation tissue proved to be the
data reported in the age groups under 70 days [9]). These posterior tympanum and the lower lateral attic. These
multiple hemorrhagias. Culture from pulmonary tissues was empty, the lower half contained secretion. Prussak’s
was negative. space was half-filled with secretion and drained to the
In the analysis of the serial sections the findings in anterior pouch. The common posterior pathway from the
cases 1–3 between 5 and 8 months of age were much more lower lateral attic ended short of Prussak’s space. The pos-
extensive than those in the infants over 1 year of age. terior pouch was superiorly a blind space containing scant
Cases 4 and 5 were apparently exposed only to a small secretion. Mucoid secretion was present in the posterior
extent to an AFCC contamination. hypotympanum and tympanic sinus.
In summary, mucoid secretion in most compartments
Case 1 indicated recent changes. Long-standing changes ap-
Right Temporal Bone peared in the superior attic as an epithelialized block of
The superior epitympanum contained epithelium-cov- granulation tissue, and as thinner strands in the tympanic
ered granulation tissue with wide bridges extending to the isthmus and in the posterior tympanum.
lateral attic wall (fig. 135). The mucosa was moderately
invaded by round cells. The granulation tissue thinned Left Temporal Bone
progressively inferiorly and at the tympanic isthmus only The superior attic was devoid of secretion. Mucoid
strands remained, which formed a bridge to the medial secretion with round cells and occasional multinucleated
mucosa. Mucoid secretion with round cells and some cells was present in all more inferior attic compartments.
multinucleated cells was present in the attic compart- The mucous membrane showed a slight round cell inva-
ments. A small portion of the tympanic isthmus and the sion with occasional portals for organization. The tensor
posterior tympanum (fig. 136) contained mucoid cell-rich fold was invaded by round cells and had a small 0.2 ! 0.2
secretion and strands of connective tissue. mm defect in the anterior portion, mucoid secretion pass-
The tensor fold had anteriorly a 0.7-mm-high mem- ing through it to the supratubal recess (fig. 137). Prussak’s
brane defect. The upper portion of the supratubal recess space was aerated from the anterior pouch (fig. 25b). Pos-
In summary, changes due to ongoing infection in- with a narrow passage showing polypoid changes. The
cluded degenerating cells and all classes of inflammatory tympanic isthmus was full of a network of epithelium-
cells in desiccated secretion undergoing organization, covered pseudocystic granulation tissue (fig. 149).
with immature granulation tissue and a polypous, in- The intact tensor fold separated the supratubal recess
flamed mucosa. Long-standing changes appeared as a from the anterior attic and showed sections of inflamma-
pseudocystic network of epithelialized granulation tissue tory thickening. The recess was nearly empty at its dome
in the tympanic isthmus and in the posterior tympanum. but the inferior sections were half-filled with secretion
(fig. 149). Prussak’s space was superiorly empty, the low
Left Temporal Bone sections contained round cell secretion (fig. 149) which
The superior epitympanum and the antrum contained continued via the posterior pouch to the mesotympanum.
a central mass of epithelium-covered granulation tissue The oval window area and the entire posterior tympanum
with pseudocysts and bridges to the mucosa (fig. 148). were filled with secretion and granulation tissue (fig. 150).
Outside this mass the adjoining open compartments con- Secretion continued to the tympanic sinus and the round
tained secretion rich in macrophages. Embryonic tissue window niche.
with a few round cells appeared in the subepithelial tissue. In summary, changes related to infection included cell-
At lower levels the secretion masses and the granulation rich secretion in the attic and in the tympanic sinus and
tissue were present in the entire area from the anterior round window niche. Extensive long-standing changes
attic to the antrum. The mucosa of the lateral attic con- appeared as pseudocystic, epithelium-covered granula-
tained partly embryonic tissue and partly old inflammato- tion tissue network which filled the superior and lateral
ry pseudocysts, causing a thickening of the whole mucosal attics, mastoid antrum, tympanic isthmus and posterior
layer. The upper and lower lateral attics were connected tympanum.
Fig. 156. Case 2, right ear, section 100. a The general view reveals a mass of epithelialized granulation tissue in the
tympanic isthmus (T); a narrow strand (oblique arrow) extends to the anterior attic (A). Secretion containing round
cells fills the anterior and lower lateral (L) attics. M = Malleus; I = incus. b The area boxed in a shows a remnant of a
disintegrating hair, one end breaking into filaments (vertical arrow) surrounded by macrophages and giant phago-
cytes. Magnification !12 (a), !250 (b).
