The History of Cystic Fibrosis by Dr James Littlewood OBE

 

See also MALABSORPTION

 

1941 Beazell JM, Schmidt CR, Ivy AC. The diagnosis and treatment of achylia pancreatica. J Am Med Assoc 1941; 116:2735-2739.
This classic paper, from Professor Ivy's group, was said to initiate the modern treatment of pancreatic insufficiency. At the start, the authors state “Since the number of well studied patients are few and since oral pancreatic enzyme therapy in man is not generally considered to be of value, a view that is inconsistent with most of the results with animal experiments, we have undertaken to determine the effectiveness of oral pancreatic enzyme therapy in human patients with achylia pancreatica”.
Four patients with pancreatic insufficiency were studied in detail with estimation of faecal fat and nitrogen and also estimation of their duodenal enzymes – which were totally absent. Enzyme replacement therapy resulted in an average reduction in daily faecal nitrogen excretion of 62% (7.84 to 2.99 g) and lipids of 63.3% (74.4 to 27.3 g). All the patients gained weight (6 to 40 lbs) over 2.5 to 36 months. The authors note that Dorothy Andersen had already shown that pancreatic enzyme therapy improved the intestinal malabsorption of children with cystic fibrosis (Andersen DH. J Pediatr 1939; 15:763).

Fig. 1: Professor Ivy. From The Paradox of the Pancreas. With permission of Professor Irvin Modlin.

Professor AC Ivy (1893-1978) (figure 1) was a famous physiologist who, with Oldberg, discovered cholecystokinin in 1928 and apparently published some 2000 scientific articles. There was considerable discussion for many years among paediatricians as to whether pancreatic replacement therapy was worth using – some paediatricians considered that the unpleasant taste of the crude pancreatic extract did more harm than good in putting children off their food to achieve only modest improvements in absorption. However, all agreed that the introduction of the acid resistant microspheres (Pancrease and Creon) in the early Eighties represented a major advance in treatment.1942 Andersen DH. Pancreatic enzymes in duodenal juice in celiac syndrome. Am J Dis Child 1942; 63:643.
The first suggestion that estimation of duodenal trypsin was a reliable diagnostic test for CF as there was a wide difference between the concentration of trypsin in CF and in other gastrointestinal conditions and controls. Andersen noted that in normal infants trypsin and lipase were present from birth and amylase from 2 or 3 months.
Trypsin was more reliable than lipase although levels of the enzymes may be reduced in marasmus or severe dehydration; this was also noted by Shwachman and was confirmed many years later as a temporary phenomenon in gluten-induced coeliac disease, when it was attributed to lack of substrate to form the enzymes as a result of severe malnutrition (Carroccio A et al, Gut 1991; 32:796-799).  

 

1943 Farber S, Shwachman H, Maddock CL. Pancreatic function and disease in early life. I. Pancreatic enzyme activity and the celiac syndrome. J Clin Invest 1943; 20:827-833. [PubMed]
One of a series of papers from Boston by Sidney Farber (figure 2) around this time discussing the differentiation between coeliac disease and cystic fibrosis – largely based on the enzyme content of the duodenal fluid to document and extend the conclusions of their preliminary report (Farber & Maddock, Am J Pathol 1941; 17:445). They used data from over 150 children – either normal controls or with various nutritional disturbances, idiopathic coeliac disease or pancreatic fibrosis.
The differentiation of pancreatic fibrosis from coeliac disease had taken on a new importance since the description of CF in 1938 as the prognosis for the two conditions was very different. They suggested differentiation of children with gastrointestinal features of coeliac syndrome into those with normal pancreatic function and those with abnormal pancreatic function – the later being a permanent fatal condition. Pancreatic achylia was a constant feature of those subsequently found to have pancreatic fibrosis at autopsy.
There was a clear separation of those with and without pancreatic fibrosis. So measurement of the pancreatic enzyme activity was considered mandatory for diagnosis of pancreatic fibrosis. Removal of 70-80% of the pancreas reduced the enzyme activity but did not produce pancreatic insufficiency (a finding later confirmed by DiMagno et al, 1973 below). Also Andersen (1942 above) had suggested duodenal trypsin as a reliable test for differentiating CF from coeliac disease.

Figure 2: Sidney Farber. From Wikipedia.

This was one of Harry Shwachman’s (1910-1986) earlier papers as co-author and he would become one of the leading authorities on CF for many years to come.

