Examining the Severity and Progression of Illicitly Manufactured Fentanyl Withdrawal: A Quasi-experimental Comparison

Abstract

Objective:

Illicitly manufactured fentanyl has largely replaced heroin throughout the United States. Characteristics of fentanyl-specific withdrawal are not well understood compared to traditional opioid withdrawal. This study examines opioid withdrawal severity among 2 cohorts of study participants who underwent identical morphine stabilization procedures before and after fentanyl was introduced to the local drug market.

Methods:

The Non-Fentanyl study (n = 103) included participants testing positive for non-fentanyl opioids, and the Fentanyl study (n = 30) included participants testing positive for fentanyl. Both studies completed a 7-day morphine stabilization protocol (30 mg subcutaneous, 4 times daily) and multiple daily self-report and observer-rated assessments of opioid withdrawal and vital signs. Two-way repeated-measures analyses of variance (ANOVAs) examined the effects of study, time, and study × time on daily peak ratings for each outcome.

Results:

There were significant elevations in self-report and observer-rated withdrawal scores among the Fentanyl versus Non-Fentanyl study (study × time, P < 0.05) during stabilization days 2–5 and days 2–6, respectively. There was a higher rate of tachycardia among the Fentanyl group compared to the Non-Fentanyl study, and peak diastolic blood pressure was greater among the Fentanyl study compared to the Non-Fentanyl study.

Conclusions:

Individuals with fentanyl exposure were less stabilized by morphine and experienced more severe opioid withdrawal via several metrics compared to persons with non-fentanyl opioid exposure. Withdrawal also remained elevated for several days despite morphine initiation. Adjustments to existing treatment induction protocols may be needed given the permeation of fentanyl into the heroin supply.

The opioid epidemic is particularly distressing because opioid use disorder is characterized with high rates of treatment discontinuation, especially early in treatment. Further complicating the opioid epidemic, heroin in some areas of the United States has been adulterated or replaced entirely with illicitly manufactured fentanyl (IMF),^1,2 a highly potent analgesic and anesthetic opioid responsible for unprecedented overdose rates³ and with unknown influence on treatment initiation and retention. Fentanyl is distinct from many other opioids in its potency and pharmacokinetic profile; specifically, it is 50 times more potent than heroin, is highly lipophilic, and crosses the blood-brain barrier more quickly than other opioids.⁴ Fentanyl swiftly penetrates highly perfused tissues, producing peak subjective effects within minutes followed by rapid redistribution and sequestration into the peripheral muscles and lipids and a delayed urinary excretion pattern.⁵

Although most information regarding the pharmacological effect of fentanyl is derived from research on therapeutic doses of fentanyl, the pharmacodynamic and pharmacokinetic profiles of IMF, for which doses are orders of magnitude larger than medicinal fentanyl, are not well understood. IMF produces severe physical dependence that may respond differently to treatment with a full (methadone) versus partial (buprenorphine) agonist.⁶ IMF exposure also leads to an extended (1–2 weeks) urinary excretion pattern,⁷ which in turn may lead to more severe spontaneous but not naloxone-precipitated withdrawal.^8,9

The present work examines whether fentanyl exposure alters the clinical presentation of patients presenting for opioid use disorder (OUD) treatment among persons enrolled in 2 identical residential research trials conducted (1) before and (2) after fentanyl was introduced to the local drug product. Participants in trial one (Non-Fentanyl study) had verified non-fentanyl opioid exposure (e.g., heroin, oxycodone), whereas participants in trial 2 (the Fentanyl study) were all verified as having been fentanyl-exposed. Participants in both studies received morphine for 7 days to stabilize them off illicit opioids, and several measures of opioid withdrawal were collected. Data provide a quasi-experimental opportunity to compare withdrawal severity following fentanyl exposure, and hypothesized persons with fentanyl exposure would express a more severe form of opioid withdrawal relative to those with non-fentanyl opioid exposure.