Mastoid Pneumatization
In cases 1–3 with severe inflammatory changes there
were slight differences in the mastoid pneumatization as
compared to the two younger age groups. The petrous
bone generally showed a full arrest, the bone marrow still
bordered the antrum (fig. 139, 144, 145) and only seldom
showed longer bone ridges. In case 3 with the most exten-
sive inflammation, the antral mucosa was separated from
the bone marrow with a dense parallel layer of bone with
minute contacts to the mucosa (fig. 144–148). The squa- Fig. 157. Case 3, right ear, section 123. A magnified view of a granu-
lation tissue polyp marked with a vertical arrow in the mastoid
mous bone in cases 2 and 3 showed distinct pneumatiza-
antrum in figure 144. A disintegrating squamous epithelial cell (verti-
tion with filling of the air cells with secretion or connec- cal arrow) is surrounded by macrophages and giant phagocytes. The
tive tissue, resulting in a slight lateral enlargement of the cell remnant is so thin that the round cells shimmer through it. Mag-
antrum (fig. 139, 140, 144, 145). On the other hand, in nification !250.
cases 4 and 5 with minor signs of old or recent inflamma-
tion, pneumatization was much more advanced. In case 4,
15 months of age, the entire squamous bone and the later-
Fig. 158. Case 4, left ear, section 31. a The general view shows secretion in the anterior attic (A) and around the
malleus (M) and incus (I). Fully epithelialized tissue strands (oblique arrows) cross the posterior attic and mastoid
antrum and contained a few dense cell clusters (boxed area). The squamous bone shows medially air cells (horizontal
arrow) while the bone near the cortex still contains connective tissue. Hemopoietic bone marrow appears in the
petromastoid bone (vertical arrow). b A magnified view of the boxed area shows that the dense cell cluster consists of
macrophages and giant phagocytes surrounded by a dense ring of round cells. They apparently represent an initial
small cluster of AFCC, of which filament-like remnants of keratin are still present (oblique arrow). An unusually thick
epithelium forms the surface of the tissue strand. Magnification !6 (a), !250 (b).
References
1 Aschoff L: Die Otitis Media Neonatorum. Ein 8 de Sa D: Mucosal metaplasia and chronic in- 14 Rüedi L: Die Mittelohrraumentwicklung vom
Beitrag zur Entwicklungsgeschichte der Pau- flammation in the middle ear of infants receiv- 5. Embryonalmonat bis zum 10. Lebensjahr.
kenhöhle. Z Ohrenheilkd (Wiesbaden) 1897; ing intensive care in the neonatal period. Arch Acta Otolaryngol 1937:(suppl 22):1–131.
312:295–346. Dis Child 1983;58:24–28. 15 Rüedi L: Mittelohrraumentwicklung und Mit-
2 Wittmaack K: Über die normale und patholo- 9 Northrop C, Piza J, Eavey R: Histological ob- telohrenentzündung. Z Ohrenheilkd 1939;45:
gische Pneumatisation des Schläfenbeines. servations of amniotic fluid cellular content in 175–213.
Jena, Fischer, 1918. the ear of neonates and infants. Int J Pediatr 16 Diamant M: Otitis and pneumatisation of the
3 Wittmaack K: Die entzündlichen Erkran- Otorhinolaryngol 1986;11:113–127. mastoid bone. A clinical-statistical analysis.
kungsprozesse des Gehörorgans; in Henke F, 10 Piza J, Gonzales M, Northrop C, Eavey RD: Acta Otolaryngol 1940(suppl 41):1–149.
Lubarsch O (eds): Handbuch der speziellen pa- Meconium contamination of the neonatal mid- 17 Dahlberg G, Diamant M: Inheritance of pneu-
thologischen Anatomie und Histologie. Berlin, dle ear. J Pediatr 1989;115:910–914. matization of the mastoid bone. Hereditas
Springer, 1926, pp 102–379. 11 Eavey R: Abnormalities of the neonatal ear: 1945;31:441–456.
4 Benner M: Congenital infection of the lungs, Otoscopic observations, histologic observa- 18 Diamant M: Chronic otitis. A critical analysis.
middle ears and nasal accessory sinuses. Arch tions, and a model for contamination of the Pract Oto-Rhino-Laryngol (Basel) 1952;14
Pathol 1940;29:455–472. middle ear by cellular contents of amniotic (suppl 1):1–190.
