Добавил:
Sekretar
kiopkiopkiop18@yandex.ru
t.me/Prokururor I Вовсе не секретарь, но почту проверяю
Опубликованный материал нарушает ваши авторские права? Сообщите нам.
Вуз:
Предмет:
Файл:Ординатура / Хирургия / Библиотека им академика М.И. Перельмана / Книга_4532_Библиотеки_им_академика_М_И_Перельмана
.pdf
Benign Neoplasms and Tumor-Like Lesions 135
ba
Fig. 8.26a–c. Juvenile angiofi broma: MR signal inten-
sity and pattern of enhancement. The juvenile angiofi broma exhibits intermediate signal, comparable
to brain parenchyma, on both TSE T2 (a) and plain
SE T1 (b) sequences. a Part of the lesion extends into
the left nasal fossa, scarcely separable on plain T1
from the middle turbinate (b, white arrows). Intense
enhancement after paramagnetic contrast injection
(c). Major intra-lesion vessels are detected on MR as
serpiginous fl ow voids (arrowheads). A small part of
juvenile angiofi broma abuts the left pterygopalatine
c
fossa (black arrows)

136
R. Maroldi et al.
Fig. 8.27 Juvenile angiofi broma. TSE T2, coronal plane.
Hyperintensity of the intra-sphenoidal component of the lesion refl ects spontaneous cystic changes
a
Fig. 8.28a–c. Juvenile angiofi broma. CT after contrast adminis-
tration, axial plane (a); TSE T2 (b) and fat saturated SE T1 (c)
after contrast administration, coronal plane. a The lesion completely fi lls the nasopharynx and the nasal fossae, hypodense
cystic areas are detectable (asterisks). CT clearly depicts erosion of the medial pterygoid plate and tumor growth into the
cancellous bone of pterygoid process (black arrows); fi ngerlike projections of the mass spread into the infratemporal fossa
(white arrows). MR shows invasion of cancellous bone on both
TSE T2 and enhanced fat-sat T1, with the lesion extending into
the pterygoid root and into the greater sphenoid wing (white
arrows). These fi nding are more clear-cut on the fat saturated
sequence (c). The lesion grows along the path of the foramen
rotundum (black arrows) and, submucosally, along the under
surface of sphenoid bone (black arrowheads), without invading cavernous sinus (single black arrowhead in b)
b
c

Benign Neoplasms and Tumor-Like Lesions 137
8.7.5.1
Pathways of Spread
Of course, the pathways of spread of juvenile angiofi broma will infl uence the choice of the surgical
approach (Fig. 8.29).
Owing to its nature, juvenile angiofi broma tends
to grow along the paths of least resistance, causing
displacement of adjacent soft tissues, rather than invasion. Adherence can be present, particularly when
the lesion contacts the dura, but dural or brain infi ltration is rare (Danesi et al. 2000; Scholtz et al.
2001).
Its peculiar dual pattern of bone involvement (remodeling and destruction) is likely to result from two
different mechanisms of interaction between juvenile
angiofi broma and bony structures. Displacement of
the periosteal invested cortical surfaces causes remodeling and thinning, with only late breakthrough
and destruction, whereas the direct growth of juvenile angiofi broma along perforating arteries into the
cancellous root of the sphenoid lets the lesion extend
into the medullary content of the sphenoid (fl oor of
sphenoid sinus, greater wing). This could explain
why the vidian canal (invested by a periosteal layer),
though the closest to the growing lesion, shows more
frequently enlargement of its anterior third, rather
Fig. 8.29. Pathways of spread of juvenile angiofi broma. A, ex-
tent through the sphenopalatine foramen into the nasal fossa
(1) and – via the choana – into the nasopharynx (2); through
the erosion of the sphenoid sinus fl oor into the sphenoid sinus
(3). B, extent from the pterygopalatine fossa along the foramen
rotundum (actually a groove or a complete bone canal) (4),
into the cancellous bone of the greater wing of the sphenoid
(5), and into the masticator space (6)
than destruction. Moreover, further posterior extension into the canal is infrequent, and usually observed
in advanced lesions.
Due to the knowledge of the elementary interactions of juvenile angiofi broma with surrounding
structures and its constant site of origin, the patterns
of spread are highly predictable (Lloyd et al. 2000b;
Schick and Kahle 2000).
From its site of origin in the pterygopalatine fossa,
the juvenile angiofi broma extends medially into the
nasal cavity and nasopharynx – the areas of least resistance – via enlargement and erosion of the sphenopalatine foramen. Growth of tumor anteriorly indents
the postero-superior maxillary sinus wall, resulting in
anterior bowing of the sinusal wall, the so-called antral
sign, described by Holman and Miller on lateral plain
X-ray (Lloyd and Phelps 1986) (Fig. 8.30) .
The lateral extent, via an enlarged pterygo-maxillary fi ssure, gives rise to infratemporal fossa spread.
Extension into this space is demonstrated by detecting the “fi nger-like projections” of the enhancing
juvenile angiofi broma characterized by sharp and
lobulated margins (Fig. 8.31). In this area, the least
resistant structure consists of the fat tissue between
the pterygoid muscles, usually splayed.
From the pterygo-maxillary fi ssure, the lesion
can also access the apex of the orbit through the inferior orbital fi ssure, and further extends into the
middle cranial fossa via the superior orbital fi ssure
(Fig. 8.32).
Posterior spread from pterygopalatine fossa is
almost certainly the most dangerous for the patient
because it enables the juvenile angiofi broma to penetrate the cancellous bone of the root of pterygoid
process. From this site, juvenile angiofi broma extends both medially, into the fl oor of the sphenoid
sinus, and laterally, into the greater wing (Lloyd et
al. 1999). As the center of growth of the intrasinusal
component of juvenile angiofi broma is located at the
intersection between the fl oor and the lateral wall,
there is evidence to suggest that the intrasinusal extension comes from the root of the pterygoid, rather
than being due to upward extension from the choana.
Lateral spread allows the juvenile angiofi broma to replace the diploe of the greater wing, usually with late
erosion of the inner table.
The key to detecting the diploic invasion consists
of differentiating its medullary content from the lesion on the basis of CT density and MR signal. On
CT, this is achieved by the strong enhancement of
juvenile angiofi broma within the diploe. On MR, the
optimal discrimination is obtained by combining a
plain T1 with a post-contrast T1 with or without fat

