Supportive Care in Cancer

, Volume 12, Issue 11, pp 762–766

Addition of a second opioid may improve opioid response in cancer pain: preliminary data


    • Anesthesia and Intensive Care Unit and Pain Relief and Palliative Care Unit, La Maddalena Cancer Center, Palliative MedicineChair of Anesthesia, Intensive Care and Emergency
  • Patrizia Villari
    • Anesthesia and Intensive Care Unit and Pain Relief and Palliative Care Unit, La Maddalena Cancer Center, Palliative MedicineChair of Anesthesia, Intensive Care and Emergency
  • Patrizia Ferrera
    • Anesthesia and Intensive Care Unit and Pain Relief and Palliative Care Unit, La Maddalena Cancer Center, Palliative MedicineChair of Anesthesia, Intensive Care and Emergency
  • Alessandra Casuccio
    • Anesthesia and Intensive Care Unit and Pain Relief and Palliative Care Unit, La Maddalena Cancer Center, Palliative MedicineChair of Anesthesia, Intensive Care and Emergency
    • Department of Hygiene and MicrobiologyUniversity of Palermo
Original Article

DOI: 10.1007/s00520-004-0650-1

Cite this article as:
Mercadante, S., Villari, P., Ferrera, P. et al. Support Care Cancer (2004) 12: 762. doi:10.1007/s00520-004-0650-1


Recent experimental data suggest a possible use of an opioid combination to improve analgesia. In cancer patients, a rapid opioid escalation due to either worsening of the pain condition or the development of tolerance is a critical phase, as this condition is associated with a negative prognosis. The aim of this study was to assess the effects of adding a second opioid at low doses in patients with a poor analgesic benefit after dose escalation. Fourteen patients receiving strong opioids who had increased their dosage more than 100% in the last week unsuccessfully were randomly chosen to add a second opioid to the first using an initial equivalent dosage of 20% of the previous therapy. The dose of the second opioid was then changed according to the clinical situation to obtain an acceptable pain control with minimal adverse effects. Pain and symptoms associated with opioid therapy were assessed, and an opioid escalation index (OEI) was calculated at weekly intervals. OEI significantly decreased after adding the second opioid, and this trend was also maintained for the following weeks. Similarly, pain intensity was significantly improved and maintained at acceptable levels for the following weeks. The second opioid did not induce significant opioid-related adverse effects. The opioid combination was able to break opioid escalation in patients with pain syndromes with a poor response to the previous opioid. It allowed for a regaining of analgesia, regardless of the opioid combination used. These preliminary observations should be confirmed in further studies.


Cancer painOpioidsToleranceSwitching


In the clinical setting of cancer pain, a rapid development of tolerance may occur. The phenomenon is really less clear than in experimental setting, because there are other driving forces able to produce a decrease in the analgesic response [13, 17]. A rapid opioid escalation due to either worsening of the pain condition or development of tolerance is a negative prognostic factor for the development of adverse effects [2] as it can limit opioid responsiveness [1]. In recent years, opioid switching has been proposed for restoring the opioid response [4] and for limiting the development of tolerance. This is based on the different biochemical characteristics and receptor activity of opioids, presenting asymmetric tolerance among them. According to published reports, equianalgesic doses with a second opioid lower than the one being used may produce a better analgesia and a reduction of the intensity of adverse effects in about 80% of patients [4, 11].

Opioid combination in clinical practice has never been recommended for general use, and is considered a marker of poor clinical judgment or fear of the patient using higher doses of a single analgesic for a variety of nonmedical reasons. However, coadministration of low doses of different opioids may produce marked antinociceptive synergy with reduced central nervous system adverse effects [18]. Recent experimental data are in favor of a possible opioid combination to improve analgesia [8].

After using this approach in several patients, based on this experimental rationale, a prospective open-label study was planned with the aim of evaluating the possible advantages of adding a second opioid at low doses to limit opioid escalation.

Patients and methods

A sample of 29 consecutive cancer patients with difficult pain control requiring escalating doses of opioids was examined. Patients who developed severe adverse effects during the phase of opioid dose titration were excluded, and a traditional opioid or route switching was offered according to the department’s policy. Fourteen patients who had unsuccessfully increased their opioid dosage more than 100% in last week (pain level more than 4 on a numerical scale 0–10) were randomly chosen to receive a second opioid in addition to the first.