METHODS

Overview

Participants who enrolled in 1 of 2 randomized controlled clinical trials were combined for these secondary analyses. Adults with OUD were recruited from the community for admission into the parent trials that examined 1) tramadol-ER, clonidine, and buprenorphine (Non-Fentanyl study),¹⁰ or 2) lofexidine, the percutaneous nerve stimulator for substance use disorders (NSS)-bridge device, or placebo/sham controls (Fentanyl study) for opioid tapering. Data for this report are taken from the pre-randomized period wherein all participants received identical dosing of immediate release morphine subcutaneous (SC) (30 mg SC, 4 times a day [QID]) for 7 days on the same residential unit. Both studies employed the same general eligibility criteria, with slight modifications for specific medication/devices being examined.

Parent Trials

The first study (Non-Fentanyl) was a 26- to 28-day residential evaluation of the relative efficacy of tramadol-ER compared to clonidine and buprenorphine for opioid tapering (NCT01188421). Participants (N = 103) were enrolled between October 2010 and June 2015. The study enrolled adults aged 18–60 years who met DSM-IV criteria for opioid dependence, provided a urine sample that showed evidence of recent opioid use, and had no significant medical/psychiatric illnesses.

The second study (Fentanyl) was a 16- to 20-day residential evaluation of a neuromodulator medical device compared to lofexidine for OUD (NCT04325659). Participant enrollment began in March 2021 and is ongoing; only participants enrolled up to February 2024 (N = 30) are included in these analyses. This study enrolled adults aged 22–65 years who met DSM-5 criteria for OUD, provided a urine sample that showed evidence of recent opioid use, and had no significant medical or psychiatric illnesses.

Both studies excluded participants for pregnancy, physical dependence on alcohol and/or benzodiazepines that required withdrawal treatment, known allergies to study medications, and current enrollment in opioid agonist treatment. Full eligibility criteria for both studies are provided in eTable 1, http://links.lww.com/JAM/A556.

Concomitant Medications

No concomitant medications were provided during the morphine stabilization period during the Non-Fentanyl trial; however, due to a pattern of uncontrolled withdrawal, the Fentanyl trial made the following concomitant medications available during stabilization: acetaminophen, antacid, clonazepam, dicyclomine, hydroxyzine, ibuprofen, loperamide, methocarbamol, milk of magnesia, ondansetron, promethazine, simethicone, and trazodone. All concomitant medications, if administered, were given after the data collection timepoint and in a protocolized fashion where study nurses provided medication if a participant endorsed a symptom related to withdrawal.

Measures

All opioid withdrawal ratings were collected 4 times daily. Participants provided self-report ratings via the Subjective Opioid Withdrawal Scale (SOWS), a 16-item Likert scale (range 0–64). Trained observers also rated withdrawal via the Clinical Opiate Withdrawal Scale (COWS), an 11-item scale (range 0–48). Primary outcomes were total SOWS summed score (continuous) and percent of participants meeting the criteria for mild (score 1–10), moderate (11–20), or severe (21–30) withdrawal each day, and total COWS summed score (continuous) and the percent of participants meeting the criteria for none/minimal (0–4), mild (5–12), moderate (13–24), moderately severe (25–36), or severe (>36) withdrawal on each study day. Primary outcomes were peak daily SOWS and COWS.

Additional secondary outcomes assessed included individual SOWS symptoms, individual COWS symptoms, heart rate, blood pressure (systolic/diastolic), and pupil diameter. Rates of tachycardia (heart rate >100 bpm at least one time) among participants were also computed.

Statistical Analysis

Demographics were assessed for the overall sample and compared across study samples using t tests for continuous and χ² test of independence for categorical data. Peak ratings for each morphine stabilization day (1–7) were compared using 2-way repeated-measures analyses of variance (ANOVAs) to understand simple main effects of total and individual symptom scores as well as physiological measures as a function of study (Non-Fentanyl, Fentanyl) and day (1–7), as well as study × day interactions. The proportion of participants in each withdrawal severity category by study day was examined using χ² test of independence. All analyses were conducted using STATA 18, and alpha was set to 0.05.

RESULTS

Participants

When collapsed together (N = 133; Non-Fentanyl study n = 103, Fentanyl study n = 30), participants were 56% Black/African American, 40% non-Hispanic White, and 4% other. Most participants were male (84%), and 79% of participants had at least a high school education. Demographic comparisons revealed that participants in the Fentanyl study were, on average, 5 years older than participants in the Non-Fentanyl study (P = 0.02) (Table 1); no additional variables differed significantly (P values > 0.05).

