Results: Outcome data were available in 92.2% of patients with an average follow-up time of 1.7 years. A positive treatment response to LDN was reported by 73.9% of the patients. Most patients experienced improved vigilance/alertness and improved physical and cognitive performance. Some patients reported less pain and fever, while 18.3% of patients did not report any treatment response to LDN. Mild adverse effects (insomnia, nausea) were common at the beginning of the treatment. Neither severe adverse effects nor long-term adverse symptoms were reported.
Conclusions: The high frequency of treatment response and good safety profile observed in this retrospective open label study could prompt prospective controlled studies to confirm the feasibility of LDN in alleviating ME/CFS symptoms.
Low-dose naltrexone in the treatment of myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS)
Received 13 Sep 2019, Accepted 11 Nov 2019, Published online: 19 Nov 2019
- Download citation
It is typical of myalgic encephalomyelitis (ME)/chronic fatigue syndrome (CFS) that exercise, which previously provided a source of well-being and an energy boost for these patients, now causes malaise and exhaustion . More generally, the physical, cognitive or stress-related exertion required during everyday life may suffice to induce pervasive fatigue in this illness, which is not relieved by sleep. Post-exertional malaise (PEM) is a symptom that is specific to ME/CFS, separating it from other fatigue states. Systemic exertion intolerance disease, SEID, is a new, more specific name proposed for the illness . The SEID diagnosis requires that compared to pre-illness levels, the ability of the patient to engage in occupational, educational, social or personal activities is substantially reduced for more than six months. In addition the diagnosis requires non-refreshing sleep and either cognitive impairment or orthostatic intolerance.
ME/CFS is not a rare disease. A large community sample collected via phone interviews indicated a prevalence of 0.42% . ME/CFS limits the functionality and work capacity of most patients. In the most severe cases, ME/CFS leads to loss of ambulation, severe cognitive symptoms and sensory oversensitivity affecting social interaction. For many years, ME/CFS symptoms were thought to have psychological origin (somatization). However, behavioral treatments such as recently reevaluated cognitive psychotherapy and exercise therapy approaches [3,4] or, alternatively, psychiatric drugs have failed to demonstrate a clear and sustainable treatment response. Currently, there are no widely accepted effective pharmaceutical or non-pharmaceutical treatments for ME/CFS.
Naltrexone is an opioid antagonist formulated as a tablet, with a mode of action similar to naloxone. It is indicated for the treatment of ethanol or opioid addiction in a dose of 50–100 mg/day. At this dosage, naltrexone is believed to block the euphoric effects of endorphins which act to sustain addiction. A low dose (3.0–4.5 mg/day, low-dose naltrexone, LDN) appears to have an opposite effect, enhancing the endorphin effect . This is thought to have relevance in illnesses such as ME/CFS, where insufficient secretion of opioid peptides affects immune response or pain control [6,7], or the release of pro-inflammatory cytokines [8,9]. Recently, LDN was shown to restore the impaired transient receptor potential melastatin 3 ion channel function in natural killer cells of ME/CFS patients . There is only anecdotal information that LDN can improve daily function and quality of life in ME/CFS . In this study, we tested the hypothesis that LDN is safe and effective to alleviate ME/CFS symptoms by retrospective examination of the clinical patient records of 218 ME/CFS patients.
Participants. The study cohort consisted of all the patients who consulted a private Unesta Sleep and Breathing Clinic during the years 2010–2014 (n = 248) who were also (1) diagnosed with or had a previous diagnosis of ME/CFS and (2) started or continued with previously started LDN treatment. The patients were sorted in alphabetical order based on surname, and the medical reports of the first 218 (surnames beginning with A to T) were analyzed further.
Most patients contacted the clinic without referral, but some were referred by their general practitioners. A typical ME/CFS patient to consult Unesta Clinic had been sick for several years and had undergone extensive batteries of diagnostic tests at the primary and secondary healthcare levels. Diseases that could have explained their symptoms had previously been ruled out and included endocrinological, metabolic, infectious rheumatological, hematological, neurological or primary psychiatric diseases. A common scenario was an ME/CFS patient previously diagnosed with depression, but multiple antidepressant therapies were either not tolerated or inefficient. Formal sleep studies were done if sleep apnea or periodic limb movement disorder had not previously been excluded or treated.
In some of the initially evaluated patients, the ME/CFS diagnosis was based on the Fukuda CFS criteria ; however, most patients were diagnosed using the Canadian criteria . Unrefreshing sleep and PEM, the key symptoms of ME/CFS, were retrospectively confirmed in all patients, even when their presence was not mandatory to fulfill the Fukuda criteria. The severity of ME/CFS symptoms ranged from mild to very severe. The most severe cases were bed-ridden or confined to a wheelchair. Most often, the patients were unable to work and their symptoms interfered with their studies, social life or everyday activities. Some patients were still gainfully employed despite symptoms that hindered their work performance.
