Preliminary In Vivo Experimental Evidence on Intratumoral Morphine Uptake. Possible Clinical Implications in Cancer Pain and Opioid Responsiveness

Vol. 24 No. 1 July 2002

Journal of Pain and Symptom Management

1

Letters

Preliminary In Vivo Experimental Evidence on Intratumoral Morphine Uptake. Possible Clinical Implications in Cancer Pain and Opioid Responsiveness To the Editor: Some patients with cancer pain demonstrate poor analgesic response to opioid drugs. The need for rapid opioid escalation (sometimes without evident disease progression) may suggest opioid tolerance. The neurophysiologic mechanisms underlying opioid tolerance and other states of poor opioid responsiveness are unknown. Among other potential explanations, imbalance in the production of morphine metabolites (morphine3-glucuronide and morphine-6-glucuronide) shift from a nociceptive to a neuropathic pain pathophysiology, opioid receptor internalization and/ or desensitization, and spinal and supraspinal hypersensitivity have been postulated.1 Based on numerous clinical observations (about 3000 patients in ten years) at Pain Center of the Regina Elena National Cancer Institute, we have hypothesized that increasing opioid tolerance or more generally opioid hyporesponsiveness also may be attributable to an intratumoral opioid uptake.2 This hypothesis is supported by experimental evidence indicating that opioid receptors expressed in the nervous system and mediating opioid analgesia are also expressed on tumor cells.3 They are functionally capable of binding morphine and their engagement results in nitric oxide (NO) release.4 To model intratumoral opioid uptake in vivo, mimicking the clinical situation of a cancer patient treated with morphine, we analyzed the binding of morphine in vivo by two human tumor cell lines, a glioblastoma (LI) and a melanoma (M14) growing in immunodepressed nude mice. Both tumor cell lines were previously tested © U.S. Cancer Pain Relief Committee, 2002 Published by Elsevier, New York, New York

for their ability to bind morphine in vitro by indirect immunofluorescence and by mass spectrometry analysis. Both were found to express morphine binding sites. Viable melanoma or glioblastoma cells (5  106) were injected into the hind leg of CD.1 nude mice. The implant gave rise to a tumor (1 cm) within ten days in all the mice. At this time, morphine (10 mg/kg) was injected intravenously and mice were sacrificed at different times after opioid administration. Time course analysis (Fig. 1) revealed that the intratumoral levels of morphine were higher at four hours after injection and progressively increased at 72 hours, as compared to normal muscle tissue from the contralateral leg. Conversely, the serum levels of morphine slowly decreased. By immunohistochemical analysis, we found that the M14 tumors were infiltrated, although at low levels (2%/mm2 tumor tissue) by neutrophils and macrophages (data not shown). These observations support our hypothesis that tumors expressing opioid receptors show a preferential morphine uptake.

Comment Our experimental in vivo model showed intratumoral morphine uptake and also raised the possibility that other components of tumor mass, such as infiltrating leukocytes and endothelial cells expressing opioid receptors, lead to morphine uptake. In the clinical situation, leukocytes infiltrating human tumors are more abundant, and it is conceivable that the functional effects induced by morphine activity on immune cells may be more relevant. Thus, the response and the responsiveness to exogenous opioids in cancer pain patients may depend on the levels of opioid receptor expression by tumor cells as well as by the degree of leukocyte infiltration. Which could be the functional consequences of morphine binding to immune vs. tumor cells 0885-3924/02/$–see front matter

2

Fig. 1. Morphine muscle () and tumor (x) concentrations at different times (4-8-24-32-72 h) after intravenous administration of the opioid (10 mg/kg) in two different groups of nude mice injected with 5  106 viable human melanoma (M-14....) or glioblastoma (LI—) cultured cells, as evaluated by mass spectrometry. Mice were injected with morphine and sacrificed when tumor diameter reached 1 cm and tumor weight was 500 mg. At the same times, the mean morphine serum concentrations assayed by HPLC (not shown in this figure for sake of simplicity) decreased from 114 (4h) to 30 (72h) ng/ml. Each point represents the mean  S.E. of at least three experiments.

and is there any relationship between the two interacting cell systems and nociception? All types (, , ) of opioid receptors have been identified on tumor cells. These receptors interact with their endogenous or exogenous opioid ligands and are capable of mediating a number of signaling and functional responses.5 In a recent study,4 we have provided evidence that morphine stimulates a more sustained release of NO by a human non-small cell lung carcinoma expressing a subtype of  receptor, as compared to that generated by stimulated neutrophils. Identification of morphine receptors on cells of hematopoietic origin has been an area of intense immunopharmacological research.6 The inhibitory action of morphine on immune functions has been widely reported and attenuation of morphine-induced immunosuppression (together with analgesia) has been described in mice lacking the  opioid receptor gene.7 Among the mechanisms underlying the morphine-mediated suppression of immune functions (and in particular leukocyte chemotaxis and phagocytosis), the opiate-induced release of NO plays a major role,

