The Management of Pain in Metastatic Bone Disease

Sorin Buga, MD; Jose E. Sarria, MD


Cancer Control. 2012;19(2):154-166. 

In This Article

Interventional Management

The WHO analgesic ladder was developed in 1982 as a global public health program to address the problem of untreated cancer pain, particularly at the end-of-life stage.[38] Prior to the release of these guidelines in 1986, numerous barriers that prevented the effective treatment of cancer pain existed, and descriptions of dying patients in pain were depicted as "a cruel and callous disgrace."[39] With advances in the understanding of opioid analgesics and the newly created specialty of palliative medicine, the WHO analgesic ladder had a major impact on the management of patients who suffered mild to severe cancer-related pain. At the core of its creation, one of the central premises was its simplicity — simple enough to be adopted even by underprivileged societies.

By 1996, however, critical reviews were highlighting drawbacks of the WHO ladder (Fig 2A), mostly the fact that it consistently failed to provide sufficient relief in 10% to 20% of patients.[40,41] In many instances, the ladder was described as an oversimplification of a complex problem. It was for these cases that interventional techniques were considered. The use of the interventional approach when systemic analgesia was unsuccessful, due to either uncontrolled pain and/or unacceptable side effects, was termed "the fourth step of the ladder" (Fig 2B).[42] Failure of systemic analgesia can be closely related to specific pain generators and amount of malignant disease burden. Pain of neuropathic origin, for example, is known to be a poor respondent to opiates and conventional adjuvant therapies.[43,44]

Furthermore, there is growing evidence of the pervasive effects caused by the chronic use of opiates. These include a complex process of progressive central sensitization known as opioid-induced hyperalgesia that may actually lead to increased perception of the experience of pain and a reduced ability to cope,[45] cognitive dysfunction, hypogonadism, intractable constipation and/or nausea, psychosocial implications such as addiction, pseudo-addiction, diversion, and abuse of controlled substances, all of which can lead to destructive behaviors and disrupt the patient's social and family support system.[46] The prevalence of substance abuse problems in the cancer patient, although lower than the general population, remains a cause for concern.[47] With the increasing number of cancer survivors and thus a heightened prevalence of chronic pain in these patients,[48] some have proposed that the long-practiced paradigm of the WHO ladder, which may limit the ability of cancer survivors to return to normal life and activities, be turned "upside down," with earlier utilization to the interventional and adjuvant therapies.[49]

Metastatic disease to the bone illustrates the concerns stated above. The nervous system is closely related to the bony structures that surround it. The management of malignant disease in the vicinity often focuses on preventing invasion of the adjacent nervous structures and treating the ominous symptoms of pain and/or neurologic deficits. The appearance of secondary malignant disease in the bone signals progression to systemic disease, and local control and palliation become priorities. In these instances, when issues regarding pain control are common, implementing the "fourth step" of the ladder should be considered. Moreover, given the predictable course of many of these lesions, a multidisciplinary approach must be undertaken early on.

Metastatic bone disease can be focal, multifocal, or generalized, and so will the procedural approach. The first two can be discussed together.

The solitary or oligofocal vertebral lesion presents with pain as the most common and earliest symptom, typically nocturnal.[50,51] At initial stages, the pain is thought to be somatic due to invasion of the receptorrich periosteum from the receptor-poor marrow. Neuropathic pain may follow when epidural extension, compression fracture, or spinal cord compression occurs. The average time frame from initial pain presentation to complications is 7 months.[52] Ominous signs include rapid progression of back pain in a crescendo pattern, radicular pain exacerbated by recumbency or strain, and neurologic deficits such as weakness, sensory loss, autonomic and sphinchteric dysfunction, and osteotendinous reflex abnormalities.[53] The initial clinical suspicion is often confirmed by imaging studies such as MRI, CT, or bone scan. An accurate pathological diagnosis is paramount for prognosis and patient survival regardless of treatments offered, especially when there is no prior history of cancer.[54] Therefore, a bone biopsy is frequently considered at different stages, depending on the presentation.

