Int J Clin Oncol DOI 10.1007/s10147-014-0708-0
ORIGINAL ARTICLE
Opioid needs of terminally ill patients with gynecologic malignancies Fumi Utsumi • Hiroaki Kajiyama • Jun Sakata • Makiko Higashi • Kaoru Niimi • Ryuichiro Sekiya • Hiroko Mitsui • Shiro Suzuki • Tomokazu Umezu Mika Mizuno • Eiko Yamamoto • Kiyosumi Shibata • Fumitaka Kikkawa
•
Received: 28 January 2014 / Accepted: 7 May 2014 Ó Japan Society of Clinical Oncology 2014
Abstract Background Little is known about patterns and predictive factors regarding opioid use for terminally ill patients with gynecologic malignancies. The aim of this study was to elucidate predictors affecting opioid requirements of endof-life patients with gynecologic malignancies. Methods A retrospective study was carried out on patients with gynecological malignancies admitted to our institute and died during the years 2002 to 2012. The association between maximum opioid dose and factors affecting opioid requirements were examined. Data extracted from medical records included age, site of primary cancer, maximum total dose of opioids prescribed over 24 h, the site of recurrence and metastasis, procedures performed during the hospital stay, total number of chemotherapy courses and overall survival. Results The study identified 189 patients. Most patients had ovarian cancer (42.3 %) followed by cervical cancer (28.0 %) and then corpus malignancy (27.0 %). Opioid requirements decreased with increasing age, especially from the 50s onward. This was particularly marked in cervical cancer patients. In addition, pelvic metastasis was associated with the maximum dose of opioids and the average opioid use was highest in patients with cervical cancer. Conclusion Young age and pelvic invasion were significant predictive factors regarding opioid requirements.
F. Utsumi H. Kajiyama (&) J. Sakata M. Higashi K. Niimi R. Sekiya H. Mitsui S. Suzuki T. Umezu M. Mizuno E. Yamamoto K. Shibata F. Kikkawa Department of Obstetrics and Gynecology, Nagoya University Graduate School of Medicine, Tsuruma-cho 65, Showa-ku, Nagoya 466-8550, Japan e-mail: [email protected]
Additionally, cervical cancer patients may require more opioids among those with gynecologic malignancies. Keywords Cancer-related pain Gynecologic malignancy Opioid intake Age Pelvic invasion
Introduction The incidence of and mortality due to cancer have steadily increased and it has been estimated that approximately 28,000 women died of gynecologic malignancies in 2013 in the United States [1]. It is well-known that cancer patients experience many symptoms that may cause suffering and have a negative impact on their quality of life. Pain is one of the most common and distressing symptoms at the end of life. Incidence of pain in patients with advanced cancer is 58–90 % [2–7]. As part of worldwide efforts to improve the quality of pain control, multiple clinical guidelines have been published on the management of cancer pain [8–11]. The World Health Organization (WHO) designed a 3-step treatment, according to patients’ needs, from non-opioid analgesics to weak opioids and then strong opioids, and it is generally accepted that for each individual the dose is titrated until a sufficient effect is provided. However, despite the fact that opioids have been widely used, pain in terminal cancer patients is still undertreated. Furthermore, predictors in patients who need high doses of opioids are not clear. Concerning this point several studies have addressed the different opioid requirements among primary cancer sites [12, 13]. In gynecologic malignancies women with uterus cancer have been reported to need more opioids than patients with many other cancers [12]. In addition, women with ovarian cancer live longer due to
123
Int J Clin Oncol
tumor reductive surgery and many chemotherapeutic agents [14, 15]. With improved survival the terminal period with pain and symptoms becomes longer. Furthermore, patients who have previously undergone chemotherapy have been reported to need higher opioid doses [16]. Therefore, palliative care, especially pain management, is more important for patients with gynecologic malignancies. Nevertheless, few studies have assessed the patterns of pain and opioid needs in gynecologic cancer patients and influence on quality of life (QOL). The objective of this study was to elucidate the predictive factors of a high opioid requirement in order to improve cancer pain management and the QOL of patients near the end of life.
