Case Scenario: Self-extraction of Intrathecal Pump Medication with a Concomitant Intrathecal Granulomatous Mass

A 38-yr-old female registered nurse presented to the pain medicine clinic for continued management of her chronic thoracic spine pain and possible malfunction of her intrathecal drug delivery system. The patient's past medical history was significant for depression, anterior cervical discectomy with fusion, and a SynchroMed EL Infusion Pump® (Medtronic Neurologic, Minneapolis, MN) placement for chronic pain related to T4 and T5 vertebral hemangiomas. The intrathecal pump was placed 9 months before her initial visit in our clinic. She was previously evaluated by multiple pain medicine specialists, with failure to attenuate her pain complaint. Upon initial evaluation she was receiving 40 mg/day of intrathecal morphine at a concentration of 50 mg/ml. At implantation she began therapy at 10 mg/day (20 mg/ml) but escalated to 40 mg/day. The high concentration, daily dose, and lack of analgesia prompted further evaluation of the system, which included cannulation of the catheter access port to evaluate patency of the intrathecal catheter. A lack of cerebral spinal fluid back-flow necessitated further evaluation, which included a catheter-access-port myelogram showing an intact catheter. Upon further inspection and injection of contrast medium, extreme back pain and the appearance at thoracic level 12 of a flame-shaped pooling of contrast at the tip of the catheter (fig. 1A) led to a diagnostic magnetic resonance imaging (MRI) scan, which confirmed the suspicion of an intrathecal catheter-tip mass (figs. 1B and C). After the diagnosis of the catheter-tip mass, further interrogation of the intrathecal pump revealed discrepancies between the aspirated residual volume and the calculated residual volume. The expected residual volume from telemetry was 7.6 ml, and the actual (aspirated) residual volume was 1.0 ml. In addition, evidence of multiple needle sticks at the reservoir fill port prompted further questioning, after which the patient admitted to accessing the pump and self-administering morphine intramuscularly. The patient was extracting 1 ml intrathecal morphine (50 mg/ml) with a standard bevel needle (nondeflected) and delivering (5 mg) diluents intramuscularly and replacing the withdrawn reservoir volume with saline. Subsequent to this admission, intrathecal medication delivery was discontinued, and the intrathecal pump was explanted, with surgical removal of the granulomatous mass without complications. The patient was administered transdermal clonidine for withdrawal prophylaxis and referred for substance abuse rehabilitation. Alternative treatment of her chronic pain included vertebroplasty at T4 and T5, transdermal fentanyl, hydrocodone/acetaminophen, and clonazepam, which provided satisfactory analgesia.

Important issues to consider in this case include the following:

Inflammatory masses can develop at the tip of an indwelling intrathecal catheter; they are typically associated with intrathecal delivery of opioids (e.g. , morphine) but also have been described with the delivery of nonopioid medications (e.g. , baclofen).11These mass lesions have often been described as granulomas, granulomatous masses, or granulomatous inflammatory masses. A granuloma is a focus of chronic inflammation consisting of a microscopic aggregation of macrophages that are transformed into epithelium-like cells, surrounded by a collar of mononuclear leukocytes, principally lymphocytes and occasionally plasma cells.12Granulomatous inflammation is a distinctive pattern of chronic inflammation that is formed to contain an offending agent.12In histopathological examinations of the inflammatory reaction in intrathecal masses, the inflammatory mass is devoid of the epithelium-like macrophages13constituting granulomatous inflammation; it is thus not a true “granuloma,” even though this term is commonly used to describe these inflammatory masses. The reaction contains signs of both acute and chronic inflammatory processes but lacks the essential elements that define granulomas.14 

The first intrathecal catheter-tip mass was reported by North et al.  in 1991,15but since then there have been numerous case studies reporting occurrences, management, and treatment outcomes of this phenomenon. Although the incidence of catheter–related granulomatous mass formation is uncertain, reports range from fewer than 1%16to 3%,17with incidence increasing with the length of therapy.18,19The incidence of self-extraction from intrathecal drug delivery systems is unknown, but these authors are aware of one other case published in the literature,20and one case of administrating illicit substances via  an intrathecal pump has been reported.21 