Conventional external-beam radiotherapy (EBRT) is the mainstay treatment of painful vertebral lesions, without mechanical instability, that do not involve the nervous system.[55] EBRT may provide profound pain relief, prevent pathological fractures, and delay neurologic dysfunction. In addition, newer radiation techniques, collectively known as stereotactic radiosurgery, may offer several advantages such as increased radiation dose to the target area with reduced incidence of radiation toxicity. It may also offer the ability to treat patients in 1 or 2 days rather than the several days needed for conventional radiation; these newer techniques may also be more efficacious for radioresistant tumors such as renal cell carcinomas and sarcomas. The conventional wisdom regarding EBRT and solitary bone lesions, although anchored on evidence that functional outcomes are comparable to surgery, has often been challenged in the literature.[54]

For patients who present with painful pathological vertebral compression fractures (VCRs) but no neurologic compromise, newer percutaneous vertebral augmentation procedures (most notably vertebroplasty and kyphoplasty) offer a novel option. These minimally invasive procedures consist of an injection of bone cement (polymethylmethacrylate) in a fractured or disrupted vertebral body via a percutaneous cannula placed in the vertebral body using a uni- or bi-pedicular approach. This provides structural support and minimizes mechanical pain. In addition, the cement may have intrinsic analgesic and antitumor properties. Kyphoplasty differs from vertebroplasty in that the injection of the bone cement occurs after creation of a cavity in the vertebral body by inflation of a balloon. This will allow a low-pressure injection, thus minimizing complications from extravasation.[56] The first vertebroplasty report, which came from France in 1987, was used for the treatment of aggressive vertebral hemangiomas.[57] With experience, two other indications were found: osteoporotic vertebral VCFs and spinal tumors. The safety and efficacy of these procedures have been acclaimed in some large, multicenter, randomized controlled trials such as the FREE study[58] and challenged in others.[59] The Cancer Patient Fracture Evaluation (CAFE) trial was a randomized controlled trial at 22 sites in Europe, the United States, Canada, and Australia. In this trial, 134 cancer patients with 1 to 3 VCFs were randomized to receive kyphoplasty vs nonsurgical management. The primary endpoint was functional status as measured by the Roland-Morris disability questionnaire ( At 1 month, a statistically significant difference was seen in favor of those who received kyphoplasty, and no complications were reported with this approach.[60] Experience with the use of vertebral augmentation procedures has allowed the expansion of their reach from the classic uncomplicated VCF to special situations such as prophylaxis against imminent fracture,[61] treatment when there is epidural involvement, and combined techniques with EBRT and radiofrequency ablation (RFA).

Whether displaying neurologic symptoms or not, vertebral lesions with epidural extension, also described as breach of the posterior cortex, have been primarily surgically managed. However, many patients with these lesions are poor surgical candidates or have a limited life expectancy. Vertebral augmentation again has a role along with XRT and RFA. Those who are no longer candidates for radiation therapy seem to receive the most benefit from RFA,[62] but vertebroplasty, in the face of epidural extension with VCFs, has been used alone[63,64] and in combination with RFA.[65,66] Combination radiosurgery and kyphoplasty has also been used, with fiducial markers for radiation placed during the kyphoplasty an average of 12 days prior.[67] Many of these patients had received XRT in the past. Intraoperative radiotherapy during kyphoplasty (kypho-IORT) is a novel approach used to deliver a single dose of 10 Gy to the spinal lesion during a kyphoplasty procedure.[68,69]

Residual pain after successful vertebral augmentation procedures is estimated to average 23%. Although there is no literature as to what are the likely pain generators, degenerative changes in the adjacent structures such as facets and discs are the logical causes, leading to persistent axial back pain and radiculopathy. Interventional procedures such as epidural corticosteroid injections, facet joint injections, trigger point injections, intercostal nerve blocks, and sacroiliac joint injections have been successfully employed for further relief of painful symptoms.[70,71,72]

Localized metastatic disease in other bones can also be painful, particularly when the original somatic pain becomes neuropathic due to invasion of adjacent neural structures. When these lesions respond poorly to XRT alone or in combination with reconstructive surgery, injection of bone cement has been evaluated with excellent results. In acetabular lesions compromising ambulation, Maccauro et al[73] presented a retrospective study of 25 patients undergoing cement acetabuloplasty when surgical reconstruction was not an option. All patients obtained marked clinical and functional improvement initially, with a mean duration of pain relief of 7.3 months. No major complications were observed. Many other sites are amenable to this technique when conventional treatments fail. In a prospective report of 50 patients, Anselmetti et al[74] successfully applied this technique to the femoral shaft, pelvis, ribs, knee, tibia, humerus, and sacrum. Seven of the 50 patients underwent RFA in the same session. No complications were reported, but at 1 month, 2 of 15 patients treated at the femoral diaphysis suffered pathological fractures. Combination cementoplasty and RFA has also been described with good results.[75]

When a neuropathic component is present, the resulting pain can be more difficult to treat and frequently fails systemic analgesia. An evaluation for XRT or surgery, while always worth exploring, often leads to the need for alternative palliative approaches. Interventional pain techniques can be beneficial in this situation.

