Although pellagra used to be a disease of epidemic proportions in the developed world, it is now a rare condition confined to certain at-risk groups. Little progress has been made in our knowledge of pellagra since it was first described by Gaspar Casal in 1735 and since Goldberger discovered that nicotinamide was a preventive factor in 1926.1,2 Proof of this is that there have been no changes in treatment or diagnostic criteria in the last 90 years. The main difference between now and 90 years ago is probably our limited current experience due to the rarity of the disease. The aim of this study was to report our experience with pellagra in a series of patients at our hospital.

We reviewed the clinical and histologic records of Hospital Doctor José Molina Orosa in Lanzarote, in the Canary Islands, Spain in search of diagnoses of pellagra and pellagra-like syndrome between 1998 and 2009. The inclusion criteria were (a) clinical findings consistent with pellagra (more or less symmetrical skin lesions in sun-exposed areas) or histologic findings consisting of pallor in the upper layers of the epidermis with ballooning of keratinocytes; (b) resolution of symptoms following administration of nicotinamide; and (c) exclusion of other diseases. To be included in the study, patients had to fulfill all 3 criteria.

We reviewed the medical records and noted data on age, sex, predisposing factors for pellagra, substance abuse (in particular excessive alcohol consumption), social situation, clinical manifestations, abnormal laboratory results, initial suspected diagnosis, treatment, and long-term outcome. We also examined the histology reports of patients who underwent biopsy.

Incidence and prevalence data were based on annual hospital records and demographic data were compiled from the council data center of Lanzarote.3

Seven patients were included in the study. Tables 1–3, respectively, show the corresponding epidemiological, clinical, and histologic data.

All the patients were men aged between 37 and 72 years at the time of diagnosis. The mean age was 55 years. The annual incidence was 0.5 cases per 100000 inhabitants. Based on data from the hospital, 1 in every 4663 new patients seen by the dermatology department was diagnosed with pellagra.

The most common predisposing factors were excessive alcohol consumption and poor diet. It is noteworthy that 2 patients were being treated for epilepsy and that this treatment was the only predisposing factor in 1 of these (Table 1).

The back of the hands and/or the forearms were affected in all 7 patients (Table 2) and all the lesions were bilateral and symmetrical (Fig. 1). The face and the upper surface of the feet and/or the legs were affected in fewer patients. The lesions on the face were mild, but there were blisters and ulcers on the feet and/or the legs (Figs. 2 and 3).

Figure 1.

Patient #1. Erythematous whitish scar-like lesions with several erosions and surrounded by brown, parchment-like skin. Note the sharp demarcation between the lesions and the normal skin in the wrist area.

(0.14MB).

Figure 2.

Patient #4. Pellagra affecting the upper surface of the feet. Note the erythematous eruption accompanied by considerable swelling, blisters, and erosions produced by the rupture of blisters.

(0.09MB).

Figure 3.

Patient #3. Note the brown scaling rash on the forearms and in particular the large blisters on both hands.

(0.11MB).

Pellagra was only suspected initially in 3 patients, even though they all had characteristic clinical manifestations. Other diagnoses considered were lupus, photocontact dermatitis, polymorphic light eruption, and pseudoporphyria cutanea tarda.

Almost all of the patients had extracutaneous symptoms, but these were mild and were only detected during a structured interview. Two patients had an impaired oral mucosa membrane (bald tongue and atrophy of the mucosa). There were also gastrointestinal symptoms (diarrhea) and neurological symptoms (bradypsychia and slight disorientation).

Laboratory tests were only of value for ruling out other diseases.

The outcome was favorable in all patients but resolution was slower in 2 patients as niacinamide treatment was not administered correctly initially.

The most common histologic findings were the presence of dilated vessels associated with extravasated red blood cells, little or no inflammatory infiltrate in the dermis, and compact or parakeratotic hyperkeratosis in the epidermis (Table 3). There was some degree of ballooning of keratinocytes and epidermal pallor in 4 cases but it was so slight that it did not raise suspicion of pellagra (Fig. 4). Three patients had subepidermal blisters. A direct immunofluorescence study performed in 1 patient (#3) was negative.

Figure 4.

Patient #5. Histologic image showing parakeratotic hyperkeratosis together with a somewhat atrophic epidermis characterized by mild spongiosis, pallor, and ballooning. Dilated vessels with extravasation of red blood cells can be observed in the dermis (hematoxylin–eosin staining, original magnification ×200).

(0.14MB).

Pellagra is endemic in parts of the developing world and there have also been reports of outbreaks in areas affected by famine or war.4 It is extremely rare in developed countries, where it is limited to a small number of at-risk groups. Its incidence in Spain is unknown as only a few isolated cases and small series have been published.5,6 The only reference is a series described by Rodríguez-Cuartero et al.6 in which 0.23% of patients admitted to a hospital had pellagra. Based on our findings, the annual incidence of pellagra is 0.5 cases per 100000 population.