Table 1presents diagnostic characteristics of intrathecal catheter-tip masses. Most lesions develop slowly. Therefore, subtle but progressive neurologic decline or new onset motor weakness, including gait difficulties, sensory loss, proprioceptive loss, hyperactive or hypoactive lower extremity reflexes, and any evidence of bowel or bladder sphincter dysfunction, may be an indication of mass effect related to an intrathecal mass.22 

The imaging modality of choice is MRI if no contraindication is present.16,23Ring enhancement has been observed with the administration of contrast medium (gadolinium) but does not appear to be critical in diagnosis.17On MRI, contrast-enhanced T1-weighted imaging typically reveals enhancement of the lesion.22,23On unenhanced T1-weighted images, the masses vary in appearance and have intermediate to mildly low signal intensity.23Variable signal intensity characteristics on T2-weighted images have been reported but peripheral ring enhancement with central hypointensity is often observed16,24(figs. 1B and C). Concern arises with respect to the magnetic field imposed during the exam; the patient may experience a slight tugging sensation at the pump implant site, but the effect is less than that due to gravity. The magnetic field of the MRI scanner will temporarily stop the rotor of the SynchroMed II pump motor and suspend drug infusion for the duration of the MRI exposure. The pump should resume normal operation upon termination of MRI exposure. Presence of the pump can potentially cause an increase of the local temperature in tissues near the pump (approximately 1°C at 1.5 T). After imaging, the clinician programmer should be used to interrogate the pump to ensure that no changes in programming occurred.§An alternative modality is a computed tomography-myelogram using nonionic, water-soluble contrast to assist in differentiating between the granulomatous mass and the spinal cord.16,22,24At initial presentation of suspected catheter malfunction or occlusion, an intraspinal catheter-accessed myelogram (dye study) may be undertaken. The injectant (nonionic, water-soluble radiographic contrast medium) may produce a “flame-tipped lesion” (fig. 1A). Upon injection, mass effect may cause the patient to experience increased pain in a radicular pattern that corresponds to the level of the catheter tip.

Risk factors associated with the development of intrathecal masses have been examined in multiple studies16,25and recommendations made for prevention of intrathecal masses.26Data from animal studies with extrapolation to human subjects have borne out that medication concentration and not dose predispose to mass formation.13,18In addition, the location, either lumbar or thoracic, plays a role in the potential for neurologic compromise. The majority of intrathecal catheters are placed in the thoracic area, near the innervating spinal segment of abdominal pain origins. In the thoracic area, the ventral subarachnoid space may have a region of low cerebral spinal fluid flow.18,27Placement of a catheter by a Tuohy needle, in a conventional manner, may cause the catheter trajectory to turn rostrally only after the catheter deflects off the ventral dura mater at the level of insertion; it will then reside within the ventral thoracic subarachnoid space. Consequently, the local concentration of infused drugs in the subarachnoid space may be higher than anticipated.18A propagating mass in this area has the potential to impinge the spinal cord and cause neurologic deficits or compromise. A lumbar catheter, placed caudal to the conus medullaris, would reside in an area of greater cerebral spinal fluid flow, but placement in the lumbar area would be caudal to the spinal cord, which may confer a greater margin of safety if a mass propagates at the catheter tip. In a study of 41 patients with intrathecal masses Coffey and Burchiel27described chronic noncancer pain conditions as an additional predisposing factor for intrathecal mass formation, presumably because of the life expectancy difference between chronic cancer and noncancer patients. In this study the mean duration of therapy was 24.5 months before either symptoms or discovery of an intrathecal mass.27 

