A 15-Year-Old Boy With Primitive Neurectodermal Tumor

ABSTRACT: Case Presentation Dr. Peter Staats presented the case of a 15-year-old, 40-kg boy with a primitive neurectodermal tumor located in his right hemipelvis, invading the lumbar plexus. No invasion in the spinal cord was noted. The boy’s pain score was 8 out of 10 on the visual-analog scale (VAS). He had a history of anxiety and fear of procedures, and he was sedated on a regimen of Neurontin (gabapentin), nortriptyline, and oxycodone at relatively low doses (5 mg q4h). He had previously received maximal radiation therapy and high doses of systemic steroids. The patient’s pain was out of control, and sedation was a major problem. Several therapeutic options included (1) working with medications, eg, decreasing nortriptyline, switching or rotating opioids, adding psychostimulants, and adding anxiolytic agents; (2) placing an epidural catheter for infusions of a local anesthetic; (3) performing a cordotomy; and (4) considering further chemotherapy and palliative therapy. Dr. Staats and his pain management team sought to gain control as quickly as possible because of their knowledge that over time this patient’s tumor would be very difficult to manage. An intrathecal catheter was placed under general anesthesia. The patient received a 1-mg bolus followed by a 1-mg infusion of morphine. Complete relief of pain was achieved and persisted the next day. A pediatric pump was implanted, and the catheter was placed at T10 where the substantia gelatinosa was processing much of his pain. All systemic opioids and anticonvulsants were stopped; tricyclic antidepressants were continued to help the boy sleep. The patient’s course after starting on oral opioids is depicted in Figure 1. Intrathecal opioids initially relieved the patient’s pain. The boy chose not to increase his intrathecal opioids until his pain reached about 3 on the VAS scale. At that point, his dose of intrathecal morphine was doubled to 2 mg/d, and his pain decreased. A burning pain in the patient’s right leg and hip developed at which time intrathecal morphine was increased to 3 mg/d and bupivacaine at 2.5 mg/d was added; his pain decreased. The boy’s pain again intensified and morphine was increased to 3 mg/d and bupivacaine to 3.5 mg/d. The patient was pain-free for approximately 1 month before he developed facial V3 pain, for which the dose of intrathecal morphine was increased to 4 mg/d and bupivacaine was increased to 5 mg/d. The oncology team started the patient on intravenous Dilaudid (hydromorphone), but he became extremely nauseated. Intrathecal morphine and bupivacaine doses were increased, and the patient’s pain decreased. He was pain-free at death 9 months later.

Discussion

Peter S. Staats, MD: This was a case where we had poor control to begin with and many options. The pain management team elected to implant an intrathecal pump in this child for several reasons, including the patient’s lumbar plexus problem. We believed he would develop significant problems and that we ought to implant the pump early. I felt that we could better manage this patient long term with bupivacaine and morphine delivered by this neuraxial route.

Russell K. Portenoy, MD: How was the boy doing psychologically, and what were some of the issues concerning his parents?

Dr. Staats: The boy’s parents were very supportive of him. The child wanted nothing more than to go back to school. He did have a lot pain, and he was anxious about procedures and needles. I had seen the boy previously because his pain was out of control, and when I explained his options, he said he really didn’t hurt that bad. The patient waited 1 to 2 months before returning to see us, at which time we discussed his concerns about the implanted devices.

Dr. Portenoy: What was the patient’s level of function when systemic opioid therapy with oxycodone was failing because of somnolence?

Dr. Staats: His function was very poor at the time because of sedation, which was really his problem. Once we placed an externalized intrathecal catheter, the sedation improved rapidly and he returned to school. One might question why we chose not to switch opioids since the boy was on a relatively low dose of opioids. Although we thought about this option, we wanted an alternative route of therapy for the local anesthetics that we knew would eventually be needed because the tumor was invading the lumbar plexus.

Decision-Making

Dr. Portenoy: The major decision point in managing this patient appears to be between providing more aggressive pharmacotherapy and undergoing a trial of intraspinal therapy. From the point of view of more aggressive pharmacotherapy, I probably would have tried an opioid switch to methadone first, and then certainly would have added a psychostimulant before opting for intraspinal therapy. This difference in selection of therapy raises important issues about how much that decision-making in pain management is based on who is available to manage a patient’s therapy.