Institutional approval and informed consent were obtained. The characteristics of patients are presented in Table 1. The second opioid was added to the previous regimen starting with an initial equivalent dosage of 20% of the previous therapy using the following ratio: oral morphine-transdermal fentanyl 100:1 [5]; oral morphine-oral methadone 5:1 [14]. Different combination were used, and the study was partially randomized. Patients on transdermal fentanyl were semiswitched to oral morphine only (group 1), because no sufficient experience exists on switching to methadone. Patients on morphine were randomly assigned, based on patients’ initial letters, to be semiswitched to transdermal fentanyl or oral methadone (groups 2 and 3, respectively). These combinations were chosen as they are commonly employed for opioid switching at our institution. The dose of the second opioid was then changed according to the clinical situation to obtain a pain intensity level less than 4 in a numerical scale from 0 to 10 and an acceptable level of opioid-induced symptoms. Attempts to reduce the first opioid were made in the presence of optimal pain control (no pain at all). All patients were examined and treated at an acute Pain Relief and Palliative Care Unit and then followed by phone or during further admissions.
Table 1

Oral morphine added to transdermal fentanyl. Opioid doses are expressed as mean milligrams/day (95% CI). OEI opioid escalation index. Pain intensity (numerical scale 0–10) (95% CI). Significancy versus week 0, unless specified




Pain intensity

0 week (n=5)

3.7 (2.9–4.6)

20.7 (17–24)

6.7 (5.6–7.6)

1 week (n=5)

3.7 (2.9–4.6)

61 (43–78)

8.6 (6.2–11.1)


3 (2.2–3.7)


2 week (n=5)

3.7 (2.9–4.6)

43 (23–62)

6.1 (3.3–8.9)


3.4 (1.9–4.9)


3 week (n=4)

3.7 (2.7–4.7)

45 (23–66)

6.3 (3.1–9.5) 7.5 (21 days)


3.6 (2.8–4.5)


4 week (n=4)

3.8 (2.6–5)

28 (-23–79)

3.9 (-3.3–11.2) 6.4 (28 days)


2.8 (0.6–5)


5 week (n=2)

3.9 (2–5.7)

22 (14–30)

p=0.03 versus week 1

p=0.04 versus week 2

3.2 (2–4.3) 5.2 (35 days)


2.7 (1.2–4.2)


The use of other drugs, such as adjuvants, was considered as well as the general treatment commonly used in palliative care (laxatives, etc.) for controlling other symptoms that eventually occurred. The following data were recorded at admission and at weekly intervals for 5 weeks: Symptoms associated with opioid therapy or commonly present in advanced cancer patients, such as nausea and vomiting, drowsiness, confusion, xerostomy, and so on, using a scale from 0 to 3 (not at all, slight, a lot, awful). Symptoms were assessed by the patients, when possible. Symptoms presenting a median value of all the observations (2–3 a week) more than 1 were considered (an intensity of 2–3 is commonly considered distressing or requiring treatment).

The functional status (Karnofsky) and pain syndromes were considered on the basis of clinical history, anatomical site of primary tumor and distant metastases, physical examination, and radiological investigations, such as CT scan, MRI, and so on. Opioid escalation index in milligrams (OEI mg) was calculated as the weekly mean increase of opioid dosage in milligrams using the following formula: (OMD-OSD)/days [15], during the previous week before starting the protocol (week 0) and then at week intervals (week 1, week 2, etc.) where OSD is the starting dose of the week taken into consideration and OMD is the dose at the end of week. At week 3 and at following intervals, OEI was recalculated for the entire period of study (21 days, 28 days, etc.).

The paired Wilcoxon signed-rank test was used to compare the differences in pain intensity scores and OEI in the three groups of patients between time intervals. The paired samples Student’s t test was used to compare mean dose of opioids at time intervals. All p values were two-sided, and p values less than 0.05 were considered to indicate statistical significance.