Patients and methods Study subjects were patients with gynecologic malignancies who died between 2002 and 2012 at our hospital. Using an institutional database we identified all patients with gynecologic malignancy diagnoses who were admitted to our hospital during the study period; the study identified 229 patients. The patients who had double cancer or unknown cancer, or whose date of death was unclear, were excluded. Consequently, a total of 189 patients were eligible. Data extracted from medical records were age, the site of the primary cancer, maximum total dose of opioids prescribed over 24 h, site of recurrence and metastasis, symptoms, procedures performed during the hospital stay which may affect the opioid dose, including ileus, carcinomatous ascites (it was defined as accumulation of ascites requiring paracentesis to palliate their symptoms), blood infusion, the total number of chemotherapy courses and overall survival (OS). Pelvic invasion was defined as the presence of a deeply invasive tumor in the pelvic organs and/or retroperitoneal cavity, excluding superficial peritoneal dissemination. Maximum opioid use was calculated as the oral morphine intake. The calculation formula to equate each equianalgesic dose to oral morphine based on 24 h opioid requirements was as follows. We used these calculated doses as oral morphine equivalents (OME). OME = 2 9 parenteral morphine, OME = 1.5 9 modified-release oxycodone, OME = 30 9 transdermal fentanyl, OME = 100 9 parenteral fentanyl, and OME = 1.5 9 morphine suppository. Statistical analysis All analyses were performed using SPSS. The Student’s t test was used to compare the age and number of chemotherapy courses between two groups. Frequency analysis was performed with a chi square test. The Mann–
123
Whitney U test was used to compare the mean OME between two groups, Kruskal–Wallis one-way analysis of variance was used to compare the mean OME among more than three groups, and OS was compared using Kaplan– Meier statistics and the log-rank test. Regression analysis was used to establish the relationship between the mean OME and age. All p values were two-sided, and p values \0.05 were considered to indicate significance.
Results The survey included 189 patients. The median patient age at the time of last admission to the hospital was 58.3 years. The most common diagnosis was ovarian cancer (42.3 %), followed by cervical cancer (28.0 %), corpus malignancy (27.0 %), and vaginal/vulvar cancer (2 %). First, we compared patients whose OME was under 40 mg/day (low dose group) with those whose OME was 40 mg/day (high dose group). We set the standard at 40 mg/day because the median OME of those who needed opioids in this study was 36 mg/day. Additionally, in general we need to increase the dose at least once or twice to reach the 40 mg/ day threshold. Therefore, patients who were placed in the low dose group were the ones whose pain could be relieved with zero or only one increase of opioid dose, and those in the high dose group needed more increases of opioid dose. Patients’ characteristics and the results are shown in Table 1. The median age of patients in the low dose group was 60.2, compared to 56.2 for those in the high dose group (p = 0.006). There were significant differences in the frequency of pelvic invasion (p = 0.006) and carcinomatous ascites (p = 0.032) between the two groups. The distribution of other clinicopathologic factors, including primary cancer, pulmonary metastasis, bone metastasis, intestinal obstruction, blood infusion, and the chemotherapy course, did not significantly differ between the two OME groups. The relationship between OME and the site of the primary cancer is shown in Table 2. Although the average OME was higher in patients with cervical cancer, it was not significant (p = 0.54). Next, to evaluate the influence of age on opioid requirements in more detail we divided patients into five groups based on age. The frequency distribution of age is shown in Table 3. This also shows a progressive decline in OME with age. In Fig. 1 this change in OME with the mean age is shown as an exponential curve. Regression analysis confirmed a significant decrease in OME with increasing age in the total patients, and in patients with ovarian and cervical cancer (p \ 0.001, 0.002, and 0.041, respectively). There was no significance in patients with endometrial cancer (p = 0.14). Selecting the four clinicopathologic indicators with a significant or marginal-significant difference, as shown in Table 1, we