5 McLellan M, Strong J, Johnson Q, Dent J: Oti- fluid. Laryngoscope 1993;103(suppl 58);1–31. 19 Ojala L: Contribution to the physiology and
tis media in premature infants. J Pediatr 1962; 12 Northrop C, Piza J, Karmody C, Eavey R: The pathology of mastoid air cell formation. Acta
61:53–57. neonatal middle ear. Indicator for future prob- Otolaryngol 1950(suppl 86):1–134.
6 Buch N, Jorgensen M: Leukocytic infiltration lems? 3rd Extraordinary Symposium on Re- 20 Tumarkin A: On the nature and vicissitudes of
in the middle ear of newborn infants. Arch Oto- cent Advances in Otitis Media, Copenhagen, the accessory air spaces of the middle ear. I.
laryngol 1964;80:141–148. June 1–5, 1997. The Hague, Kugler, 1999, pp Facts. II. Theories. J Laryngol Otol 1957;71:
7 de Sa D: Infection and amniotic aspiration in 321–325. 65–99.
stillbirths and neonatal deaths. Arch Dis Child 13 Eckert-Möbius A: Die pathologisch-anato- 21 Tumarkin A: On the nature and vicissitudes of
1973;48:872–880. mische Untersuchungstechnik und die normal- the accessory air spaces of the middle ear. III.
histologische Grundlage; in Henke F, Lubarsch Experiments. IV. Arguments. J Laryngol Otol
O (eds): Handbuch der speziellen patholo- 1957;71:137–161.
gischen Anatomie und Histologie. Berlin,
Springer, 1926, pp 1–101.
Microsurgical Approaches to
Inflammatory Ear Disease
91
Part 3
OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
Microsurgical Approaches to
Inflammatory Ear Disease
93
case leading to the development of a papillary ingrowth ear canal, he used the canal-wall-up-type surgery for all
cholesteatoma and in another to a prestage of a retraction forms of cholesteatoma. Bone removal from the lateral
pocket cholesteatoma (fig. 31). attic was continued until the zygoma so that the working
In one ear in this series [7], there was an absence of the axis towards the tensor fold became frontally parallel with
tensor fold which provided the attic with a large direct the axis of the ear canal. This allowed nearly vertical drill-
drainage and aeration pathway to the supratubal recess ing in front of the malleus head through bone, which was
and eustachian tube. The epitympanic pathology was called the ‘attic plate’ and was described as the separating
slight and in marked contrast to the left ear with the tensor structure between the anterior attic and the supratubal
fold present. In many other cases, studied in serial sec- recess. The misconception here was that the attic plate in
tions, epitympanic secretion was seen to drain via a defect reality is the transverse crest which from this lateral
in the tensor fold membrane directly to the supratubal approach gave an illusory appearance as a dividing struc-
recess, and thus to the orifice of the eustachian tube. How- ture between the two compartments. From the two pre-
ever, even this additional direct attic aeration route and ceding parts of this atlas it is clear that the dividing struc-
the use of a ventilation tube did not prevent the destruc- ture that exists and matters is not the transverse crest but
tion of the epithelium in Prussak’s space and its gradual the tensor fold.
fibrotic obliteration.
These documented observations, arrived at by com-
bining data obtained by microdissection with those ob- Microsurgical Methods in Surgery for Retraction
tained by studying serial sections of diseased temporal Pockets
bones, made us search for new microsurgical approaches
with a better guarantee of improving aeration and drain- Obstruction of the aeration pathway either via the pos-
age to and from the epitympanic compartments than the terior pouch, or from the lower lateral attic directly, may
currently used techniques [8]. Initially we tried to arrive at lead to gradual obliteration of Prussak’s space (fig. 24, 27,
methods which would prevent the development of the 141). If the tympanic isthmus functions normally the
serious disease processes behind Shrapnell’s membrane main attic remains aerated as shown in CT scans [10]. A
and in Prussak’s space, and later focused our attention on deeply indrawn Shrapnell’s membrane may then appear
surgical methods for tensor fold removal. as the only sign of this initially limited pathology. This
state will remain resistant to all modes of conservative
treatment. An indrawn membrane may remain inactive
Early Attempts to Improve Epitympanic Aeration when there is a thick layer of mature connective tissue
underneath, allowing no retraction to occur towards the
In the late 1980s removal of the tensor fold had neck of the malleus. The epidermis then becomes resting,
become a routine step in our canal-wall-down surgery loses its papillae and this condition does not indicate any
with mastoid obliteration and there were no technical interference (fig. 30, 32).