Selective diagnostic nerve blocks that offer short-term relief are used as conduits leading to ablative procedures such as RFA, cryoablation, and phenol and alcohol neurolysis, seeking long-term analgesia. The central premise of these neuroablative procedures is their ability to achieve selective C and Aδ fiber (pain fibers) neurolysis in a given nerve, preserving to a higher or lesser degree the anatomical integrity of the peri-, epi-, and endo-neurium (which will allow future reinervation), as well as sensory and motor fiber function. This is possible by taking advantage of the smaller diameter and relative lack of myelin of the pain fibers. Autonomic fibers usually cannot be spared since they are small and unmyelinated.

While almost any nerve may be subject to this approach, those controlling the motor function of the extremities are treated with more caution due to the potential for loss of limb function. Consequently, these techniques have been most commonly described for axial pain such as intercostal nerve-mediated pain from rib metastases or postthoracotomy pain and postamputation pain. Particular attention has been given to pulsed radiofrequency of the dorsal root ganglion and nerve roots. RFA has been the preferred neurolytic technique, given its ability to control the size of the lesion by tissue temperature feedback control.[76–80]

In other cases, the risk of loss of limb function, shortened life expectancy, or a possible endpoint to the source of the pain by active oncologic treatment may warrant a different approach. The placement of a temporary catheter for the continuous infusion of local anesthetic (regional analgesia) is a reliable, safe, and feasible option, particularly in end-of-life care. Infusions of local anesthetics are the most common. Volume and concentration dictate the depth of the nerve blockade — specifically, the development of anesthesia and motor block vs analgesia. This allows a titration range that can accommodate the changing needs of each patient. By tunneling these catheters under the skin, infection risk is acceptable so they can be left in place for extended periods of time. This also leads to more stability of the catheter, thus reducing the risk of migration.[81–83]

Femoral/sciatic nerve and brachial plexus as well as epidural catheters have been used successfully.[84–88] The drawback associated with these catheters is the need for constant care. The infusate solution requires frequent refills, the mobility of these patients is restricted to some degree in every case, and complications of obstruction and catheter migration are common. Infections are rare, but the bacterial colonization rate is significant in these catheters, even after short-term (48-hour) infusions.[89]

Epidural catheters deserve special attention. With the discovery of spinal opioid receptors in the 1970s,[90] the field of neuraxial analgesia found an alternative to local anesthetic administration that could minimize the side effects of motor blockade and autonomic dysfunction. Epidural opioids, more specifically hydrophilic opioids such as morphine and hydromorphone, can provide segmental analgesia when placed close to the spinal level corresponding dermatomes. Their use in the hospice setting has been well documented,[91] and despite advances in other forms of neuraxial analgesia such as intrathecal (IT) infusions that require less labor-intensive follow-up, they continue to have a place in the care of intractable pain in cancer patients at the end of life. The lack of randomized controlled trials for these techniques can be explained by several factors: only a small percentage of cancer patients require their use, randomization can be difficult to implement due to concerns of informed consent and ethics, primary endpoints are difficult to define in these patients, and study participant cohorts would be highly heterogeneous.

Additional interventional alternatives are available to manage localized bone cancer pain, such as neurostimulation and spinal cord stimulation.

Neurostimulation is a field of neuromodulation with the potential to offer high levels of analgesia to patients with neuropathic pain in whom the steps of the WHO analgesic ladder have been insufficient. Dorsal column stimulation was originally described in a case report in 1967,[92] and it was based on the "gate control theory" proposed 2 years earlier.[93] The basic concept of spinal cord stimulation is centered on the early findings that, in spinal transmission, when an increased input of a sensory modality is applied, it can "close the gate" to other modalities, effectively modulating the conveyance of these signals to higher centers in the central nervous system. More specifically, electrical stimulation of the dorsal column has a neuromodulatory effect on the activity of the ascending pain pathways.