Treatment. The patients were directed to introduce LDN in a morning dose of 1.5 mg for one week, and then continue with a daily dose of 3.0 mg (1.5 mg twice daily). After six weeks on LDN they were allowed to increase the daily dose to 4.5 mg. Patients were informed about the most common adverse symptoms (nausea, insomnia, worsening of pain) that may appear during the initiation of LDN therapy. They were also told that the efficacy and safety of LDN had not been assessed in systematic studies, and the medication might have adverse effects that have not been previously described. If no treatment effect was observed within the first 6 months, LDN was discontinued.
Treatment responses and adverse effects were registered on control visits or upon renewal of prescriptions. Detailed information about both treatment responses and adverse effects collected during LDN treatment was compiled from patient medical reports. For some patients, feedback about treatment response was unavailable or there were only data from a short follow-up period.
The treatment responses of patients were recorded based on their spontaneous reporting without any structured questionnaire or rating scales. Therefore, only the modality, not the degree of improvement could be analyzed. The magnitude of the treatment response was scored on a scale of 0–6, based on the number of symptom modalities for which they reported any improvement. Improved vigilance/alertness or physical performance and alleviation of cognitive impairment, pain or fever produced one point each. One or more treatment responses outside the above-mentioned symptom modalities yielded one additional point. The treatment responses were calculated using all the patients (n = 218) as the denominator.
Adverse symptom effects were calculated with a similar scoring system. If no adverse effects were reported, the adverse effects score was zero. Insomnia, nausea, dizziness, headache or nightmares scored one adverse symptom point each. Any additional adverse symptoms yielded one adverse symptom point each. The sample size used for calculating the frequencies of adverse symptoms was the number of patients with follow-up data available (n = 201).
The average follow-up period was 1.7 years (range 0.1–6.8 years). Women were overrepresented (77%) compared to men. The demographic data of the cohort are presented in Table 1. Any treatment response to LDN was experienced by a minimum of 73.9% of the patients. The treatment response scoring used in the study is presented in Figure 2. About half of the patients experienced a response with respect to at least two different ME/CFS symptoms. In responders, the extent of the response averaged 2.06 symptoms (men 2.19, women 2.03 symptoms). 18.3% of the sample did not report any response to LDN. 13.8% discontinued LDN for lack of effectiveness after an average of 5.9 months of use. 4.6% were unable to gain benefit because adverse symptoms forced them to discontinue the treatment during the introductory phase before the treatment response could be assessed (after an average of 1 month [1–57 days] of use).
Table 1. Demographic data for the ME/CFS patients at initiation of treatment.
Nearly half of the group (45.8%) did not experience any adverse symptoms from LDN treatment. The other 54.2% reported adverse symptoms, which were mostly mild and temporary. The adverse symptoms occurred only at the beginning of the therapy, or after dose increases. Typically, adverse symptoms were present for only a few days or a few weeks. In the worst cases, the adverse symptoms did not disappear until 1–2 months on treatment. Only one patient reported adverse symptoms during long-term treatment but preferred to stay on LDN in spite of mild nausea.
Serious adverse effects (SAEs) were not reported, although 7.3% of the patients discontinued treatment because of adverse symptoms. The most common reason for discontinuation was nausea (5 patients, 2.5%). Increased anxiety was the reason for early termination in two patients (1.0%).
In this retrospective analysis of clinical patients, 73.9% reported some degree of alleviation of ME/CFS symptoms after initiation of the LDN therapy. The proportion of patients benefiting from the treatment was surprisingly high for an illness which currently lacks a widely accepted effective medical treatment. Based on our retrospective data, LDN may show some potential to alleviate a wide spectrum of ME/CFS symptoms, suggesting that its effect may be targeted towards the core illness processes of ME/CFS (specificity). The high proportion of patients who achieved a treatment response is surprising given that ME/CFS is commonly considered a heterogeneous or multi-etiological illness (sensitivity). If verified in a controlled efficacy trial, a treatment response to LDN could offer the possibility of phenotyping ME/CFS patients according to the different illness processes behind the symptoms. Adverse symptoms were reported by more than half (54.2%) of the patients, but only for a short period during treatment initiation. Most adverse symptoms were mild and only in a small minority of cases (7.5%) resulted in discontinuation of treatment. One patient stayed on LDN for its benefits despite mild long-standing nausea.