problems when the incus and the head of the malleus were Patients with a distinct indrawing of Shrapnell’s mem-
removed. In retraction pockets without mastoid patholo- brane with even minimal signs of retained keratin should
gy, with an intact ossicular chain and still with normal have otomicroscopy once a year to provide a sufficiently
hearing, no methods were available that allowed direct early diagnosis of a permanent, increasing collection of
vision of the tensor fold. Only in spacious anterior attics keratin in the initially tiny pocket. Surgery at this stage is
with a good safety margin to the head of the malleus, bone still a minor procedure and should be undertaken to pre-
could be removed to enter into the supratubal recess. vent serious future problems.
However, because the working angle was not comparable
to the superior approach in microdissection but was Surgery for Incipient Retraction Pockets
through the ear canal, the tensor fold remained hidden Both the endaural and postauricular incisions can be
behind the neck of the malleus and had to be destroyed used but we prefer the postauricular one because it gives a
blindly with angled hooks. When the anterior attic was better angle to the anterior portions of Prussak’s space.
short, there was no safe way of excising the fold. Canal skin is cut transversely 5 mm above the annulus
At the same time, Morimitsu [9] started to use a basi- and the incision continued superiorly over the vascular
cally similar approach that he called ‘anterior tympanoto- strip to the anterior edge of the notch of Rivinus. Canal
my’. While we worked with limited disease through the skin, superior portion of the posterior pars tensa and
ment remain fixed to both malleal fragments after cutting, removal. We were initially perplexed by the term ‘attic
sideward shifting of the manubrium fragment could pull plate’ used by their group [9, 12] which was said to sepa-
the head laterally and dislodge it. A further division of the rate the anterior epitympanum from the supratubal re-
ligament on the malleus handle is then made with laser, or cess. Since the days of Siebenmann [13] and Hammar
with a sharp sickle knife. The surgeon’s touch should be [14], more than a hundred years ago, it has been recog-
gentle in all phases, and a few temporal bones should be nized that the separation is done by the tensor fold. We
dissected in the laboratory before applying the method to then understood that in the lateral approach the trans-
actual surgery. verse crest of the tegmen gave the authors an illusion of a
If mastoidectomy is indicated, or the retracted skin separation of the epitympanum and the recess by the
envelopes the neck of the malleus, or the pocket extends transverse crest. In superior microdissections, and in hori-
superiorly, we regularly remove the incus and cut the head zontal serial sections, the transverse crest is seen to play
of the malleus. In our experience, whenever it is unclear no role except at times it divides the anterior epitympa-
whether or not the squamous epithelium has been re- num superiorly into two sections (fig. 67, 114).
moved totally, the procedure should be carried out. This We use the frontolateral atticotomy in generalized cho-
exposure reveals the contents of the whole attic and allows lesterol granulomatous mastoiditis and epitympanitis
removal of the diseased tissues. The view of the tensor with a canal-wall-up surgery for excision of the tensor
fold becomes unobstructed, allowing its removal with a fold, combined with preservation of the ossicular chain.
full examination of the supratubal recess. Reconstruction As Morimitsu [9] suggested, bone in the atticotomy has to
of the stapes is done with the patient’s own tissues, incus be removed as far as the zygoma so that the approach
or cortical bone, following the established principles. allows for transversely continued drilling in front of the
head of the malleus (fig. 165). At the end of the atticotomy
we also break the lateral incudomalleal fold blindly, with a
Frontolateral Atticotomy long right angle hook, by gentle movements of the hook
between the incus short process and the lateral attic bone.
After having read Morimitsu’s manual [9] on what he This opens one additional aeration route to the attic com-
called ‘anterior tympanotomy’ for cholesteatoma removal bining the air spaces of the upper and lower lateral attics
we thought the approach worth trying in tensor fold (fig. 39).
