Introduction

Fibromyalgia (FM) is a commonly recognized syndrome characterized by pain, sleep disturbance, and fatigue combined with a general increase in medical symptoms, including problems of memory or thinking, and often psychological distress [1]. The pharmacological treatment of FM results in only partial relief [2]. This condition is frequently associated with depression and irritable bowel syndrome.

Celiac disease (CD) is a frequent disease, affecting about 1 % of the population that can be diagnosed at all ages under different clinical settings. The classic presentation of CD is chronic diarrhea, positive anti-tissue transglutaminase antibodies (anti-tTG), and villous atrophy with intraepithelial lymphocytosis observed by duodenal biopsy. CD often does not conform to the classical clinical description, and up to 50 % of patients with adult CD lack prominent gastrointestinal symptoms [3, 4]. Clinical manifestations of CD include muscle and osteoarticular pain, fatigue, and neurological and psychological symptoms [5, 6].

Non-celiac gluten sensitivity is increasingly recognized as a frequent clinical condition with symptoms similar to CD in the absence of the diagnostic features of CD [7, 8]. Intraepithelial lymphocytosis without villous atrophy and without specific celiac antibodies has been shown to be part of the spectrum of non-celiac gluten sensitivity. These mild gluten-related lesion have been given different names, such as gluten-sensitive enteropathy or mild celiac enteropathy [912]. The aim of this report is to describe selected patients with FM, non-celiac gluten sensitivity, duodenal intraepithelial lymphocytosis, and clinical response to gluten-free diet, in order to raise the hypothesis that non-celiac gluten sensitivity could be a treatable cause of FM.

Methods

As part of a wider research project approved by the local ethics committee, patients diagnosed with FM and referred to a widespread chronic pain rheumatology unit were given the option to undergo fibrogastroscopy with duodenal biopsies, despite negative anti-tTG assay results, and gave informed consent for fibrogastroscopy knowing the negative results of the CD serology. After the pathology report ruled out CD, consented to participating in an open gluten-free diet trial. Currently, we are following 246 patients of these patients. Clinical response has been observed in 90 of them. This is a report of the first patients for whom duodenal intraepithelial lymphocytosis and a clinical response to a gluten-free diet were observed. These patients also gave additional consent for their cases to be published, including their genetic data.

Duodenal biopsies were obtained from the second duodenal portion before beginning the GFD. These samples were retrospectively evaluated by three pathologists from two academic centers with anti-CD3 T-cell receptor immunohistochemical stains. The identification of intraepithelial lymphocytosis was agreed upon by the three pathologists and considered to be present when the intraepithelial lymphocyte count was more than 25 per 100 enterocytes [1315]. HLA-DRB1, HLA-DQA1, and HLA-DQB1 alleles were determined from blood samples.

A trial of a strict gluten-free diet was conducted with the help and training of the CD Patients Association of Madrid. In addition, iron, vitamin D, and multivitamins supplements with oligoelements were prescribed. A lactose-free diet was also recommended when lactose intolerance was suspected on clinical grounds or demonstrated by a breath hydrogen test after lactose administration. Clinical response was defined as the achievement of at least one of the following scenarios: remission of FM pain criteria, return to work, return to normal life as judged by the patient, or opioid discontinuation.

Results

Twenty patients with FM are described for whom duodenal intraepithelial lymphocytosis and a clinical response to a gluten-free diet were observed. Table 1 shows the clinical and HLA descriptions of each patient. All subjects were female. The mean subject age was 46 years (range 25–73). The mean duration of FM evolution was 11.9 years (range 3–20). Eight patients had been diagnosed with chronic fatigue syndrome. Seventeen patients had a gastrointestinal disorder (irritable bowel syndrome, gastroesophageal reflux, constipation, dyspepsia, hiatal hernia, lactose intolerance, or nonspecific colitis). Other diagnoses were depression (eight patients), migraine (eight patients), and hypothyroidism without improvement in FM with thyroid hormone substitution (three patients). Five patients had features of undifferentiated spondyloarthritis, and one patient had psoriatic arthritis.

Table 1 Descriptions of the twenty patients with fibromyalgia and duodenal IEL and their responses to a gluten-free diet

Eleven patients carried either the DQ2 (DQA1*05DQB1*02) or DQ8 (DQA1*0301DQB1*0302) heterodimers. Seven patients had only one allele of the DQ2 heterodimer, either DQB1*02 or DQA1*05. Two patients did not carry either DQ2 alleles or DQ8.

The mean follow-up period for the gluten-free diet was 16.4 months (range 5–31). Eight patients were also on a lactose-free diet. For five of these eight patients, a lactose-free diet had been attempted before the gluten-free diet, resulting in partial relief of the gastrointestinal symptoms but no improvement in the FM symptoms. For three patients (#4, 10, and 20), a lactose-free diet was started concurrently with the gluten-free diet.

