PAMORAs in Opioid-Induced Constipation: Are We Following the Guidelines?

Abstract

INTRODUCTION:

Peripherally acting u-opioid receptor antagonists (PAMORAs) are a safe and efficacious medication for treating opioid-induced constipation (OIC). This study evaluates the utilization of PAMORAs in patients with laxative-refractory OIC.

METHODS:

Patients diagnosed with OIC from 2018 to 2023 were retrospectively identified at a tertiary care center. As recommended by the guidelines, patients were considered eligible for PAMORAs if they had failed at least 2 laxatives. Data on PAMORA prescription, demographics, clinical characteristics, and specialty of the managing provider were collected and analyzed using logistic regression and χ² analysis.

RESULTS:

Of 281 patients diagnosed with OIC, 204 (73%) were eligible for a PAMORA, of which only 58 (28%) were prescribed one. There were no differences in demographic or clinical characteristics between those who were offered PAMORAs vs not. There was a significant difference in prescribing patterns based on provider specialty (P < 0.001). Although gastrointestinal (GI) specialists saw less than half of the eligible patients, they offered PAMORAs at the highest proportion of 59% compared with non-GI specialists (13–16%) or primary care providers (18%). Multivariable logistic regression analysis showed 12.7-fold increased odds of being offered a PAMORA if the provider was from the GI department compared with all non-GI providers (P < 0.001).

DISCUSSION:

PAMORAs are underutilized in patients with OIC. GI prescribers offer PAMORAs at a higher proportion than other departments, and being seen in GI independently increases the likelihood of a patient receiving a PAMORA. This suggests that underutilization may be due to a lack of awareness and further education about PAMORAs is warranted.

INTRODUCTION

Opioid use is common and increasing, with studies estimating that 5%–11% of the US population uses prescription opioids regularly (1–3). The most common adverse effect of opioids on the gastrointestinal (GI) tract is opioid-induced constipation (OIC), caused by activation of u-opioid receptors in the gut (4). OIC affects 1.7% of the US general population and 40%–80% of people taking chronic opioids (2,5,6). In addition, OIC has a significant impact on quality of life, with studies showing increased physical discomfort and psychosocial distress compared with nonopioid-treated patients (7).

Management of OIC often includes lifestyle changes, laxatives (including stool softeners, osmotic laxatives, and stimulants), and peripherally acting mu-opioid receptor antagonists (PAMORAs). PAMORAs work by blocking u-opioid receptors in the gut but do not enter the central nervous system, thus maintaining the analgesic effects of opioids. Given their safety and efficacy in treating OIC, the US Food and Drug Administration has approved 3 PAMORAs (methylnaltrexone, naldemedine, and naloxegol) for the management of OIC (8). Despite these options, OIC continues to be difficult to manage and is often underdiagnosed and undertreated (9–11). To improve patient outcomes in patients with OIC, several societies have made recommendations regarding the medical management of OIC. Many societies including American Gastroenterology Association (AGA), the American Academy of Pain Medicine, the European Association for Palliative Care, and other multidisciplinary expert panels recommend laxatives as first-line pharmacological therapy for patients with OIC and reserve PAMORAs for patients with laxative-refractory OIC (2,12–14). Other societies such as the United European Gastroenterology recommend PAMORAs as first-line treatment in patients with isolated OIC (15).

To date, no study has evaluated the adherence and implementation of these guidelines in clinical practice. We aim to study the utilization of PAMORAs in patients with laxative-refractory OIC at a tertiary care center. Furthermore, we aimed to determine patient characteristics associated with a lack of satisfactory response to laxatives in OIC and factors determining PAMORA utilization in this population.

METHODS

This was a single-center, retrospective study.