The actual approach to the tensor fold necessitates a dures for disease in Prussak’s space and removal of the
small diamond drill in an angled handpiece for removing tensor fold, all of the meso- and epitympanic polypoid
the lateral portion of the transverse crest in front of the granulation tissue should also be removed after early
malleus head. The anterior malleal ligament in front of removal of the incus. The mucoperiosteum together with
the malleus head leads directly to the tensor fold which thin areas of mucosa should be conserved to allow epithe-
has the appearence of an oblique thin membrane ante- lial regeneration from the remaining small stretches of a
romedial to the surgical route (fig. 166). Thin folds can be near-normal epithelium and to keep the development of
excised with a sickle knife but laser is generally preferred, granulation tissue formation on raw surfaces at a mini-
and it is mandatory in thick folds (fig. 167). If a major mum. Much of the excision of the granulation tissues
portion of such folds is not evaporated, a simple incision should be done with laser, especially around the stapes. In
quickly heals leaving no permanent membrane defect. If the absence of the incus columellization with the head of
the tensor fold has become inserted into the crest itself, in the malleus on the stapes is the best method for creating
around 10% of our cases, the surgical route through the space as it leads to a large tympanoattic middle ear com-
crest also destroys the fold directly when the lateral por- partment with unhindered drainage and aeration. A
tion of the crest is drilled away. short-term cortisone therapy with antibiotic cover during
the postoperative period in such cases would contribute to
Extensive Attic and Mesotympanic Disease in Chronic quicker healing of the mucosal layer.
Otitis media
In recurring inflammatory ear disease, both the epi-
tympanum as well as the mesotympanum may present Spread of Cholesteatoma from Prussak’s Space
with an extensive amount of granulation tissue. In addi-
tion to the blockade of both the tympanic isthmus and the There are four routes for the spread of cholesteatoma
aeration pathway to Prussak’s space, granulomatous obli- from Prussak’s space, the posterior, inferior, superior and
terating tissue and polyps may envelope the incus so that anterior, and each gives a different type of involvement of
preservation of an intact ossicular chain becomes too ris- the neighboring compartments. These cholesteatomas
ky (fig. 52, 53). In addition to the above-described proce- start from an epithelial papillary projection, penetrating
the basal lamina, into adjacent granulation tissue in Prus- touch either the fold insertion ring or mucosa of the supra-
sak’s space. There the advancing front portion of the tubal recess, which are well outside the area of cholestea-
papilla becomes a minicyst and begins to enlarge due to toma.
trapped keratin and transforms into a typical ball-type
cholesteatoma. Common to them all is the squamous epi- Inferior Central Route
thelium in the inner lining of the cystic formation, con- In around 36% of ears Prussak’s space is aerated
nective tissue covering the outside. However, any granu- directly from the lower lateral attic and the posterior
lation tissue adjacent to the connective tissue may contain pouch forms a superiorly blind sac without any connec-
papillae from the squamous epithelium, regardless of the tion to Prussak’s space [15]. In such ears the aeration
type of cholesteatoma. pathway leads to the lower lateral attic and the superior
mesotympanum immediately posterior to the short pro-
Posterior Route cess of the malleus. From here the cholesteatoma sac has a
This is the common way of extension because the sac tendency to enlarge into the anterior direction under the
enlarges along the regular aeration pathway through the handle of the malleus and it is impossible to diagnose it at
posterior pouch, gradually fills and expands it so that it an early stage before perforation occurs in Shrapnell’s
occupies the entire space below the lateral incudomalleal membrane. We recently studied a temporal bone with this
fold. When it arrives at the mesotympanum, an early sign type of involvement [7], combined with Shrapnell’s mem-
is a slightly bulging whitish area in the posterior upper brane perforation (fig. 168) and complicated during the
segment of the pars tensa; in later stages this area extends terminal stage of septicemia with extensive mastoid gran-
to the lower posterior segment. Generally the sac has ulomatous infection. When we started with the anterior
become so large that it reaches the long process of the microdissection we found that there had been a disrup-
incus and the stapes crura, necessitating partial ossicular tion of the cholesteatoma sac and the keratin mass had
resection. There is a good chance for total removal spread anteriorly as far as the bony eustachian tube
because the upper lateral attic and the superior attic are (fig. 169).
primarily intact. Excision of the membranous portion of
the tensor fold is made routinely, but there is no need to
Subject Index
The subject index contains terms which are of major importance in finding the passages which deal with the term
sought for. Due to the frequent appearance of some of the terms, all occurrences have not been printed; mainly those
are included which contain additional information. Those page numbers which contain essential basic information of
anatomy or pathology are printed in bold.
