3 Diabetes secondary to tropical calcific pancreatitis C . S . YAJNIK
Diabetes secondary to calcific pancreatitis was reported in the tropics in the early twentieth century but assumed importance after its description by Zuidema (1959). Geevarghese (1968) highlighted its significant contribution to diabetes in the young from South India and succinctly described its natural history: 'abdominal pain in childhood, diabetes in youth and death at the prime of life'. Different aspects of 'tropical calcific pancreatitis' (TCP) have been extensively discussed in many of the classic reviews (Geevarghese, 1968, 1986; Viswanathan, 1980; Narendranathan, 1981; Pitchumoni, 1984; AbuBakare et al, 1986; Balakrishnan, 1987; Assan et al, 1988; Mohan V e t al, 1988c; Mohan and Alberti, 1990; Yajnik, 1990, 1992). Recent revival of interest in this condition can be ascribed to the report of a W H O Study Group (1985) who named diabetes secondary to TCP 'fibrocalculous pancreatic diabetes (FCPD)' and classified it as one of the two subgroups of the so-called malnutrition-related diabetes mellitus (MRDM), the other group being protein-deficient-pancreatic diabetes (PDPD or PDDM). The W H O study group also felt that it was justifiable to include M R D M as a major clinical subclass of diabetes 'in recognition of its clinical distinctiveness and severity, and its high prevalence in some tropical countries'. This has stimulated research and raised new controversies. The first international meeting on M R D M was held in London under the auspices of the W H O and the Wellcome Trust. The proceedings of this meeting are in press and contain a wealth of information. I have devoted this review mostly to the relatively 'new' aspects of TCP which are likely to occupy the thoughts of workers in this field for some time. DEFINITION AND THE SPECTRUM OF TCP
Until recently there were no uniform diagnostic criteria for TCP; nor have there been any studies of its prevalence in the community. The W H O terminology of FCPD is now used by most but a potential confusion is its classification as a subgroup of MRDM, implying that malnutrition is aetiologically and diagnostically important. As we will see later, the role of malnutrition in the aetiology of TCP is still not fully clear. Some patients Bailli~re's Clinical Endocrinology and Metabolism-777 Vol. 6, No. 4, October 1992 Copyright © 1992, by Bailli~re Tindall ISBN 0-7020-1621-7 All rights of reproduction in any form reserved
778
C.S. YAJNIK
Table 1. Diagnostic criteria for fibro-calculouspancreatic diabetes (FCPD) (Mohan et al, 1988b, 1988c). 1. Occurrence in a 'tropical' country. 2. Diabetes by WHO (1985) criteria. 3. Evidenceof chronic pancreatitis: pancreatic calculi on X-ray or at least three of the following: (a) abnormalpancreatic morphologyby sonography, CT-scan or ERCP; (b) chronicabdominal pain since childhood; (c) steatorrhoea; (d) abnormalpancreatic function test. 4. Absenceof other causes of chronic pancreatitis, i.e. alcoholism, hepatobiliary disease or primary hyperparathyroidism, etc. Features like clinicalmalnutrition, youngage at onset and absenceof ketosis are usefuladjuncts but not diagnosticallyessential. with TCP/FCPD do not show any evidence of malnutrition, and rarely FCPD may be associated with obesity (Mohan V et al, 1990a). We now know that the hallmark of this condition is a chronic calcific pancreatitis which occurs at a relatively young age where alcohol is not an aetiological factor. Mohan reviewed the literature and proposed a set of diagnostic criteria for FCPD which are the most comprehensive to date (Table 1). It should be noted that young age and low BMI ('malnutrition') are of relatively 'minor' significance and not 'absolute' criteria as many previous workers suggested them to be (Ahuja, 1985) and as the W H O classification of FCPD under M R D M assumes. To date there is no specific marker for TCP except pancreatic calculi, although we know that 'non-calcific' cases occur. In the absence of such a criterion, diagnosis remains geographic, partly historic, and 'exclusive'. Clinicians in rural areas have to depend principally on clinical-historical criteria and at the most could have an X-ray of the abdomen. Sonography and pancreatic enzyme measurements (the more sensitive indices) are not available in the