Annals of Oncology 16 (Supplement 4): iv132 – iv135, 2005
doi:10.1093/annonc/mdi922
World Health Organization guidelines for cancer pain:
a reappraisal
S. Mercadante1 & F. Fulfaro2
1
Anesthesia & Intensive Care Unit and Pain Relief and Palliative Care Unit, La Maddalena Clinic for Cancer, Palermo, Italy; 2Department of Oncology,
University of Palermo, Italy
Pain is a prevalent symptom experienced by at least 30% of
patients undergoing an oncological treatment for metastatic
disease and by more than 70% of advanced cancer patients
[1]. In 1986 the World Health Organization [2] published a set
of guidelines for cancer pain management based on the three-
step analgesic ladder [2]. The main aim of WHO guideliens
was to legitimize the prescribing of strong opioids, arising
from evidence of poor management of cancer pain, due to
reluctance of health care professionals, institutions, and gov-
ernment to use opioids because of fears of addition, tolerance
and illegal abuse.
Its application is reported to achieve satisfactory pain relief
in up to 90% of patients with cancer pain. Despite the large
experience proving the feasibility and efficacy [3 –5], in the
years of evidence-based medicine, the three-step ladder has
been criticized for the lack of robust data supporting this
approach. Studies validating the WHO analgesic ladder had
methodologic limitations including the circumstances during
which assessment were made, small sample size, retrospective
analyses, high rate of exclusions and dropout, inadequate
follow-up, and a lack comparison with levels of analgesia
before the introduction of the analgesic ladder [6].
Thus, different problems are unresolved due a lack of con-
trolled studies on this subject. These problems include, for
example, a better definition of the role of NSAIDs, the pro-
longed use of NSAIDs in cancer pain, and the utility of step 2.
Moreover, the indications for using different strong opioids
and alternate routes of administration to improve pain relief in
difficult pain situations are not well established. The pro-
portion of patients who do not benefit from these treatments
remain unclear, and how the opioid response may be improved
with the use of adjuvants is also uncertain. Finally, different
countries apply the WHO ladder approach differently depend-
ing on the availability of drugs.
*Correspondence to: Dr Sebastiano Mercadante, Anesthesia
& Intensive Care, La Maddalena Cancer Center Palermo, Via S Lorenzo
Colli 312 90146 Palermo, Italy. Tel: +39-09-16808111;
E-mail: terapiadeldolore@la-maddalena.it
q 2005 European Society for Medical Oncology
First step
The first step in the WHO analgesic ladder involves the use of
a nonopioid with or without an adjuvant analgesic. Studies of
various non-opioid analgesics, combined or not with opioids
were recently assessed. Heterogeneity of study designs and
outcomes, and short length of study precluded meta-analyses.
From data available non-opioid analgesics appeared more
effective than placebo. No superior safety or efficacy was
demonstrated for one of this class of drugs, and slight advan-
tages where found in trials of combinations of an non-opioid
analgesic with an opioid, compared with either single entity
[7, 8]. Some studies suggest a different place of non-opioid
analgesics, administered in patients already on opioids, given
first, to reinforce analgesia in patients with difficult pain con-
trol or who tend to develop adverse effects with increasing
doses of opioids [9]. The conclusions of this study are intri-
guing, because the reluctance of North American physicians to
use NSAIDs and conversely the extensive, and sometime
exaggerated, use in European countries may find a compro-
mize based on the data pointed out in this work. Patients could
be started on opioids alone and then added non-opioid analge-
sics, for example in conditions where pain is particularly sen-
sible to this class of drugs, or to reduce the tendency to further
opioid escalation, when adverse effects tend to develop.
NSAIDs are considered be effective in some specific cancer
pain syndromes, such as bony metastases, although data to
support this do not exist. Recent studies have shown that
NSAIDs are equally effective in both visceral and somatic
pain syndromes [10].
Although adverse effects occurred infrequently in previous
large experiences, these studies did not specifically assess this
issue. The development of ulcer or renal toxicity might not be
apparent with short-term dosing [7], and the specific long-
term safety profile has never been established in a randomized
studies.
Second step
The role of so-called ‘weak opioids’ in the treatment of mod-
erate cancer pain has been questioned, and it has been specu-
lated that this step could be by-passed. A meta-analysis [11]

reported that no significant differences in pain relief were
noted when the use of non-opioids alone was compared to
non-opioids plus opioids for moderate pain. However, these
results were based on single-dose studies or studies involving
a small number of patients, and a regular clinical use would
be more effective than would be predicted on data involving
single-dose administration. Previous studies underlined the
role of opioids for moderate pain (namely codeine, dextropro-
poxyphene, and codeine), in comparison to morphine in terms
of efficacy and adverse effects. In opioid-naive-patients, a
more favorable balance between side effects and analgesia
occurred when step 2 opioids were administered compared to
low doses of morphine used to omit Step 2 of the analgesia
ladder [12, 13].