Fourteen patients received an opioid combination after unsuccessful escalating opioid doses. The mean age of patients was 61.5 (range 44–73) years; nine patients were men. Almost all patients had a nociceptive-neuropathic mechanism. The principal primary diagnoses were lung (four patients), breast (three patients), and colon cancer (two patients); five patients had other diagnoses. Five patients on transdermal fentanyl escalating doses received oral morphine, five on oral morphine received transdermal fentanyl, and four taking oral morphine received oral methadone. All patients had a Karnofsky status at admission of 40 or more. Data were available for 5 weeks in seven patients. Four patients died, and data from the others were not complete or available. Data regarding these patients are presented in Tables 1, 2, and 3, respectively. OEI decreased after adding the second opioid, independently of the drug combination used (about 30% of OEI calculated the previous week with a single opioid), after a week. In the following weeks, no significant increases in OEI were found.
Table 2

Transdermal fentanyl added to oral morphine. Opioid doses are expressed as milligrams (95% CI). OEI opioid excalation index. Pain intensity (numerical scale 0–10) (95% CI). Significancy versus week 0, unless specified




Pain intensity

Week 0 (n=5)

396 (191–600)

35.3 (15–55)

6.4 (5–7.8)

Week 1 (n=5)

396 (191–600)

0.96 (0.5–1.3)

13.6 (7.8–19.5)




Week 2 (n=5)

344 (165–522)

0.96 (0.5–1.3)

3 (1.7–4.2)


Week 3 (n=4)

305 (94–515)

1.35 (0.4–2.2)

6.37 (-0.4–13.1) 4.5 (21 days)

4.2 (2.7–5.7)

Week 4 (n=4)

305 (94–515)

1.50 (0.9–2)

p=0.0005 v week 1

p=0.0005 v week 2

2.12 (-4.6–8.9) 3.9 (28 days)

3.7 (2.9–4.5)

Week 5 (n=3)

293 (19–566)

1.80 (0.3–3.3)

p=0.04 v week 1

p=0.04 v week 2

p=0.0005 v week 3

8.50 (8.5–8.5) 4.18 (35 days)

3.3 (1.9–4.7)

Table 3

Oral methadone added to oral morphine. Opioid doses are expressed as milligrams (95% CI). OEI opioid escalation index. Pain intensity (numerical scale 0–10) (95%CI). Significancy versus week 0, unless specified




Pain intensity

Week 0 (n=4)

320 (11–628)

25.6 (5–46)

6.2 (5.4–7)

Week 1 (n=4)

320 (11–628)

12.2 (3.9–20.5)

8.7 (2.8–14.5)

3.5 (2.6–4.4)

Week 2 (n=3)

366 (−150 to 883)

6.6 (−7 to 21)

4.7 (−5.4 to 14.9)

2.6 (1.2–4.1)

Week 3 (n=3)

366 (−150 to 883)

15 (2.5–27)

p=0.04 versus week 2

10.6 (1.8–19.4) 8.3 (21 days)

3 (0.5–5.48)

Week 4 (n=3)

366 (−150 to 833)

16.6 (9.5–23)

11.8 (6.8–16.9) 9.2 (28 days)

3.3 (1.9–4.7)

Week 5 (n=3)

366 (−150 to 883)

13.3 (−15 to 42)

7.3 (−0.7 to 15.4) 9.2 (35 days)

3 (0.5–5.9)

This trend was also maintained for the following weeks with global long-term OEI ranging from 3.9 to 8.3. The result differences were significant when adding oral morphine to transdermal fentanyl, probably due to the limited number of patients in the other groups. Similarly, pain intensity was significantly improved and maintained at acceptable levels for the following weeks.

In group 1, oral morphine doses tended to decrease, and transdermal fentanyl doses tended to increase in group 2 in the last weeks of study. In group 3, oral methadone doses did not presented relevant changes over time. The addition of the second opioid did not significantly increase the intensity of opioid-related adverse effects in comparison with the previous treatment.


The rapid need to escalate opioid doses for the presence of apparent pain intensity increase is challenging for physicians and represents a critical phase for patients who have poor pain control despite receiving progressively increasing doses of opioids [12]. The incidence of neuroexcitatory adverse effects, including hyperalgesia, allodynia, myoclonus, and seizures in patients administered large doses of systemic morphine or its structural analogues to relief uncontrolled cancer pain has been increasing with a more liberal use of opioids [3, 9]. This could result in an exacerbation rather than an attenuation of excitatory behaviors. Clinical reports suggest that opioids, intended to abolish pain, can unexpectedly produce abnormally heightened pain sensations, which are characterized by a lowering of the pain threshold (hyperalgesia), and pain elicited by normally innocuous stimulation (allodynia) [10, 19]. A rapid opioid escalation is considered a negative predictive factor, as more adverse effects are likely to occur [13]. In some cases, these adverse effects are associated with clinical manifestations of neuronal excitability, considered an expression of hyperalgesia.