remote areas where the disease could be more prevalent. No reliable data on the prevalence of this condition are therefore available. Most of the data are from referral clinics and perhaps biased in favour of more advanced cases. In the South Indian state of Kerala where the largest number of cases have been described so far, hospital statistics suggests that TCP/FCPD is on the decline (Geevarghese, 1986). However in the neighbouring state of Tamil Nadu, Mohan has collected the largest 'active' series of TCP/FCPD patients. A high index of suspicion is important because the classic clinical manifestations of malnutrition are seen infrequently. The true diagnosis is perhaps missed in many patients. A study of pancreatic enzyme concentrations in diabetic patients attending a hospital clinic suggested that exocrine pancreatic involvement is much more common in a tropical country than clinically apparent (Yajnik et al, 1990b), The classic description of patients with large pancreatic calculi may represent only an extreme end of the spectrum ('end stage' pancreatitis) (Yajnik, 1990). One point of interest is the relationship between 'tropical' calcific pancreatitis and non-alcoholic pancreatitis seen in the Western countries. Diabetes in the latter is not called 'FCPD' and is classified as 'secondary
DIABETES SECONDARY TO TROPICAL CALCIFIC PANCREATITIS
779
diabetes'. Both Braganza (1993) and Pitchumoni et al (1993) have proposed unifying hypotheses for the pathogenesis of chronic pancreatitis in these two different situations. Similarly, the relationship of TCP to PDPD (if any) is not known. AETIOLOGY AND PATHOGENESIS We know very little about the aetiology of TCP. There are many hypotheses and many likely candidates. Probably the aetiology is multifactorial. Geographical distribution (by definition) in the tropical belt, mostly in the developing countries, suggests a significant role for environmental factor(s). Two prime suspects are 'malnutrition' and cyanogenic alkaloids (cassava or other food items). Childhood infections, diarrhoeas and attendant changes in pancreatic secretion have also been suggested as significant factors. Recently, 'oxidant stress' damage has been proposed as an important mechanism in pancreatic damage. It is not agreed whether diabetes in TCP is secondary to pancreatitis or a disease entity in its own right. It is well known, however, that pancreatitis precedes diabetes by many years. Recent reports suggest that [~-cell reserve in TCP parallels the exocrine pancreatic reserve, implying that hyperglycaemia is 'secondary' to exocrine pancreatic damage. It could also suggest independent damage to both pancreatic components by a common (or related) aetiological agent(s). Additional susceptibility factors could operate in the evolution of hyperglycaemia. Another vexing problem is the possible role of ductal stones in the pathogenesis of TCP. Many authorities ascribe a primary aetiological role to calculous obstruction in the pathogenesis of 'pancreatitis'. However, a non-calculous 'variety' certainly exists and as a rule calculi seem to develop after 'pancreatitis' is well established. Whether stones contribute to further pancreatic damage by ductal obstruction, or are an inconsequential epiphenomenon is not known. Malnutrition
Severe malnutrition at presentation in many of these patients suggested to many workers a causative role for nutritional deficiencies (Zuidema, 1959; Shaper, 1964; Ahuja, 1985). Pros and cons of a nutritional aetiology for TCP/FCPD have been extensively discussed (Pitchumoni, 1984; Balakrishnan, 1987; Balakrishnan et al, 1993; Bajaj and Subba Rao, 1988; Mohan Vet al, 1988c, 1988d; Yajnik, 1990). Clinical histories in these patients rarely show any evidence of kwashiorkor in childhood or unusually low intake of protein. Malnutrition at presentation could be secondary to severe exocrine and endocrine pancreatic deficiency; a case for pancreatitis-related and diabetes-related malnutrition rather than malnutrition-related diabetes. On the other hand, protein malnutrition has long been known to affect pancreatic function and cause pancreatic fibrosis (Pitchumoni, 1973; Durie et al, 1989). Children with kwashiorkor and adults from famine areas show pancreatic acinar atrophy, reduced enzyme output, and fibrosis of the gland.