TABLE 1.

Demographic Characteristics

	Non-Fentanyl Study (n = 103) n (%)	Fentanyl Study (n = 30) n (%)	P
Gender			0.255
Female	15 (14.6%)	7 (23.3%)
Male	88 (85.4%)	23 (76.7%)
Race
Black/African American	57 (55.3%)	17 (56.7%)	0.222
White/Caucasian	43 (41.8%)	10 (33.3%)
Other	3 (2.9%)	3 (10.0%)
Hispanic			0.297
Yes	5 (4.9%)	3 (10.0%)
No	98 (95.2%)	27 (90.0%)
Marital status			0.267
Never married	66 (64.1%)	18(60.0%)
Married	8 (7.8%)	4 (13.3%)
Divorced/Separated	26 (25.2%)	8 (26.7%)
Widowed	3 (2.9%)	0 (0%)
Education			0.595
Less than high school	22 (21.6%)	5 (17.2%)
High school/GED	54 (52.9%)	17 (58.6%)
Some college	21 (20.6%)	7(24.1%)
Bachelor’s degree	5 (4.9%)	0 (0%)
	M (SD)	M (SD)
Age	41.17(10.24)	46.03 (10.11)	0.021

All participants in the Non-Fentanyl study tested positive for opioids on a general opioid assay using a morphine standard, which confirms non-fentanyl opioid exposure, whereas all participants in the Fentanyl study tested positive for fentanyl. Urine samples available for a small sample of participants (n = 3) in the Non-Fentanyl study (all of whom were enrolled during the final study year) that were tested using liquid chromatography with tandem mass spectrometry revealed no evidence of fentanyl or norfentanyl (below the lower limit of quantitation, <5 ng/mL), suggesting that the Non-Fentanyl study was indeed conducted prior to full penetrance of fentanyl into the local market.

Self-Reported Withdrawal (SOWS)

Total Peak Scores

The model (F(1,6) = 12.44, P < 0.01) revealed a significant main effect of study (Non-Fentanyl, Fentanyl), day, and study × day on peak SOWS self-reported withdrawal total scores. The study × day interaction showed that participants in the Fentanyl study displayed significantly higher peak SOWS on stabilization days 2 through 6 (P < 0.05; see Fig. 1A) compared to participants in the Non-Fentanyl study. Participants in the Fentanyl study also demonstrated an inverse-U pattern whereby total SOWS scores peaked on day 2 before slowly returning to the levels experienced by the Non-Fentanyl study by day 7. In contrast, participants in the Non-Fentanyl study experienced an early and steady decline in withdrawal severity (eTable 2, http://links.lww.com/JAM/A556).

FIGURE 1.

A, Mean ± SEM total Subjective Opiate Withdrawal Scale (SOWS) peak scores. B, Total Clinical Opiate Withdrawal Scale (COWS) peak scores by study group (Non-Fentanyl and Fentanyl) across days. Groups statistically significantly differ on study days where markers are black and do not statistically significantly differ on days that are white.

Withdrawal Categorical Ratings

A significant effect of study and day was observed on SOWS severity categories such that a higher proportion of participants in the Fentanyl study rated their withdrawal in the severe range compared to participants in the Non-Fentanyl study during days 2 (48% vs 16%), 3 (47% vs 8%) 4 (37% vs 6%), and 5 (27% vs 6%); this was also evident for the proportion of participants scoring in the moderate withdrawal range on days 2 (31% vs 20%), 4 (20% vs 14%), 5 (20% vs 13%), and 6 (17% vs 16%) (Fig. 2) relative to the Non-Fentanyl study. A lower proportion of Fentanyl-study participants rated their withdrawal in the mild range on days 2 (17% vs 58%), 3 (33% vs 60%), 4 (40% vs 70%), 5 (47% vs 61%), and 6 (50% vs 63%) compared to Non-Fentanyl study participants. Days 1 and 7 did not vary significantly (eTable 3, http://links.lww.com/JAM/A556).