Subjects

We identified all patients who were newly diagnosed with OIC between January 1, 2018, and December 31, 2023. We excluded patients diagnosed before 2018 as AGA guidelines on OIC were published in 2018, and the American Academy of Pain Medicine updated their guidelines on OIC in 2017 (2,13). Patients were selected if they had an institutional diagnosis code of OIC. We used an institutional diagnosis code of OIC (Epic Systems Diagnosis ID 1257985 and 1450190), which was highly specific (99% of charts were OIC-related). We also performed searches using the International Classification of Diseases, Tenth Revision (ICD-10) codes for “drug-induced constipation,” code K59.03, but found that this code was not specific (e.g., among the first 30 charts screened used this ICD code, less than 20% had ever been prescribed an opiate). We also evaluated other ICD codes including “other adverse effect of opioid” with code T40.2X5A and “other constipation” with code K59.09, but found them to be highly nonspecific as well. Once patients were identified using the institutional diagnosis code (see above), detailed manual chart review of provider notes and medication history was performed to ensure that patients were on opiates at the time of diagnosis. We then confirmed the presence of constipation symptoms with key words such as “constipation,” “hard stools,” “lumpy stools,” “straining with defecation,” “incomplete evacuation” and “digitalization.” Patients were followed from the time of diagnosis to 1 year after diagnosis or first follow-up appointment. Patients were excluded if they were diagnosed with OIC before 2018, did not have a follow-up, or had a history of bowel obstruction or structural cause of constipation (e.g., unresected colon cancer, colonic stricture).

Data collection

Data regarding the patients were captured using electronic medical record query and retrospective chart review. Demographics including age, body mass index (BMI), sex, race, insurance provider, and Elixhauser comorbidity index were collected (16). Prescribed opioid frequency and dose were collected and reported as morphine milligram equivalents (MMEs). Patients who were prescribed “as-needed” doses were counted as if they were taking the opioids scheduled. Scheduled laxatives that were trialed, including polyethylene glycol, docusate, magnesium hydroxide, magnesium oxide, magnesium citrate, lactulose, mineral oil, bisacodyl, sodium picosulfate, senna, psyllium, methylcellulose, milk of magnesia, lubiprostone, prucalopride, linaclotide, and plecanatide, were collected. The managing provider's department was collected.

PAMORA eligibility

While there is no standardized definition for nonresponse to laxatives in OIC, previous studies in OIC have defined inadequate laxative response as a continued symptom of constipation despite the use of 2 or more laxative classes (17). Thus, we defined eligibility for PAMORAs as (i) ongoing symptoms despite 2 or more classes of laxatives with at least one as a scheduled laxative and (ii) a trial of scheduled use of 1 laxative for >6 months with no improvement in symptoms, which is consistent with recommendations from recent AGA guidelines (2). Ongoing constipation symptoms were identified through a qualitative review of progress notes. Key words in the follow-up encounter such as “continued,” “unchanged,” “no relief,” “minimal relief,” and “no or minimal improvement” in relation to the above key words used to confirm constipation symptoms (constipation, hard or lumpy stools, straining, incomplete evacuation) indicated ongoing symptoms or no improvement. Similarly, if the provider escalated therapy by adding additional daily medication, it was considered that the patient had ongoing symptoms. Eligible patients were followed for 1 year after diagnosis of OIC, and data on whether a PAMORA was offered were collected.

Statistical analysis

Descriptive statistics (mean and standard deviation) and two-tailed t-tests of demographic variables, MMEs, laxatives trialed, and treating departments were calculated among the patients who were eligible for PAMORAs vs those who were not eligible. Among patients who were eligible, they were further divided into patients who were offered PAMORAs and those who were not. Demographic variables and clinical variables were calculated and compared using the two-tailed t-test among the offered vs not-offered group. Multivariable logistic regression was performed, with PAMORA offering as the dependent variable and variables of interest as the independent variables. Interaction terms between variables of interests were also analyzed. χ² analysis was used to determine whether different departments have different proportions of PAMORA eligibility and prescription. The alpha is set to 0.05 for all statistical analyses. All analyses were performed using Matlab R2023b (MathWorks, Natick, MA).

RESULTS

Eligibility for PAMORAs

Of 328 patients diagnosed with OIC during the study period, 281 were included (Figure 1). Of these, 204 (73%) were eligible for PAMORAs. One hundred eighty-eight patients (92%) had trialed 2 or more classes of laxatives (with at least 1 scheduled laxative) but had ongoing OIC symptoms. Sixteen patients (8%) trialed 1 scheduled laxative for >6 months with no improvement.

Figure 1.

Flow diagram of the sample studied. OIC, opioid-induced constipation.

