Keywords
costal cartilage - nasal septal perforations - rhinologic procedures - diced cartilage
- fascia graft
Even though the prevalence of nasoseptal perforations has decreased, therapeutic approaches
for closure are repeatedly in the focus of research. For a long time, closure of nasoseptal
defects has seemed like squaring the circle as countless methods have been attempted
and described. As early as 1929, Jackson recommended enlarging small perforations
in order to reduce whistling and crusting by moving the posterior edge of the defect
farther posterior where humidification of the inhaled air is higher.[1] In the following decades, ear nose throat (ENT) and plastic surgeons described intranasal
muco (perichondrial) flaps, including rotational and advancement mucosal flaps (combined
with reduction rhinoplasty) as well as lateral nasal wall and inferior turbinate flaps.[2]
[3]
[4] Due to the shortage of nasal mucosa in large perforations, multistage labial sulcus
flaps combined with skin grafts were suggested.[5]
[6]
[7]
Also, there exist advocates of external rhinoplasty and midfacial degloving, approaches
that allow complete exposure of the septum and enable different repair options.[2]
[8] Moreover, nonsurgical methods such as customized obturator buttons have been described
and are still in use today.[9]
[10] At about the same time, connective tissue autographs and their combination with
local mucosa flaps have been experimented with, including free septal, auricular and
costal cartilage grafts, vomerine bone, and mastoid periosteum.[2]
[3]
[11] It was in 1980 that Fairbanks introduced temporalis fascia for septal reconstruction,
a graft offering various amenities, especially when combined with cartilage.[12]
[13] When in 2000 diced cartilage was brought back into use by the work of Erol and his
colleagues, diced cartilage in fascia (DC-F) grafts became an essential tool in corrective
rhinoplasty.[14]
[15]
[16]
[17] However, DC-F grafts have not yet been used for closure of septal perforations.
Even though the surgical toolbox seems limitless, every approach has its pitfalls,
and we are still facing difficulties including flap ischemia, nasal obstruction due
to the volume of the graft, as well as remaining or reperforations with loss of support
and saddling of the nose, necessitating revision surgery. Many patients wearing septal
prostheses complain of discomfort, crusting, and movement of the button and finally
ask for surgical closure. Having in mind the positive characteristics of a diced or
crushed cartilage in fascia graft as proven on extranasal application (no limitations
due to graft size or thickness, no absorption due to foreign body reaction, and sufficient
stability, independent of a vascular pedicle), we believe that this might also be
an adequate procedure for the closure of septal defects.
In the following, we describe and prove the feasibility of closing septal perforations
in what we call the “German ravioli technique” on 18 patients from three cooperating
institutes.
Materials and Methods
In cooperation with the otorhinolaryngology departments of three hospitals, we performed
surgical repair of nasoseptal perforations using either an autogenous diced cartilage
combined with human fibrin glue or crushed cartilage in fascia graft (DC-F) in 18
patients from 2020 until beginning of 2022. We included patients independent of the
size and site of the defect. When an inflammatory etiology was suspected, a histopathologic
exam of the septal mucosa was performed to rule out specific inflammatory processes
such as granulomatosis with polyangiitis. These cases were excluded from our study
as well as patients having undergone radiation including the midface or nasal region.
Patients mainly presented with crust formation, recurring epistaxis, and nasal blockage.
In some of the cases, cartilaginous and/or osseous nasal deformity was present in
addition to the septal pathology. The group consisted of nine males and nine females
with a median age of 43.9 years and age range from 24 to 73 years. For further patient
characteristics see [Table 1].