Int J Clin Oncol Table 1 Patient characteristics Characteristic
Total
N
%
Table 2 Mean and median OME of each primary tumor site Low dose group (0–39 mg/ day)
High dose group ([40 mg/ day)
N
%
N
%
60.3
75
39.7
Total
189
114
Age (median)
58.3
60.2
56.2
Mean ± SE
0.006
Ovary
80
42.3
52
45.6
28
37.3
Cervix
53
28.0
30
26.3
23
30.7
Corpus
51
27.0
30
26.3
21
28.0
Vagina/vulva
5
2.6
2
1.8
3
4.0
0.583
121
64
69
60.5
52
69.3
68
36
45
39.5
23
30.7
Yes
100
52.9
51
44.7
49
65.3
No
89
47.1
63
55.3
26
33.7
74
39.2
45
39.5
29
38.7
No
115
60.8
69
60.5
46
61.3
Corpus
84.6 ± 23.4
30.0
Vagina/vulva
54.8 ± 15.6 170.1 ± 58.7
31.2 24
OME oral morphine equivalent, SE standard error
0.006
Yes
13
6.9
5
4.4
8
10.7
No
176
93.1
109
95.6
67
89.3
0.911
Bone metastasis 0.095
Carcinomatous ascites Yes
28
14.8
22
19.3
6
8.0
No
161
85.2
92
80.7
69
92.0
0.032
Intestinal obstruction 19
10.0
8
7.0
11
14.7
170
90.0
106
93.0
64
85.3
Yes
19
10.0
9
7.9
10
13.3
No
170
90.0
105
92.1
65
86.7
13.2
12.4
14.3
0.32
No. of chemotherapies (mean courses)
20.0 30.0
Discussion
Yes
No Blood infusion
76.0 ± 22.1 405.2 ± 203.3
0.217
Pulmonary metastasis
Yes
Ovary
OS among the 189 patients is shown in Fig. 3. There was no significant difference in OS between the two OME groups (p = 0.86).
Recurrence No Pelvis invasion
Median
Cervix
Total
Primary cancer
Yes
OME (mg/day)
p value
0.087
0.224
compared the mean OME between the presence and absence of pelvic invasion, bone metastasis, intestinal obstruction and carcinomatous ascites using the Mann– Whitney U test (Fig. 2). As we expected, patients with pelvic invasion required higher doses of OME than those without this factor (p \ 0.01). Additionally, consistent with results of the chi square test, patients without ascites accumulation also needed higher opioid doses than those who required paracentesis for their carcinomatous ascites (p = 0.017). Bone metastasis and intestinal obstruction were not significantly correlated with OME (p = 0.17 and 0.435, respectively). Furthermore, the impact of OME on
The incidence and prevalence of cancer has increased with an aging population and improving cancer treatment outcomes. It has become much more common for people to have a prolonged period of cancer and disability prior to death. Pain is one of the most common and distressing symptoms of cancer patients despite growing knowledge of pain management and the use of various opioids. Although studies addressing palliative care are becoming more prevalent, literature regarding gynecologic oncology is very limited and predictors of opioid needs for terminally ill patients are conflicting. Some studies suggest that age has an inverse correlation with opioid needs [12, 13, 16–19]. Roes retrospectively examined opioid doses prescribed for British hospice patients using regression analysis and found that older patients were prescribed significantly lower opioid doses (p = 0.007) [12]. Zerzan et al. [19] conducted a secondary analysis of randomized trial data and found that use of all opioid types of pain medication decreased with age [odds ratio (OR): 0.75]. Our results were consistent with these previous studies. The reason for this finding is still not completely understood. It is probably in part due to an escalation in the pain threshold that comes with age. In addition, reduced metabolic capacity resulting from altered renal and hepatic function appears to be associated with pharmacokinetic factors and metabolic activity. Systemic clearance of opioids is reduced in patients older than 50 years [20]. However, there is a study reporting that age was only a minor contributor to the variability of morphine, M3G and M6G serum concentrations [21], so the reason why elderly patients need less opioids than younger patients is still controversial. The relationship between opioid requirements and the site of the primary cancer has been examined in several reports. Although the actual dose of opioids differed widely among studies, most reported that cervical or gynecological