FIGURE 2.

Percentage of participants in the Non-Fentanyl study (A) and Fentanyl study (B) who experienced Subjective Opiate Withdrawal (SOWS) scores that fell into “none,” “mild,” “moderate,” or “severe” categories across study days. Percentage of participants in the Non-Fentanyl study (C) and Fentanyl study (D) who experience Clinical Opiate Withdrawal Scale (COWS) scores that fell into “none/minimal,” “mild,” and “moderate” categories across study days.

Individual Symptom Ratings

No individual symptoms emerged as uniquely driving self-reported withdrawal differences between the Non-Fentanyl and Fentanyl studies. Withdrawal severity was elevated across several self-reported symptoms, with significant study and day main effects as well as significant study × day interactions for the following symptoms: anxiety (P < 0.001), perspiration, tearing, runny nose, goosebumps, shaking, hot flashes, cold flushes, bone and muscle aches, restlessness, nausea, vomiting, muscle twitches, and stomach cramps (P values < 0.05). Symptoms of yawning and the urge to use showed a significant main effect of day (P < 0.001) and significant study × day interactions (P < 0.001), but no main effects of study (P > 0.05). Consistent with the total SOWS scores, between-study differences in individual symptom severity did not emerge until day 2, at which point participants in the Fentanyl study demonstrated an inverted-U withdrawal curve that peaked on day 2 and participants in the Non-Fentanyl study demonstrated a steady, linear decline in withdrawal symptoms after day 1 (Table 2).

TABLE 2.

Mean ± SD Individual SOWS Symptom Scores Across Days of Morphine Stabilization and Study Groups