Improvement in vigilance/alertness was reported by 51.4% of the patients. Physical performance improved in 23.9% and cognitive dysfunction diminished in 21.1% (Table 2). LDN reduced symptoms of ME/CFS in most patients, but its effect was usually insufficient to eliminate them. In some less symptomatic patients, LDN reportedly made it possible to remain in gainful employment. In patients with more severe ME/CFS symptoms, even a slight symptomatic improvement may improve the quality of life. The initiation of LDN in the most severely ill patients is challenging, as they experienced the highest numbers of initial adverse symptoms associated with LDN. A subset of the patients in this study did not achieve a therapeutic response, because adverse symptoms during treatment initiation made them unable to continue the LDN. Their adverse symptoms could perhaps have been avoided if their LDN therapy been started at even a lower initial dose or if the dose was increased at a slower rate.
Table 2. Most common treatment responses to LDN therapy in ME/CFS (N = 218).
Adverse symptoms of LDN treatment
LDN therapy is particularly intriguing because, with a low dose as used in this study, one might expect a low frequency of adverse symptoms . However, due to small study populations in prior studies, it has been difficult to assess the safety and efficacy of LDN in different illnesses . Currently, there is evidence for the safety of LDN in multiple sclerosis , fibromyalgia  and inflammatory bowel disease [18,19]. In studies in fibromyalgia  and Crohn’s disease , LDN was associated with fewer adverse effects than placebo. Temporary insomnia, bizarre dreams and headache are the most commonly reported adverse effects .
The most common adverse symptoms from LDN treatment experienced by ME/CFS patients were consistent with adverse symptoms reported when treating other illnesses with LDN. However, in ME/CFS, adverse symptoms were more common than expected during the initial phase of treatment. On the other hand, long-term adverse symptoms were practically absent. The most common adverse symptoms (Table 3) in both women and men were insomnia (15.3%) and nausea (15.3%). In women, dizziness (9.0%) and headache (6.5%) were also common. In men, gastrointestinal symptoms (8.5%) and nightmares (6.4%) were most frequent. Less common adverse symptoms are listed in Tables 4–5. The most common adverse symptom that led to discontinuation was nausea (5 patients, 2.5%, Table 6). For the most part, adverse symptoms were mild and only rarely (15 patients, 7.5%) led to discontinuation of treatment. With the exception of one patient, the adverse symptoms disappeared within a few days/weeks or in a few months at the latest.
Table 3. Prevalence of adverse symptoms during the initial phase of LDN treatment (N = 201).
Table 4. Adverse symptoms that occurred rarely (2% or less) during the initiation of LDN (N = 201). These symptoms are common in untreated ME/CFS population.
Table 5. Adverse symptoms that were reported only once (0.5%/N = 201). Connection with LDN therapy unsure.
Table 6. Nausea was the most common reason for discontinuing LDN. Altogether, 15 patients (7.5%) discontinued LDN due to adverse symptoms (N = 201).
Of interest, adverse symptoms included a wide spectrum of autonomic symptoms which were also characteristic of the ME/CFS cohort without treatment. The initial aggravation followed by rapid alleviation of the ME/CFS symptoms raises the possibility that the pharmacological effect of LDN targets an underlying autonomic nervous system dysfunction. In some patients, initial adverse symptoms did not diminish even after several weeks or they reappeared after the medication was resumed following a temporary break. Some patients discontinued therapy for relatively mild adverse symptoms (Table 6). In those cases, it remains unclear whether the adverse symptoms could have been alleviated by a temporary break or a temporary reduction of the dose.
The most severely ill (bedbound) patients had the greatest likelihood and severity of adverse symptoms associated with LDN initiation. In those patients, LDN therapy should be introduced in markedly lower doses and the dose titrated slowly. In the most severely affected ME/CFS patients, the dose titration protocol as well as the efficacy and safety of LDN should be studied separately.
Limitations of the study
The study findings should be considered preliminary as they are based on retrospective data from medical reports compiled as part of a clinical practice the time-period is long of 4–8 years. Also for one of the clinicians, a significant learning process regarding ME/CFS was required. Another study limitation was that no validated fatigue scale or clinical global impression (CGI) rating was used to assess treatment response. It was possible only to count the number of symptoms reported to have improved. This symptom counting measure could perhaps be regarded as a surrogate of the patient global impression (PGI) rating. For instance, one patient may have experienced a substantial improvement in quality of life with a treatment response in one symptom, while another patient could have reported improvement in multiple symptoms, but not necessarily in the one symptom most affecting the functional capacity of the patient.