Comparison of patients eligible for PAMORA vs not eligible for PAMORAs

There were no differences in age, sex, race, insurance provider, comorbidity index, or mean opiate dose (measured using MMEs) between patients eligible for a PAMORA vs those not (Table 1). Patients eligible for PAMORAs had significantly lower BMI than patients not eligible for PAMORAs (P = 0.01). There were no differences in age, sex, race, BMI, insurance provider, and comorbidity index of those who were offered a PAMORA vs not (Table 2). The mean number of laxatives trialed and dose of opiate were also similar between the 2 groups (Table 2). However, the proportion of patients offered PAMORAs varied significantly with the provider department (P < 0.001), with gastroenterology providers being much more likely to prescribe PAMORAs compared with other departments (59% vs 8%–13%, respectively, Table 2).

Table 1.

Demographic and clinical characteristics stratified by PAMORA eligibility

	Yes (n = 204)	No (n = 77)	P
Age in years (mean ± std)	57.0 ± 16.3	56.6 ± 15.4	0.85
BMI (mean ± std)	27.4 ± 7.0	29.9 ± 8.22	0.01
ECI (mean ± std)	7.6 ± 4.9	8.0 ± 4.3	0.58
Sex (male:female)	88:116	34:43	0.88
Race (Caucasian:other)	167:37	57:20	0.15
Insurance (private:other)	77:127	26:51	0.54
MME (mean ± std)	111.91 ± 146.96	114.00 ± 145.90	0.92
Laxatives trialed (mean ± std)	2.86 ± 1.09	1.66 ± 0.70	<0.001
Department			0.003
Palliative, n (%)	33 (80)	8 (20)
Primary care, n (%)	38 (58)	28 (42)
Gastroenterology, n (%)	68 (85)	12 (15)
Oncology, n (%)	50 (68)	23 (32)
Other, n (%)	15 (71)	6 (29)

P values less than 0.05 are given in bold. “Primary care” department includes internal medicine and family medicine primary care. “Other” department includes PMR, surgery, OBGYN, neurology, psychiatry, and medicine subspecialties including pulmonology, nephrology, rheumatology, geriatrics, and endocrine.

Abbreviations: BMI, body mass index; ECI, Elixhauser comorbidity index; MME, morphine milligram equivalent; PAMORA, peripherally acting u-opioid receptor antagonists.

Table 2.

Demographic and clinical characteristics stratified by PAMORA offered

	Offered (n = 58)	Not offered (n = 146)	P
Age in years (mean ± std)	57.9 ± 14.7	56.7 ± 17.0	0.63
BMI (mean ± std)	27.8 ± 7.1	27.2 ± 7.0	0.64
ECI (mean ± std)	7.7 ± 5.5	7.6 ± 4.7	0.89
Sex (male:female)	21:37	67:79	0.24
Race (Caucasian:other)	50:08	117:29	0.31
Insurance (private:other)	22:36	55:91	0.96
MME (mean ± std)	121.42 ± 199.20	108.10 ± 120.57	0.56
Laxatives trialed (mean ± std)	2.84 ± 1.32	2.87 ± 0.99	0.89
Department			<0.001
Palliative, n (%)	6 (18)	27 (82)
Primary care, n (%)	6 (16)	32 (84)
Gastroenterology, n (%)	40 (59)	28 (41)
Oncology, n (%)	4 (8)	46 (92)
Other, n (%)	2 (13)	13 (87)

P values less than 0.05 are given in bold. “Primary Care” department includes internal medicine and family medicine primary care. “Other” department includes PMR, surgery, OBGYN, neurology, psychiatry, and medicine subspecialties including pulmonology, nephrology, rheumatology, geriatrics, and endocrine.

BMI, body mass index; ECI, Elixhauser comorbidity index; MME, morphine milligram equivalent; PAMORA, peripherally acting u-opioid receptor antagonists.

Multivariable logistic regression for being offered PAMORAs

Providers in the gastroenterology department were significantly more likely to offer a PAMORA compared with providers from other departments by multivariable logistic regression (odds ratio = 12.7, 95% confidence interval 5.42–29.7, Table 3). Demographic characteristics, opiate dosage, number of laxatives trialed, or the interaction term between the number of laxatives trialed and provider specialty (P = 0.09) were not associated with the likelihood of being offered a PAMORA.