Spinal cord stimulation and, more recently, peripheral nerve stimulation have unique mechanisms of action in the treatment of neuropathic pain and may be the only alternative available when all other therapeutic interventions have failed. Implementation of this therapy requires an initial placement of percutaneous leads with electrical contacts at the target neural structure (typically the dorsal column) for a trial. The position of these contacts is somatotopic and requires the patient to be awake to offer feedback on where the stimulation is felt. The patient then uses an external pulse electrical generator at home for an average of 5 days. If there is more than 50% pain relief, along with some objective measures such as a decrease in opioid consumption and improvement in activities of daily living, a permanent implant can be scheduled. This entails creating a subcutaneous pocket to place a pulse generator unit similar to a pacemaker and anchoring the new leads to prevent them from moving, since precise position is critical for continued benefit (Fig 3).

Figure 3.

Implementation of spinal cord stimulation, which requires placement of percutaneous...

Successful use of neurostimulation in cancer patients has been documented in the literature.[94–96] Proceeding with this option remains an individualized decision, particularly in the setting of a critical or often terminal illness. A good patient-physician relationship as well the assistance of other supportive services can help in making the transition to neurostimulation.

Metastatic bone disease can be widespread and may call for more than localized, targeted approaches when interventions for pain are needed. When optimized systemic analgesia fails, intrathecal infusions might be considered.

The first use of opioids infused in the cerebrospinal fluid (CSF) for cancer patients was reported in 1979 by Wang et al.[97] Since then, numerous advances have been made in the indications for implantable infusion pumps, drugs used, and patient selection. There are several advantages with this therapy. The potency of intrathecal opioids is multiplied by a factor of 1:300 compared to oral administration.[98] Additional benefits are minimized side effects and the ability to use combination infusates with drugs that are approved only for IT administration (ziconotide) or that are more effective through this route (eg, local anesthetics, clonidine). Because the CSF courses throughout the entire central nervous system, IT therapy is not segmental in principle and may provide analgesia virtually anywhere in the body.

Although the use of implanted IT therapy in chronic nonmalignant pain remains controversial, its use in cancer is rarely argued. In a multicenter randomized trial comparing analgesia delivered via an intraspinal implantable drug delivery system to comprehensive medical management in 201 patients with refractory cancer pain, IT therapy was significantly superior in clinical effectiveness (defined as at least 20% pain level rating decrease).[99] Side effects were similar; however, decreased rates of depression and mental status changes were reported, as well as improved survival (53.9% alive at 6 months in the IT therapy group compared with 37.2% in the medical management group).

IT infusions can be administered through tunneled percutaneous catheters or implantable drug delivery systems (IDDSs) that now come with computerized programmable features, including patient-controlled dosing (Fig 4). IDDS insertion is recommended when the patient's life expectancy is longer than 3 months.[100]

Figure 4.

Example of intrathecal (IT) infusion. These can be administered through tunneled percutaneous catheters or implantable drug delivery systems that are now available with computerized programmable features, including patient-controlled dosing. Reprinted with permission.

The list of drugs that can be used in the IT space, whether on- or off-label, continues to expand ( Table ) and reflects the ongoing efforts to combine different mechanisms of action that may act synergistically against neuropathic pain.

Ziconotide is one of a small number of drugs that are approved by the US Food and Drug Administration for use in IT therapy. Ziconotide, a synthetic peptide derived from the sea snail Conus magus, selectively blocks N-type voltage calcium channels at presynaptic terminals of the dorsal horn and is used for IT administration only. Validated for cancer pain by Staats et al,[101] ziconotide is an effective drug for neuropathic pain but is associated with many possible side effects.

The excellent analgesic properties of intrathecal ketamine has also been demonstrated in cancer patients, although evidence of neurotoxicity has limited its clinical application. However, in terminal cancer patients with a short life expectancy, it is a valid alternative.[102]

A combination of several interventional therapies may be needed for each individual. For example, a patient may have widespread malignant disease to the spine that is initially well controlled with systemic analgesia. The patient may then develop a pathological VCF that benefits from vertebral augmentation and epidural corticosteroid injection if radicular symptoms are present. This patient is also likely to progress to opioid tolerance and ultimate failure of systemic analgesics, thus requiring IT therapy to improve quality of life.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.