780
c . s . YAJNIK
Endocrine pancreatic function (insulin secretion) is also known to be diminished in malnutrition (Baig and Edozien, 1965; Becket et al, 1971; Smith et al, 1975). There is usually a complete recovery of these functions after refeeding, though some studies have shown a long-term defect (Cook, 1967; James and Coore, 1970). Recent animal work has suggested that even mild to moderate protein malnutrition in utero (Snoeck et al, 1990), or in early childhood (Swenne et al, 1987, 1988; Crace et al, 1989) or even later (Khardori et al, 1980) can lead to [3-cell dysfunction and hyperglycaemia in later life. Similar long-term studies are not available in humans. It is possible that a pancreas 'weakened' by 'early life' malnutrition could be more susceptible to many 'toxic' influences to which it might be exposed in later life. Interesting reports from the UK have shown an association of low birth weight (an indicator of intrauterine growth retardation) with 'Type 2' diabetes, hypertension and coronary heart disease in adult life (Barker, 1990); no information is yet available on exocrine pancreatic function. Prospective studies of malnourished children in the developing countries are urgently needed. Our classical thinking on 'malnutrition' is usually restricted to deficiency of calories and proteins. However, micronutrient deficiencies (selenium, vitamins, etc.) could set in motion a series of pathological processes in the pancreas which could culminate in the final picture of TCP. Relative deficiency of certain elements or excess of others might produce deleterious effects. These avenues have not been explored in relation to TCP. An international comparative study of dietary factors suggested that high protein intake was associated with alcoholic pancreatitis and low fat intake with tropical pancreatitis (Durbec and Sarles, 1978). This observation was confirmed recently in Kerala (South India) and Marseille (France) (Balakrishnan et al, 1988b). The biochemical significance of this observation is not clear. Nwokolo and Oli suggested that recurrent infections and a highcarbohydrate, low-protein diet (mostly from cassava) in children in the tropics lead to pancreatic stasis, lamination of secretions and inspissation of mucus, ultimately leading to calcification in the pancreatic duct and chronic pancreatitis (Nwokolo and Oli, 1980). In summary, nutritional factors are probably important in the aetiotogy of this condition but the actual food components and mechanisms of damage are not clear. We also need to know more about the interactions of nutritional factors with other possible aetiological agents (Rao, 1988).
Cyanogenic alkaloids The geographical distribution of TCP in areas where cassava is grown and consumed as a staple food suggest a causative role for cassava in the aetiology of this condition (McMillan and Geevarghese, 1979; McMillan, 1993). A further observation that cassava causes goitres in iodine-deficient areas (elevated concentrations of thiocyanate from cassava in blood inhibit iodine uptake by the thyroid gland) but that goitre and FCPD almost never co-exist suggested to McMillan that a failure of detoxification of cyanide could be responsible. Deficiency of dietary sulfur amino acids (methionine,
DIABETES SECONDARY TO TROPICAL CALCIFIC PANCREATITIS
781
cystine) could hamper detoxification of cyanide to thiocyanate and lead to high levels of 'free cyanide' which might be toxic to [3-cells. Acute cyanide poisoning is known to be associated with hyperglycaemia and McMillan produced hyperglycaemia in rats by feeding cyanide (McMillan and Geevarghese, 1979). The relevance of these animal experiments to the clinical situation is still unclear, especially the role of cyanide in causing exocrine pancreatic damage. Recently Pitchumoni (1988) has suggested a universal role for 'cyanide toxicity' in the pathogenesis of pancreatitis; he proposes that 'alcoholic' pancreatitis in Western countries results from concomitant exposure to cyanide in cigarette smoke while tropical pancreatitis results from cyanogenic alkaloids in the diet, especially from cassava. Braganza (1993) has proposed that the toxicity of cyanogenic alkaloids might operate through increased oxidant stress. However, recent studies in Africa failed to find TCP or FCPD in areas where cassava is eaten as a staple food (Teuscher et al, 1987). These reports cast doubt over the