Aditus ad antrum 3, 10, 12, 100 Cochleariform process 38, 47, 96 – obturator 8
Aeration pathway 9, 14, 17–19, 21–27, 30, Cog 38 – posterior incudal 30
31, 40, 46, 66, 70, 75, 80, 93, 94, 99 Cornelius 4 – stapedial 8, 32
Aimi 10, 40 Cutting of malleus neck 96 – tensor 6–8, 11, 37, 38, 39, 40, 42, 43,
Air sac 46, 47, 62, 66, 69, 94
– – anterior 7, 8, 16 Dahlberg 59 – trailing medial surface of malleus to
– – medial 7, 8, 16, 31 Diamant 59, 60 tensor fold 42
– – posterior 7 Döderlein 87 – trailing superior malleal ligament 8, 41
– – superior 7, 16 Drainage pathway → Aeration pathway Foreign body 18, 56, 57, 60
Amniotic fluid 59, 61 Duplicate folds 7 Fossa incudis 30, 66
– – cellular content 21, 61–66, 93 Friedmann 60, 88
Animal model of pneumatization 60, 61 Eavey 57 Frontolateral atticotomy 97
Annular bone 4, 8 Eckert-Möbius 58
Anterior attic → Epitympanum Endaural incision 94 Giant cells 69, 70, 73, 84, 86
– membrane of Prussak’s space 13, 16, Epidermal papillae 26 Glomus tumor 95
19, 46, 95 Epitympanic diaphragm 11, 12 Goldenhar’s syndrome 75, 79
– pouch 4, 23, 27, 46, 48, 95 – folds → Folds Granulation tissue 21, 23, 25, 34–36, 42,
– tympanic isthmus 12 – ligaments → Ligaments 43, 56–58, 67, 69, 70, 72–74, 81,
Arnold 4 – sinus 38, 39 85–87, 98
Aschoff 55, 56, 59 Epitympanum Guluya 9
Attic → Epitympanum – anterior 11, 27, 28, 31, 36–44, 58, 61,
– bone 5, 15, 30, 31 94, 97 Hammar 4, 5, 7–9, 13, 14, 32, 40, 45, 97
– roof 8, 28, 30, 31 – medial 9, 11, 12, 28, 31–33, 38, 42, 64 Hammer-Amboss-Schuppenraum 7
Atticotomy 97 – posterior 12, 27, 28–36, 42, 101 Helmholtz 4, 5, 13
Auxiliary aeration pathway 12, 30 – superior 8, 28, 31, 61, 62, 76, 77, 81, 88 Henle 5
Eustachian tube 7, 40, 48, 59, 66, 73, 93, Hypotympanum 35, 56, 61
von Bardeleben 3 94
Beaumont 60, 61, 88 Eysell 3 Incudal fossa 30, 66
Benner 56 Incudostapedial articulation 32, 35, 66
Bezold 3 Fetal development 7 Inferior tympanic artery 4, 12
Buch 56 Fibrosis 60 Interatticotympanic diaphragm 11
Folds (variable, position-changing)
Chatellier 11 – chordal 8 Jörgensen 56
Cholesteatoma 26, 27, 37, 57, 94, 95, 98–100 – cruris longi incudis 8
Cholesterol crystals 27, 36, 43, 55, 93 – incus intercrural 32 Kuppelraum 3
– granuloma 57, 61, 86, 97 – interossicular 47
Chorda tympani nerve 4, 5, 8, 12, 17, 42, – lateral incudomalleal 8, 11, 12, 15, 20, Labyrinth capsule 3
72, 95 28, 30, 32, 62 Labyrinthine bone 58
Chordal fold 8 – medial ossicular 32 Lanugo hair 57, 75, 86
103
Lateral attics 8, 12, 62, 69 – – infant 56–58, 60 Sheehy 38
– attic bone 5, 12 – – infectious 55 Shrapnell 5
– incudomalleal fold → Folds – – neonatorum 32, 55, 59 Shrapnell’s membrane 3–5, 14–16, 18, 19,
– malleal space 5, 6, 8, 9, 12, 13, 15, 16, – – recurrent 87–89, 93 21, 24–27, 75, 93, 94
18–21, 27–29, 38, 64, 66, 69 – – secretory 5, 6, 11, 27, 34, 36, 60, 93 Siebenmann 3, 6, 7, 9, 37, 38, 40, 42, 49,