Table 1
Patient characteristics
Patient
|
Age
|
Sex
|
Etiology
|
Symptoms
|
Previous surgery
|
Risk factors
|
1
|
51
|
f
|
Decongestive NS
|
Epistaxis, crusts, blockage
|
None
|
Diabetes mellitus, 10py nicotine
|
2
|
32
|
m
|
Iatrogenic
|
Epistaxis, crusts, blockage
|
Septoplasty, nasal bone reposition
|
Colitis ulcerosa, MTX
|
3
|
25
|
f
|
Frequent cauterization
|
Epistaxis, crusts, blockage
|
None
|
None
|
4
|
24
|
m
|
Iatrogenic
|
Crusts, blockage
|
Septoplasty, FESS
|
None
|
5
|
25
|
f
|
Unknown
|
Blockage, saddle nose
|
None
|
None
|
6
|
51
|
m
|
Posttraumatic iatrogenic
|
Blockage
|
Undetermined procedure involving the nose
|
Factor-V-Leiden
|
7
|
58
|
m
|
Iatrogenic
|
Blockage
crusts
|
Rhinoplasty
septal prosthesis
|
Diabetes II, coronary bypasses
|
8
|
35
|
m
|
Iatrogenic
|
Blockage, epistaxis
|
Septoplasty
|
None
|
9
|
73
|
m
|
Iatrogenic
|
Blockage, crusts, epistaxis, whistling sound of breath
|
Septoplasty
|
None
|
10
|
56
|
f
|
Unknown
|
epistaxis
|
None
|
None
|
11
|
34
|
m
|
Posttraumatic
|
Blockage, epistaxis
saddle nose
|
None
|
None
|
12
|
33
|
f
|
Posttraumatic
iatrogenic
|
Blockage, crusts, epistaxis
|
None
|
None
|
13
|
55
|
f
|
Unknown
rhinitis sicca
|
Blockage
|
None
|
None
|
14
|
47
|
f
|
Unknown
|
Blockage
|
None
|
None
|
15
|
29
|
f
|
Iatrogenic
|
Crusts
|
Septoplasty
|
None
|
16
|
33
|
m
|
Unknown
|
Epistaxis
|
None
|
15py nicotine
|
17
|
58
|
f
|
Privinism
|
Crusts, epistaxis
|
None
|
Diabetes II
|
18
|
71
|
m
|
Iatrogenic
|
Crusts
|
Septoplasty
|
None
|
Preoperative
Using nasal endoscopy, the defect was measured in its antero-posterior and cranio-caudal
dimensions and classified as small (up to 1 cm diameter), medium sized (1 to 2 cm
diameter), and large (more than 2 cm in diameter).[18] The site of the defect was documented according to the commonly used anatomic areas
I to V.[19] Additionally, further information on previous treatment measures (nasal irrigation,
moisturizing, and occlusion) or previous surgery involving the nose, paranasal sinuses,
and midface was obtained. Potential risk factors such as previous and ongoing nicotine
abuse and impaired wound healing due to medication and/or systemic disease were documented.
After presenting the technique to be used in detail, informed consent was obtained,
specifically clarifying the experimental character of the method as well as determining
the surgical approach and donor site for the fascia (temporalis fascia or fascia lata)
and cartilage grafts (auricular, septal, or costal; see [Table 2]). In patients with a low hairline or long hair temporalis fascia was preferred due
to cosmetic reasons. Preference was independent of the size of the graft needed for
septal repair.
Table 2
Surgical specifications
Patient
|
Defect size (cm)
|
Defect localization (area)
|
Surgical approach
|
Splints (days)
|
1
|
1 × 1
|
II–III
|
Open
|
20
|
2
|
4 × 4
|
II–V (complete cartilaginous septum)
|
Closed
|
20
|
3
|
0.8 × 0.8
|
I–II
|
Closed
|
20
|
4
|
1 × 1
|
II–III
|
Closed
|
20
|
5
|
2.5 × 2.2
|
II–III
|
Open, combined rhinoplasty
|
25
|
6
|
1 × 1
|
I–III
|
Open, combined rhinoplasty
|
42
|
7
|
3 × 4
|
II–IV
|
Open, combined with nasal valve reconstruction
|
42
|
8
|
3 × 4
|
I–III
|
Open, combined with nasal tip graft
|
42
|
9
|
1 × 1
|
II–III
|
Closed
|
21
|
10
|
2 × 1.5
|
II–III
|
Closed
|
49
|
11
|
2 × 1
|
II–III
|
Open, combined rhinoplasty
|
36
|
12
|
4 × 2
|
II–III
|
Open, combined rhinoplasty
|
39
|
13
|
0.5 × 0.5
|
I–II
|
Open, combined rhinoplasty
|
47
|
14
|
2 × 1
|
II–III
|
Closed, septal exchange
|
42
|
15
|
1.8 × 1.2
|
II–III
|
Closed
|
21
|
16
|
4 × 4
|
II–IV
|
Closed
|
28
|
17
|
1 × 1
|
II
|
Closed
|
21
|
18
|
1.5 × 2
|
I–II
|
Closed
|
42
|
Intraoperative
The procedure was performed under general anesthesia as an inpatient procedure. Generally,
an endonasal approach was chosen and all relevant septal pathologies were corrected.