aetiological role of cassava in TCP, but the absence of a 'susceptibility factor' in these populations could also explain the results. Occurrence of TCP in areas where cassava is not eaten is also cited as evidence against an aetiological role for cyanide. However, other cyanidecontaining foodstuffs (ragi, sorghum etc.) could be responsible (McMillan and Geevarghese, 1979; Nagalotimath, 1980; Kakrani et al, 1985; Patole et al, 1988). If cyanogenic alkaloids are indeed responsible, processing of foods to reduce cyanide content or development of low-cyanide varieties will be a simple preventive measure for this condition. The verdict on the role of cassava or other cyanogenic foodstuffs in the aetiology of TCP and FCPD must remain 'open'. Genetic, familial and immunological factors Although the geographical distribution of TCP is highly suggestive of an environmental aetiology, familial clustering has also been reported from South India. Families with more than one TCP patient, sometimes in successive generations, have been described (Pitchumoni, 1970; Geevarghese, 1986; Balakrishnan, 1987). A systematic study from South India looked at oral glucose tolerance, abdominal X-ray, ultrasound of the pancreas, and stool chymotrypsin concentrations in families of 38 FCPD patients (Mohan V et al, 1989a). Twelve percent of parents and 21% of siblings showed some evidence of exocrine pancreatic involvement, and 21% of parents and 11% of siblings revealed previously undiagnosed 'Type 2' diabetes. Unless a specific marker is available it is difficult to decide whether these results reflect genetic transmission or a strong environmental influence. However, consanguinity was common in families with TCP in successive generations, and this might indicate a genetic connection. Genetic markers for Type 1 and Type 2 diabetes (HLA-DQ gene and the 'insulin gene' polymorphism respectively) have been studied in the same population (Kambo et al, 1989). Approximately 40% of FCPD patients showed presence of an HLA-DQ-B marker (p
Diabetes secondary to calcific pancreatitis was reported in the tropics in the early twentieth century but assumed importance after its description by Zuidema (1959). Geevarghese (1968) highlighted its significant contribution to diabetes in the young from South India and succinctly described its natural history: 'abdominal pain in childhood, diabetes in youth and death at the prime of life'. Different aspects of 'tropical calcific pancreatitis' (TCP) have been extensively discussed in many of the classic reviews (Geevarghese, 1968, 1986; Viswanathan, 1980; Narendranathan, 1981; Pitchumoni, 1984; AbuBakare et al, 1986; Balakrishnan, 1987; Assan et al, 1988; Mohan V e t al, 1988c; Mohan and Alberti, 1990; Yajnik, 1990, 1992). Recent revival of interest in this condition can be ascribed to the report of a W H O Study Group (1985) who named diabetes secondary to TCP 'fibrocalculous pancreatic diabetes (FCPD)' and classified it as one of the two subgroups of the so-called malnutrition-related diabetes mellitus (MRDM), the other group being protein-deficient-pancreatic diabetes (PDPD or PDDM). The W H O study group also felt that it was justifiable to include M R D M as a major clinical subclass of diabetes 'in recognition of its clinical distinctiveness and severity, and its high prevalence in some tropical countries'. This has stimulated research and raised new controversies. The first international meeting on M R D M was held in London under the auspices of the W H O and the Wellcome Trust. The proceedings of this meeting are in press and contain a wealth of information. I have devoted this review mostly to the relatively 'new' aspects of TCP which are likely to occupy the thoughts of workers in this field for some time. DEFINITION AND THE SPECTRUM OF TCP
Until recently there were no uniform diagnostic criteria for TCP; nor have there been any studies of its prevalence in the community. The W H O terminology of FCPD is now used by most but a potential confusion is its classification as a subgroup of MRDM, implying that malnutrition is aetiologically and diagnostically important. As we will see later, the role of malnutrition in the aetiology of TCP is still not fully clear. Some patients Bailli~re's Clinical Endocrinology and Metabolism-777 Vol. 6, No. 4, October 1992 Copyright © 1992, by Bailli~re Tindall ISBN 0-7020-1621-7 All rights of reproduction in any form reserved