On the other hand, other studies assessed the use of strong
opioids in opioid naive patients, that is skipping the second
step drugs. Doses of 25 mg/h of fentanyl have been used suc-
cessfully [14, 15], although it means to administer equivalent
doses of 60 mg/day of oral morphine, which is a considerable
dosage for opioid-naive patients, having a high risk to produce
adverse effects and reduce patient’s compliance. This obser-
vation was confirmed in a study where transdermal fentanyl
used at doses of 25 mg/h, equivalent to 60 mg of oral mor-
phine, was better tolerated in codeine using patients rather
than in opioid-naive patients [16].
In a more recent study the sequential treatment proposed by
WHO was compared with a direct administration of strong
opioid as a first step. Treatments appeared equally effective,
although the group treated with opioids first had a better pain
relief and patients’ satisfaction, and less number of therapeutic
interventions. However, nausea was more frequently reported.
Only 50% of patients in control group needed strong opioids
in doses similar to those used in patients who received strong
opioids first [17]. Unfortunately initial doses of strong opioids
were not mentioned, making evaluation of data difficult. The
choice of the initial doses makes the difference in terms of
compliance, efficacy, and tolerability. Doses of morphine, or
other strong opioids should as flexible and low as those in the
range commonly used of opioids for moderate pain.
Personal experience on the use of low dose morphine
Morphine used at very low doses in opioid-naive patients may
offer different advantages, including a greater tolerability
while providing analgesia. The rationale was to replace
opioids for moderate pain with morphine used in doses equiv-
alent to the range of doses commonly prescribed at flexible
doses in clinical practice. This approach has been prospec-
tively evaluated in a sample of consecutive advanced cancer
patients, including very old patients and preliminary data are
presented. Forty-five opioid-naive patients with a pain inten-
sity in the range of 4 –7 on a numerical scale from 0 to 10,
received initial doses of 12 mg divided in 5–6 doses per day
(patients aged m > 70 years received 10 mg/day). Patients
receiving this approach were titrated and achieved stabiliz-
ation in a few days, and doses were maintained relatively con-
stant one month after starting the therapy, with an mean final
iv133
doses of 40 mg, and a calculated escalation index (OEI) <5.
OEI is the mean increase of opioid dosage percentage using
the following formula:
[(last opioid dose –starting dose)/starting dose] /
days100).
The value reported is considered devoid of risk in previous
experiences with chronic opioid dosing, that is a limited need
to escalate the dose [18]. Of interest tolerability was excellent
with a low number of patients who discontinued morphine and
required alternative drugs (4 patients, <10%). This approach
has been proved feasible, effective and well tolerated, and not
associated with abnormal requirements of increasing doses,
although this finding should be confirmed in controlled
studies. Of course, patients with severe pain are already candi-
dates to receive strong opioids at a consistent dose of about
60 mg/day of oral morphine equivalents, according to the
WHO guidelines, so that this approach should be reserved to
patients with moderate pain, potentially requiring step drugs.
Third step
Morphine is the most frequently used opioid in cancer pain
management. Although morphine remains a cornerstone for
the management of cancer pain, due to the largest experience
existing among physicians and widely availability in a variety
of formulation, no clear data exist about the superiority of one
opioid over another [19]. Individualization of therapy has
been emphasized to minimize the side effects and to improve
the opioid response. A substantial minority of patients treated
with oral morphine (10–30%) do not have a successful out-
come because of excessive adverse effects, inadequate analge-
sia, or a combination of both adverse effects along with
inadequate analgesia [20]. It is now recognized that individual
patients vary greatly in their response to different opioids.