123
Int J Clin Oncol
Age
Total N
Fig. 1 Age versus OME (mg/ day). A line graph showing the mean OME (mg/day) of patients with each primary tumor site
OME
N
Cervix OME
N
Corpus OME
N
Vagina/vulva OME
OME
18
999.2
6
427.7
8
1854.9
3
166.7
1
80
40–49
27
212.0
12
55.2
10
415.1
5
182.0
0
–
50–59
55
55.2
31
40.5
18
85.8
5
42.4
1
24
60–69
55
61.1
19
45.3
11
78.6
23
67.3
2
45
C70
34
60.1
12
61.5
6
13.3
15
76.4
1
80
1200
All malignancies
1000 800 600 400 200 0 30
40
50
60
70
500
Ovary
400 300 200 100 0
80
30
40
50
2000
Cervix 1500 1000 500 0 30
40
50
60
Age
cancer patients needed more opioids than patients with many other cancers [12, 22]. Terminal cervical cancer patients are usually younger than those with other gynecologic malignancies. That may be one explanation. However, Hall et al. found the primary tumor site was associated with opioid dose independently of age [13]. High opioid needs of patients with cervical cancer may be due to mechanical problems involving nerve roots. A pelvic tumor can directly invade the lumbosacral plexus. Different from bone metastasis, radiation therapy is difficult for tumors in the pelvis because of the intestine around the tumor. In this study, although we could not confirm the significance of comparing OME with the primary cancer site among gynecologic malignancies, pelvic metastasis was significantly correlated with opioid requirements. This supports the aforementioned theory. In our investigation there was a marked dissociation between the mean and median OME. This may be explained by the presence of several patients in each group who received an extremely high dose of opioids. Although we could not identify the kind of pain this time, neuropathic pain, which is difficult
60
70
80
Age Mean OME (mg/day)
Mean OME (mg/day)
Age
123
N
\40
Mean OME (mg/day)
OME Oral morphine equivalent
Ovary
Mean OME (mg/day)
Table 3 Age distribution and mean OME (mg/day)
70
80
200
Corpus 150 100 50 0 30
40
50
60
70
80
Age
to manage with opioids, was likely to account for these cases. Moreover, gene polymorphism of the opioid receptor has been considered to be responsible for a poor reaction to opioids [23, 24]. However, cases with an extremely high dose of opioids mostly involved patients under 50 years. Thus, these theories are far from explaining all of these situations. In addition to pelvic metastasis, bone or brain metastasis has been reported to have some impact on opioid requirements. However, in this gynecological study there was no significance. On the other hand, carcinomatous ascites showed an inverse correlation with OME. Tumors that produce ascites tend to form peritoneal dissemination rather than an invasive tumor, and carcinomatous ascites has been suggested to have a negative correlation with pelvic invasion. We investigated the situation and predictive factors for opioid use in terminally ill patients with gynecologic malignancies. We elucidated that the OME was inversely associated with age and the strongest association was in patients with cervical cancer. In addition, pelvic invasion was another predictor for opioid requirements. Further
Int J Clin Oncol
Carcinomatous ascites
350 300 250 200 150
**
100 50
Mean OME (mg/day)
Mean OME (mg/day)
Pelvis invasion 400
0
300
200 150 100 50 0
Positive
Negative
Positive
Intestinal obstruction
Negative
Bone metastasis
700
250
600 500 400
N.S.
300 200 100 0
Mean OME (mg/day)
Mean OME (mg/day)
*
250
N.S.