778
C.S. YAJNIK
Table 1. Diagnostic criteria for fibro-calculouspancreatic diabetes (FCPD) (Mohan et al, 1988b, 1988c). 1. Occurrence in a 'tropical' country. 2. Diabetes by WHO (1985) criteria. 3. Evidenceof chronic pancreatitis: pancreatic calculi on X-ray or at least three of the following: (a) abnormalpancreatic morphologyby sonography, CT-scan or ERCP; (b) chronicabdominal pain since childhood; (c) steatorrhoea; (d) abnormalpancreatic function test. 4. Absenceof other causes of chronic pancreatitis, i.e. alcoholism, hepatobiliary disease or primary hyperparathyroidism, etc. Features like clinicalmalnutrition, youngage at onset and absenceof ketosis are usefuladjuncts but not diagnosticallyessential. with TCP/FCPD do not show any evidence of malnutrition, and rarely FCPD may be associated with obesity (Mohan V et al, 1990a). We now know that the hallmark of this condition is a chronic calcific pancreatitis which occurs at a relatively young age where alcohol is not an aetiological factor. Mohan reviewed the literature and proposed a set of diagnostic criteria for FCPD which are the most comprehensive to date (Table 1). It should be noted that young age and low BMI ('malnutrition') are of relatively 'minor' significance and not 'absolute' criteria as many previous workers suggested them to be (Ahuja, 1985) and as the W H O classification of FCPD under M R D M assumes. To date there is no specific marker for TCP except pancreatic calculi, although we know that 'non-calcific' cases occur. In the absence of such a criterion, diagnosis remains geographic, partly historic, and 'exclusive'. Clinicians in rural areas have to depend principally on clinical-historical criteria and at the most could have an X-ray of the abdomen. Sonography and pancreatic enzyme measurements (the more sensitive indices) are not available in the remote areas where the disease could be more prevalent. No reliable data on the prevalence of this condition are therefore available. Most of the data are from referral clinics and perhaps biased in favour of more advanced cases. In the South Indian state of Kerala where the largest number of cases have been described so far, hospital statistics suggests that TCP/FCPD is on the decline (Geevarghese, 1986). However in the neighbouring state of Tamil Nadu, Mohan has collected the largest 'active' series of TCP/FCPD patients. A high index of suspicion is important because the classic clinical manifestations of malnutrition are seen infrequently. The true diagnosis is perhaps missed in many patients. A study of pancreatic enzyme concentrations in diabetic patients attending a hospital clinic suggested that exocrine pancreatic involvement is much more common in a tropical country than clinically apparent (Yajnik et al, 1990b), The classic description of patients with large pancreatic calculi may represent only an extreme end of the spectrum ('end stage' pancreatitis) (Yajnik, 1990). One point of interest is the relationship between 'tropical' calcific pancreatitis and non-alcoholic pancreatitis seen in the Western countries. Diabetes in the latter is not called 'FCPD' and is classified as 'secondary
DIABETES SECONDARY TO TROPICAL CALCIFIC PANCREATITIS
779
diabetes'. Both Braganza (1993) and Pitchumoni et al (1993) have proposed unifying hypotheses for the pathogenesis of chronic pancreatitis in these two different situations. Similarly, the relationship of TCP to PDPD (if any) is not known. AETIOLOGY AND PATHOGENESIS We know very little about the aetiology of TCP. There are many hypotheses and many likely candidates. Probably the aetiology is multifactorial. Geographical distribution (by definition) in the tropical belt, mostly in the developing countries, suggests a significant role for environmental factor(s). Two prime suspects are 'malnutrition' and cyanogenic alkaloids (cassava or other food items). Childhood infections, diarrhoeas and attendant changes in pancreatic secretion have also been suggested as significant factors. Recently, 'oxidant stress' damage has been proposed as an important mechanism in pancreatic damage. It is not agreed whether diabetes in TCP is secondary to pancreatitis or a disease entity in its own right. It is well known, however, that pancreatitis precedes diabetes by many years. Recent reports suggest that [~-cell reserve in TCP parallels the exocrine pancreatic reserve, implying that hyperglycaemia