Patients who obtain poor analgesic efficacy or tolerability with
one opioid will frequently tolerate another opioid. A shift
from one opioid to another is recommended when the adverse
effect/analgesic equation is skewed towards the side effect
component, despite an aggressive adjuvant treatment. Opioid
rotation has been shown to be useful in opening the thera-
peutic window and in establishing a more advantageous
analgesia/toxicity relationship. By substituting opioids and
using lower doses than expected (according to equivalency
conversion tables), it is possible in most cases to not only
reduce or relieve the symptoms of opioid toxicity and to man-
age patients who are highly tolerant to previously used
opioids, but also to improve analgesia and thus the opioid
responsiveness. This strategy uses much lower doses of
alternative opioids in patients who are unresponsive to high
doses of morphine. Drugs commonly used for opioid rotation
include hydromorphone, oxycodone, and methadone [21]. The
biological basis for the individual variability in sensitivity to
opioids is multifactorial and has been described elsewhere,
although some aspects remain unclear [22]. Although the con-
version ratio between opioids in such circumstances remains
unpredictable, morphine, oxycodone, and hydromorphone
iv134
seem to be more manageable in the clinical context. Of inter-
est, differently from other opioids, methadone, which is
chemically distinct from morphine, oxycodone, or hydromor-
phone, shows some particularities which make this drug
unique, particularly looking for the conversion ratio to choose
when switching. Individual response varies remarkably from
opioid to opioid, due to asymmetric tolerance, different effica-
cies, pharmacokinetic profiles, and extra-opioid effects, like
an anti-NMDA effect, although the clinical ‘weight’ remains
to be established in cancer patients. The correct conversion
ratio from morphine to methadone is quite complex and par-
ticularly debated in literature. It seems that the previous dose
of morphine is determinant in the choice of the following dose
of methadone in a proportional way [23, 24]. For example
patients taking less than 90 mg of oral morphine, should take 14
of this dose as methadone, patients taking 90 –300 mg should
take 1/8 of this dose as methadone, and patients taking more
than 300 mg should use higher ratio (1:12 or more). However,
most of these studies focused on the concept of equi-analgesia,
that is patients with adequate pain control presenting adverse
effects who were switched and titrated to reach the same level
of analgesia. Unfortunately, these concepts are hard to apply
in the clinical settings where patients suffer adverse effects
from opioids with poor analgesia in most circumstances, a
critical condition which requires immediate intervention. The
use of these ratios, proportional to the previous dose of mor-
phine, may result in underdosage of methadone in the first
days and a slower recovery form the clinical condition for
which switching was indicated. Methadone is characterized by
a slow onset, due to its large volume of distribution, so that it
requires a priming dose to develop an effect to be maintained
with subsequent doses. For these reasons a dose of methadone
of 1/5 of the previous morphine dose, which could appear
high in some circumstances where patients are taking hun-
dreds milligram of morphine, may represent an effective load-
ing dose to be corrected in the following days. In fact,
problems related with methadone accumulation occur not
immediately, but after 2–3 days. Rotation at risk includes
patients on high doses of morphine or those that are highly
tolerant, or taking morphine for a long period. Patients on
very low doses of opioids mainly do not present relevant pro-
blems and can be safely converted using the same methadone-
morphine ratio of 1/5 on an outpatient basis [25].
Adjuvants
Adjuvants comprehend a wide range of drugs, commonly pre-
scribed to improve opioid analgesia. Antidepressants and anti-
epileptics are most frequently administered as co-analgesics in
the presence of neuropathic pain, in combination with opioids.
Despite the relative efficacy demonstrated by this class of
drugs in chronic non-cancer conditions, evidence of the benefit
of such combination in cancer pain management is weak.
Gabapentin proved to be effective in reducing the mean global
score and dysesthesia, but not allodynia, compared to placebo
in people with neuropathic cancer pain that is not adequately
controlled by systemic opioids [26]. However, magnitude and
duration of benefit for patients remains questionable in clinical
practice. Similarly, amitryptiline added to opioids was able to
reduce the worst pain only, while producing more central
adverse effects, in comparison to patients receiving opioids
alone [27].
Corticosteroids are widely used as co-analgesic for different
pain syndromes however further randomized studies are
warrant to better define their role [28].
The role of adjuvants remains poorly defined, at least in
advanced cancer patients receiving opioids, and requires
further data from well designed and robust studies. In particu-
lar, the timing of starting these drugs, potentially they should
be started at the first step when non-opioid analgesics,
considered poorly effective for neuropathic pain are given
first.
Conclusion
The WHO method remains of paramount importance and
should continued to be encouraged when approaching
advanced cancer patients with pain, for the high chances of
success, ranging between 70 and 90%. Despite the lack of
strong evidence to produce unbiased estimates of the pro-
portion of patients in whom the ladder produces satisfactory
results and the fact that no controlled studies with other
methods have been conducted to assess its validity, there is
the risk to underestimate the educational meaning of this
simple approach. Although these guidelines can be
implemented, currently the correct use of the WHO method
can lead to adequate long-term pain control in most patients
with advanced cancer disease.
References
1. Cleeland CS, Gonin R, Hatfield AK et al. Pain and its treatment in out-
patients with metastatic cancer. N Engl J Med 1994; 330: 592–596.
2. World Health Organization. Cancer pain relief and palliative care.
Geneva: WHO 1990.
3. Ventafridda V, Tamburini M, Caraceni A, De Conno F, Naldi F.
A validation study of the WHO method for cancer pain relief. Cancer
1987; 59: 850–856.
4. Zech DFJ, Grond S, Lynch J, Hertel D, Lehmann KA. Validation
of World Health Organization guidelines for cancer pain relief: a
10-year prospective study. Pain 1995; 63: 65 –76.
5. Mercadante S. Pain treatment and outcome in advanced cancer
patients followed at home. Cancer 1999; 85: 1849–1858.
