Blackwell Publishing IncMalden, USAPMEPain Medicine1526-2375American Academy of Pain Medicine200682199203Case

ReportUse of an Intrathecal Pump in Tuberous SclerosisByrd et al.

PA I N M E D I C I N E
Volume 8 • Number 2 • 2007

CASE REPORT

Chronic Pain and Obstetric Management of a Patient with

Tuberous Sclerosis

Downloaded from https://academic.oup.com/painmedicine/article/8/2/199/1848574 by Universidad de Antioquia user on 17 October 2020

Louise M. Byrd, MRCOG,* B. Jadoon, MB ChB,† I. Lieberman, MBBS, FRCS,‡ and
Tracey Johnston, MD, MRCOG§
*Specialist Registrar Obstetrics and Gynecology; †SHO Obstetrics and Gynecology; §Consultant Fetal/Maternal Medicine,
St Mary’s Hospital; ‡Consultant Anesthetist, Consultant in Pain Medicine and Anesthesia, South Manchester University
Hospital, Manchester, UK

ABSTRACT

ABSTRACT Chronic nonmalignant pain is very disabling and carries a heavy financial strain on the individual
and society as a whole. This case describes a woman with tuberous sclerosis, in her fourth pregnancy.
Approximately 18 months prior to pregnancy, intractable left loin pain, thought to be secondary to
hemorrhage within a tuberous lesion in the left kidney, had led to the siteing of an intrathecal
morphine pump. The risks of system failure (dislodgement, dislocation), escalating dosage, infec-
tion, use in labor, and neonatal opioid withdrawal are all explored and discussed. While data are
limited, with increasing use of intrathecal opioids for nonmalignant pain, such patients may be seen
more regularly in obstetric clinics. With a multidisciplinary team approach, risks can be minimized
and outcome for mother and baby optimized.

Key Words. Pregnancy; Intrathecal Pump; Morphine; Chronic Pain

hronic nonmalignant pain is very disabling and kidneys. She had intermittent hematuria and
C and carries a heavy financial strain on the
individual and society as a whole. Intrathecal opi-
chronic iron-deficiency anemia, and her pain was
thought to be secondary to hemorrhage within
oid delivery for pain management in a patient with tuberous lesions in the left kidney. Prior to the
tuberous sclerosis has, to the authors’ knowledge, referral, she had been tried on maximal doses of
never before been described. Case reports of the gabapentin, venflaxine, carbamazepine, amitrip-
obstetric management of patients with indwelling tyline, and nortryptyline oral and intravenous
intrathecal opioid pumps are similarly rare and, to morphine, to little avail. It had been noted that
our knowledge, have only been described once while the pain seemed to be opioid sensitive, the
before [1]. sedative, cognitive, and constipatory side effects of
the morphine at the doses she required to have any
effect (400–600 mg per day) were significant and
Case Report
dose limiting. This pain was so severe that she was
A 33-year-old woman was referred to a consultant considered for left nephrectomy, and in an attempt
in pain medicine in May 2001 because of intracta- to avoid this, a trial of pain control using epidural
ble pain in the left loin. At the time of referral, she blockade (diamorphine at a dose of 20 mg per day)
had been an inpatient for 2 months. She had was undertaken. Diamorphine (C17H17NO
tuberous sclerosis with lesions in the brain, liver, (C2H3O2)2) is available on a physician-only-
prescription basis in the UK. It is the diacetyl ester
Reprint requests to: Louise M. Byrd, MRCOG, 9, Orchard of morphine (i.e., morphine has two of its hydro-
Green, Alderley Edge, Cheshire SK9 7DT, UK. Tel/Fax: gen atoms replaced with CH3CO). It is approxi-
01625-586-515; E-mail: tahghighibyrd@aol.com. mately twice as potent as morphine (equianalgesic

© American Academy of Pain Medicine 1526-2375/07/$15.00/199 199–203 doi:10.1111/j.1526-4637.2006.00140.x