Dr. Staats: That is a very good point, and something that I’ve struggled with over time. How aggressively does one work to manage a patient medically to get the same level of function? I am not convinced that this was the absolute way to go. There are two viable approaches in this case: 1) opioid switch and 2) intraspinal therapy. I believe that you would have worked a lot harder and that the patient would have been taking a lot more pills by opting for the opioid switch than by intervening very early with intraspinal therapy as we did. But, there are no randomized trials to confirm this, which is one of the reasons we’re having this discussion.

Michael H. Levy, MD, PhD: At Fox Chase Cancer Center we don’t see patients younger than 18 years, so for the sake of discussion, I would like to assume that the patient was 20 years old. We probably would have opted to try different opioids, as Dr. Portenoy discussed. Other than first trying different opioids, our management would not have differed much from that of Dr. Staats in this patient. Even in our hypothetical 20-year-old patient, implanting a pump and administering medications intrathecally to manage most of the pain so that he was not taking multiple drugs (eg, laxatives, coanalgesics, stimulants, opioids) could be in the patient’s best interest and could improve his quality of life.

Anticipating the Course Of Pain

Stuart Du Pen, MD: With severe neuropathic pain in a young patient, I support neuraxial analgesia through a temporary device. Once pain is under control, then other options may be evaluated.

Dr. Staats: This case is very unusual, which is one reason I chose to present it for discussion. My belief was that some of you might disagree with implanting a device in a patient who was receiving only 5 mg of oxycodone q4h. However, my thought process in selecting intraspinal therapy included anticipation of what was foreseeable over the life of this patient.

In a much more typical case at Johns Hopkins, we would have tried two or three opioid rotations. This case, however, demonstrated some principles worth discussing. One principle is to think through what you anticipate the problem is going to be, for example, the issue of neuropathic pain vs nociceptive pain, and the addition of local anesthetics. This is not a static disease, and the oncologist or pain specialist needs to work with the patient and the course of his or her pain. You cannot just implant a device and hope that 1 mg/d of intrathecal morphine will effectively relieve the patient’s pain until he or she dies.

Dr. Portenoy: After the pump was implanted, why did you add bupivacaine when the patient was actually still taking very low doses of intrathecal morphine?

Dr. Staats: The quality of the pain was changing to more of a burning pain that was neuropathic. The additional local anesthetic was what I believed to be in the child’s best interest. Clearly, another option could have been to increase the intrathecal morphine.

Dr. Portenoy: It seems like that decision was similar to what we noninterventionists do when we’re choosing adjuvant analgesics. The decision is completely idiosyncratic. Some clinicians will choose to increase the morphine to some arbitrary ceiling based on the fear of hyperalgesia syndrome, whereas others will increase the morphine until patients get side effects; still others will add bupivacaine early. Do you feel the decision is basically clinician-specific without any justification?

Elliot S. Krames, MD: This is the art of medicine. We do not have randomized controlled studies to guide us to the appropriate time to add bupivacaine. My art of medicine would be to increase the dose of morphine. I basically give the opioid the benefit of the doubt. When I approach, in my art, 20 mg/day, then I add bupivacaine and/or clonidine.

Dr. Portenoy: There seems to be a general feeling among you that a painful lumbosacral plexopathy would not be as responsive to opioids as would, for example, a somatic pain. I only bring up this point for the oncologists who may be reading this because it does vary with the guideline for systemic opioid therapy discussed in my article on page 25. If you accept the concept that you select an opioid drug, and then you individualize the dose by escalating until you get to treatment-limiting toxicity, which defines the responsiveness to that drug by that route, then the appropriate intervention would always be opioid dose titration first, before addition of another drug.

Dr. Staats: If a patient presented with a lumbar plexopathy with tumor invading the right hip and you believed that this lumbar plexopathy neuropathic component might respond well to Neurontin or to carbamazepine (Tegretol), or to something else if there was a shooting or lancinating component, you might start that agent. And if the tumor was invading the spine, I might use opioids or steroids or something else to manage this nociceptive component. Conceptually, that’s what I believed I was doing in this case.

Systemic Pharmacologic Approach

Dr. Portenoy: A problem with very early use of combination therapy is the possibility of additive side effects. If a person has a painful lumbosacral plexopathy, our teaching about systemic therapy would be to increase the systemic opioid first, and to do so very quickly. Once the patient starts to become somnolent without adequate pain relief, you could reduce the opioid dose and quickly administer adjuvant analgesics. There is just no evidence behind the rationale for adding adjuvant analgesics when the opioid dose is very low and is not associated with any toxicity because you assume that the efficacy of the opioid won’t be adequate to treat that pain.