However, when relevant deformity or deviation of the external nasal framework needed
to be addressed, an open rhinoplasty approach was performed. Either a right-sided
transfixion incision or a transcolumellar inverted-V-incision (combined with an intercartilaginous
incision) was used. Via upper and lower tunneling, the mucoperichondrium was dissected
from the underlying cartilage to expose the septal defect. This design facilitated
a correct placement of the graft.
Depending on the patient's hairline position, the fascia graft was harvested either
in the temporal or the distal lateral thigh region as reported several times.[20]
[21]
[22] The cartilage graft was harvested either from the posterior septum, the auricular
region, or the rib as described previously[23]
[24]
[25]
[26]. Dicing or crushing of the cartilage was performed as recommended in previous publications
and blended with human fibrin glue.[27]
[28]
[29] The diameter of the cartilage dices measured about 1 to 1.5 mm. The crushed cartilage
segments were adapted to the individual septal defect. The strip of fascia was folded
into the form of a pad. This pad was filled with the diced or crushed cartilage in
fibrin glue, resembling a “ravioli.” The amount of cartilage used corresponded to
the diameter of the defect in order not to produce obstruction. The open sides of
the fascia graft were closed by a continuous suture with 5.0 Vicryl or 6.0 Prolene.
The obtained DC-F pad graft was then placed into the defect with a fascia overlap
of approximately 5 mm and fixed on to its edges in an underlay technique using 5.0
or 6.0 Vicryl or 4.0 Monocryl mattress sutures penetrating two layers of fascia and
two layers of mucoperichondrium at the edge of the defect (see [Fig. 1a–d]).
Fig. 1 (a) Diced cartilage. (b) Diced cartilage in fascia with fibrin glue. (c) Suturing of the DC-F graft (“ravioli”). (d) Placing of the DC-F graft.
Finally, individually adapted septal splints were placed on either side, ensuring
full coverage of the graft on all sides (see [Figs. 2a–d] and [3a–d]). They were fixed with mattress sutures. The used incisions were closed in the usual
manner. We used hemostyptic gelatin tampons (Gelita by B. Braun Melsungen AG) for
nasal packing.
Fig. 2 (patient 4). (a) Septal defect in area II–III (left lateral view). (b) DC-F graft (left lateral view). (c) DC-F graft (right lateral view). (d) Septal splint covering the graft.
Fig. 3 (patient 7). (a) Septal defect in area II–IV (right lateral view). (b) Septal defect in area II–IV (left lateral view). (c) DC-F graft (right lateral view). (d) Septal splint covering the graft.
Postoperative
The procedure was performed as an inpatient procedure. Analgesics were administered
following the department̀s standard protocol. In our institute, no prophylactic antibiotic
therapy was conducted. In the cooperating hospitals, patients were started on prophylactic
antibiotic therapy with either ampicillin/sulbactam 3 g or ceftriaxon 2 g intravenously
while hospitalized, continued orally with amoxicillin/clavulanic acid 875/125 mg or
cefuroxim 500 mg after dismissal for a duration of 7 days in total. The drainage from
the lateral thigh and the thoracic region was removed after 2 days. The gauze swabs
on the auricle were removed after 5 days. Patients were specifically counseled on
nasal irrigation using saline solution and moisturizing measures using soft nasal
ointments. Also, we recommended temporary partial or complete nasal occlusion using
a semipermeable 3M Micropore fleece tape (3M, St. Paul, Minnesota, ) as previously
published by Wirsching et al.[30] The sutures or skin staples in the harvesting site were removed after 7 to 10 days
in the auricular and temporal region as well as after 10 days in the lateral thigh
and the thoracic region. The septal splints were removed after an interval von 20
to 49 days (mean of 33.1 days), depending on the postoperative aspect of the graft.
Results
The procedure was feasible in all of the 18 patients and was performed by four surgeons
with comparable skill levels and experience in rhinologic procedures (two surgeons
being chief of department and two being experienced ENT, head and neck attendings).