Inflammatory masses have been reported with intrathecal administration of all medications except sufentanil and rarely with fentanyl.26,28Through returned product analysis and in vitro  testing, Medtronic Neuromodulation has confirmed “nonindicated drugs” that result in intrathecal catheter occlusion include compounded medication of baclofen and morphine, admixtures of baclofen with clonidine, baclofen mixed with other drugs, and admixtures for chronic pain therapy containing morphine, baclofen, hydromorphone, clonidine, bupivacaine, fentanyl, and/or sufentanil.§The propagation of the inflammation with mast cell degranulation, histamine release, and other inflammatory mediators results in the mass formation. As the mass increases in size, medication outflow may be impeded, and a “mass effect” may impinge on neural structures. Although the patient in our report was not experiencing neurologic symptoms or deficits, the continued propagation of the lesion could have led to increased mass effect and subsequent neurologic compromise. In fact, occult granulomatous masses may be present without neurologic deficits.17As stated earlier, sudden increases in intrathecal dose requirements for pain attenuation are common in patients with intrathecal catheter-tip masses. In addition, new onset neurologic symptoms, particularly at the level of the catheter tip, changes in pain quality, and higher than predicted residual volume remaining in the reservoir at pump refills should raise suspicion for catheter-related complications. Although increasing dose requirements and new neurologic deficits are suspicious for inflammatory masses, these changes may be related to disease progression, new disease processes, catheter malfunction, and pump malfunction.17The symptoms usually develop gradually,16but acute presentations have been reported.27 

Microscopic pathology of intrathecal morphine-related granulomas often reveals necrotic tissue surrounded by macrophages, plasma cells, eosinophils, or lymphocytes.16In this patient the surgical pathology was consistent with fibrous tissue with old hemorrhage, necrotic tissue, and chronic inflammation. No neoplastic cells were seen, and all cultures were negative. Speculation regarding the etiology of the masses at various times has included the chemical characteristics, concentration, or immunologic effects of the infused medications; impurities in, or contamination of, medication sources; bacterial infection, pyrogens, or endotoxins; silicone allergy; catheter design features; and/or delayed effects of traumatic catheter implantation.18The risk of infection during medication refills has been examined; under sterile conditions, refills of intrathecal implanted pumps do not seem to be a risk.29In a study utilizing MRI in dogs, substitution of saline for morphine subsequent to the development of a granuloma led to a regression of the granuloma and the conclusion that mass formation was dependent on the local concentration, not the dose, of morphine.14Alternatively, studies in large-animal models have shown that these masses reflect either a dose- or concentration-dependent effect, and the effects of dose versus  concentration cannot be clearly distinguished.18,30 

New onset neurologic deficits at the time of mass discovery necessitate a neurologic exam and may require neurosurgical consultation. Prompt surgical intervention to excise the mass and/or the catheter may result in rapid clinical improvement and restoration of neurologic function or prevent further neurologic deterioration.22,27If signs of spinal cord atrophy or necrosis are observed, residual deficits are likely.16The masses are not neoplastic, and several cases have reported that the postoperative residual mass gradually contracted or disappeared over time.18,22,27If no acute neurologic deficits are present at mass discovery, conservative measures may be employed. Case records and anecdotal reports suggest that patients with a mass that does not fill the spinal canal or cause neurologic impairment sometimes can be treated by discontinuing/emptying the drug infusion pump or by refilling it with preservative-free normal saline to infuse at a minimal rate.27Contraction, reduction, or disappearance of masses has been described on follow-up imaging studies after an interval of 2–5 months.22,27,31Substantial decrement of the intrathecal mass seen on imaging studies is necessary before resumption of intrathecal treatment should be considered. In addition, treatment options include removal of the catheter and placement of a new catheter under monitored anesthesia care at the time of removal or at a later date. Another viable option is disconnecting the catheter from the pump, obliterating the lumen of the catheter, leaving the intraspinal segment undisturbed, and placing a new intraspinal catheter segment at the same operation or at a later date.17,22In all management options, the infused medication is discontinued. Catheter revision with medication rotation allows intrathecal treatment to continue with minimal interruption in medication delivery. Many physicians may opt for this approach because there is cessation of the offending agent and the mass is removed. The medication rotation should include medications at doses recognized as less likely to precipitate intrathecal masses. The placement of the catheter can be caudal to the conus medullaris, which may decrease the “mass effect” of an intrathecal mass if formation occurs. The catheters should not be placed preferentially such that they reside in the ventral intrathecal space, where the greatest area of low-cerebral spinal fluid flow occurs.18,26Theoretically, a dorsally placed catheter may avoid this low-flow segment, but it has been reported that dorsal or ventral placement of catheters does not appear to influence the formation of inflammatory masses.22 

