Keywords
Tarlov cysts - perineural cysts - spine endoscopy
Introduction
Tarlov cysts or perineural cysts (PC) were described in an autopsy in 1938.[1] PCs are sacs filled with cerebrospinal fluid (CSF), formed within the posterior
nerve root between the endoneurium and perineurium and distally to the ganglion. They
are usually located in the sacral spine and have a variable communication with the
spinal subarachnoid space (SSS) that causes their filling with CSF. Inside the cyst
sac and/or its walls, nerve fibers and/or ganglion cells are located.[2]
[3]
[4]
Numerous theories have tried to explain their origin, and although it is believed
that they are acquired, no theory has fully demonstrated that. Most PCs are asymptomatic,
and they are usually considered radiologic findings. Nowadays, the diagnosis is easily
done with magnetic resonance imaging (MRI). The images show their typical morphological
characteristics, allowing their differentiation from other spinal cysts.[2]
[3]
[4]
Tarlov cysts have a prevalence of 1.5%, but only 13% of them are symptomatic.[5] This probably occurs when some of them fill and distend with CSF via the communication
with the SSS.
The aim of the present study is to describe the endoscope-assisted obliteration of
the communication between PCs and the SSS, and to analyze the outcome.
Methods and Materials
Population
Between 2007 and 2011, we treated six cases (four men and two women) with PCs. Ages
ranged between 32 and 52 years (median: 45 years).
Patients complained of lumbar pain (six cases), sciatic pain (two cases), perineal
pain (vaginal-perianal) (five cases), vesical tenesmus (one case), and/or rectal tenesmus
(one case). The neurologic examination was normal in all cases. The pain increased
in upright position and decreased in decubitus position.
Diagnosis
The diagnostic protocol included a neurologic, proctologic, urologic, and gynecologic
examination and a lumbosacral MRI. The cysts were hypointense in T1-weighted images
and hyperintense in T2-weighted images. Computed tomography (CT) scan was not used
routinely. All cysts were located in the midline sacral spine, between S1 and S3,
and were 4 to 7 cm in length and 2 to 3 cm in width.
Endoscope-Assisted Surgery
In these midline cysts, a 2- to 3-cm skin incision and a small sacral opening over
the caudal portion of the cyst were performed. This allowed the introduction of a
rigid endoscope with a 0 degree optic through the lower cyst pole ([Fig. 1]).
Fig. 1 (A) A 3-cm skin incision at the level of the caudal portion of the cyst. (B) Small
sacral bone window and visualization of the pole of the cyst. (C) Entry point and
direction of a rigid endoscope.
A controlled and complete aspiration of the CSF from inside the cyst was performed
to work with the endoscope in a dry field. Then, with a 30 to 45-degree inclination
and through the longitudinal axis of the cyst, the communication with the SSS was
searched ([Fig. 2]).
Fig. 2 Endoscopic view of the communication of the midline cyst with the spinal subarachnoid
space.
In some cases, the communication was difficult to see at a first glance because no
CSF entered the cyst, and it was only possible to see the entry of the roots through
the superior pole of the cyst ([Fig. 3]). In these cases a Valsalva maneuver was performed. Then the CSF entered easily
into the cyst, allowing the surgeon to see the communication. Once visualized, it
was plugged with muscle, fat tissue, Gelfoam, and fibrin glue, completing the obliteration
of the SSS. All cases had an external lumbar drain for 3 days postoperatively to assure
the plugging.
Fig. 3 Endoscopic view of the superior pole of the midline cyst with roots inside it and
without a clearly identifiable communication.
Evaluation
Postoperative clinical outcome was measured with the Odom scale in four categories:
excellent, good, fair, and poor.[6] All PCs were controlled with a postoperative MRI.
Results
After a median follow-up of 25.8 months (9–52 months) the postoperative outcome of
the six cases was excellent in four cases, good in one case, and poor in one case.
The postoperative MRI showed that the cysts did not disappear completely, but were
reduced in size. Their intensity in T1- and T2-weighted images changed ([Figs. 4] and [5]).
Fig. 4 Lumbosacral magnetic resonance imaging with a sacral cyst that shows (A) hypointensity
in a T1-weighted sagittal image, (B) hyperintensity in a T2-weighted sagittal image,
and (C) hypointensity in a T1-weighted axial image. Postoperative control shows cyst
size changes in (D) T1-weighted sagittal image, (E) T2-weighted sagittal image, and
(F) T1-weighted axial image.
Fig. 5 Lumbosacral magnetic resonance imaging (MRI) with (A) preoperative hyperintensity
in a T2-weighted sagittal image, (B) preoperative hyperintensity in a T2-weighted
axial image, (C) postoperative changes in a T2-weighted sagittal image, and (D) postoperative
changes in a T2-weighted axial image.
In the case with a poor outcome, the obliteration did not work, and after 30 days,
the cyst was filling again with CSF seen in the MRI. The patient refused another procedure.