The operating time ranged from 80 to 145 minutes. No relevant postoperative complications,
necessitating acute revision surgery resulting from (septal) hematoma, abscess formation,
or postoperative bleeding, occurred. Analgetic medication following standard protocol
was sufficient. Most of the patients merely complained of nasal blockage due to swelling
of the mucosa, the gelatin tampons, and the splints. When patients were dismissed,
they were again counseled on postoperative care using nasal irrigation to clean out
mucous and the gelatin tampons, moisturizing soft nasal cremes, and protection measures
for the fascia and cartilage harvesting site. The sutures in the harvesting sites
(auricular, temporal, and thoracic region) were to be removed by the patient̀s general
practitioner. The fascia and cartilage harvesting sites in all patients showed good
wound healing with cosmetically acceptable scarring.
First follow-up was set 7 to 10 days after dismissal. The remaining gelatin tampons
and the septal splints were removed after 20 to 49 days after surgery (mean 33.1).
After splint removal, subsequent follow-up intervals were set to 3, 6, 9, and 12 months
postoperatively (see [Figs. 4a, b] and [5a–c]).
Fig. 4 (patient 5). Graft region marked in yellow. (a) Follow-up endoscopy right lateral view (6 months after surgery). (b) Follow-up endoscopy left lateral view.
Fig. 5 (patient 18). Graft region marked in yellow. (a) Follow-up endoscopy right lateral view (3 months after surgery). (b) Follow-up endoscopy left lateral view. (c) Follow-up endoscopy right lateral view (6 months after surgery).
In all cases, at first follow-up endoscopy showed vital grafts and complete closure
of the septal perforation, even in the subtotal defect in patient 3. After the splints
were removed and the nasal cavity was cleaned, all patients reported comfortable nasal
breathing. The volume of the graft did not seem to disturb the nasal airflow. However,
in three patients (patient 3, 4, and 9), a septal reperforation has occurred over
time. In patient 3, a perforation of 7 mm in diameter was observed when seeing her
for second follow-up and splint removal 20 days postoperatively. In this case, the
remaining graft itself was vital but must have detached from the fragile cranial edge
of the initial defect zone which had already shown advanced mucoperichondrial atrophy
before surgery. We will further follow-up in this case and potentially consider revision
surgery depending on the symptoms. On the 3-months follow-up of patient 4, a reperforation
of less than 1 mm in diameter was observed in the posterior septum, the graft being
vital and showing no signs of infection. The patient has not reported any symptoms
so far. Also, in this case, we will further follow-up to evaluate a potential progression
of the defect size and reoccurrence of symptoms.
In patient 9, a bulging in the lower portion of the DC-F graft was observed underneath
the splints on the 1-week follow-up after surgery. When the splints were removed after
21 days, the cranial portion of the graft was still attached. However, the cartilaginous
part (in this case crushed septal cartilage without fibrin glue) had presumably slipped
off to the nasal floor, leaving the abovementioned bulge in the inferior portion.
When seen again 6 weeks later, a reperforation of 5 mm in diameter had occurred in
the cranial portion of the graft. Obviously, the mattress suture was not stable enough
to hold the cartilaginous part of the graft in place. In addition, as according to
the surgical procedure report, the overlap of the DC-F graft was less than 5 mm in
the cranial portion, which may be ultimately responsible for its detachment. Revision
surgery is planned due to the reoccurrence of symptoms.
So far, we can conclude that reconstruction of septal defects with a DC-F graft leads
to long-term reduction of crust formation and reduction of epistaxis in 16 out of
the 18 patients when seen for their 12-month follow-up. Also, all 16 of these patients
report significant improvement in nasal breathing. Only in two out of three patients,
the reperforation became symptomatic over time.
Discussion
When addressing nasoseptal perforations, one is confronted with several problems and
tasks.
First, in the narrow space of the nasal cavity, the radius of operation is limited
which requires a surgeon's dexterity and attention to detail, especially in placing
and suturing delicate pedicled flaps or grafts of any sort. This may be one of the
reasons for the high variability in the success rates of closing septal defects.