An inclusion bias may have resulted from the fact that severely ill patients who had not benefited from previous treatment, such as cognitive behavioral therapy or graded exercise, were likely to be overrepresented in our clinical patient cohort. On the other hand, females represented 77% of our study population, which is in line with the epidemiological studies showing the female predominance of ME/CFS. This suggests that both males and females with ME/CFS are searching for medical help and corroborates successful recruitment of a representative population. The age range of our patients was large (19–86 years). The 86-year-old patient without significant co-morbidities went through the same procedures of ME/CFS diagnosis and exclusions as the other patients. We found no support in the literature to exclude this patient just because of her age.
Based on this uncontrolled study, it is not possible to distinguish among subjectively experienced treatment effects, placebo effects and the pharmacological effects of LDN. Also the validating diagnosis for patients’ ‘medically unexplained’ symptoms and feeling understood and accepted by the doctor likely had an impact on patients’ expectations of effectiveness of the prescribed treatment. The effectiveness of LDN was suggested by the return of symptoms to their previous level when LDN therapy was experimentally paused followed by reacquisition of the previous treatment response when LDN treatment was resumed. The symptoms of ME/CFS had a tendency to return gradually, mild at first, but as the pause continued, gradually worsening towards the severity level that experienced before starting the therapy. Interruptions in treatment did not raise any concern about the potential development of addiction or temporary rebound symptoms associated with discontinuation. According to a meta-analysis of 29 studies, the response to placebo treatment in patients with ME/CFS is lower than expected and lower than in several other illnesses . In fibromyalgia, the symptoms of which partially overlap with those of CFS/ME, the placebo response to painkillers decreases the longer the pain persists . Based on the observations in this study and in light of previous findings, the contribution of the placebo effect to the LDN treatment response of ME/CFS patients is likely to be low.
Currently, there is no widely accepted, illness-specific treatment for ME/CFS. As a result, patients are left without effective treatments or exposed to incorrect diagnoses and ineffective or even harmful therapies. To our knowledge, this is the first report on the effectiveness and safety of LDN in treating symptoms of ME/CFS. Despite methodological shortcomings of open and retrospective study design, preliminary safety data collected in a clinical setting may serve to support the planning and approval of prospective randomized controlled studies. Our preliminary effectiveness data and good safety profile of LDN in treating ME/CFS symptoms may encourage prospective placebo-controlled studies with objective outcome measures such as changes in the levels of cytokines, metabolites or other biomarker candidates. Figures 1–4.
Figure 1. The follow-up periods of individual patients who responded to LDN and continued with the treatment. The average follow-up period was 1.7 years (0.1–6.8 years). 91 patients (63%) were followed up for more than one year.
Figure 2. LDN alleviated the symptoms of ME/CFS in a minimum of 73.9% of the patients. Despite symptom alleviation, 1.8% discontinued with LDN because of adverse symptoms. 18.3% were non-responders. In a subset of them (5.5% of the entire cohort) the discontinuation was caused by adverse symptoms (Table 6). 7.8% of patients were lost to follow-up. These patients were included in the non-responder group.
Figure 3. The quantified treatment response in 201 patients on a scale of 0–6. Points (0–5) were scored for improved vigilance/alertness, improved physical performance, improved cognition, pain reduction, less fever. One additional point was scored if the patient experienced improvement in some (one or more) other ME/CFS symptoms of ME/CFS. Patients 1–40 (18.3%) were non-responders.
Figure 4. The frequency of adverse symptoms in 201 individual patients during the initiation of LDN (N = 201).
Ms Maija Haavisto and Dr Derek Pheby are cordially acknowledged for translation and checking the manuscript for the English language.
No potential conflict of interest was reported by the authors.
Notes on contributors
Olli Polo, MD, PhD, is a former professor of Respiratory Medicine at the University of Tampere and a former director of the Unesta Clinic and Research Center. His research focuses on sleep disorders, including sleep apnea, restless legs, insomnia, fatigue and daytime sleepiness. Since 2007 he diagnosed over 600 patients with ME/CFS and developed biomedically-based interventions to control dysautonomia, which is one of the key mechanisms behind the ME/CFS symptoms.
Pia Pesonen is a research assistant with a life-long experience in ME/CFS. She has an important role in research data collection and in dissemination of new biomedical research findings to ME/CFS stakeholders and health professionals in Finland.
Essi Tuominen, who recently graduated as MD, started her research into ME/CFS during her medical studies at the University of Tampere.
- IOM Report. Committee on the diagnostic criteria for myalgic encephalomyelitis/chronic fatigue syndrome, Board on the health of Select populations, Institute of Medicine. Beyond myalgic encephalomyelitis/chronic fatigue syndrome: Redefining an illness. Washington (DC): The National Academies Press (US); 2015.