200
ratio of 1:2), and is considerably more lipid soluble
than morphine so crosses the blood brain barrier
faster, giving a more rapid onset after systemic
administration. However, this has less relevance
when it is administered intrathecally as there is no
blood brain barrier to cross. It was chosen here as
it is available in a preservative-free powder that is
readily soluble in saline, allowing high concentra-
tions of the drug to be delivered in low volumes
of solution. The initial trial lasted 10 days via
a tunneled epidural catheter (Portex, Portex Ltd.,
Canterbury, UK) and was repeated for a further
2 weeks by a second catheter when the first fell
out.
Following successful abolition of her pain by
this method, the possible insertion of a permanent
intrathecal catheter, to dispense morphine, via a
programmable pump, was discussed and agreed. A
SynchroMed (Medtronic, Hertfordshire, UK)
pump was implanted under general anesthesia on
July 18, 2001. The intrathecal space was accessed
at L3/L4 level and the catheter screened to the
level of the body of T8. The pump was implanted
between the superficial and deep facia in the left
iliac fossae. The pump was charged with 10 mg
per mL of preservative-free morphine. The initial
rate of infusion was 2 mg per day. Her postopera-
tive recovery was complicated by a persistent post-
dural puncture headache that was managed with
two epidural blood patches, fluid, and bed rest.
She was discharged after 3 weeks postimplant. At
this time, the pump was set to dispense 5.5 mg per
day, and she described herself as 95% pain free. In
order to maximize the length of time between
refills, the reservoir had been refilled with 40 mg
per mL of preservative-free morphine.
Between the summer of 2001 and March 2003,
the patient attended for regular review and refill
of her pump. The pump itself remained functional
and trouble free. During this time, the dose of
morphine dispensed varied, cycling between a
minimum of 5.5 mg per day and a maximum of
14 mg per day, depending on the “flares” of her
left loin pain.
In March 2003, the patient was referred by her
pain consultant to St. Mary’s hospital for tertiary
obstetric care. She was then 18 weeks pregnant in
her fourth pregnancy. Her three previous preg-
nancies had resulted in live births of children, all
of whom had tuberous sclerosis. Genetic counsel-
ling had been offered and declined. When seen at
the antenatal clinic, the patient was noted to have
chronic iron-deficiency anemia. Renal and liver
function tests were normal. A fetal anomaly scan
Byrd et al.
showed no gross abnormality; Down’s syndrome
screening was declined.
At 21 weeks’ gestation, the patient presented to
her local Accident and Emergency department
with acute lower back pain. The concern was that,
her pump had disconnected and she was suffering
from a combination of pain and opioid with-
Downloaded from https://academic.oup.com/painmedicine/article/8/2/199/1848574 by Universidad de Antioquia user on 17 October 2020
drawal. A plain X-ray indicated the line to be
intact. It is recognized that dye studies are a better
functional assessment of the intrathecal (IT) deliv-
ery system, and these studies were considered.
However, following transfer to the obstetric unit
for further management, assessment indicated that
the pain in her lower back radiated down the left
leg and was associated with both paraesthesia and
urinary retention. In view of this, a morphine
bolus of 20 mg was administered, and urgent mag-
netic radiation imaging (safely performed with the
pump switched off) was arranged. This confirmed
acute lumbar disc prolapse at the level of L5/S1 as
the source of pain. The patient was therefore
transferred to the nearest neurosurgical unit
where she underwent total laminectomy, per-
formed under general anesthesia. The fetal heart
was assessed using sonicaid, both pre- and post-
operatively, and found to be satisfactory
(155 bpm). Postoperatively her pain was con-
trolled by a combination of oral and intravenous
opioids in addition to the intrathecal pump, non-
steroidal anti-inflammatory drugs (NSAIDs), and
paracetamol. At the time of discharge from hospi-
tal, she was on a delayed release oral morphine
preparation (morphone sulphate 100 mg per day)
in addition to the intrathecal morphine pump
which was dispensing 14 mg per day. Oral opioids
were gradually withdrawn, achieved without
recourse to any further increase in the IT dose.
While rehabilitating, the patient developed signs
suggestive of a deep vein thrombosis, managed
with therapeutic levels of low molecular weight
heparin (LMWH). Following a venous Doppler,
which demonstrated patent veins, this was reduced
to prophylactic doses of LMWH, maintained
throughout the remainder of the pregnancy
because of a persistent reduction in mobility.
At 32 weeks’ gestation, general pruitis heralded
the onset of obstetric cholestasis, and she was
monitored accordingly. At 36 weeks’ gestation,
labor was induced (using vaginal prostaglandins)
because of a nonreassuring nonstress test.
The intrathecal morphine infusion was main-
tained during labor and Entonox (gaseous mixture
of 50% O2 and 50% N2O), provided additional