Multiple features are involved in the development of intrathecal catheter-tip masses. The medication delivered to the intrathecal space, medication concentration, and dose appear to be the most important factors associated with mass formation. Morphine is the most commonly implicated medication in the development of intrathecal granulomatous masses, especially at higher concentrations.16,18As treatment duration increases, many patients develop tolerance, requiring dose escalations to maintain analgesic efficacy. The concentration of the medication is increased to maintain refill intervals. The maximum concentration and dose recommended by the Polyanalgesic Consensus Panelists are represented in table 2. In many reported cases of catheter-tip masses, the concentration of morphine has been at or above 40 mg/ml. If the concentration requirements reach this level, alternative strategies should be utilized. In surveys of implanting physicians, up to 35% of patients using morphine alone fail the therapy.17Clonidine has been described as having protective properties with respect to inflammatory mass formation,18,22but it has also been reported to fail in protection against inflammatory mass events.31Nevertheless, if morphine is the agent of choice, clonidine may present synergistic analgesic effects, thus allowing a lower dose of morphine through the course of treatment. Alternatively, hydromorphone can be delivered to the intrathecal space and may have advantages over morphine because it is a significantly more potent analgesic than morphine. It has been reported that equivalent pain control can be achieved with much lower intrathecal doses of hydromorphone than morphine (approximately 20% of morphine doses), with a lower potential for undesirable side effects.32,33Medication rotation and a polyanalgesic approach of adding a second or third medication before the concentration or daily dose of any one medication reaches the upper limit of recommendations may prove prudent in reducing the incidence of intrathecal mass propagation.

The potential for significant neurologic compromise that may occur with an intrathecal catheter-tip inflammatory mass and the adverse consequences of self-extraction from an intrathecal pump necessitate aggressive assessment and management. The goals of intrathecal therapy are analgesia, functional restoration, and minimizing morbidity from the therapy. Achieving these goals begins with patient evaluations for potential psychosocial issues that may preclude implantation. When initiating intrathecal therapy or assuming care of an intrathecal pump, an initial neurologic assessment is paramount. A screening imaging study is not essential, but when assuming care for a previously implanted pump this should be considered. Medication concentration and dose recommendations serve as a reference for titration of intrathecal medications. In addition, neurologic assessment should be performed at each refill interval, along with assessment of treatment goals and adverse effects of the medications. Furthermore, assessment of expected and actual residual volumes should be assessed at every refill visit. The flow rate should be within ± 14.5% of the programmed rate; a flow rate error of ± 25% is a marker of significant discrepancy between expected and actual residual volumes.§

Patients with chronic nonmalignant pain may benefit from psychologic evaluation before intrathecal therapy. Although not an absolute contraindication to intrathecal therapy, risk factors such as untreated substance abuse and unrealistic expectations of intrathecal therapy may preclude intrathecal trial/implantation. If self-extraction is suspected or confirmed, intrathecal therapy should be discontinued and alternative treatment along with addiction therapy should be initiated. Optimal outcome requires well-coordinated multidisciplinary care. Table 3summarizes common diagnostic and treatment issues and provides recommendations to avoid their potentially serious sequelae.

Intrathecal granulomatous masses and self-extraction from an intrathecal pump have both been reported in the literature, but not concomitantly. It is unclear whether the high morphine dose and concentration were solely causative in the propagation of the mass or whether repeated entrance to the pump by the patient under nonsterile conditions introduced nonsterile, potentially inflammatory material. Although the inciting event that precipitated the granulomatous mass formation is unclear, the cause is most likely multifactorial. Several studies demonstrate the hazards of high-dose intrathecal opioids, but additional studies and recommendations are needed to clarify predisposing factors and device safety measures to prevent intrathecal medication substance abuse. The use of intrathecal opioids should, theoretically, confer less abuse potential than enteral or parenteral opioids. Through this case scenario, noteworthy findings are that the potential for substance abuse remains in patients with intrathecal pumps and that although consensus recommendations are readily available, based on the best available evidence, responsible medication prescribing is still lacking. Continued investigations will be necessary to determine the best practice (i.e. , polyanalgesia, nonopioid medications) for the management of intrathecal medication delivery in chronic nonmalignant pain patients. Finally, specific attention should be paid to patients who exhibit characteristics of substance abuse before intrathecal trial/implantation.