Dr. Levy: In a systemic pharmacologic approach, it is much like what Dr. Staats did in this case. Our experience indicates that it is uncommon for a patient’s burning pain to go away; the typical response to opioids is that the pain is reduced in intensity, but the burning sensation is not gone. And if you choose to use a tricyclic antidepressant, it will take weeks to reach a therapeutic dose with minimal side effects. We do a lot of anticipatory use of coanalgesics, particularly of the neuropathic drugs, to perhaps prevent or minimize the side effects of increasing doses of opioids. We use combination chemotherapy in oncology in the same way. Moderate doses of several agents with different mechanisms of action work better and with less toxicity than high doses of just one agent. We don’t have the data, but I believe our experience, particularly with neuropathic pain, is that it’s the rare patient in whom most symptoms can be controlled with just opioids. And because it takes some titration to reach the full effect (even Neurontin can take days, or weeks), we do that same kind of anticipatory approach that Dr. Staats used in this case.

Dr. Krames: Many patients with neuropathic pain do not respond to intrathecal morphine at 1 mg/d but may respond to 2.5 mg/d or even higher. Once you’ve started therapy, unless you’ve titrated to either efficacy or side effects, you’re not really giving that therapy an adequate trial. In some patients we’ll add Neurontin or desipramine. With intrathecal therapy, however, there is evidence that higher doses of opioids are better in neuropathic pain states, and I believe you should titrate until efficacy is achieved or side effects occur. In a patient with severe neuropathic pain, I’ll increase the dosage by 50%, and if the patient is still in pain, will increase another 50% a day or two later, and still again until toxicity occurs or efficacy is achieved. I’ll stop once dosage has reached 20 mg/d and then consider the addition of a nonopioid such as clonidine or bupivacaine.

Dr. DuPen: In my practice, we strongly support the use of opioids and adjuvant drugs through an algorithm. However, this is a 15-year-old child with out-of-control pain who is not responding to opioids. My therapeutic choice is intravenous sedation, with a temporary epidural catheter with bupivacaine to get pain control, then a progressive effort to use drug trials and adjuvant therapy to determine the role for long-term intrathecal therapy.

Best Clinical Judgment

Dr. Portenoy: I think it is very interesting to hear disagreement among the experts, especially given the data Dr. Staats showed that oncologists don’t really know how to use adjuvant analgesics . The bottom line is that everybody agrees that adjuvant drugs can be very helpful, including adjuvant bupivacaine combined with an intrathecal opioid. The particular drugs selected and the timing of use, however, are very much based on best clinical judgment. Some clinicians would tend to use them early based on an experience that suggests that there’s a relatively low likelihood of a good response to the higher doses of the opioids. Others, myself included, would tend to say we can titrate fast enough so that we’d like to see what the higher dose opioids do rather than run the risk of additive toxicity from the premature use of a combination regimen. That same controversy exists whether you’re talking about systemic drugs or about intraspinal drugs.

Dr. Staats: Conceptually, I agree. If someone else had presented this case and had increased morphine in double doses, I would have thought that appropriate. What we’ve learned in anesthesia is that if we use a little bit of this agent and a little of that agent, the side effects of either agent are less. It’s a term called “balanced anesthesia” in the operating room. I try to apply those principles whenever I think it’s appropriate intrathecally, and have termed this balanced analgesia. It’s not the only way to do it, but I believe that this was a good way, and in this case, the patient did well with this therapeutic regimen.

Dr. Krames: I think that it’s inappropriate to call these medicines adjuvant medicines. Bupivacaine and clonidine are really not adjuvant medications. They are specific agents that are analgesic in the spinal canal and work differently than opioids. Rather than adjuvant medications I would call them coanalgesic medications.

Samuel J. Hassenbusch, MD: I think the day is close when we will identify syndromes for which we may in fact use coanalgesia, or coadministration of drugs, for spinal delivery up front in the sense that this approach will be more effective than either agent alone across the board.