Letters

Vol. 24 No. 1 July 2002

and NO synthase inhibitors abolish the opiatemediated inhibitory effects.8 This finding indicates that morphine can induce similar functional responses in neoplastic vs. immune cells, and further suggests that opioid-induced NO release may affect tumor growth by exerting a direct tumoricidal effect and/or regulating anti-tumor immune response. We can envisage that in a particular microenvironment, such as the tumor where immune and neoplastic cells cohabit and likely communicate, the “theft” of exogenous opioids, morphine in particular, by the different tumor components may interfere with nociception in at least two different and opposing ways: –antinociceptive, due to the potentiation of peripheral pain regulatory mechanisms that integrate the opioid and cytokine-mediated antinociceptive function of immune cells;9 –pronociceptive due to the NO release induced by morphine binding on tumor and immune cells, triggering the cascade of biochemical events that have been suggested to link the clinical phenomena of hyperalgesia, opioid tolerance and neuropathic pain.1 In conclusion, we hypothesize that in many painful situations of cancer patients, opioid responsiveness may change during tumor progression as the result of an unpredictable balance between different morphine-mediated effects. Edoardo Arcuri, MD Regina Elena National Cancer Institute, Rome, Italy Patrizia Ginobbi , PhD Walter Tirelli, MD Hospice Sacro Cuore, Rome, Italy Rino Froldi, PhD University of Macerata, Macerata, Italy Gennaro Citro, PhD Regina Elena National Cancer Institute, Rome, Italy Angela Santoni, PhD University of La Sapienza, Rome, Italy PII S0885-3924(02)00425-6

Acknowledgments This work was supported by grant No. 0201R2 from Minister of Health.

References 1. Mercadante S, Portenoy RK. Opioid poorly-responsive cancer pain. Part 2: Basic mechanisms that could

Vol. 24 No. 1 July 2002

shift dose response for analgesia. J Pain Symptom Manage 2001;21:255–264. 2. Arcuri E. Can tumors act as opioid traps, mimicking opioid tolerance? J Pain Symptom Manage1998; 16:78–79. 3. Maneckjee R, Biswas R, Vonderhaar BK. Binding of opioids to human MCF-7 breast cancer cells and their effects on growth. Cancer Res 1990;50:2234–2238. 4. Fimiani C, Arcuri E, Santoni A, et al. 3 opiate receptor expression in lung and lung carcinoma: ligand binding and coupling to nitric oxide release. Cancer Lett 1999; 146: 45–51. 5. Zagon IS. Opioid receptors and endogenous opioids in diverse human and animal cancers. J Natl Cancer Inst 1987; 79: 1059–1065. 6. Sibinga NES, Goldstein A. Opioid peptides and opioid receptors in cells of the immune system. Ann Rev Immunol 1988; 6: 219–249. 7. Gaveriaux-Ruff C, Matthes HWD, Peluso J, Kieffer BL. Abolition of morphine-immunosuppression in mice lacking the -opioid receptor gene. Proc Natl Acad Sci USA 1998; 95: 6326–6330. 8. Kolesnikov YA, Pick CG, Ciszewska G, Pasternak GW. Blockade of tolerance to morphine but not to k opioids by a nitric oxide synthase inhibitor. Proc Natl Acad. Sci USA 1993;90: 5162–5166. 9. Mousa S.A, Machelska H, Shafer M, Stein C. Coexpression of beta-endorphin with adhesion in a model of inflammatory pain. J Neuroimmunol 2000; 108:160–170.

Block of Interdigital Nerves for Cancer Pain To the Editor: Tumor growth is the most common cause of cancer pain.1 Although opioid therapy has been recognized as the mainstay of treatment for advanced cancer patients,2 there is a selected group of patients who can benefit from other interventions, including neural blockade. Neural blockade techniques form an important part of a multidisciplinary approach to cancer pain management. Many minor procedures that do not require the availability of radiography may be undertaken in a hospice setting.3 Well-localized pain can be relieved more easily, while large painful areas innervated by multiple nerves may require more central blocks. The following case illustrates the benefits that can occur from a peripheral block in the context of palliative care.

Letters

3

Case Report A 65-year-old man had been admitted to the pain relief and palliative care unit for the consequences of cerebral radiotherapy performed in another hospital to treat cerebral metastases secondary to bladder cancer. At admission he was drowsy and severely compromised. He received supportive treatment, including care of skin lesions, dietary advice, and psychological help, and was discharged home six days later with a good level of consciousness, an acceptable performance status and nutritional intake. One month later, he underwent radiotherapy again and his condition rapidly worsened. He was admitted to the unit for diffuse edema, especially in the legs, low nutritional intake, severe dehydration, drowsiness, lumbar pain, and pain in the finger. The finger pain was due to a bone metastasis, which had already been irradiated unsuccessfully. The evaluation revealed the presence of lumbar metastases and relapse of cerebral metastases. He was somnolent (2 on a scale of 0–3, commonly used to monitor symptoms in the unit), but collaborating. He was receiving tramadol 300 mg a day, which controlled his lumbar pain but not the finger pain. Supportive measures were started. Considering the impairment of consciousness due to the cerebral metastases, the option of opioid escalation was discarded. A block of the interdigital nerves was proposed and accepted by patient to control the pain. Three ml of alcohol, preceded by a local anesthetic block with lidocaine, were injected in each side of interdigital angle, where the interdigital nerves are presumed to pass. This produced immediate and complete anesthesia in the finger. The procedure was performed at bedside and required a few minutes. After stabilization of the supportive therapy, the patient was discharged home for compassionate home care. He died one week later with good pain control.

Comment In the palliative care setting, anesthetic blocks are rarely performed. Anesthesiologists with knowledge and experience in these techniques are seldom involved in the care of these patients.4 This is unfortunate because there is a range of simple anesthetic blockade techniques that are associated with few adverse effects and may provide symptomatic relief for long periods. Neurolysis can produce a deafferentation pain syndrome that may be worse than the original pain,