is 'secondary' to exocrine pancreatic damage. It could also suggest independent damage to both pancreatic components by a common (or related) aetiological agent(s). Additional susceptibility factors could operate in the evolution of hyperglycaemia. Another vexing problem is the possible role of ductal stones in the pathogenesis of TCP. Many authorities ascribe a primary aetiological role to calculous obstruction in the pathogenesis of 'pancreatitis'. However, a non-calculous 'variety' certainly exists and as a rule calculi seem to develop after 'pancreatitis' is well established. Whether stones contribute to further pancreatic damage by ductal obstruction, or are an inconsequential epiphenomenon is not known. Malnutrition
Severe malnutrition at presentation in many of these patients suggested to many workers a causative role for nutritional deficiencies (Zuidema, 1959; Shaper, 1964; Ahuja, 1985). Pros and cons of a nutritional aetiology for TCP/FCPD have been extensively discussed (Pitchumoni, 1984; Balakrishnan, 1987; Balakrishnan et al, 1993; Bajaj and Subba Rao, 1988; Mohan Vet al, 1988c, 1988d; Yajnik, 1990). Clinical histories in these patients rarely show any evidence of kwashiorkor in childhood or unusually low intake of protein. Malnutrition at presentation could be secondary to severe exocrine and endocrine pancreatic deficiency; a case for pancreatitis-related and diabetes-related malnutrition rather than malnutrition-related diabetes. On the other hand, protein malnutrition has long been known to affect pancreatic function and cause pancreatic fibrosis (Pitchumoni, 1973; Durie et al, 1989). Children with kwashiorkor and adults from famine areas show pancreatic acinar atrophy, reduced enzyme output, and fibrosis of the gland.
780
c . s . YAJNIK
Endocrine pancreatic function (insulin secretion) is also known to be diminished in malnutrition (Baig and Edozien, 1965; Becket et al, 1971; Smith et al, 1975). There is usually a complete recovery of these functions after refeeding, though some studies have shown a long-term defect (Cook, 1967; James and Coore, 1970). Recent animal work has suggested that even mild to moderate protein malnutrition in utero (Snoeck et al, 1990), or in early childhood (Swenne et al, 1987, 1988; Crace et al, 1989) or even later (Khardori et al, 1980) can lead to [3-cell dysfunction and hyperglycaemia in later life. Similar long-term studies are not available in humans. It is possible that a pancreas 'weakened' by 'early life' malnutrition could be more susceptible to many 'toxic' influences to which it might be exposed in later life. Interesting reports from the UK have shown an association of low birth weight (an indicator of intrauterine growth retardation) with 'Type 2' diabetes, hypertension and coronary heart disease in adult life (Barker, 1990); no information is yet available on exocrine pancreatic function. Prospective studies of malnourished children in the developing countries are urgently needed. Our classical thinking on 'malnutrition' is usually restricted to deficiency of calories and proteins. However, micronutrient deficiencies (selenium, vitamins, etc.) could set in motion a series of pathological processes in the pancreas which could culminate in the final picture of TCP. Relative deficiency of certain elements or excess of others might produce deleterious effects. These avenues have not been explored in relation to TCP. An international comparative study of dietary factors suggested that high protein intake was associated with alcoholic pancreatitis and low fat intake with tropical pancreatitis (Durbec and Sarles, 1978). This observation was confirmed recently in Kerala (South India) and Marseille (France) (Balakrishnan et al, 1988b). The biochemical significance of this observation is not clear. Nwokolo and Oli suggested that recurrent infections and a highcarbohydrate, low-protein diet (mostly from cassava) in children in the tropics lead to pancreatic stasis, lamination of secretions and inspissation of mucus, ultimately leading to calcification in the pancreatic duct and chronic pancreatitis (Nwokolo and Oli, 1980). In summary, nutritional factors are probably important in the aetiotogy of this condition but the actual food components and mechanisms of damage are not clear. We also need to know more about the interactions of nutritional factors with other possible aetiological agents (Rao, 1988).