SOWS Individual Symptoms (0–4)	Study Group	Morphine Stabilization Day (M [SD])
		1	2	3	4	5	6	7
I feel nauseous*	Non-fentanyl	0.9 [1.0]	0.6 [0.8]	0.4 [0.8]	0.3 [0.8]	0.3 [0.6]	0.2 [0.6]	0.2 [0.5]
	Fentanyl	0.9 [1.1]	1.5 [1.5]	1.2 [1.3]	1.2 [1.5]	0.6 [0.8]	0.5 [0.8]	0.4 [0.8]
I am perspiring*	Non-fentanyl	0.9 [1.0]	0.6 [0.8]	0.4 [0.7]	0.4 [0.6]	0.3 [0.6]	0.4 [0.7]	0.3 [0.6]
	Fentanyl	0.9 [1.0]	1.6 [1.5]	1.2 [1.3]	0.9 [1.1]	0.9 [1.0]	0.9 [1.0]	0.4 [0.6]
I feel restless*	Non-fentanyl	1.7 [1.2]	1.0 [1.0]	0.7 [0.8]	0.7 [0.8]	0.6 [0.8]	0.6 [0.8]	0.5 [0.8]
	Fentanyl	1.5 [1.4]	1.9 [1.2]	2 [1.4]	1.9 [1.4]	1.3 [1.1]	1.1 [1.0]	1.0 [1.1]
My bones and muscle ache*	Non-fentanyl	1.6 [1.2]	1.0 [1.0]	0.8 [0.9]	0.7 [0.9]	0.7 [0.8]	0.6 [0.8]	0.5 [0.8]
	Fentanyl	1.3 [1.2]	1.7 [1.3]	1.6 [1.5]	1.5 [1.5]	1.3 [1.3]	1.0 [1.0]	0.9 [1.1]
My nose is running*	Non-fentanyl	1.3 [1.0]	0.7 [0.9]	0.5 [0.8]	0.5 [0.8]	0.4 [0.7]	0.4 [0.7]	0.3 [0.6]
	Fentanyl	1.4 [1.3]	1.6 [1.2]	1.4 [1.2]	1.0 [1.1]	0.8 [0.9]	0.6 [0.7]	0.4 [0.6]
I feel like vomiting*	Non-fentanyl	0.6 [1.0]	0.4 [0.7]	0.2 [0.6]	0.2 [0.6]	0.2 [0.5]	0.1 [0.5]	0.1 [0.4]
	Fentanyl	0.5 [0.9]	1.3 [1.5]	0.9 [1.2]	0.9 [1.5]	0.5 [0.8]	0.4 [0.7]	0.3 [0.8]
I am shaking*	Non-fentanyl	0.7 [1.0]	0.5 [0.8]	0.3 [0.7]	0.3 [0.6]	0.2 [0.6]	0.2 [0.5]	0.2 [0.4]
	Fentanyl	0.6 [1.1]	1.2 [1.3]	0.8 [1.3]	0.8 [1.2]	0.6 [1.0]	0.6 [0.8]	0.4 [0.8]
I feel like yawning†	Non-fentanyl	1.7 [1.1]	1.0 [1.0]	0.7 [0.9]	0.6 [0.8]	0.6 [0.8]	0.6 [0.7]	0.5 [0.7]
	Fentanyl	1.6 [1.1]	1.6 [1.3]	1.3 [1.1]	1.2 [1.1]	0.7 [0.8]	0.7 [0.7]	0.4 [0.6]
I feel anxious*	Non-fentanyl	1.5 [1.1]	0.9 [1.0]	0.6 [0.8]	0.5 [0.7]	0.5 [0.7]	0.5 [0.7]	0.4 [0.5]
	Fentanyl	2.0 [1.4]	2.1 [1.3]	2.0 [1.3]	1.7 [1.3]	1.6 [1.3]	1.4 [1.2]	1.1 [1.2]
I have goosebumps*	Non-fentanyl	0.9 [0.9]	0.5 [0.8]	0.4 [0.7]	0.3 [0.7]	0.3 [0.7]	0.2 [0.6]	0.2 [0.5]
	Fentanyl	0.8 [1.0]	1.3 [1.2]	0.9 [1.2]	0.9 [1.1]	0.6 [0.9]	0.6 [0.8]	0.2 [0.4]
My muscles twitch*	Non-fentanyl	0.9 [1.0]	0.4 [0.7]	0.3 [0.6]	0.3 [0.6]	0.2 [0.6]	0.2 [0.5]	0.2 [0.5]
	Fentanyl	0.6 [0.9]	1.3 [1.3]	1.1 [1.3]	0.9 [1.3]	0.6 [1.0]	0.4 [0.8]	0.3 [0.7]
My eyes are tearing*	Non-fentanyl	1.1 [1.0]	0.6 [0.8]	0.5 [0.8]	0.4 [0.7]	0.4 [0.7]	0.4 [0.7]	0.3 [0.6]
	Fentanyl	0.8 [0.9]	1.3 [1.3]	1.2 [1.2]	0.9 [1.0]	0.9 [0.9]	0.6 [0.7]	0.4 [0.6]
I have hot flashes*	Non-fentanyl	0.9 [1.0]	0.7 [0.9]	0.5 [0.8]	0.4 [0.7]	0.3 [0.7]	0.3 [0.6]	0.3 [0.5]
	Fentanyl	1.1 [1.3]	1.7 [1.4]	1.3 [1.3]	1.2 [1.4]	0.9 [0.9]	0.7 [0.9]	0.6 [0.9]
I have cold flushes*	Non-fentanyl	1.0 [1.0]	0.7 [1.0]	0.6 [0.9]	0.4 [0.7]	0.4 [0.8]	0.3 [0.6]	0.3 [0.5]
	Fentanyl	1.2 [1.3]	1.6 [1.3]	1.3 [1.4]	1.3 [1.3]	0.9 [1.0]	0.7 [0.9]	0.5 [0.7]
I feel like using†	Non-fentanyl	2.5 [1.4]	1.7 [1.4]	1.5 [1.4]	1.5 [1.4]	1.3 [1.4]	1.2 [1.4]	1.2 [1.3]
	Fentanyl	2.0 [1.5]	2.1 [1.5]	2.1 [1.6]	1.9 [1.6]	1.6 [1.5]	1.0 [1.2]	1.1 [1.4]
I have stomach cramps*	Non-fentanyl	1.1 [1.2]	0.7 [0.8]	0.5 [0.8]	0.4 [0.9]	0.4 [0.7]	0.3 [0.6]	0.4 [0.7]
	Fentanyl	1.0 [1.2]	1.6 [1.2]	1.5 [1.5]	1.2 [1.5]	0.8 [1.1]	0.6 [0.8]	0.6 [1.1]

^*Significant effect of group, day, and group × day.