Opioid switching has been reported to restore the opioid response, especially in patients with relevant adverse effects that limit further dose escalation. This approach is commonly reserved for patients presenting substantial opioid toxicity and in whom it is necessary to remove the cause. However, opioid switching may pose problems due to asymmetric tolerance among opioids, and the response is not easily preventable [1]. Moreover, patients who do not present relevant adverse effects could be not strictly indicated for opioid switching, although they should be considered at risk due to rapid escalating opioid doses. However, as opioid escalation is a critical phase, important decisions must be made, for example, whether to continue opioid escalation and face the risk, or whether to definitively switch to another opioid or route of administration.

Anecdotally, multiple opioids are often simultaneously administered for different reasons, although it should be considered a nonsense approach. However, recent experimental data have offered a rationale for using different opioids in an attempt to improve analgesia or to limit the development of tolerance in difficult conditions, such as the need for rapid escalation of pain relief. An integration of basic knowledge and clinical aspects is useful to formulate hypotheses, to produce data useful for experimental studies, or to apply specific alternative approaches in daily activities when such difficult conditions occur [13]. Activation of different opioid receptors by coadministration of low doses of different opioids has been shown to induce a synergic analgesic effect [18].Other than the obvious disparity in effecting different receptor subgroups and specificity, as reported in an explanation of the benefits of opioid switching [11], there is a significant variability among opioid drugs in inducing endocytosis of opioid receptors. This is an independent functional property that distinguishes clinically important analgesic drugs, such as morphine, methadone, and fentanyl.

Recently, the regulation of opioid receptors by endocytosis has been hypothesized to have protective functions in reducing the development of tolerance. Agonist activity and receptor endocytosis have opposing effects on receptor-mediated signaling, and the final result is a function of both processes (named RAVE). Morphine, in comparison with other opioids, has an high activity-endocytosis ratio and has an enhanced propensity to prolonging signals with prolonged drug exposure. Molecular events, such as desensitization and endocytosis, would reduce this response. It has been experimentally demonstrated that endocytosis-promoting agonists may facilitate morphine-induced receptor endocytosis, reducing the compensatory adaptive cellular changes that lead to upregulation of the cAMP pathway [8]. Thus, a combination of opioids with different characteristics may reciprocally alter their RAVEs, thus reducing the potential for the development of tolerance. On the other hand, morphine treatment can produce adaptational changes that can attenuate high-efficacy agonist-mediated desensitization and internalization of G-protein-coupled receptors [6].

The administration of small doses of a second opioid in patients with an unfavorable response during escalation with the prior opioid has been found effective in both directions (morphine to fentanyl and vice versa), as well as with methadone. Dose increments used for opioid titration commonly range between 30–50%. We used lower increases of equivalent doses of the second opioid, less than 20%, to better demonstrate a clinical benefit.

Physicians should be aware that there are no clear guidelines about the equivalence between morphine, methadone, and fentanyl, and the ratios reported in the literature are presumptive. An acceptable conversion ratio between oral morphine and transdermal fentanyl is 100:1 [5]. The conversion ratio is even more complicated between oral morphine and oral methadone. Although a methadone-morphine ratio of 1:5 is not universally accepted, this is the common approach used at our institution, at least for initiating methadone therapy, based on the need to prime the system to reach a significant plasma concentration with a drug having a slow onset that requires several days to reach the steady state [7, 14].

Global OEI calculated in the following weeks after starting the treatment was maintained at levels considered as acceptable—about five on average. Of interest, the relatively low doses of the second opioid administered did not produce adverse effects of significant intensity while improving the analgesia.

Obvious limitations of this study are represented by the lack of control, as in most studies of cancer patients with uncontrolled pain on high doses of opioids. Given the preliminary data of this report, no comparison is possible with previous data in the literature. Nevertheless, suggestions from experimental studies seem to find support from the data presented in this open-label study performed in a raw clinical setting. A decrease and then a maintenance of similar levels of OEI in the following weeks was found to be of benefit for advanced cancer patients requiring increasing doses of opioids. According to these observations, opioid “semiswitching” could be a new therapeutic option in patients requiring escalating doses of opioids. A better knowledge of molecular mechanisms that modulate the opioid response may offer alternative treatments, which should be tested in the clinical setting. These data, supported by recent molecular investigation, should be considered as preliminary and should be confirmed in a larger number of patients in controlled studies.

Copyright information

© Springer-Verlag 2004