200 150 100 50 0
Positive
Negative
Positive
Negative
Bar: SE, *P
ORIGINAL ARTICLE
Opioid needs of terminally ill patients with gynecologic malignancies Fumi Utsumi • Hiroaki Kajiyama • Jun Sakata • Makiko Higashi • Kaoru Niimi • Ryuichiro Sekiya • Hiroko Mitsui • Shiro Suzuki • Tomokazu Umezu Mika Mizuno • Eiko Yamamoto • Kiyosumi Shibata • Fumitaka Kikkawa
•
Received: 28 January 2014 / Accepted: 7 May 2014 Ó Japan Society of Clinical Oncology 2014
Abstract Background Little is known about patterns and predictive factors regarding opioid use for terminally ill patients with gynecologic malignancies. The aim of this study was to elucidate predictors affecting opioid requirements of endof-life patients with gynecologic malignancies. Methods A retrospective study was carried out on patients with gynecological malignancies admitted to our institute and died during the years 2002 to 2012. The association between maximum opioid dose and factors affecting opioid requirements were examined. Data extracted from medical records included age, site of primary cancer, maximum total dose of opioids prescribed over 24 h, the site of recurrence and metastasis, procedures performed during the hospital stay, total number of chemotherapy courses and overall survival. Results The study identified 189 patients. Most patients had ovarian cancer (42.3 %) followed by cervical cancer (28.0 %) and then corpus malignancy (27.0 %). Opioid requirements decreased with increasing age, especially from the 50s onward. This was particularly marked in cervical cancer patients. In addition, pelvic metastasis was associated with the maximum dose of opioids and the average opioid use was highest in patients with cervical cancer. Conclusion Young age and pelvic invasion were significant predictive factors regarding opioid requirements.
F. Utsumi H. Kajiyama (&) J. Sakata M. Higashi K. Niimi R. Sekiya H. Mitsui S. Suzuki T. Umezu M. Mizuno E. Yamamoto K. Shibata F. Kikkawa Department of Obstetrics and Gynecology, Nagoya University Graduate School of Medicine, Tsuruma-cho 65, Showa-ku, Nagoya 466-8550, Japan e-mail: [email protected]
Additionally, cervical cancer patients may require more opioids among those with gynecologic malignancies. Keywords Cancer-related pain Gynecologic malignancy Opioid intake Age Pelvic invasion
Introduction The incidence of and mortality due to cancer have steadily increased and it has been estimated that approximately 28,000 women died of gynecologic malignancies in 2013 in the United States [1]. It is well-known that cancer patients experience many symptoms that may cause suffering and have a negative impact on their quality of life. Pain is one of the most common and distressing symptoms at the end of life. Incidence of pain in patients with advanced cancer is 58–90 % [2–7]. As part of worldwide efforts to improve the quality of pain control, multiple clinical guidelines have been published on the management of cancer pain [8–11]. The World Health Organization (WHO) designed a 3-step treatment, according to patients’ needs, from non-opioid analgesics to weak opioids and then strong opioids, and it is generally accepted that for each individual the dose is titrated until a sufficient effect is provided. However, despite the fact that opioids have been widely used, pain in terminal cancer patients is still undertreated. Furthermore, predictors in patients who need high doses of opioids are not clear. Concerning this point several studies have addressed the different opioid requirements among primary cancer sites [12, 13]. In gynecologic malignancies women with uterus cancer have been reported to need more opioids than patients with many other cancers [12]. In addition, women with ovarian cancer live longer due to
123
Int J Clin Oncol
tumor reductive surgery and many chemotherapeutic agents [14, 15]. With improved survival the terminal period with pain and symptoms becomes longer. Furthermore, patients who have previously undergone chemotherapy have been reported to need higher opioid doses [16]. Therefore, palliative care, especially pain management, is more important for patients with gynecologic malignancies. Nevertheless, few studies have assessed the patterns of pain and opioid needs in gynecologic cancer patients and influence on quality of life (QOL). The objective of this study was to elucidate the predictive factors of a high opioid requirement in order to improve cancer pain management and the QOL of patients near the end of life.