Dr. Portenoy: I’d also make the point that as drugs become available, we tend to use those agents earlier that seem to have very good maximal efficacy with a big therapeutic window. In my own case, for example, the use of gabapentin has become much more aggressive than the use of amitriptyline, because I can titrate it quickly. It’s a relatively safe drug, and if it’s going to work, it works at a dose range that I can reach in a few days. I think that it’s good for the clinicians who are not experts in pain management to hear that there is a diversity of opinion about how these coanalgesic compounds are used, whether they’re given systemically or intrathecally.

Practical Management

Dr. Portenoy: Dr. Staats, could you comment on the practical management of this 15-year-old boy for a period of 7 or 8 months with the pump? Who becomes the primary care provider with respect to the pain therapy, and with respect to the cancer therapy?

Dr. Staats: We have an attending physician in my practice who is a pediatric anesthesiologist pain specialist who does not implant devices. I implant the devices with her and then she manages the patient. She does not have the same level of experience that I have with infusion therapy, but she’s learning. The pediatric anesthesiologist pain specialist receives the first phone calls from the patient’s oncology team and/or the patient’s caregiver; if she has concerns, she calls me. In this particular case, I made a lot of the management decisions. At one point, in response to a phone call, a decision was made by the oncology team to put the patient back on 30 mg of intravenous Dilaudid a day. I don’t think it was the right decision, and in fact the boy became quite ill, but it was the only decision available to the team at that time. This is a point in the patient’s care where I believe the best decision was not made because I could not be reached to address the issue. Once the oncology team did reach me, I advised that the patient be taken off of the intravenous Dilaudid and that his intrathecal infusion be increased.

Dr. Portenoy: Who was responsible for the management of other symptoms like nausea?

Dr. Staats: In that one situation, the oncology team tried to get the nausea under control and called me after multiple therapies for nausea failed. Together we decided to decrease the systemic load of the opioid by giving it intrathecally, and that’s what worked. So, to some extent there was an interaction between the oncology team and the pain management team saying, “Let’s try different approaches.”

A Broader Model of Palliative Care

Dr. Portenoy: Basically you want to have the most sophisticated approach to pain management available as is necessary, but you want to have that approach integrated into a broader model, which could mean bringing nonphysicians into that model. You might have to bring the hospice team, occasionally, even if there is an issue of reimbursement.

Dr. Staats: Those issues of reimbursement are important to address early and up front in a patient’s care. Once this pump is in place, early intervention with this child, or with any patient, makes subsequent care very inexpensive to maintain. And it’s actually in hospice’s best interest to get the pain management team’s evaluation about implanting a device before a patient is converted to hospice care. The patient can continue with a good quality of life, and when things aren’t going so well, return to hospice. Hospice will be able to take care of most of the patient’s medical and analgesic requirements via the implanted device, and the other hospice needs can still be provided. It’s not an either/or situation.

Dr. Portenoy: I’m a hospice medical director and these issues about what we can and cannot afford come up all the time. I would really view this implanted device similar to having a pathologic fracture of the hip repaired. No hospice can afford to routinely send patients to surgery to repair pathologic fractures. The patient in this case has to decertify from hospice, have his hip fracture repaired, and then, if in the patient’s best interest, elect hospice care again.

Dr. Staats: On average, patients have 2 weeks of life left when they enter hospice, and that’s too late. We need to get oncologists to think about these issues of pain management and hospice care well in advance of when we say, “Let’s go to hospice.”

The System Worked

Dr. Portenoy: This was a case in which the patient’s pain was an overriding impediment to a good quality of life and a good quality of dying. Consequently, implanting the pump was the primary modality for palliative care. So, the system worked. Obviously, the type of palliative care needed by another patient might be much more complex. A patient’s suffering may only be partly determined by the pain. It might be related to psychological factors, family distress, financial concerns, social disarray, and/or spiritual issues. Given the complexity of the pain management, could this case be integrated into a broader model of palliative care that also addressed the psychological, spiritual, and family issues that occurred as this child was dying? Or did the complexity of the pain management work against comprehensive palliative care?

Dr. Staats: Conceptually, in my opinion, that should be part of the palliative care program. There is palliative care and there is interventional pain management, and the oncologist needs to decide which way the patient should be directed. That’s not how things should be, however. I believe that we should have a cooperative group, and when it’s clear that maximal intrathecal therapy is not working, then another systemic therapy should be tried. One advantage of intrathecal therapy is that it improves pain relief with a lower side-effect profile, but it’s not perfect. For this child, implanting a pump was clearly the right decision because what he wanted most was to go back to school. He died happy because he was able to spend the last 8 months at school.