138
R. Maroldi et al.
ba
Fig. 8.30a,b. Juvenile angiofi broma. SE T1 after contrast administration, sagittal plane. a The lesion infi ltrates the medullary
bone of clivus, the intracranial cortical boundary being detectable only in its inferior part (white opposing arrows). Reactive
thickening of the adjacent dura of prepontine cistern (white arrowhead). Submucosal grow th along the undersurface of the
sphenoid bone is appreciated (black arrows). b The lesion indents the posterior wall of the maxillary sinus, reaches the inferior
orbital fi ssure (black arrows) and spreads along the foramen rotundum (white arrows)
ba
Fig. 8.31a,b. Juvenile angiofi broma. CT (a) and SE T1 (b) after contrast administration, both in the axial plane. a Finger-like
projections of the juvenile angiofi broma grow into the infratemporal fossa. The posterior wall of the maxillary sinus is remodeled and interrupted (black arrows). A small part of the lesion extends posteriorly to the pterygoid plates (into the pterygoid
fossa) along the medial pterygoid muscle (white arrows). b The enhancing juvenile angiofi broma occupies the left masticator
space (black arrows). Laterally it borders the temporalis muscle, posteriorly it reaches the foramen ovale. An enlarged middle
meningeal artery is detected (white arrowheads). Remodeling of posterolateral maxillary sinus wall is seen (white arrows)