Patients and methods Study subjects were patients with gynecologic malignancies who died between 2002 and 2012 at our hospital. Using an institutional database we identified all patients with gynecologic malignancy diagnoses who were admitted to our hospital during the study period; the study identified 229 patients. The patients who had double cancer or unknown cancer, or whose date of death was unclear, were excluded. Consequently, a total of 189 patients were eligible. Data extracted from medical records were age, the site of the primary cancer, maximum total dose of opioids prescribed over 24 h, site of recurrence and metastasis, symptoms, procedures performed during the hospital stay which may affect the opioid dose, including ileus, carcinomatous ascites (it was defined as accumulation of ascites requiring paracentesis to palliate their symptoms), blood infusion, the total number of chemotherapy courses and overall survival (OS). Pelvic invasion was defined as the presence of a deeply invasive tumor in the pelvic organs and/or retroperitoneal cavity, excluding superficial peritoneal dissemination. Maximum opioid use was calculated as the oral morphine intake. The calculation formula to equate each equianalgesic dose to oral morphine based on 24 h opioid requirements was as follows. We used these calculated doses as oral morphine equivalents (OME). OME = 2 9 parenteral morphine, OME = 1.5 9 modified-release oxycodone, OME = 30 9 transdermal fentanyl, OME = 100 9 parenteral fentanyl, and OME = 1.5 9 morphine suppository. Statistical analysis All analyses were performed using SPSS. The Student’s t test was used to compare the age and number of chemotherapy courses between two groups. Frequency analysis was performed with a chi square test. The Mann–
123
Whitney U test was used to compare the mean OME between two groups, Kruskal–Wallis one-way analysis of variance was used to compare the mean OME among more than three groups, and OS was compared using Kaplan– Meier statistics and the log-rank test. Regression analysis was used to establish the relationship between the mean OME and age. All p values were two-sided, and p values \0.05 were considered to indicate significance.
Results The survey included 189 patients. The median patient age at the time of last admission to the hospital was 58.3 years. The most common diagnosis was ovarian cancer (42.3 %), followed by cervical cancer (28.0 %), corpus malignancy (27.0 %), and vaginal/vulvar cancer (2 %). First, we compared patients whose OME was under 40 mg/day (low dose group) with those whose OME was 40 mg/day (high dose group). We set the standard at 40 mg/day because the median OME of those who needed opioids in this study was 36 mg/day. Additionally, in general we need to increase the dose at least once or twice to reach the 40 mg/ day threshold. Therefore, patients who were placed in the low dose group were the ones whose pain could be relieved with zero or only one increase of opioid dose, and those in the high dose group needed more increases of opioid dose. Patients’ characteristics and the results are shown in Table 1. The median age of patients in the low dose group was 60.2, compared to 56.2 for those in the high dose group (p = 0.006). There were significant differences in the frequency of pelvic invasion (p = 0.006) and carcinomatous ascites (p = 0.032) between the two groups. The distribution of other clinicopathologic factors, including primary cancer, pulmonary metastasis, bone metastasis, intestinal obstruction, blood infusion, and the chemotherapy course, did not significantly differ between the two OME groups. The relationship between OME and the site of the primary cancer is shown in Table 2. Although the average OME was higher in patients with cervical cancer, it was not significant (p = 0.54). Next, to evaluate the influence of age on opioid requirements in more detail we divided patients into five groups based on age. The frequency distribution of age is shown in Table 3. This also shows a progressive decline in OME with age. In Fig. 1 this change in OME with the mean age is shown as an exponential curve. Regression analysis confirmed a significant decrease in OME with increasing age in the total patients, and in patients with ovarian and cervical cancer (p \ 0.001, 0.002, and 0.041, respectively). There was no significance in patients with endometrial cancer (p = 0.14). Selecting the four clinicopathologic indicators with a significant or marginal-significant difference, as shown in Table 1, we
