Successful Reduction of Hepatocellular Haemosiderin Content in Toco toucans (Ramphastos toco) with Haemochromatosis by Dietary Modification

*Bristol Zoo Gardens, Guthrie Road, Clifton, Bristol BS8 3HA, United Kingdom (Redrobe).

**Department of Pathology and Infectious Diseases, The Royal Veterinary College, Hawkshead Lane, Hatfield, Hertfordshire AL9 7TA, United Kingdom

Author for correspondence:

Sharon Redrobe. BSc (Hons) BVetMed CertZooMed MRCVS, Head of Veterinary Services, Bristol Zoo Gardens, UK. Email: sredrobe@bristolzoo.org.uk

Abstract: This project was conducted to determine whether dietary modification can result in significant reductions in hepatocellular haemosiderin content, by computerised image analysis of serial liver biopsy specimens. Two adult Toco Toucans (Ramphastos toco) previously diagnosed with haemochromatosis, were used in this study. Liver biopsy specimens were taken before the birds were placed on a low iron diet and again 8 months later; one bird died from other causes at 12 months, and a post-mortem specimen was also analysed. The post-diet liver biopsies from both birds showed significant reductions in hepatocellular haemosiderin content (P<0.001). Computerised image analysis is a valid means of quantifying relative iron content in serial liver biopsies, and can be used to monitor how well a bird is responding to treatment over time.

Key words: hemochromatosis, toucans, Ramphastos toco, image analysis, iron

Hemochromatosis is defined as excessive accumulation of iron in the liver and other major organs, with associated functional or morphologic evidence of iron toxicity (Lowenstine and Munson 1999). This disease may be primary (inherited) or secondary to other factors including diet, infections, toxicities, and certain types of anaemia (Lowenstine and Munson 1999). Certain avian species have a high incidence of iron storage disease, including toucans (Ramphastidae), mynahs (Sturnidae), quetzals, and birds of paradise (Paradisaeidae) (Lumeij 1994). Genetic factors have been suspected but not positively identified, and there has been some debate as to the contribution of dietary iron levels to development of the disease.

Of the Ramphastidae, toco toucans (Ramphastos toco) are one of the species particularly susceptible to hemochromatosis, which is a common cause of death of captive birds and a possible significant factor in the infrequency of successful captive breeding (Cornelissen and Ritchie 1994, Lowenstine and Munson 1999). Birds may die acutely or present with more chronic signs including emaciation, dyspnoea, ascites, equilibrium disturbances and feather picking (Roels 1996). It has been suggested but not proven that species such as toucans that originate from tropical areas consume food in the wild that is very low in iron, and have therefore adapted highly efficient mechanisms for sequestration of this mineral. Normal hepatic iron levels in toucans are not well established, in particular how levels in captive birds compare with those of wild populations.

Treatment of hemochromatosis involves removal of iron from the erythron by phlebotomy and from the tissues by chelation (Lowenstine and others 1986). Daily injection of iron-chelating agents has had variable clinical results, however one group reported marked reductions of hepatic iron levels with deferoxamine treatment (in conjunction with low iron diets) in 2 channel-billed toucans (Cornelissen et al. 1995, Roels et al. 1996). Both phlebotomy and injection of chelation agents are stressful procedures for captive birds. Dietary modification is not of use in humans in which the disease is due to a genetic defect in small intestinal absorption. For species in which the genetic basis is not known, reduction of dietary iron may have an effect. Citrus fruits or other foods high in Vitamin C are excluded, as ascorbic acid is known to increase iron bioavailability in many species. Tea has been added to food or water of captive ramphastids as it is believed that wild birds drink rainwater from tree cavities which has been leached by plant tannins; controlled studies on the actual effect of tea on iron bioavailability in these species have not been published.

Levels of total serum iron, plasma iron and total iron-binding capacity do not correlate with tissue iron content in toucans (Worell 1993). Definitive diagnosis of hemochromatosis is therefore dependent on histologic and/or toxicologic examination of liver biopsy specimens (Lowenstine and Munson 1999). Iron accumulates in the cytoplasm of hepatocytes and epithelioid macrophages; the latter form nodules and are thought to arise from Kupffer cells (Worrell 1993). The success or failure of treatment can only be objectively monitored by quantitative examination of serial liver specimens, which has included image analysis of histologic sections (Cornelissen et al. 1995, Roels et al. 1996). The effect of dietary modification alone has not been monitored in this way. The aim of the current study was determine whether reductions occurred in hepatocellular iron levels in two toco toucans with hemochromatosis over an 8 month period of dietary modification, by histologic image analysis of liver biopsy specimens.