The level of widespread chronic pain improved dramatically for all patients; for 15 patients, chronic widespread pain was no longer present, indicating remission of FM. Fifteen patients returned to work or normal life. In three patients who had been previously treated in pain units with opioids, these drugs were discontinued. Fatigue, gastrointestinal symptoms, migraine, and depression also improved together with pain. Patients #2 and #3, both with oral aphthae, went into complete remission for psoriatic arthritis and undifferentiated spondyloarthritis.

For some patients, the clinical improvement after starting the gluten-free diet was striking and observed after only a few months; for other patients, improvement was very slow and was gradually observed over many months of follow-up. For eight patients (# 2, 3, 8, 9, 12, 14, 17, 19, and 20), the intake of gluten was followed by clinical worsening, which subsided after returning to a strict gluten-free diet.

Discussion

Here, we report the cases of patients with severe longstanding FM in whom CD was ruled out, but had duodenal intraepithelial lymphocytosis and exhibited clinical responses to a gluten-free diet. The observation for FM patients of a histopathologic lesion associated with a treatable disease opens new perspectives.

The concept of non-celiac gluten sensitivity arises from clinical observations of patients whose symptoms clearly improve or resolve with a gluten-free diet even though CD has been ruled out. Non-celiac gluten sensitivity is currently included in the wider concept of gluten-related disorders, along with CD and wheat allergy [16].

In non-celiac gluten sensitivity, specific autoantibodies are not present, and duodenal biopsy only reveals intraepithelial lymphocytosis or no pathologic changes. Only approximately half of patients with GS carry the HLA DQ2 or DQ8 heterodimers, in contrast to CD in which almost all patients carry the HLA DQ2 or DQ8. The symptoms of non-celiac gluten sensitivity include behavioral changes, bone or joint pain, muscle cramps, leg numbness, weight loss, chronic fatigue, and a foggy mind. Oral aphthae and psoriasis can also be due to non-celiac gluten sensitivity [7]. Intraepithelial lymphocytosis has been the hallmark of gluten intestinal lesions since the first descriptions of the intestinal pathology of CD [17]. In the current classification criteria, a Marsh 1-type lesion is defined as the presence of more than 25 intraepithelial lymphocytes per one hundred enterocytes in a duodenal biopsy. Intraepithelial lymphocytosis can be overlooked or not considered relevant in a pathology report focused on villous atrophy or can be missed if anti-CD3 staining is not performed [13, 14]. Although intraepithelial lymphocytosis is associated with several other conditions, such as Helicobacter pylori disease, NSAID intake, Crohn’s disease, or parasitic infestation, gluten enteropathy must always be considered [1820]. For our patients, intraepithelial lymphocytosis was found by experienced pathologists after a specific search. Two examples of intraepithelial lymphocytosis are shown in Fig. 1, which presents images of CD3-stained preparations. The high frequency of HLA alleles related to susceptibility to CD in our patients is similar to that described in non-celiac gluten sensitivity.

Fig. 1
figure 1

a CD3 staining of intraepithelial lymphocytes in the duodenal villi of patient #8. b CD3 staining of intraepithelial lymphocytes in the tips of the duodenal villi of patient #1

The spectrum of symptoms of our patients is wide and complex, as is typical for FM and described in non-celiac gluten sensitivity. The reduction in the level of pain was accompanied by improvements in asthenia and gastrointestinal and neurological symptoms, suggesting a common underlying cause related to gluten. The clinical response definition for this report was decided upon after the initial observations of impressive improvement in some patients after starting the gluten-free diet. We chose demanding outcome measures, such as the remission of FM or return to normal life, rather than changes reflected by questionnaires. To our knowledge, this degree of improvement has not been reported previously for FM. It is unlikely that our results were caused by a placebo effect. The majority of patients had severe distress and disability and were unable to cope normally with daily activities or on sick leave, despite having been treated with many different regimens for years. In FM clinical trials, the placebo effect has been very modest and transient, in contrast to the great long-term improvement of our patients. Lactose-free diet and the addition of vitamins could be confounders, but the majority of the patients with lactose-free diet had tried it before without improvement of FM, and in our experience FM does not have relevant clinical improvement just adding vitamins and minerals. Furthermore, the reintroduction of gluten, for 7 patients, was followed by FM relapse; when re-introduced, the strict gluten-free diet led to clinical improvement.

This report describes the first patients for which we have observed FM with intraepithelial lymphocytosis and the response to a gluten-free diet. Currently, we are following 246 patients with FM on a strict gluten-free diet, and the degree of clinical improvement described in this report has been observed in 90 of them. We still do not have long-term follow-up and CD3 staining data for all of our patients. However, it is worth communicating our first observations, which open a wide field for research. Indeed, non-celiac gluten sensitivity as a cause of FM, and the role of duodenal intraepithelial lymphocytosis, needs to be tested in a controlled trial.

Conclusion

This report shows that remarkable clinical improvement can be achieved with a gluten-free diet in patients with FM, even if CD has been ruled out, suggesting that non-celiac gluten sensitivity may be an underlying treatable cause of FM syndrome. The presence of intraepithelial lymphocytosis in the duodenal biopsies of these selected patients further supports this hypothesis.