^†Significant effect of day and group × day.

SOWS indicates Subjective Opiate Withdrawal Scale.

Observed Withdrawal (COWS)

Total Peak Scores

The model revealed a significant main effects of study, day, and study × day interaction (F(1,6) = 7.76, P < 0.001) on peak COWS total scores. Participants in the Fentanyl study evinced significantly higher peak total COWS scores on days 2 through 7 (P < 0.05; see Fig. 1B) relative to Non-Fentanyl study participants. Consistent with self-reported withdrawal ratings, differences between groups emerged on day 2, wherein participants in the Fentanyl study demonstrated an inverted-U trend that peaked on days 2 and 3 before slowly remitting, and the Non-Fentanyl study participants demonstrated a steady linear decline in withdrawal between day 1 and day 4 (eTable 2, http://links.lww.com/JAM/A556).

Withdrawal Categorical Ratings

No participants evinced COWS scores that met the criteria for moderately severe or severe withdrawal (25+) during morphine stabilization, and the number of participants in both studies categorized as having moderate withdrawal was too small (n’s < 5) to interpret. A nominally larger proportion of participants in the Fentanyl study experienced moderate withdrawal compared to participants in the Non-Fentanyl study during days 1 through 4 (Fig. 2). The proportion of those with observed mild withdrawal was greater for participants in the Fentanyl versus Non-Fentanyl studies on days 2 (73% vs 33%, respectively), 3 (70% vs 13%), 4 (73% vs 18%), 5 (60% vs 15%), 6 (53% vs 18%), and 7 (40% vs 14%). Fewer participants in the Fentanyl versus Non-Fentanyl study had withdrawal categorized as none/minimal on days 2 (20% vs 66%), 3 (20% vs 88%), 4 (23% vs 81%), 5 (40% vs 85%), 6 (47% vs 82%), and 7 (60% vs 86%) (eTable 3, http://links.lww.com/JAM/A556).

Individual Symptom Ratings

No single COWS symptom emerged as driving increased observed withdrawal severity between studies. Several individual COWS symptoms significantly differed as a function of study, day, and/or study × day, including anxiety, joint aches, runny nose, and sweating. Gastrointestinal upset and gooseflesh both had significant day and study × day effects (P < 0.001), but no significant main effect of study (P > 0.05). Observed pupil size, resting heart rate, and yawning differed by study (P < 0.05), but not day or study × day (P > 0.05). Tremor had a significant study × day interaction (P < 0.05), but no study or day main effects (P > 0.05). Restlessness was the only symptom to not demonstrate any significant effects. Consistent with total scores, groups diverged in individual symptom severity on day 2, with the participants in the Fentanyl study displaying an inverted U syndrome that peaked over 4 days before remitting and participants in the Non-Fentanyl study demonstrating a steady linear decline in individual symptom severity following a peak on day 1 (see eTable 4, http://links.lww.com/JAM/A556).

Physiological Measures

Heart Rate

There was a significant main effect of day on heart rate (F(1,6) = 16.12, P < 0.001), where participants experienced an increase in heart rate over the stabilization period. There was a significant study × time interaction (F(1,6) = 16.12, P = 0.007) (Fig. 3; eTable 2, http://links.lww.com/JAM/A556) where the Fentanyl group experienced a steeper and sustained increase in heart rate over the stabilization period compared to the Non-Fentanyl group. There was not a significant effect of study (F(1,6) = 16.12, P = 0.416). Changes in heart rate followed the same pattern as withdrawal, whereby participants did not vary significantly on day 1 but diverged thereafter with participants in the Fentanyl study experiencing increased heart rate over days 1–5, followed by decreases on days 6–7. Forty-three percent and 19% of participants in the Fentanyl and Non-Fentanyl studies had >1 incidence of tachycardia during the evaluative period, P < 0.05, respectively.

FIGURE 3.

Mean ± SEM heart rate (A), diastolic blood pressure (B), and systolic blood pressure (C) by study group (pre- and post-fentanyl) across study days.