Discussion
Spinal meningeal cysts are usually classified into three types according to pathologic
criteria.[7] PCs or Tarlov cysts are considered type II cysts because they have roots inside
of them.[8]
Because these midline cysts are bigger (≥ 4 cm), there is enough space to maneuver
with the endoscope and, with a proper inclination, to find the communication with
the SSS. In addition, because they were single, surgery can be performed only with
one small opening in skin and bone, being much less invasive. This technique is not
suitable for the more common bilateral PCs because it would be necessary to use more
than one entry point. In addition, because they are smaller, there is not enough space
to maneuver the endoscope in the cyst.
PCs usually contain only a few millimeters of CSF, but cysts containing up to 2.5 L
have been described.[9] The communication with the SSS is not always evident. All the cysts contained CSF,
and once emptied at surgery, their filling was not immediate. In some cases the communication
became visible only by increasing the CSF hydrostatic pressure with a Valsalva maneuver.
This showed that cyst filling by a valve mechanism could be a sound theory in symptomatic
cysts.[2]
[3]
[4]
PCs are usually located in the sacrum. Rarely they are seen in other regions of the
spine. They can erode the bony walls of the spinal canal as could be seen in some
of our cases. Huge cysts can erode the sacrum leading to a fracture.[2]
Sometimes they may grow through a sacral foramen and expand in front of the sacrum
and inside the pelvis, simulating adnexal masses in the echographic examination.[10]
As stated earlier, most PCs are asymptomatic; only 13% are symptomatic.[5] Tarlov himself was the first to consider this possibility that most PCs are asymptomatic.[11] Pain is the most frequent symptom, usually presenting as low back pain, sciatica,
and/or coccygodynia. Perineal pain located in the vagina and/or anus is more characteristic;
it may be associated with genital paresthesias.[2] In large presacral cysts, the pain can be abdominal.[9]
[10]
When the cyst erodes the sacrum, the pain may be caused by a fracture.[12] If the sacral fracture ruptures the cyst, fat tissue can penetrate the subarachnoid
space, causing cerebral fat embolism and mental confusion.[13]
Frequently, patients also have dysfunctional bladder and/or rectal symptoms: tenesmus,
frequency, and /or incontinence.[2]
[14] Sometimes, sexual dysfunctions such as dyspareunia or infertility can be present.[4]
[15]
The appearance of symptoms probably depends on the cyst size or its capability to
compress nearby structures. Our cases had a size of at least 4 cm. More than the size,
the increased hydrostatic pressure in the cyst by CSF filling seems the more likely
way how PCs compress adjacent structures and become symptomatic. This mechanism possibly
explains why in some patients the pain increases when standing and decreases when
lying.
Before assuming that PCs are responsible for the symptoms, the more common proctologic,
urologic, and gynecologic pathologies and lumbar degenerative disc disease must be
excluded by clinical and radiological examination.[16]
MRI is the gold standard for the diagnosis of PCs.[5] They are located in the sacrum along the course of a nerve root and are of variable
size, with an intensity similar to CSF. MRI allows to differentiate PCs from the other
types of spinal cysts.[7]
[8]
A thorough obliteration from the SSS is probably the best treatment for symptomatic
PCs.[14] By obliteration, PCs would not fill any longer with CSF, eliminating intracystal
hydrostatic pressure increase and distension as well as compression of neighboring
structures. However, the published therapeutic strategies are numerous, and each one
has advantages and disadvantages.
When PCs are ≤ 1.5 cm in size, treatment with nonsteroidal anti-inflammatory drugs,
steroids, amitriptyline, gabapentin, transcutaneous electrical stimulation, and/or
a rehabilitation program can alleviate symptoms.[17] Usually, surgery is performed in PCs ≥ 2 cm in size.
Total removal is possible, and the outcome reported has been favorable.[8]
[11] However, because PCs contain nerve roots within or in their walls, total removal
can cause neurologic deficits. That is why less invasive techniques have been developed:
cyst opening with partial removal and wall plication, with or without using microsurgical
techniques.[18]
[19]
[20] Several times we tried to use these techniques, but the fragility of the PC walls
precluded a safe closure. Therefore, we preferred cyst opening and plugging with fat
and fibrin glue.
Percutaneous CSF aspiration and percutaneous injection of fibrin glue were reported
as less invasive treatments.[4]
[21] Both procedures only have a transient positive effect because the PCs fill with
CSF again spontaneously and/or after fibrin glue absorption.
As an alternative minimally invasive technique, we used an endoscope-assisted obliteration
with the SSS in a midline PC. A similar approach has not been reported before. There
were no complications. The outcome after surgery was very favorable.
Recently, lumboperitoneal shunting was reported for symptomatic cases, with a favorable
response after a follow-up of 3 months.[22] However, in the long run, lumboperitoneal shunting leads to an acquired Chiari malformation
in 30% of the cases, as a result of the pressure gradient.[23]
[24] Lumboperitoneal shunting therefore seems to be an inferior surgical strategy to
treat PCs.
Conclusions
The presented endoscope-assisted obliteration of the communication between the cysts
and the SSS resulted in good to excellent outcomes and represents a simple and minimally
invasive technique.