Two Toco toucans (Ramphastos toco) aged unknown, had been maintained in captivity for 10 years. They were presented for a health check as no breeding had been observed. Each animal was examined under general anaesthesia. General anaesthesia was achieved using isoflurane 5% via mask for induction then intubation and maintenance on 1.5% isoflurane in oxygen. Survey radiographs were taken. Blood samples were taken from the ulnar vein using aseptic technique, transferred to a lithium heparin blood tube and analysed for iron, and total iron binding capacity (Greendale Laboratories, Surrey, UK). Liver biopsy specimens were collected using aseptic technique via endoscopic biopsy (Stortz 2.8mm endoscope using 5mm biopsy forceps). The biopsy specimens were submitted in 10% neutral buffered formalin to the Department of Pathology and Infectious Diseases, The Royal Veterinary College (UK).

Dietary modification was instituted after the first biopsy. The previous diet had consisted of a ‘usual diet’ of a mixture of pulses, vegetables and fruits and a ‘summer diet’ that included a mouse pup to improve condition. The diet was modified to reduce iron, lower vitamin C and add increase tannins. Hence the new diet involved the exclusion of all citrus fruits, and soaking the diet in tea every other month. Feeding mouse pups to the birds during the summer months was also stopped; these had been added by the keepers to help improve condition, but were not intended as part of the original diet. A second biopsy specimen was collected after the toucans had been maintained on the lower iron diet for eight months. One bird died due to other causes four months after the second liver biopsy, and the post-mortem specimens were also included in this study.

Each diet was analysed for iron content using the Zootrition software (Zootrition, 1999: Wildlife Conservation Society, USA) and 100g of the new diet was analysed at a commercial laboratory for confirmation (Promar Laboratories, Surrey, UK).

The biopsy specimens were submitted in 10% neutral buffered formalin to the Department of Pathology and Infectious Diseases, The Royal Veterinary College (UK) and a diagnosis of haemochromatosis was made based on the histological examination.

Fixed biopsy specimens were processed routinely and embedded in paraffin wax. For image analysis, five micrometer sections were cut and stained with either hematoxylin and eosin (HE) or Perls’ Prussian blue stain (PB). Histopathologic lesions were noted in the HE-stained sections.

Image ProPlus software (4.0 Windows, Media Cybernetics) was used to analyse PB-stained sections, at a 200 X magnification. The same microscope light intensity was used for all slides. Gray-scale 8 digital images of representative fields were obtained from each section, or set of sections. An AOI (Area of Interest) box measuring 0.1 x 0.1 mm was created, which allowed selection of areas containing hepatocellular cords only. Areas containing nodular macrophage aggregates or artefactual changes were not analysed. Maximal possible numbers of AOIs for each biopsy specimen were analysed, which ranged from 20-60; 100 were analysed in the post-mortem specimen from toucan 2. For each AOI, an intensity histogram operating on a scale 0 - 255 with zero being black (more intensely stained with PB) and 255 white was generated. From the numerical data for each histogram, a mean intensity value was calculated, termed the "intensity score".

Intensity score distributions for each biopsy specimen were compared with all others from the same bird and those from the other bird by the two-way analysis of variance (ANOVA) method (SPSS 10.0 for Windows, SPSS Inc., Chicago, IL). Statistical difference was indicated when P < 0.05.

Previous summer diet (including mouse pup and citrus fruits) = 291.50 mg/kg

Previous Usual diet (summer diet without mouse pup) = 110.20 mg/kg

Modified lower iron diet (exclusion of citrus fruits and mouse pup) =98 mg/kg

In all sections, there was severely distorted hepatic architecture with numerous, dense nodular clusters of epithelioid macrophages. The cytoplasm of hepatocytes was swollen, and both hepatocytes and macrophages were heavily loaded with brown pigment, which stained positively with PB (hemosiderin).

The intensity scores for each biopsy specimen had a normal distribution, with a skewness statistic of <1.0. There was a highly significant difference between pre-diet and post-diet biopsy specimens in both toucans (P < 0.001).

This study demonstrated a quantitative reduction in hepatocellular haemosiderin content in 2 toco toucans over an 8-12 month period. Only the hepatocellular hemosiderin levels were measured as it has been suggested that monitoring those levels is the most effective method of determining whether there has been a significant response to treatment (Cornelissen et al. 1995).

Iron initially accumulates in hepatocytes, and ultrastructurally is seen to be both membrane-bound and free in the cytosol (Lowenstine and Petrak 1980, Worrell 1993). Epithelioid macrophage accumulations begin in the sinusoids (Worrell 1993), and are more chronic lesions with less functional relevance.

A histological analysis method was chosen for several reasons. Chemical methods are expensive, and it is generally more convenient for veterinarians to fix and submit biopsy specimens rather than arranging for transport of fresh tissue to the laboratory. Also, the specimens are preserved on slides and in the paraffin blocks for future evaluation. An additional advantage of histological analysis is that it indicates where in the liver the iron is located; initial and significant reductions in iron levels in treated birds have been shown to occur in the hepatocytes, with significant amounts still remaining in the non-functional, macrophage nodules (Cornelissen et al. 1995, Roels et al. 1996).