Lemoine 11 – – sterile 32 50, 97
Ligaments (ligamental folds, position-fixed) Otosclerosis 43 Sinus tympani → Tympanic sinus
– anterior malleal 4, 5, 8, 12, 15, 42, 46, 48 Spina capitis mallei 5
– anterior malleal suspensory 12 Paukenhöhle 3 Squamous bone 3, 61, 66, 74, 85–88
– lateral malleal 5, 9, 12, 13, 15 Periosteal bone 88 – epithelial cells 55, 75, 79, 85
– posterior incudal 8, 10, 28, 30, 31, 36 Petrotympanic fissure 12, 42 Stapedial fold → Folds
– posterior malleal 5, 12, 13, 29 Petrous bone 3, 66, 73, 85–88 Stapedius tendon 5
– superior malleal 28, 38 Plica cruris longi incudis 8 Superior attic → Epitympanum
Lower lateral attic 8, 13, 15, 17, 21, 22, – transversa 3, 7 – pouch 4, 9
28–30, 29, 63, 66, 74, 95 Pneumatic cell 7 Supratubal recess 3, 6–8, 27, 39, 40, 42, 43,
Pneumatization 36, 38, 47, 57–61, 66, 73, 45–51, 62, 63, 66, 69, 94
McLellan 56 85–88
Malleal ligaments 4, 5 Politzer 5, 6, 14, 15, 17, 22 Tensor fold → Folds
– spine 5 Portal for organization 26, 62, 68 – tendon 3, 6, 10, 16, 31, 36, 38, 40, 47, 48
Margo tympanicus 5 Position-changing folds 7 Tos 6
Mastoid antrum 3, 12, 59, 61, 62 Position-fixed folds 7 Transverse crest 3, 6, 38, 39, 94
Mastoidectomy 97 Postauricular incision 94, 95 von Tröltsch 3, 5, 12, 18
Mastoiditis 60 Posterior attic → Epitympanum Trommelfelltasche 7
Meconium 55, 57, 61 – pouch 4, 9, 17, 18, 21, 23, 25, 95 Tubal orifice 5
Medial attic → Epitympanum – tympanic isthmus 12, 30 – tegmen 38
– ossicular fold → Folds – tympanotomy 21 Tumarkin 60, 88
Mesotympanum 4, 17, 18, 21, 26, 28, 40, – tympanum 12, 27, 30, 59, 61, 72, 74 Tympanic cavity 3, 9, 49, 58, 64, 79
45, 56, 98 Potter 57 – glomus tumor → Glomus tumor
Microsurgery 94–97 Proctor 9, 12, 13, 16, 18, 31, 32, 38, 46 – isthmus 9, 10, 12, 28, 31–33, 34–36, 40,
Morimitsu 94, 97 Protympanic recess 12 48, 64, 67, 72, 93, 94
Mucosal fibrosis 58 Prussak 4, 5, 9, 17, 22 – membrane 4, 5, 17, 23, 57
– hyperplasia 58, 59 Prussak’s space 5, 6, 9, 11–27, 46, 49, 63, – sinus 65, 67, 73
– polyps 56, 69, 73 64, 66–70, 72, 93–95, 98 – spine
Multinucleated cells 64, 79, 81, 82, 86 Pseudocystic granulation 67, 78–81, 84 – – anterior 5, 12, 38
Pyramidal process 10, 31 – – posterior 5, 17, 23
Northrop 57 Tympanostomy tube 27, 34, 75, 89, 93
Notch of Rivinus 3, 94, 95 Recessus epitympanicus 3 Tympanotomy
Retraction pocket 25, 27, 75, 93–95 – anterior 11, 94
Obturator fold → Folds Riding columella 95 – posterior 11, 94, 95
Ojala 60, 87 Round window membrane 65, 73, 84
Organization 74, 80–83, 87 – – niche 73, 79 Upper lateral attic 8, 12, 13, 28, 38, 62
Otitis media Rüedi 59, 60, 68
– – acute 5, 19, 59 Wildberg 4
– – adhesive 34, 89, 93 de Sa 56 Wittmaack 4, 26, 55–61, 87, 88
– – chronic 26, 37, 40, 48, 59, 60, 93, 98 Schuknecht 9 Wullstein 6, 31
– – hyperplastic 58 Schwartze 3