Benign Neoplasms and Tumor-Like Lesions 139
Figure 8.32. Juvenile angiofi broma. Enhanced SE T1 in the
coronal plane. The juvenile angiofi broma invades the sphenoid sinus through the fl oor. A second component reaches the
cavernous sinus (white arrows) through the foramen (groove)
rotundum, running above the maxillary nerve (arrowhead)
saturation (Fig. 8.33). The latter makes it possible
to easily distinguish the hyperintense enhanced juvenile angiofi broma from the suppressed signal of
the normal bone marrow. An alternative option to
Fig. 8.33. Juvenile angiofi broma. Enhanced SE T1 in the coronal plane. The juvenile angiofi broma completely replaces the
cancellous bone of both the left pterygoid root and the greater
sphenoid wing. The inferior orbital fi ssure is reached through
a defect of the lateral wall of sphenoid sinus (thin black ar-
rows). Intracranial growth is appreciated along the fl oor of
middle cranial fossa (white arrows). The extracranial component of the lesion invades the infratemporal fossa. The lateral
pterygoid muscle is inferiorly displaced (thick black arrows)
reduce marrow signal on T1 sequences consists of
decreasing the TR and selecting thinner sections:
while the signal of bone marrow greatly diminishes,
juvenile angiofi broma maintains its hyperintensity.
be found in most cases, making complete removal
feasible (Danesi et al. 2000).
Replacement of the cancellous structure of the clivus
can be observed in advanced lesions that completely
fi ll the sphenoid sinuses and displace both the lateral
8.7.5.2
Angiography
walls and the roof.
Juvenile angiofi broma shows two different types
of intracranial invasion: extent along a canal, and
spread through bone destruction. It is interesting to
note that even medium size lesions may gain access
into the middle cranial fossa by growing along the foramen rotundum, and running lateral to the cavernous sinus to reach the anterior aspect of the Meckel’s
cave. Generally, the second pattern occurs when huge
lesions break through the inner table of the greater
wing or the lateral sphenoid sinus walls.
Regardless of the pattern of intracranial access, infi ltration of the dura is very rare. In fact, it has been
recently reported that even when cross sectional imaging suggests cavernous sinus invasion or internal
carotid artery involvement, a dissection plane can
Surgical resection is currently the most widely accepted treatment for juvenile angiofi broma. Due to its
high vascularization, surgical removal can sometimes
be diffi cult because of signifi cant intraoperative hemorrhage, resulting in incomplete resection and higher
rate of persistence. Pre-operative embolization was
introduced in 1972 to obtain lesion devascularization and facilitate complete excision of the tumor
(Roberson et al. 1972). Nowadays, the availability of
intra-arterial digital subtraction angiography, microcatheters, and embolic agents – such as PVA particles
– makes superselective embolization of feeders easier
and safer (Valavanis and Christoforidis 2000).
Though some authors questioned the usefulness of
this procedure, as in their experience no signifi cant

140
R. Maroldi et al.
difference in surgical bleeding was observed, there is
increasing evidence that embolization is a safe and
effective method to reduce intra-operative blood loss
(Siniluoto et al. 1993; Moulin et al. 1995; Li et al.
1998). Nevertheless, the shrinkage of lesion achieved
by embolization has been indicated as a contributory cause to incomplete excision of juvenile angiofi broma by McCombe et al. (1990).
At present, the role of angiography is to provide a
detailed map of feeders, demonstrating the recruitment of internal carotid artery, vertebral or contralateral external carotid artery branches, and to obtain
preoperative devascularization.
According to Lasjaunias et al. (1980), the angio-
graphic fi ndings of juvenile angiofi broma consist of
moderate enlargement of feeding arteries, intense
“parenchymal” blush, absence of large arteriovenous
shunts, or early venous return (Fig. 8.34).
The pattern of arterial feeders recruited is predictable in most cases. It is strictly related to the pathways of spread, but not to the actual size of the lesion,
though most large lesions are multi-compartmental.
In our experience of 15 patients treated by exclusive
endonasal excision, there was no correlation between
the volume of juvenile angiofi broma and the number of feeding vessels (Nicolai et al. 2003). However,
recruitment of internal carotid artery, vertebral
branches was signifi cantly more frequent among
lesions with several external carotid artery feeders. Notably, though internal carotid artery feeders
Fig. 8.34a,b. Juvenile angiofi broma. Enhanced SE
T1 in the coronal plane
(a); intra-arterial DSA (b).
a On MR, the lesion shows
a prevalent endoluminal
growth within the nasopharynx. Upwards extension in the pterygoid root
and within the sphenoid
sinus is also appreciated.
b DSA demonstrates vascular feeders arising from
the distal part of the sphenopalatine artery; a more
prominent blush is observed at the level of the
nasal part of the lesion
a
b
Fig. 8.35a,b. Juvenile an-
giofi broma. Enhanced SE
T1 in the sagittal plane
(a); intra-arterial DSA (b).
a The lesion fi lls the sphenoid sinus. Infi ltration of
the clivus is demonstrated
by the encroachment of
both its cortical layers
and replacement of the
medullary content. b DSA
demonstrates the several
subtle feeders from the
internal carotid artery, not
embolized
ba