Blood analysis was not useful. Although the iron level of the female bird reduced, it increased in the male; this result is not echoed in the histological analysis. Further, these values were within normal limits yet the hepatic biopsies diagnosed severe hemochromatosis.

In this study only the relative difference in hepatocellular iron content was measured between serial samples from the same bird. In two previous studies where iron levels were monitored in liver biopsy specimens from channel-billed toucans, good correlation was shown between chemical and histological analysis methods (Cornelissen et al. 1995, Roels et al. 1996). We did not correlate intensity scores with iron concentration values. The absolute values for intensity scores may change when sections are analysed under different lighting conditions, and could differ between staining runs. Also, normal hepatic iron concentrations in wild and captive toucans are not known, although in most species they range between 100 and 300 ppm (Worell 1993). The aim of this study was to detect any improvement in response to the dietary changes.

Two individual channel-billed toucans (Ramphastos vitellinus) treated by daily injection of the chelating agent deferoxamine and modification of the diet showed significant reductions in hepatic iron levels over 4 and 6 months respectively (Cornelissen et al. 1995, Roels et al. 1996). In both of those cases, numbers of iron granules in liver cells decreased to the point that they had almost disappeared. Hepatocellular iron levels in the birds in the current study did not reduce as markedly, however they were not subjected to the stress of daily injections of a chelating substance and/or weekly phlebotomy. Stresses associated with these frequent and invasive procedures were not considered to be acceptable due to the marked loss of appetite noted following capture and handling.

Otten (2001), analysed food normally eaten by keel-billed toucans in the wild for iron content, and compared those results to a commercially prepared diet specifically marketed as "low iron". He found that the commercial diet to be markedly higher in iron than the normal diet, and suggested that this might contribute to the development of iron storage diseases seen in captive birds and other sensitive species. Another study, looking at twelve commercial mynah feeds, found only three of the twelve were low enough in iron to prevent overload (Schoemaker and others 2001). Furthermore, in all feeds Schoemaker examined, the iron content was at least three times greater than the currently suggested iron requirement of 25 mg/kg. Ernest (1989) suggested that since pelleted foods often contain too much iron, feeding a diet of cooked rice: canned cat food in a 2: 1 ratio was acceptable. Schoemaker (2001), however, pointed out that providing cooked rice or fruit is not enough to significantly reduce the iron intake if commercial feed is also offered as an adjunct. The only factor known to have altered following the initial biopsy procedure in these 2 birds was modification of the diet to reduce iron content. The iron level in this diet was lowered from 291.5 mg/kg to 98 mg/kg and was associated with a significant reduction in hepatocellular haemosiderin content.

The significant relative reduction in hepatocellular iron content in the toucans in this study in response to dietary modification alone is encouraging. Further work will be required to investigate the long-term effects of various diets in this species, and computerised image analysis of histological sections from liver biopsy specimens is an effective and objective monitoring method.

Thanks to all the keepers at Bristol Zoo Gardens who care for these toucans, especially Nigel Simpson (bird keeper) and Duncan Bolton (Curator). Thanks to Kellie Wyatt (Veterinary Nurse) and John Partridge (Assistant Curator) for assisting with the procedures. Thanks to Greendale Laboratories, Uk for analysing the blood samples and to Promar Laboratories, UK for analysing the diet.

Lumeij JT. Hepatology. In: Ritchie BW, Harrison GJ, Harrison LR, eds. Avian Medicine: Principles and Application. Lake Worth, FL: Wingers Publishing Inc.; 1994:522-537.

Cornelissen H, Ritchie BW. Ramphastidae. In: Ritchie BW, Harrison GJ, Harrison LR, eds. Avian Medicine: Principles and Application. Lake Worth, FL: Wingers Publishing Inc.; 1994:1276-1283.

Cornelissen H, Ducatelle R, Roels S. Successful treatment of a channel-billed toucan (Ramphastos vitellinus) with iron storage disease by chelation therapy: sequential monitoring of the iron content of the liver during treatment period by quantitative chemical and image analysis. J Avian Med Surg 1995;9:131-137.

Lowenstine, LJ, Munson L. Iron overload in the animal kingdom. In: Fowler ME, Miller RE, eds. Zoo and Wild Animal Medicine Current Therapy 4. Philadelphia, PA: WB Saunders; 1999:260-268.

Roels S, Ducatelle R, Cornelissen H. Quantitative image analysis as an alternative to chemical analysis for follow-up of liver biopsies from a toucan with haemochromatosis. A technique with potential value for the follow-up of haemochromatosis in humans. Anal Quant Cytol Histol 1996;18:221-224.

Lowenstine, LJ, Petrak ML. Iron pigment in the livers of birds. In: Montali RJ, Migaki G, eds. The Comparative Pathology of Zoo Animals. Washington, D.C.: Smithsonian Institution Press; 1980:127-135.

Worrell, AB. Further investigations in Rhamphastids concerning hemochromatosis. Proc Assoc Avian Vet 1993:98-107.