Int J Clin Oncol Table 1 Patient characteristics Characteristic
Total
N
%
Table 2 Mean and median OME of each primary tumor site Low dose group (0–39 mg/ day)
High dose group ([40 mg/ day)
N
%
N
%
60.3
75
39.7
Total
189
114
Age (median)
58.3
60.2
56.2
Mean ± SE
0.006
Ovary
80
42.3
52
45.6
28
37.3
Cervix
53
28.0
30
26.3
23
30.7
Corpus
51
27.0
30
26.3
21
28.0
Vagina/vulva
5
2.6
2
1.8
3
4.0
0.583
121
64
69
60.5
52
69.3
68
36
45
39.5
23
30.7
Yes
100
52.9
51
44.7
49
65.3
No
89
47.1
63
55.3
26
33.7
74
39.2
45
39.5
29
38.7
No
115
60.8
69
60.5
46
61.3
Corpus
84.6 ± 23.4
30.0
Vagina/vulva
54.8 ± 15.6 170.1 ± 58.7
31.2 24
OME oral morphine equivalent, SE standard error
0.006
Yes
13
6.9
5
4.4
8
10.7
No
176
93.1
109
95.6
67
89.3
0.911
Bone metastasis 0.095
Carcinomatous ascites Yes
28
14.8
22
19.3
6
8.0
No
161
85.2
92
80.7
69
92.0
0.032
Intestinal obstruction 19
10.0
8
7.0
11
14.7
170
90.0
106
93.0
64
85.3
Yes
19
10.0
9
7.9
10
13.3
No
170
90.0
105
92.1
65
86.7
13.2
12.4
14.3
0.32
No. of chemotherapies (mean courses)
20.0 30.0
Discussion
Yes
No Blood infusion
76.0 ± 22.1 405.2 ± 203.3
0.217
Pulmonary metastasis
Yes
Ovary
OS among the 189 patients is shown in Fig. 3. There was no significant difference in OS between the two OME groups (p = 0.86).
Recurrence No Pelvis invasion
Median
Cervix
Total
Primary cancer
Yes
OME (mg/day)
p value
0.087
0.224
compared the mean OME between the presence and absence of pelvic invasion, bone metastasis, intestinal obstruction and carcinomatous ascites using the Mann– Whitney U test (Fig. 2). As we expected, patients with pelvic invasion required higher doses of OME than those without this factor (p \ 0.01). Additionally, consistent with results of the chi square test, patients without ascites accumulation also needed higher opioid doses than those who required paracentesis for their carcinomatous ascites (p = 0.017). Bone metastasis and intestinal obstruction were not significantly correlated with OME (p = 0.17 and 0.435, respectively). Furthermore, the impact of OME on
The incidence and prevalence of cancer has increased with an aging population and improving cancer treatment outcomes. It has become much more common for people to have a prolonged period of cancer and disability prior to death. Pain is one of the most common and distressing symptoms of cancer patients despite growing knowledge of pain management and the use of various opioids. Although studies addressing palliative care are becoming more prevalent, literature regarding gynecologic oncology is very limited and predictors of opioid needs for terminally ill patients are conflicting. Some studies suggest that age has an inverse correlation with opioid needs [12, 13, 16–19]. Roes retrospectively examined opioid doses prescribed for British hospice patients using regression analysis and found that older patients were prescribed significantly lower opioid doses (p = 0.007) [12]. Zerzan et al. [19] conducted a secondary analysis of randomized trial data and found that use of all opioid types of pain medication decreased with age [odds ratio (OR): 0.75]. Our results were consistent with these previous studies. The reason for this finding is still not completely understood. It is probably in part due to an escalation in the pain threshold that comes with age. In addition, reduced metabolic capacity resulting from altered renal and hepatic function appears to be associated with pharmacokinetic factors and metabolic activity. Systemic clearance of opioids is reduced in patients older than 50 years [20]. However, there is a study reporting that age was only a minor contributor to the variability of morphine, M3G and M6G serum concentrations [21], so the reason why elderly patients need less opioids than younger patients is still controversial. The relationship between opioid requirements and the site of the primary cancer has been examined in several reports. Although the actual dose of opioids differed widely among studies, most reported that cervical or gynecological