Benign Neoplasms and Tumor-Like Lesions 141
were demonstrated in approximately 47% of cases
(Fig. 8.35), intracranial extent was present only in
13%.
During the last decade the availability of small
particles and microcatheters has made it possible to
reach even peripheral small branches of the external
carotid artery, preserving adjacent normal vessels
from being devascularized by the more proximal
occlusion obtained by Gelfoam or Spongel embolization (Fig. 8.36). In fact, the goal of embolization
is to achieve vessel occlusion at the capillary level.
Consequently, polyvinyl-alcohol particles with a
minimal size of 150 µm have been suggested, as
signifi cant arteriovenous shunts have been demonstrated for particles of 50 µm or less by nuclear medicine techniques (Schroth et al. 1996).
a
The rate of minor and major complications for
embolization of external carotid artery branches
is negligible, approximately 4% (Ungkanont et al.
1996).
It is evident that the major challenge to angiography regards the management of juvenile angiofi broma vascularization by internal carotid artery
feeders. Advanced lesions with intracranial extent
have been successfully excised after pre-operative
embolization of external carotid artery branches
with acceptable blood loss, despite involvement of
the internal carotid artery branches in the blood
supply. Devascularization by direct tumor puncture and embolization, advocated by Casasco et al.
(1999), entails an unacceptable risk of major neurologic complications. Balloon occlusion and sacrifi ce
of the internal carotid artery is required in rare cases
(Casasco et al. 1999).
8.7.6
Follow-Up
Fig. 8.36a,b. Juvenile angiofi broma. DSA, before (a) and after
(b) embolization. The juvenile angiofi broma is fed by terminal branches of the sphenopalatine artery, it occupies the nasal fossa and the nasopharynx; upwards it extends into the
sphenoid sinus, downwards it reaches the superior part of the
oropharyngeal lumen. Embolization permits complete devascularization of the mass
b
There are two different types of persistent lesions:
those left intentionally because their resection would
require unacceptable neural damage, and those left
due to intraoperative oversight (Figs. 8.37, 8.38). In
both cases, proper management mandates a precise
assessment of site, size, and extent of the lesion.
Recurrences are a peculiar characteristic of juvenile angiofi broma. In fact, most authors doubt
this theory and consider more likely the hypothesis of incomplete excision leaving lesion remnants
(Andrews et al. 1989; Chagnaud et al.1998; Lloyd
et al. 2000b; ).
To support this hypothesis is the fact that most
persistent lesions occur within months or a few years
of primary treatment and more commonly are found
in anatomical areas diffi cult to reach at surgery. This
presumption is consistent with Chagnaud et al.
(1998) who demonstrated that lesion remnants were
already detectable at the fi rst follow-up with cross
sectional imaging.
Persistent lesions are more frequent when the primary juvenile angiofi broma invades the infratemporal fossa, sphenoid sinus, pterygoid root, clivus, and
cavernous sinus (Herman et al. 1999; Howard et al.
2001). According to McCombe et al. (1990), “recurrences” are also related to primary juvenile angiofi broma size, being more frequent in large lesions.

142
Fig. 8.37a,b. Follow-up of juvenile angiofi broma. SE T1 before (a) and after (b) contrast administration, axial plane. At 4 years
after incomplete surgical resection a residual submucosal lesion is detected within the greater wing of sphenoid bone (black
arrows). b After contrast administration, the lesion shows the typical enhancement of juvenile angiofi broma. Posteriorly it is
in close contact with the carotid canal (black arrowhead). Only a smooth re-epithelized mucosa is visible from the nasal fossa
(white arrows)
R. Maroldi et al.
ba
a
Fig. 8.38a–c. Juvenile angiofi broma. Same patient as in Fig. 8.37. Enhanced T1 sequences in the coronal plane. Pretreatment
examination for planning sub-total resection of the residual juvenile angiofi broma (a). Follow-up study after 13 months (b) and
18 months (c) after surgery. During surgery, a part of the lesion was intentionally left within the greater wing of the sphenoid
bone, due to its relevant lateral extension and due to middle cranial fossa extradural invasion. Both the superior (black arrows)
and medial (white arrows) limits of the lesion appear rather concave on preoperative and fi rst follow-up MR examination,
whereas 18 months after surgery their surface results more or less convex. This change in shape indicates progression of the
lesion
Depending on the clinical condition (known or
high-risk remnant lesion vs low-risk), follow-up will be
scheduled: every few months for known remnants/high
risk patients and every 6 months during the fi rst year,
then yearly, for the others, respectively (Chagnaud et
al. 1998; Ro ger et al. 2002). In the fi rst case, the goal
of imaging is to detect changes in size of known remnants in order to decide the proper treatment strategy
or to identify lesions arising from potential sites of
b
persistence. In the second case, follow-up should be
extended until adulthood, even though date supporting this strategy have not been provided yet.
Of course, unexpected persistent lesions show
the same imaging characteristics as the primitive
ones (Fig. 8.39), but their detection may be hampered by postoperative changes. These consist of
altered bony and soft structures, due to surgical
resection, healing, and chronic inflammatory re-
c