Against the common view, in 1994, Meyer claimed that with his two techniques that
he would reliably close defects of any size.[11] Depending on the size and localization of the defect, he performed either a one-step
procedure including bi-pedicled mucoperichondrial advancement flaps or a three-step
procedure including a composite buccal flap. In 52 out of 55 patients, perforations,
including those over 4 cm in diameter, were closed successfully. However, the sole
use of any local intranasal tissue is prone to flap ischemia due to tension on the
suture lines in the margins of the defect zone, resulting in reperforation over time.
Moreover, many of these procedures are multistep procedures requiring general anesthesia
and hospitalization each time, reducing patient comfort. Accordingly, interpositional
grafts (materials that are placed in between two mucosal flaps) have proven to act
as an interface for vascular ingrowth and mucosal repair.[31]
[32] Contrary to the traditional model of wound healing, fascia has the exceptional ability
to provide a matrix for revascularization and regeneration of original tissue in a
wound area.[33] This approach allows the advancement of regenerated ciliated mucosa across imperfections
in the repair zone instead of replacing it with collagenous scar tissue. For decades,
there has been disagreement on the role of cartilage in nasoseptal repair. The general
opinion was that adding cartilage would not contribute to the stability of the repair
if both mucoperichondrial layers were intact. Today, it is commonly understood that
the use of cartilage in septal defects increases stability in any repair technique
and reduces the risk of reperforation. As observed in patient 9, where the cartilaginous
component of the graft had detached, leaving a mere double layer of fascia, re-perforation
reoccurred in exactly this area 6 weeks after splint removal. This underlines the
importance of adding cartilage to the graft.
In their work, Toriumi and colleagues successfully use costal perichondrium as an
interpositional graft, emphasizing its strategic role when septal defect reconstruction
is combined with rhinoplasty.[32] In our opinion, in patients with a need for moderate to none correctional measures
on the septum or exterior nasal framework, one would avoid harvesting cartilage or
perichondrium from the coastal region due to higher donor site morbidity (more painful,
more obvious scarring) and associated risks (pneumothorax). Only in 1 patient (patient
5) costal cartilage was necessary to ensure sufficient graft material for all pathologies
(closure of large septal defect, septal extension, augmentation of the dorsum, and
augmentation of the supratip break).
Second, understanding the complexity of nasoseptal defects is crucial to their successful
closure. The repair requires stability in order to support the soft tissue and cartilaginous
framework of the outer nose and it must sustain the continuous airflow through the
nasal cavity while not impairing it and must not interfere with maintaining the intranasal
environment.
Looking back at many years of experimentation with fascia and cartilage grafts for
different kinds of indications in the nasal region, their advantages as compared to
stand-alone local intranasal flaps become again apparent in this context. With its
low vascular requirements and its framework for fibroblast growth, fascia is a viable
graft.[34] In combination with cartilage, volume and stability are increased. In 1997, Hussain
and Murthy published their work on a modified tragal cartilage-temporoparietal and
deep temporalis fascia sandwich graft claiming a 100% success rate in closing defects
up to 4 cm in diameter.[13] The limitations to their technique are the defined size of tragal cartilage, the
thickness, and the low malleability of the graft possibly impairing intranasal airflow.
In order to address the limitations of a full-thickness cartilage graft (either coming
from the septal, auricular, or costal region), diced cartilage grafts have been investigated
for soft-tissue reconstruction since the mid-20th century.[35]
[36]
[37] In the following decades, however, the method had been abandoned, until brought
back to our attention by the results published by Erol in 2000.[16] It was assumed that significant absorption of the diced cartilage in the preantibiotic
era and the trend to allograft materials might have been the reason for the draw back.[38] In 2003, Daniel and Calvert showed that the absorption of cartilage can be prevented
by wrapping it into a sleeve of autogenous fascia.[17] Meanwhile, the diced cartilage in fascia graft (often combined with human fibrin
glue) has become a standard surgical procedure, especially in relining and smoothing
of the nasal dorsum or in revision rhinoplasty.[14]
[15]
[39] We have used crushed instead of diced cartilage in three of our patients. In two
of these patients, reperforation has occurred over time. Histopathologic studies have
shown that crushing of cartilage compromises chondrocyte viability to a greater extent
than dicing.[40]
[41] We agree with other works that dicing the cartilage, adding fibrin glue, and wrapping
it in fascia is crucial for graft survival in this method. In this context, we used
cartilage dices of approximately 1.0 to 1.5 mm in diameter due to their high viability
and stability as shown by experimental work of Dong et al.[42]
Third, to restore the integrity of the septum using a free graft, wound healing and
graft survival are the most important factors.