analgesia. With this combination, the patient
Use of an Intrathecal Pump in Tuberous Sclerosis
managed until late second stage when she com-
plained of severe left-sided pain around the site of
the pump. While alternative pain relief was sought
(morphine bolus), a live male infant was delivered,
weighing 2,941 g. During the following 48 h, the
baby showed some mild jitteriness in association
with morphine withdrawal, but fed well. This
settled without the need for medication; an elec-
troencephalogram was unremarkable.
The patient’s left-sided pain persisted following
delivery. At the time of presentation, the possibil-
ity of uterine scar rupture was briefly considered
and excluded, as the site of pain was more laterally
placed, there was no direct tenderness over the
scar, and she remained hemodynamically stable
throughout. The most likely cause of pain was
bleeding into a renal angiomyolipoma/cyst. Fol-
lowing delivery, an ultrasound scan of the kidney
failed, however, to show any additional hemor-
rhage (see Discussion section). The plain X-ray
confirmed an intact pump line.
With rest and regular analgesia, the pain grad-
ually settled, and she was discharged home on day
10. She remained under review in the pain clinic.
Discussion
Tuberous sclerosis is an autosomal dominant dis-
order, first described by Bournville in 1880. The
term “sclerose tubereuse” was coined to indicate
the superficial resemblance of the cerebral lesions
to a potato [2]. The disorder arises because of a
mutation that occurs within two genes TSC1 and
TSC2, on chromosomes 9 and 16, respectively [3].
This results in the growth of benign tumors within
various organs of the body. Diagnosis is based on
clinical, radiological, and histopathological assess-
ment. While the classic triad is of convulsions
(90%), skin lesions, and mental retardation (60%)
(our patient had none of these), the specific man-
ifestations, disease severity, clinical course, and
prognosis are extremely variable [4], as this case
demonstrates.
Renal lesions, either cysts or angiomyolipomas,
occur in between 40% and 80% of patients with
tuberous sclerosis [5]. Cysts are usually small, lim-
ited in number, and cause no problems. However,
in 2% of cases, the cysts may be large. Angiomy-
olipomas are composed of varying amounts of adi-
pose tissue, smooth muscle, and thick walled blood
vessels. They tend to be multiple, bilateral, and
extend into the perinephric space. The major
complication is local hemorrhage, which can be
severe, particularly in lesions greater than 4 cm
201
[6]. Bleeding can occur acutely and may be fatal.
Alternatively, bleeding can occur chronically over
many years, presenting as anemia. The increase in
renal blood flow seen within pregnancy [7,8] may
increase the risk of hemorrhage. Hormonal influ-
ences also stimulate tumor growth during preg-
nancy [9]. Prepregnancy assessment would be
Downloaded from https://academic.oup.com/painmedicine/article/8/2/199/1848574 by Universidad de Antioquia user on 17 October 2020
pertinent, with large lesions being considered for
prophylactic embolization or renal sparing surgery
[10].
The acute pain experienced by this patient dur-
ing the second stage of labor was most likely
secondary to hemorrhage within a renal angiomy-
olipoma. Ultrasound, while safe, is only capable of
demonstrating small lesions, as the mix of fatty
and vascular tissue is surrounded by normal paren-
chyma. This acts as a window for penetration and
provides some degree of contrast. However, larger
lesions are predominantly fatty––this “reflects”
sound, limiting penetration, and hence visualiza-
tion. Computed tomography is capable of differ-
entiating between fat, soft tissue, and blood
vessels, and is therefore more suitable for the
assessment of lesion size and any associated com-
plication, including hemorrhage. It is therefore
considered the diagnostic imaging modality of
choice [11].
The use of intrathecal morphine has, for a long
time, been used in the management of malignant
pain [12]. More recently it has gained popularity
for use in managing nonmalignant pain [13]. The
system comprises two components––the infusion
pump and an intrathecal catheter. The pump is
programmable and battery operated; it stores and
dispenses drugs according to the instructions
received from an external programmer (see
Figure 1). It is placed abdominally in a subcutane-
ous pocket under the ribs and consists of a main
reservoir (where the medication is stored), a res-
ervoir fill port (used for filling and emptying the
pump), and a catheter access port (which allows
the pump to be bypassed, sending medication
direct to the catheter). The catheter, a soft silicone
tube, is inserted into the intrathecal space of the
spine, tunneled under the skin, and then con-
nected to the pump (see Figure 2).
The administration of intrathecal morphine
is an effective means of providing pain relief,
allowing lower dosages than the equivalent oral
regimes and hence fewer side effects. Tolerance
may develop and opioid dose requirements may
increase. Despite concerns, there is little evidence
of opiate withdrawal in the infant [14] when such
systems are used in pregnancy.
202 Byrd et al.

