In his letter to the editor, Dr. McMillan discusses several important issues about intrathecal granulomas. Among the points he raises, he emphasizes the need to determine the prevalence and incidence of intrathecal granulomas. Currently, several physicians who treat large numbers of “pump” patients are obtaining magnetic resonance image (MRI) scans on all patients receiving intrathecal opioids in their practices. Preliminary results from these nonselective screening protocols indicate that granulomas are uncommon (Timothy Deer, M.D., Charleston, West Virginia, personal communication, September 2003 via  e-mail), consistent with the low estimated risk of granuloma formation (< 2% over 6 yr) derived from existing clinical data.1The final results of these studies will help practitioners to determine who to screen, and how to screen how often to screen for granulomas.

Dr. McMillan refers to the association between “commercially prepared morphine sulfate&rdquo and the occurrence of intrathecal granulomas. Practitioners should be aware that the risk of granuloma formation may be even greater with off-label use of compounded opioid agents. The only commercially available opioid preparation approved by the U.S. Food and Drug Administration for intrathecal administration via  implanted pump is morphine in concentrations of 10 and 25 mg/ml. Many practitioners use morphine compounded in much higher concentrations (frequently 50 mg/ml). Clinical and laboratory data indicate that administration of morphine compounded in these high concentrations increases the risk of granuloma formation relative to the use of lower concentration preparations that are available commercially.

There is good rationale for Dr. McMillan’s recommendation that all patients receiving intrathecal opioid undergo magnetic resonance imaging of the spine to detect granulomas before they become symptomatic. Identification of asymptomatic lesions would enable physicians to treat these masses conservatively (e.g. , by cessation of opioid infusion) and might prevent the development of neurologic deficits. Radiographic screening of all patients may be desirable in some practices based on patient population, “aggressiveness” of intrathecal analgesic therapy (e.g. , use of high-dose, high-concentration opioids), and local medicolegal environment but may not be practical or cost effective on a wide-scale basis. In a 2000 survey of intrathecal analgesia practices,213,542 patients were reported as being under active management. Using this as an estimate of number of patients currently receiving intrathecal analgesics and using my own institution’s facility fee and radiologist reading fee of $3,389 for a spine MRI scan with gadolinium, $45,893,838 would be spent on the first set of MRI scans for screening these patients. We do not know how often screening studies need to be repeated to detect granulomas before they become symptomatic. Granulomas have been reported to occur within months of initiation of therapy; based on this knowledge, one can argue that MRI scans would need to be repeated at least every 6 months if granulomas are to be detected before they become symptomatic, resulting in a cost of greater than $90 million/yr. Not all patients can undergo MRI scanning. These individuals can be screened using computed tomography/ myelography. However, as an invasive procedure, and repeating it at regular intervals, the risks associated with this study might soon outweigh a patient’s risk of granuloma formation. Individual practitioners should decide whether the costs and risks associated with obtaining MRI scans or computed tomography scans/myelograms for all patients, with the expectation of identifying granulomas in a small percentage of patients, are warranted. A practical compromise between nonselective radiographic screening of all patients and radiographic evaluation only when patients become symptomatic might be to offer routine MRI scanning to patients at increased risk of granuloma formation, e.g. , those receiving relatively high doses or high concentrations (or both) of intrathecal opioid or those patients with a history of granuloma who elect to continue intrathecal opioid therapy after treatment of the previous granuloma.

A consensus panel convened in 2002 to discuss the clinical evaluation and management of intrathecal granulomas3specifically considered the role of “screening” MRI scans for all patients receiving intrathecal opioid. For a variety of reasons, and recognizing that patients with granulomas typically present with prodromal symptoms that should alert the managing physician to the presence of a granuloma before the onset of frank neurologic deficit (e.g. , loss of pain relief, rapidly escalating dose requirements, neurologic symptoms such as numbness), the consensus panel did not believe that existing data supported routine radiographic surveillance of all patients. The panel emphasized the need for vigilance, regular assessment (including neurologic evaluation), and a high index of suspicion for granulomas to detect them before the onset of neurologic deficit.

Intrathecal analgesic therapy, despite its seemingly benign nature, can have serious side effects. Physicians and patients accept these risks for the benefit derived from the therapy. Dr. McMillan appropriately expresses concern for the safety of patients receiving intrathecal opioid analgesics. Some physicians may choose, quite reasonably, to offer screening MRI scans to all patients receiving intrathecal opioid or to those patients who are at increased risk of granuloma development (e.g. , high daily opioid dose). Other physicians may elect, also quite reasonably, to monitor patients clinically, with the understanding that close attention must be paid to symptoms suggestive of granuloma formation, with expeditious radiographic evaluation of such symptoms should they arise. Regardless of their approach to monitoring patients for granuloma formation, physicians who treat patients receiving intrathecal opioids will do well to emulate Dr. McMillan’s level of awareness and concern about intrathecal granulomas and exercise due diligence in monitoring patients for the development of these lesions.

University of Iowa Hospitals and Clinics, Iowa City, Iowa.

Yaksh TL, Hassenbusch S, Burchiel K, Hildebrand KR, Page LM, Coffey RJ: Inflammatory masses associated with intrathecal drug infusion: A review of preclinical evidence and human data. Pain Med 2002; 3:300–12
Hassenbusch SJ, Portenoy RK: Current practices in intraspinal therapy: A survey of clinical trends and decision making. J Pain Symptom Manage 2000; 20:S4–11
Hassenbusch S, Burchiel K, Coffey RJ, Cousins MJ, Deer T, Hahn MB, Du Pen S, Follett KA, Krames E, Rogers JN, Sagher O, Staats PS, Wallace M, Willis KD: Management of intrathecal catheter-tip inflammatory masses: A consensus statement. Pain Med 2002; 3:312–23