Cyanogenic alkaloids The geographical distribution of TCP in areas where cassava is grown and consumed as a staple food suggest a causative role for cassava in the aetiology of this condition (McMillan and Geevarghese, 1979; McMillan, 1993). A further observation that cassava causes goitres in iodine-deficient areas (elevated concentrations of thiocyanate from cassava in blood inhibit iodine uptake by the thyroid gland) but that goitre and FCPD almost never co-exist suggested to McMillan that a failure of detoxification of cyanide could be responsible. Deficiency of dietary sulfur amino acids (methionine,
DIABETES SECONDARY TO TROPICAL CALCIFIC PANCREATITIS
781
cystine) could hamper detoxification of cyanide to thiocyanate and lead to high levels of 'free cyanide' which might be toxic to [3-cells. Acute cyanide poisoning is known to be associated with hyperglycaemia and McMillan produced hyperglycaemia in rats by feeding cyanide (McMillan and Geevarghese, 1979). The relevance of these animal experiments to the clinical situation is still unclear, especially the role of cyanide in causing exocrine pancreatic damage. Recently Pitchumoni (1988) has suggested a universal role for 'cyanide toxicity' in the pathogenesis of pancreatitis; he proposes that 'alcoholic' pancreatitis in Western countries results from concomitant exposure to cyanide in cigarette smoke while tropical pancreatitis results from cyanogenic alkaloids in the diet, especially from cassava. Braganza (1993) has proposed that the toxicity of cyanogenic alkaloids might operate through increased oxidant stress. However, recent studies in Africa failed to find TCP or FCPD in areas where cassava is eaten as a staple food (Teuscher et al, 1987). These reports cast doubt over the aetiological role of cassava in TCP, but the absence of a 'susceptibility factor' in these populations could also explain the results. Occurrence of TCP in areas where cassava is not eaten is also cited as evidence against an aetiological role for cyanide. However, other cyanidecontaining foodstuffs (ragi, sorghum etc.) could be responsible (McMillan and Geevarghese, 1979; Nagalotimath, 1980; Kakrani et al, 1985; Patole et al, 1988). If cyanogenic alkaloids are indeed responsible, processing of foods to reduce cyanide content or development of low-cyanide varieties will be a simple preventive measure for this condition. The verdict on the role of cassava or other cyanogenic foodstuffs in the aetiology of TCP and FCPD must remain 'open'. Genetic, familial and immunological factors Although the geographical distribution of TCP is highly suggestive of an environmental aetiology, familial clustering has also been reported from South India. Families with more than one TCP patient, sometimes in successive generations, have been described (Pitchumoni, 1970; Geevarghese, 1986; Balakrishnan, 1987). A systematic study from South India looked at oral glucose tolerance, abdominal X-ray, ultrasound of the pancreas, and stool chymotrypsin concentrations in families of 38 FCPD patients (Mohan V et al, 1989a). Twelve percent of parents and 21% of siblings showed some evidence of exocrine pancreatic involvement, and 21% of parents and 11% of siblings revealed previously undiagnosed 'Type 2' diabetes. Unless a specific marker is available it is difficult to decide whether these results reflect genetic transmission or a strong environmental influence. However, consanguinity was common in families with TCP in successive generations, and this might indicate a genetic connection. Genetic markers for Type 1 and Type 2 diabetes (HLA-DQ gene and the 'insulin gene' polymorphism respectively) have been studied in the same population (Kambo et al, 1989). Approximately 40% of FCPD patients showed presence of an HLA-DQ-B marker (p