While intrathecal opioids can provide complete

analgesia in early labor, somatic sacral blockade
(either a neuraxial sacral block or pudendal nerve
block) with local anesthetics is required for the late
first and second stages of labor. It therefore follows
that, even if the dose of intrathecal opioid is
increased during labor, there is no reason to

Downloaded from https://academic.oup.com/painmedicine/article/8/2/199/1848574 by Universidad de Antioquia user on 17 October 2020

believe that this alone will provide satisfactory
analgesia for most women. Local anesthetics, opi-
oids, other adjuvants, and combinations of these
drugs at different doses could be injected into the
epidural space, although “scarring” within the
space may result in a patchy block. While this may
be overcome by use of a larger bolus [20], this
particular patient, in addition to the placement of
an intrathecal catheter, had also undergone a
recent laminectomy at L5/S1. This might have
interfered with the placement of an epidural cath-
eter and/or further increased the likelihood of
ineffective epidural analgesia.

Figure 1 The SynchroMed (Medtronic) programmable

pump system.

Infection is a concern with any implantable

device, with a risk of amount 0.77 per 1,000 cath-
eters in work by Byers et al. [15], although the
main risk seems to be at around the time of inser-
tion [15]. It has not been established whether labor
necessitates the administration of prophylactic
antibiotics, although we would consider that the
potential benefits outweigh any risks, and recom-
mend administration a sensible precaution.
Pump system failure may occur, with the cath-
eter being the most vulnerable component [16].
Whether this is more likely in pregnancy is again
not known. However, it is not inconceivable that
an expanding abdomen, secondary to the enlarg-
ing, dextro-rotated uterus, could predispose to
tension and hence dislocation of the pump hous-
ing, dislodgement, catheter kinks, and/or perfora-
tions. Such “stretching” may be reduced by left
lateral placement. Orientation of the pump–cath-
eter neck superior-medially rather than laterally
or inferiorly, and ensuring significant distance
between the pump–catheter neck and the iliac
crest may also reduce possible friction-induced
catheter breakdown [17]. A plain X-ray is usually
enough to demonstrate a dislocation, dislodge-
ment, or a kink [18]. Radionulclide flow studies Figure 2 This demonstrates the positioning of both the
external pump/programmer, placed abdominally in a sub-
may be necessary to evaluate patency of the cath- cutaneous pocket, and the silicone catheter, sited within the
eter [19], although this should be avoided in intrathecal space, and then tunneled under the skin and
pregnancy. connected to the pump.
Use of an Intrathecal Pump in Tuberous Sclerosis
It would be intrinsically appealing to re-prime
the indwelling pump [1] with a combination of
opioid and local anesthetic. Nevertheless, this is
not without its hazards. The pump reservoir and
tubing contain highly concentrated opioids that
can be inadvertently injected into the subarach-
noid space. Prior planning and precise documen-
tation may assist those who are unaccustomed
with the system. And this should have been set in
place here, following liaison with pain specialists
and obstetric anesthetists. However, it is still pos-
sible that unfamiliarity and differences in systems
may mean that alternative analgesia will be sought
[1].
Chronic nonmalignant pain is very disabling
and carries a heavy financial strain on the individ-
ual and society as a whole. It affects millions of
people worldwide and is a major healthcare
expense [21]. With increasing use of intrathecal