The approach of leaving the splints for the abovementioned intervals follows the principle
of reducing air flow along the septal repair. This results in a significant reduction
of crust formation as well as less vascular trauma, caused by a strong airflow, allowing
better restoration of the intranasal lining. The principle has been shown in previous
studies on patients with hereditary hemorrhagic telangiectasia where nasal occlusion
creates a humid and warm chamber.[30] It is hypothesized that the healing after tympanoplasty using the cartilaginous
palisade technique follows the same mechanism.[43] However, we cannot give any recommendations on a definite time interval for keeping
the splints in place since some of our follow-up intervals are yet too short to evaluate
for long-term results. As demonstrated by Bertlich et al, a 4-week interval of septal
splints seemed to have been sufficient to ensure graft viability. However, again,
the follow-up interval in their clinical trial has been rather short to allow for
a direct conclusion on long-term success rates.[44] Furthermore, the patient number in this study is too small to allow reliable conclusions.
Also, the study design is not intended to compare the three small patient groups from
three institutes. So far, we can report about successful closure and good healing
of the graft after 12 months in 11 out of 14 patients. In this context, longer follow-ups
using the same follow-up method (e.g., videoendoscopic visualization of the graft)
over defined time intervals will provide a more realistic statement on the success
rate of nasoseptal closure.[45]
Furthermore, one must consider that any synthetic material may be the substrate of
selective bacterial growth causing inflammatory processes that might impair wound
healing. The current Sk2 guideline on functional and aesthetic rhino surgery gives
no general recommendation on perioperative systemic antibiotic therapy. Elimination
or prevention of bacterial growth in the nasal cavity may imbalance the physiological
mucous membrane flora causing selective overgrowth of facultative pathogenic bacteria.
This is the reason why our institution dispensed from perioperative systemic antibiotics.
At our cooperating institutes antibiotic therapy was administered a priori because
septal splints were kept in place for longer time intervals. In our experience, in
patients with hereditary hemorrhagic telangiectasia where splints may be left for
an even longer period (even more than 6 weeks) to prevent recurrent epistaxis prophylactic
antibiotic therapy is not necessary. The necessity of this regime should be discussed
in further studies.
Finally, our experience from these cases shows that it is crucial to ensure a sufficient
size of the harvested fascia in order to ensure a tension-free suture. As shown by
Calvert and Kwon, the fascia graft tends to shrink over time.[22] In accordance with our patient collective the overlap of the DC-F graft over the
edge of the defect zone should not be any less than 5 mm to facilitate a stable suture.
This may be especially true in cases of large perforations and advanced atrophy in
the edge area as observed in patient 3.[22] Also, reperforation seems to occur especially in the portion of the graft lacking
cartilage. It is our understanding that it is the cartilaginous portion of the graft
providing stability for the fascia. The fascia itself serves as a framework for fibroblast
growth from the surrounding mucoperichondrium and as a durable attachment for the
graft.[34] In patient 9, the crushed cartilage had slipped off to the nasal floor with the
fascia still being attached to the cranial edge of the defect zone. Over time reperforation
occurred in the cartilage-free portion of the graft. A common factor of all three
cases of reperforation was the iatrogenic etiology of the initial nasoseptal defect.
This may explain the atrophic nature of the mucoperichondrium and its high vulnerability
to suture dehiscence. Also, all three defects were localized in the anterior septum,
an area prone to high turbulences in the nasal airflow a priori.
Conclusion
A DC-F graft following the “German ravioli” technique proved to be a reliable and
reproducible method for the closure of nasoseptal perforations of variable sizes of
different locations and of different etiologies. Patient age, previous surgery, and
comorbidity do not appear to influence the outcome of the procedure significantly.
A sufficient size of the DC-F graft with a minimum overlap of 5 mm allowing a tension-free
fixation within the defect seems to be crucial. Also, dicing of the cartilage seems
to be superior to crushing. Adding fibrin glue not only improves graft survival but
also optimizes adherence of the cartilaginous portion within the fascia sheets. However,
due to the small number of cases and relatively short postoperative observation period,
the results should be considered preliminary. Further studies should be performed
to evaluate long-term results on a larger patient collective.