6. Jadad AR, Browman GP. The WHO analgesic ladder for cancer pain
management. J Amer Med Assoc 1995; 274: 1870– 1873.
7. McNicol E, Strassels S, Goudas L, Lau J, Carr D. Nonsteroidal anti-
inflammatory drugs, alone or combined with opioids, for cancer pain:
a systematic review. J Clin Oncol. 2004; 22: 1975– 1992.
8. McNicol E, Strassels S, Goudas L, Lau J, Carr D. NSAIDS or parace-
tamol, alone or combined with opioids, for cancer pain. Cochrane
Database Syst Rev 2005 Jan 25 (1): CD005180.
9. Mercadante S, Fulfaro F, Casaccio A. A randomised controlled study
on the use of anti-inflammatory drugs in patients with cancer on
morphine therapy: effect on dose-escalation and pharmacoeconomic
analysis. Eur J Cancer 2002; 38: 1358–1363.

10. Mercadante S, Casuccio A, Agnello A et al. Analgesic effects of non-
steroidal anti-inflammatory drugs in cancer pain due to somatic or
visceral mechanism. J Pain Symptom Manage 1999; 17: 351–356.
11. Eisenberg E, Berkey C, Carr DB, Mosteller F, Chalmers C. Efficacy
and safety of non steroidal antinflammatory drugs for cancer pain:
a meta-analysis. J Clin Oncol 1994; 12: 2756–2765.
12. Mercadante S, Salvaggio L, Dardanoni G, Agnello A, Garofalo S.
Dextropropoxyphene versus morphine in opioid-naive cancer patients
with pain. J Pain Symptom Manage 1998; 15: 76–81.
13. Grond S, Radbruch L, Meuser T et al. High-dose tramadol in com-
parison to low-dose morphine for cancer pain relief. J. Pain Symptom
Manage 1999; 18: 174–179.
14. Mystakidou K. Pain management of cancer patients with transdermal
fentanyl: a study of 1828 step I, II & III transfers. J Pain 2004; 5:
119 –132.
15. Mystakidou K, Tsilika E, Kouloulias V et al. Long-term cancer pain
management in morphine pre-treated and opioid naive patients with
transdermal fentanyl. Int J Cancer 2003; 107: 486–492.
16. Vielvoye-Kerlmeer A, Mattern C, Uitendaal M. Transdermal fentanyl
in opioid-naive cancer patients: an open trial using transdermal fenta-
nyl for the treatment of chronic cancer pain in opioid-naive patients
and a group using codeine. J Pain Symptom Manage 2000; 19:
185 –192.
17. Marinangeli F, Ciccozzi A, Leonardis M et al. Use of strong opioids
in advanced cancer pain: a randomized trial. J Pain Symptom Manage
2004; 27: 409– 416.
18. Mercadante S, Dardanoni G, Salvaggio L et al. Monitoring of opioid
therapy in advanced cancer pain patients. J Pain Symptom Manage
1997; 13: 204– 212.
iv135
19. Hanks GW, De Conno F, Cherny N et al. Morphine and alternative
opioids in cancer pain: the EAPC recommendations. Br J Cancer
2001; 84: 587–593.
20. Cherny NJ, Chang V, Frager G et al. Opioid pharmacotherapy in the
management of cancer pain. Cancer 1995; 76: 1288–1293.
21. Portenoy RK, Mercadante S. World Health Organization guidelines:
problem areas in cancer pain management. Cancer Control 1999; 6:
191–197.
22. Bruera E, Pereira J, Watanabe S et al. Opioid rotation in patients with
cancer pain. A retrospective comparison of dose ratios between
methadone, hydromorphone, and morphine. Cancer 1996; 78:
852–857.
23. Ripamonti C, Groff L, Brunelli C et al. Switching from morphine to
oral metadone in treating cancer pain: what is the equianalgesic dose
ratio? J Clin Oncol 1998; 16: 3216–3221.
24. Mercadante S, Casuccio A, Calderone L. Rapid switching from mor-
phine to metadone in cancer patients with poor response to morphine.
J Clin Oncol 1999; 17: 3307–3312.
25. Mercadante S. Opioid rotation in cancer pain. rationale and clinical
aspects. Cancer 1999; 86: 1856–1866.
26. Caraceni A, Zecca E, Bonezzi C et al. Gabapentin for neuropathic
cancer pain: a randomized controlled trial from the Gabapentin
Cancer Pain Study Group. J Clin Oncol 2004; 22: 2909–2917.
27. Mercadante S, Arcuri E, Tirelli W, Villari P, Casuccio A. Amitripty-
line in neuropathic cancer pain in patients on morphine therapy: a ran-
domized placebo-controlled, double-blind crossover study. Tumori
2002; 88: 239–242.
28. Mercadante S, Fulfaro F, Casuccio A. The use of corticosteroids
in home palliative care. Supportive Care Cancer 2001; 9: 386–389.