Table 3.

Multivariable logistic regression for being offered a PAMORA

	Odds ratio	95% CI	P
Intercept	0.03	0.00–0.37	0.01
Private insurance	1.28	0.60–2.78	0.52
Caucasian race	0.82	0.27–2.46	0.72
Female sex	0.98	0.44–2.18	0.96
ECI	1.04	0.96–1.12	0.34
GI provider	12.7	5.42–29.7	<0.01
BMI	1.00	0.95–1.06	0.95
Age	1.03	1.00–1.06	0.07
MME	1.00	1.00–1.00	0.05
Laxatives trialed	0.90	0.65–1.26	0.53

P values less than 0.05 are given in bold.

BMI, body mass index; CI, confidence interval; ECI, Elixhauser comorbidity index; GI, gastrointestinal; MME, morphine milligram equivalent; PAMORA, peripherally acting u-opioid receptor antagonists.

Follow-up of patients who were offered PAMORAs

Of the 58 patients who were eligible for and were offered a PAMORA, 5 patients declined initiation. Fifty-three patients (91.3%) were prescribed PAMORAs. The most prescribed PAMORA was naloxegol (38, 71.7%), followed by methylnaltrexone (9, 17.0%) and naldemedine (6, 11.3%). Eight patients were not able to start therapy because of insurance barriers, and 1 patient did not start as constipation improved before the start of therapy.

DISCUSSION

PAMORAs are a safe and effective medication for OIC and are recommended by several societies as the treatment of choice for laxative-refractory OIC (2,12–15). We found that PAMORAs are underutilized and more than 70% of patients who are eligible for PAMORAs are not offered one. Treatment by a provider in the GI department was associated with 12.7 times increased odds of being offered a PAMORA. Demographic factors, comorbidities, the amount of opioid use, and the number of laxatives trialed were not associated with PAMORA offering. Finally, most of the patients who were offered a PAMORA started the treatment, suggesting that provider practices are the most important barrier to initiating PAMORAs in patients with laxative-refractory OIC.

First, regarding eligibility, the only differences between those who were or were not eligible for a PAMORA were laxative use, BMI, and provider department. A difference in laxatives is expected as that was the criterion for selection, with those eligible for PAMORAs having a higher mean number of laxatives trialed. Those eligible for PAMORAs had a lower BMI compared with those not eligible for PAMORAs. This may be related to the fact that a lower BMI is associated with constipation in the US general population and those with a lower BMI may be more likely to have mixed constipation (not as responsive to laxatives) as opposed to pure OIC that may be more laxative-responsive (18). However, more research is needed to fully elucidate this relationship. Interestingly, there were no differences in the amount of opiate used between those eligible for PAMORAs vs not. Some studies suggest that higher amounts of opioids have a larger side effect profile, whereas others have noted that OIC is not dose-dependent and can vary based on a patient's clinical history or pharmacogenetic variation. Given that OIC has a large impact on quality of life, some patients or providers may skip or reduce opioid dosing in attempts to alleviate constipation, which is not recommended (9,19). Our study suggests that higher MME does not necessarily correlate with a lack of response from traditional laxatives; however, further research is needed to clarify this relationship. We also found a difference in eligibility between primary care vs GI (58% vs 85%, respectively). Thus, it may signify that patients seen in the GI department have more severe constipation that has been refractory to a trial of laxatives, whereas those who present to primary care clinic might have less severe constipation.

Overall, we found that the utilization of PAMORAs is low. While a high percentage (73%) of those diagnosed with OIC are eligible for a PAMORA, only 28% of those eligible were ever offered one. The only factor independently associated with patients being offered PAMORA was having a GI provider. This might be due to increased awareness of PAMORAs and indications for use among GI specialists. Despite GI providers having the highest rate of offering, the rate was only 59% of eligible patients. Thus, even in GI, PAMORAs are underutilized, and further education targeting all specialties is needed. In addition, our study showed that oncology has the lowest proportion of offering at 8%, with all other specialties ranging from 13% to 18%. This is consistent with a study by Prenmanth et al which showed that the oncology department offered PAMORAs at a lower rate than other specialties (20). This may be due to the concern of drug interactions, with naldemedine and naloxegol particularly that are moderate to strong inducers/inhibitors of P-450 (CYP450) isozymes. For example, antifungal drugs for treating oral mucositis frequently derived from chemotherapy regimens are potent inhibitors (21).