Pellagra is caused by a deficiency of the coenzymes nicotinamide adenine dinucleotide (NAD) and NAD phosphate (NADP), which participate in numerous processes including glycolysis, the metabolism of amino acids and proteins, and the generation of high-energy phosphate bonds. This would explain why pellagra mainly affects tissues with a high rate of cell turnover, such as the skin and the digestive tract, and tissues with high energy needs, such as the brain.2,7,8 Humans synthesize NAD and NADP from niacin (or its amide niacinamide), which is absorbed directly through the intestinal wall or synthesized from the essential amino acid tryptophan.

Table 4 shows the predisposing factors for pellagra in the developed world. Alcoholism is the most important factor and patients with excessive alcohol intake probably have dietary deficiencies in addition to a greater need for niacin.9 In our series, 4 of the 5 patients who consumed excessive amounts of alcohol had a poor diet. Depletion of intracellular NAD levels has been observed in patients with human immunodeficiency virus (HIV) infection,10 and these patients also probably have a greater need for niacin.11 Nonetheless, and despite the publication of isolated reports describing an association between niacin depletion and HIV,12,13 HIV-positive patients do not appear to have a higher incidence of pellagra.13,14

Other important factors that predispose to pellagra detected in our series were dietary deficiencies and use of antiepileptic drugs. Our findings also suggest that social situation plays an important role in the development of pellagra; 3 of the patients lived alone and a fourth lived with his son, and they were all responsible for preparing their own meals and managing their diet (Table 1).

Although pellagra was traditionally associated with the 3 D's (dermatitis, diarrhea, and dementia)—and possibly with a fourth (death)—this classic clinical pattern is now rare and only some of these manifestations are present.15 In our series, just 1 patient had cutaneous, digestive, and neurological symptoms. Because the digestive and neurological symptoms are varied and nonspecific, pellagra is generally diagnosed on the basis of skin manifestations, which are much more specific.16

The clinical manifestations in our series were quite characteristic of pellagra but this diagnosis was only suspected in 3 of the 7 patients initially. The presence of photosensitivity was recorded in all cases but it raised suspicion of conditions that are more common in our setting. The extracutaneous manifestations, which can help to reach a diagnosis, were not reported spontaneously and were only detected during a structured interview. This is probably because the patients did not associate these symptoms with their skin lesions and did not consider them to be important because of their mild nature.

There are currently no tests or laboratory markers of diagnostic value in pellagra. Indirect signs of malnutrition can help to confirm a diagnosis, but in our case, these were only evident in patient #3, who had marked intestinal malabsorption. Abnormal 5-OH-indoleacetic acid levels in a 24-hour urine test have been proposed as a potential marker for pellagra. Elevated levels may be due to a carcinoid tumor, which is a known risk factor for pellagra, and reduced levels could indicate tryptophan sequestration for the synthesis of niacin, causing depleted serotonin levels. Nonetheless, no cutoff points have been established and in our series at least, determination of 5-OH-indoleacetic acid levels was not of value. Blood levels of NAD and NADP have not been found to correlate well with tissue levels in patients with pellagra as they are not reduced.17 Seal et al.4 measured the levels of the niacinamide metabolites 1-methylnicotinamide and 1-methyl-2-pyridone-5-carboxymide in relation to blood creatinine levels in an Angolan population with and at risk of pellagra. While the metabolites were reduced in patients with pellagra, they were also reduced in 29.4% of the general population, suggesting that determination of these metabolites might be useful for identifying vulnerable populations in developing countries. It remains to be determined, however, whether such testing might be of diagnostic value in the developed world.

The histologic features of pellagra vary throughout the course of the disease and are nonspecific in many cases.18,19 Suggestive signs are prominent pallor and extensive ballooning of keratinocytes in the upper third of the epidermis.20 None of our patients had characteristic histologic features of pellagra. The most common changes observed were hyperkeratosis of the epidermis and dilation of the superficial vascular plexus associated with extravasated red blood cells in the dermis. Subepidermal blistering was observed in 3 cases. It is important to note that blisters in pellagra are generally intraepidermal due to excessive ballooning of keratinocytes,20 but that subepidermal blisters can also form,18,19,21 possibly through a mechanism similar to that seen in sunburn.

Treatment essentially consists of correcting predisposing factors (e.g., suppression of alcohol consumption and prescription of nicotinamide supplementation). Nicotinamide is preferred to nicotinic acid as the latter can cause intestinal disorders, hyperuricemia, and vasomotor disorders.18 The starting dose should be 100mg/6h, followed by 50mg/8h when the patient improves. Zinc, pyridoxine, and magnesium supplements should also be prescribed as these have been seen to act as cofactors in niacin synthesis.2,9

Interestingly, there might be an increased risk of pellagra in the Canary Islands due to the confluence of several factors, including high alcohol consumption, considerable sun exposure, and a diet rich in gofio (toasted cornflour). It should be recalled that the replacement of wheat with corn as a staple food in Spain was the main cause of the pellagra epidemic that swept the country in the 18th century.

Few conclusions can be drawn from our study due to its retrospective design and small sample size, but a number of interesting observations can be made: (1) pellagra is rare but does exist in certain at-risk groups such as individuals with chronic alcoholism or dietary deficiencies, or patients taking certain drugs; (2) it should be included in the differential diagnosis of any condition with photosensitivity; (3) its detection requires a high degree of diagnostic suspicion; and (4) social situation should be considered in the assessment of patients.

The authors declare that they have no conflicts of interest.