Benign Neoplasms and Tumor-Like Lesions 143
ba
c d
Fig. 8.39a–d. Recurrent/persistent juvenile angiofi bromas in two different patients. CT after contrast administration, axial plane
(a); endoscopic view (b); enhanced VIBE sequence in sagittal and coronal planes (c,d). a A thin plaque of non-enhancing scar
tissue is detected at the level of sphenopalatine foramen (white arrow). The submucosal relapse (black arrows) has its epicenter
at the pterygoid root level where the medullary bone is replaced by the enhanced juvenile angiofi broma (asterisk). Erosion of
pterygoid bony boundaries is seen (arrowheads). The encircled area corresponds to the endoscopic view (b), which does not
show any abnormality. c,d In a different patient, a relapsing juvenile angiofi broma is detected 2 years after endonasal surgery
(arrows) located submucosally within residual sphenoid sinus fl oor
action of the sinonasal mucosa (Chagnaud et al.
1998).
Bony defects of the sinonasal framework and skull
base mainly depend both on the specifi c surgical approach and on the extent of the erosion caused by
juvenile angiofi broma.
However, some constant postoperative fi ndings
appear to be the partial or total excision of the posterior, and medial antral walls, of the pterygoid root
and of the fl oor of the sphenoid sinus, detectable in
almost half of the patients, regardless of the surgical
approach (Fig. 8.40).
The scar tissue replacing the site previously occu-
pied by juvenile angiofi broma – pterygopalatine fossa,
inferior orbital fi ssure, infratemporal fossa – appears
usually hypointense on both T1 and T2 and does not
enhance on CT or MR after contrast agent administration. Conversely, enhancement is shown either by
the thickened infl amed sinonasal mucosa or by lesion
remnants. Of course, infl ammatory mucosal changes
appear hyperintense also on T2 sequences, whereas
juvenile angiofi broma has intermediate signal intensity. In addition, signal voids may be observed, their
detection probably depending on the overall size of
the persistent juvenile angiofi broma.
Nevertheless, it is not infrequent to image enhancing submucosal areas fi lling previous site(s) of the lesion. Serial examinations appear necessary to effectively estimate their growth, a fi nding that should be
consistent with residual lesions (Fig. 8.41). However,
known lesion remnants – left in place during surgery
– may not show increase in size (Deschler et al.
1992). Angiography has been advocated to obtain a
defi nite diagnosis, but it is hampered by false positive
results (Bremer et al. 1986).
Owing to this limitation, and because the surgical
treatment of small residuals – particularly those intracranially located – is rather controversial, the role

144
R. Maroldi et al.
ba
c
Fig. 8.40a–d. Juvenile angiofi broma: postsurgical changes after endonasal surgery, normal fi nd-
ings. Enhanced SE T1 in the coronal plane (a,b), TSE T2 in the axial plane (c), endoscopic view
(d). Thickening of the mucosal layer of both maxillary and sphenoid sinuses and ballooning of
the mucosa (with liquid content) is quite a common fi nding in early postoperative examinations.
The encircled area corresponds to the endoscopic view. NS, nasal septum; SS, sphenoid sinus; P,
mucosa investing the posterior maxillary sinus wall; M, inferior border of the widened maxillary
ostium; IT, inferior turbinate
of angiography remains unclear (Chagnaud et al.
1998).
In effect, small submucosal residuals may be dem-
onstrated by imaging modalities in otherwise asymptomatic patients with negative endoscopy. To adopt a
proper treatment strategy, it would be necessary to
know more about their spontaneous evolution: recurrent symptoms, regression of residual lesion, or
persistent asymptomatic, residual complications.
Moreover, spontaneous regression of juvenile angiofi broma residues has been reported (Stansbie and
Phelps 1986; Dohar and Duvall 1992).
A systematic postoperative imaging follow-up is,
consequently, suggested. The use of MR is preferable
because it avoids further radiation exposure of the
young patients.
d
Соседние файлы в папке Библиотека им академика М.И. Перельмана