opioids for nonmalignant pain, such patients may
be seen more regularly within the pregnant pop-
ulation, but to date, use is limited. While there is
a theoretical risk of complications, as previously
alluded to, a multidisciplinary team approach
minimizes these risks, so optimizing outcome
for mother and baby.
References
1 Tarshis J, Zuckerman JE, Katz N, Segal S, Mushlin
PS. Labour pain management in a parturient with
an implanted intrathecal pump. Can J Anaesth
1997;44(12):1278–81.
2 Bourneville DM. Sclerose tubereuse des circonvo-
lutions cerebrales: Contribution a e’iodiote: Idiotic
et epilepsie hemiphlegique. Arch Neurol (Paris)
1880;1:81.
3 Sancak O, Nellist M, Goedbloed M, et al. Muta-
tional analysis of the TSC1 and TSC2 genes in a
diagnostic setting: Genotype–phenotype correla-
tions and comparison of diagnostic DNA techniques
in tuberous sclerosis complex. Eur J Hum Genet
2005 June;13(6):731–41.
4 Petrikovsky BM, Vintzileos AM, Cassidy SB, Egan
JFX. Tuberous sclerosis in pregnancy. Am J Perin-
atol 1990;7:133–5.
5 Stillwell TJ, Gomez MR, Kelalis PP. Renal lesions
in tuberous sclerosis. J Urol 1987;138:477–81.
6 Van Baal JG, Smits NJ, Keeman JN, et al. The
evolution of renal angiomyolipomas in patients with
tuberous sclerosis. J Urol 1994;152:35–8.
203
7 Yanai H, Sarasagawa I, Kubota Y, et al. Spontaneous
haemorrhage during pregnancy secondary to renal
angiolipoma. Urol Int 1996;56:188–91.
8 Klein M, Matal M, Wagner C, Beck A. Obstetrical
management in maternal tuberous sclerosis (Prin-
gle–Bourneville syndrome)—A case report. Z
Geburtshilfe Perinatol 1993;197:238–9.
Downloaded from https://academic.oup.com/painmedicine/article/8/2/199/1848574 by Universidad de Antioquia user on 17 October 2020
9 Henske EP, Ao X, Short MP, et al. Frequent
progesterone receptor immunoreactivity in tuber-
ous sclerosis-associated renal angiomyolipomas.
Mod Pathol 1998;11:665–8.
10 Steiner MS, Goldman SM, Fishman EK, Marshall
FF. The natural history of renal angiomyolipoma. J
Urol 1993;150:1782–6.
11 Summers TA, Fultz PJ, Sacks D. Worsening right
flank pain over a 24-hour period. Acad Radiol Octo-
ber 1994;1(2):194–8.
12 Oberlander TF, Robeson P, Ward V, et al. Prenatal
and breast milk morphine exposure following
maternal intrathecal morphine exposure. J Hum
Lact 2000;16:137–42.
13 Chambers FA, MacSullivan R. Intrathecal morphine
in the treatment of chronic intractable pain. Ir J
Med Sci 1994;163:318–21.
14 Kumar K, Kelly M, Pirlot T. Continuous intrathe-
cal morphine treatment for chronic pain of non-
malignant etiology: Long term benefits and efficacy.
Surg Neurol 2001;55:79–88.
15 Byers K, Axelrod P, Michael S, Rosen S. Infections
complicating tunnelled intraspinal catheter systems
used to treat chronic pain. Clin Infect Dis
1995;21:403–8.
16 Penn RD, York NM, Paice JA. Catheter systems for
intrathecal drug delivery. J Neurosurg 1995;83:215–
7.
17 Dickerman RD, Schneider SJ. Recurrent intrathecal
baclofen pump catheter leakage: A surgical obser-
vation with recommendations. J Pediatr Surg
2002;37:E17–9.
18 Pasquier Y, Cahana A, Schnider A. Subdural cathe-
ter migration may lead to baclofen pump dysfunc-
tion. Spinal Cord 2003;41:700–2.
19 Rosenson AS, Ali A, Fordham EW, Penn RD.
Indium 111 DTPA flow study to evaluate surgically
drug pump delivery system. Clin Nucl Med
1990;15:154–6.
20 Fontane P, Adam P, Svendsen KH. Should intrath-
ecal narcotics be used as a sole labor analgesic? A
prospective comparison of spinal opioids and epidu-
ral bupivacaine. J Fam Pract 2002;51:630–5.
21 Daley MD, Rolbin SH, Hew EM, Morningstar BA,
Stewart JA. Epidural anaesthesia for obstetrics after
spinal surgery. Reg Anesth 1990;15:280–4.