Blood Pressure

There was a significant main effect of day on systolic blood pressure (F(1,6) = 13.91, P < 0.001), indicating a slight mean increase in systolic blood pressure over the stabilization period. There was not a significant study (F(1,6) = 13.91, P = 0.504) or study × day interaction (F(1,6) = 13.91, P = 0.300) (eTable 2, http://links.lww.com/JAM/A556). A significant effect of day (F(1,6) = 10.53, P < 0.001) was observed for diastolic blood pressure where mean diastolic blood pressure increased over the stabilization period. There was a significant effect of study where Fentanyl group had a higher mean systolic blood pressure than the Non-Fentanyl group (F(1,6) = 10.53, P = 0.003). A significant study × day interaction (F(1,6) = 10.53, P = 0.022) was observed for diastolic blood pressure where the Fentanyl group experienced a steeper and sustained increase in heart rate over the stabilization period compared to the Non-Fentanyl group (see Fig. 3, eTable 2, http://links.lww.com/JAM/A556). Participants in the Fentanyl study evinced increases in systolic and diastolic blood pressures from day 1 to day 2 and decreases from day 2 to day 6, whereas participants in the Non-Fentanyl study evinced decreases in systolic blood pressure from day 1 to day 2 that stabilized between days 2 and 4 and then decreased on days 4 through 7; diastolic blood pressure remained stable throughout the stabilization period.

Pupil Diameter

There was a significant effect of study (F(1,6) = 34.56, P < 0.001), day (F(1,6) = 34.56, P = 0.032), and study × day (F(1,6) = 34.56, P = 0.0495) on pupil diameter as determined by pupilometer, whereby participants in the fentanyl study exhibited smaller pupil size across all days compared to Non-Fentanyl participants (eTable 2, http://links.lww.com/JAM/A556).

DISCUSSION

These data are the first empirical demonstration that participants with fentanyl exposure experience more severe and sustained opioid withdrawal compared to persons without fentanyl exposure. Participants in these studies were stabilized from non-fentanyl opioids or fentanyl using an identical procedure, affording a quasi-experimental opportunity to assess opioid withdrawal severity following fentanyl exposure. Participants from both studies presented with similar initial withdrawal ratings; however, participants with fentanyl exposure experienced an inverted-U type pattern of withdrawal that peaked during days 2 and 3 and took >7 days to remit, whereas participants with non-fentanyl opioid exposure demonstrated a linear decline in withdrawal severity. The fact that withdrawal remained elevated among participants with fentanyl exposure despite receiving morphine and concomitant medications demonstrates the profound clinical impact fentanyl exposure has for patients presenting for opioid treatment. More participants with fentanyl exposure also met the criteria for severe self-reported withdrawal and tachycardia, further elevating the level of clinical care necessary for their management.

Withdrawal severity was increased across most individual symptoms following fentanyl exposure, and gastrointestinal distress, anxiety, and bone/muscle aches were more prominently elevated during the first 4 days for persons with fentanyl versus non-fentanyl opioids. GI distress was also evident from the number of concomitant medications for GI symptoms needed and may be related to heavy sequestration of fentanyl in the gut.⁵ Protracted withdrawal observed here could be related to fentanyl’s peripheral sequestration and extended renal clearance,⁷ known to be extended in persons who have higher body mass indices (suggesting that lipid and/or muscle storage may confer clinically significant consequences for fentanyl excretion).⁸ These properties could also explain the pattern observed in pupil diameter; opioids naturally constrict pupil size and fentanyl effects pupillary reflex dilation, and it is plausible that the peripheral sequestration of fentanyl could cause continued constriction even without continued use.¹¹ A reliable relationship between urinary fentanyl concentrations and withdrawal expression has not been established, though lower levels of urinary fentanyl have been correlated with more severe withdrawal.¹²

Understanding the unique symptom expression during withdrawal from fentanyl is important for determining potential pharmaceutical targets. For instance, the H2 receptor antagonist cimetidine is routinely used to prevent enteral sequestration of fentanyl during surgery,¹² and empirical examinations of cimetidine for fentanyl withdrawal may be warranted. Fentanyl is also subject to extensive hepatic metabolization by CYP3A4 that mediates fentanyl’s analgesic efficacy.^13–16 Given the extensive population-level polymorphic variations in CYP3A4,¹⁷ research examining this relationship may inform withdrawal variability and treatment strategies.