Notably, there were no demographic or clinical differences between those offered vs not, including the amount of opiates used, the number of laxatives trialed, and BMI. This further emphasizes that the offering of PAMORAs is driven by the provider and not patient characteristics. This suggests that the number of laxatives previously tried does not significantly influence clinical decision-making regarding the offering of PAMORAs. Regarding insurance, our study shows that insurance (private vs public) is not significantly different between those offered vs not offered a PAMORA. Of the 53 patients who were prescribed a PAMORA, 8 (15%) were not able to initiate therapy because of lack of insurance coverage. Premnath et al showed that from 2014 to 2018, there was a significant and increasing Medicare expenditure on PAMORAs, with certain institutions/payers enforcing checks to prevent overprescription (20). More longitudinal studies are needed to elucidate the effect of insurance on PAMORA prescription and compliance.

There are limitations to this study. First, subjects were selected if they were given an institutional diagnosis code of OIC in the electronic medical record. Thus, the diagnosis is made at provider discretion, and given the retrospective nature of this study, we are unable to standardize this diagnosis using diagnostic criteria such as Rome IV for OIC (4). The true number of patients with OIC is likely higher than that our study captures, and the proportion of utilization is likely lower than what our study shows (as prescribing a PAMORA will likely require entering a diagnosis of OIC). The study criteria for eligibility are also based on the scheduled use of laxatives, which might not be accurate. In addition, MME data are likely overestimated because all “as-needed” prescriptions were counted as if the patient took them as scheduled. Furthermore, our results may be skewed by the fact that GI may be more likely to see more severe constipation and more likely to recognize OIC as a diagnosis as suggested by the fact that the largest proportion of patients in our study were from the GI department. There are likely also patient population differences between departments, with palliative/oncology potentially seeing patients with more comorbidities or other interacting medications. Thus, our study likely overestimates the difference in departments. Finally, we were also unable to determine the impact of switching to PAMORAs given the retrospective nature of the study.

In conclusion, PAMORAs are underutilized where only 28% of patients with laxative-refractory OIC were offered one. Provider and not patient characteristics were associated with being offered a PAMORA in this patient population. Being managed by a GI provider significantly and independently increases the likelihood of being offered a PAMORA. However, up to 40% of the patients with OIC eligible for PAMORAs seen by a GI provider are not offered one. More education and awareness about PAMORAs among general practitioners and all specialists are needed. In addition, more studies are needed to elucidate further barriers to awareness and insurance and patient barriers to PAMORA usage.

CONFLICTS OF INTEREST

Guarantor of the article: Allen A. Lee, MD.

Specific author contributions: All authors have approved the final draft submitted. K.L.: planning, collecting and interpreting data, drafting the manuscript. A.C.: collecting and interpreting data, drafting the manuscript. P.R.: collecting and interpreting data, drafting the manuscript. G.B.: collecting and interpreting data, drafting the manuscript. A.A.L.: planning, interpreting data, drafting the manuscript. P.S.: planning, interpreting data, drafting the manuscript

Financial Support: P.S. was supported by NIDDK grant number K23DK129327. A.A.L. was supported by NIDDK grant numbers 1K23DK124567 and 1R03DK139095.

Potential competing interests: None.

Ethics approval: The study was approved by the University of Michigan Institutional Review Board (HUM 00246597).

Study Highlights.

WHAT IS KNOWN

• ✓ Peripherally acting u-opioid receptor antagonists (PAMORAs) are recommended for laxative-refractory opioid-induced constipation.

• ✓ No study has evaluated the utilization of PAMORAs in clinical practice.

WHAT IS NEW HERE

• ✓ PAMORAs are underutilized in opioid-induced constipation.

• ✓ Being managed by a GI provider increases the likelihood of being offered a PAMORA.