123
Int J Clin Oncol
Age
Total N
Fig. 1 Age versus OME (mg/ day). A line graph showing the mean OME (mg/day) of patients with each primary tumor site
OME
N
Cervix OME
N
Corpus OME
N
Vagina/vulva OME
OME
18
999.2
6
427.7
8
1854.9
3
166.7
1
80
40–49
27
212.0
12
55.2
10
415.1
5
182.0
0
–
50–59
55
55.2
31
40.5
18
85.8
5
42.4
1
24
60–69
55
61.1
19
45.3
11
78.6
23
67.3
2
45
C70
34
60.1
12
61.5
6
13.3
15
76.4
1
80
1200
All malignancies
1000 800 600 400 200 0 30
40
50
60
70
500
Ovary
400 300 200 100 0
80
30
40
50
2000
Cervix 1500 1000 500 0 30
40
50
60
Age
cancer patients needed more opioids than patients with many other cancers [12, 22]. Terminal cervical cancer patients are usually younger than those with other gynecologic malignancies. That may be one explanation. However, Hall et al. found the primary tumor site was associated with opioid dose independently of age [13]. High opioid needs of patients with cervical cancer may be due to mechanical problems involving nerve roots. A pelvic tumor can directly invade the lumbosacral plexus. Different from bone metastasis, radiation therapy is difficult for tumors in the pelvis because of the intestine around the tumor. In this study, although we could not confirm the significance of comparing OME with the primary cancer site among gynecologic malignancies, pelvic metastasis was significantly correlated with opioid requirements. This supports the aforementioned theory. In our investigation there was a marked dissociation between the mean and median OME. This may be explained by the presence of several patients in each group who received an extremely high dose of opioids. Although we could not identify the kind of pain this time, neuropathic pain, which is difficult
60
70
80
Age Mean OME (mg/day)
Mean OME (mg/day)
Age
123
N
\40
Mean OME (mg/day)
OME Oral morphine equivalent
Ovary
Mean OME (mg/day)
Table 3 Age distribution and mean OME (mg/day)
70
80
200
Corpus 150 100 50 0 30
40
50
60
70
80
Age
to manage with opioids, was likely to account for these cases. Moreover, gene polymorphism of the opioid receptor has been considered to be responsible for a poor reaction to opioids [23, 24]. However, cases with an extremely high dose of opioids mostly involved patients under 50 years. Thus, these theories are far from explaining all of these situations. In addition to pelvic metastasis, bone or brain metastasis has been reported to have some impact on opioid requirements. However, in this gynecological study there was no significance. On the other hand, carcinomatous ascites showed an inverse correlation with OME. Tumors that produce ascites tend to form peritoneal dissemination rather than an invasive tumor, and carcinomatous ascites has been suggested to have a negative correlation with pelvic invasion. We investigated the situation and predictive factors for opioid use in terminally ill patients with gynecologic malignancies. We elucidated that the OME was inversely associated with age and the strongest association was in patients with cervical cancer. In addition, pelvic invasion was another predictor for opioid requirements. Further
Int J Clin Oncol
Carcinomatous ascites
350 300 250 200 150
**
100 50
Mean OME (mg/day)
Mean OME (mg/day)
Pelvis invasion 400
0
300
200 150 100 50 0
Positive
Negative
Positive
Intestinal obstruction
Negative
Bone metastasis
700
250
600 500 400
N.S.
300 200 100 0
Mean OME (mg/day)
Mean OME (mg/day)
*
250
N.S.
200 150 100 50 0
Positive
Negative
Positive
Negative
Bar: SE, *P