The data collected here also provide valuable information for clinicians and persons seeking to treat the fentanyl withdrawal syndrome. A recent review of premature hospital discharges found significant increases in attrition among opioid-exposed persons relative to the general population within the first 3 days of treatment,¹⁸ aligning closely with the data reported here. Understanding that withdrawal may initially be nonlinear and may extensively feature gut, anxiety, and restlessness symptoms can be used to help allay concerns from persons whose prior withdrawal experiences derived from non-fentanyl opioids. The degree to which elevated withdrawal severity contributes to challenges inducting persons from fentanyl onto buprenorphine is also understudied,¹⁹ and the inverted-U withdrawal observed here provides initial support for short-term bridging with opioid agonists as part of treatment initiations.¹⁸ Short-term opioid bridges could be supported under the “72-hour” rule, a strategy recently used to bridge patients with methadone outside of an opioid treatment setting.²⁰

These findings have some limitations. First, study cohorts were enrolled 10 years apart and additional non-fentanyl factors may be contributing to outcomes. Second, relative to the stabilization opioid morphine, fentanyl shows greater bias toward β-arrestin (vs G-protein) signaling, which has also been shown to mediate internalization of the mu opioid receptor.²¹ Fentanyl also binds to alternate sites on the mu opioid receptor and has a slower dissociation compared to morphine; collectively, these features could render morphine uniquely poor at stabilizing patients with fentanyl exposure.^22–25 Third, it was not possible to analyze urine samples from all of the Non-Fentanyl study participants to rule out any fentanyl exposure. However, the requirement to test positive on a traditional opioid urine assay with a morphine-standard (which does not cross-react with fentanyl) indicates opioid use in the non-fentanyl group was at least partially non-fentanyl, whereas all fentanyl participants were confirmed to be fentanyl exposed. The Drug Enforcement Administration forensic testing reports and the City’s Health Department have also indicated low but increasing rates of synthetic opioid penetration prior to 2016,^26–28 further mitigating the concern that the Non-Fentanyl study cohort had fentanyl use. Fentanyl study outcomes could also be influenced by the recent permeation of the potent α-2 adrenergic agonist xylazine into the local fentanyl supply. Although no data on xylazine withdrawal exist, abstinence from similar α-2 adrenergic agonists produces “rebound” hypertension that could explain the diastolic blood pressure changes observed here. Fourth, observers who rated COWS symptoms were not necessarily blinded to fentanyl exposure but were following standard protocols to rate opioid withdrawal severity that should mitigate the potential for bias on study outcomes. Sample size for the Fentanyl group was also relatively small, which may limit power to detect effects between groups leading to type II error. Lastly, given the ever-changing drug supply, it is not possible to conclude with complete confidence that fentanyl is the only factor responsible for the worsening of opioid withdrawal as seen in these analyses. Other synthetic opioids like nitazines and other high potency synthetic opioids may exist in the supply and influence the severity of withdrawal.

CONCLUSION

In summary, these data illustrate how fentanyl is complicating treatment presentation among individuals with OUD. Relative to non-fentanyl opioid-exposed historical controls, fentanyl-exposed individuals experienced a non-linear, more severe, prolonged, and qualitatively unique presentation of opioid withdrawal. This increased withdrawal severity despite receiving morphine and concomitant medications suggests that new strategies to treat fentanyl exposure are necessary and supports research into novel and mechanistically informed methods for opioid withdrawal mitigation. Opioid withdrawal has more recently served as a major barrier and deterrent to treatment entry,²⁹ and the severity of withdrawal experienced following fentanyl is likely to further discourage individuals with OUD from attempting treatments unless rapid treatment innovations are made to address the findings shown here.

Supplementary Material

Supplemental digital content is available for this article. Direct URL citation appears in the printed text and is provided in the HTML and PDF versions of this article on the journal’s Web site (www.journaladdictionmedicine.com).