 

1943 Shohl AT, May CD, Shwachman H. Studies of nitrogen and fat metabolism on infants and children with pancreatic fibrosis. J Pediatr 1943; 23:267-279.
A study of nine patients with cystic fibrosis. The striking feature of the infants with CF was the excess of nitrogen found in the faeces when normal diets were eaten, the patients remaining in negative nitrogen balance. Average weight of dried faeces/day - Normal infants 7.3 g - CF infants 33.4 g. Average nitrogen content of faeces/day - Normal infants 0.39 g - CF infants 1.7 g. Average fat content of dried faeces/day - Normal infants 3.1 g - CF infants 16.8 g. On a fat-free diet of casein hydrolysate and glucose, nitrogen retention was in the normal range and excretion normalised. The authors concluded that hydrolysed casein could be utilised in contrast to whole protein.
Subsequently feeds containing hydrolysed protein (e.g. Pregestimil) were shown by others to prove useful in improving malabsorption in some infants with CF where control of the malabsorption has proved difficult – particularly after meconium ileus operations and with cow’s milk intolerance. The concept of an elemental diet, which required less digestion, to improve absorption was also later pioneered by Jimmy Allen, a general paediatrician in Macclesfield, England (Allan JD et al, 1970 and 1973 below) and created considerable interest in the Seventies when it was almost impossible to maintain a normal nutritional state as the children became older and their chest became progressively worse.  

 

1943 Maddock CL, Farber S, Shwachman H. Pancreatic function and disease in early life. II. Effect of secretin on pancreatic function of infants and children. Amer J Dis Child 1943; 66:370 - 375.
These workers from Boston noted that the hormone secretin (discovered by Bayliss and Starling in 1902) had been used to stimulate pancreatic secretion only in adults up to that time. So they performed 26 secretin stimulation tests on 21 young patients and found trypsin was absent from the duodenum after intravenous stimulation by secretin in the 7 children with cystic fibrosis. They noted that in non-CF children “the increased flow of duodenal juice that followed the injection of secretin occurred in 3 – 5 minutes. Often the pancreatic flow was so copious that the bile coloring of the duodenal fluid was diluted so that fluid was colourless by the end of twenty minutes at which time the flow rate diminished. The most striking feature was the small volume of pancreatic juice obtained in patients with pancreatic fibrosis and the apparent failure of secretin to stimulate pancreatic flow in these patients in the “pancreatic fibrosis” group. (figure 3).

Fig. 3: Effect of "Pancreotest" on the volume of duodenal contents.

More extensive studies on pancreatic function in CF with secretin and pancreozymin were later reported by others – notably by Hadorn et al, 1968 (below) and these are usually quoted when discussing the subject; even more recently the subject was studied by Kopleman et al in Toronto (N Eng J Med 1985; 312:329-334 below). All have found markedly reduced fluid and bicarbonate output in all patients with CF and in the majority a reduced enzyme output.

 

1943 Shwachman H, Farber S, Maddock CL. Pancreatic function and disease in early life. III. Methods of analysing pancreatic enzyme activity. Am J Dis Child 1943; 66:418.
This is a description of the two methods of measuring tryptic activity – the tube method and gelatin method. One of series of papers with Sydney Farber regarding the investigation of pan

 

1948 Sinclair W Jr. Cystic fibrosis of the pancreas. Ohio Med J 1948; 44:1024.[PubMed]
This paper is mentioned as the first report of partial pancreatic involvement in a later report by Gibbs et al, 1950 (below). Autopsy diagnosis of an infant aged five months with CF in whom approximately one half of the pancreas, and that mostly in the tail, appeared to be involved. The infant had a normal concentration of trypsin in the duodenal aspirate and a normal output of stool fat. The usual bronchitis and bronchopneumonia expected with CF were the cause of death.  

 

1949 Shwachman H, Patterson PR, Laguna J. Studies in pancreatic fibrosis: Simple gelatin film test for stool trypsin. Pediatrics 1949; 4:222-230. [PubMed]
In this paper previous work on stool trypsin was reviewed, even as far back as the first duodenal catheterisation by Hess AF (Am J Dis Child 1912; 4:205). Faecal suspensions were placed upon unexposed and unfixed gelatin film and the extent of removal (digestion) of the gelatin film noted. The gelatin was digested by the trypsin in the stool in more than 95% of non-CF infants whereas 209 of 220 (95%) of stool specimens from untreated infants with CF showed no tryptic activity whereas less than 5% of 500 controls lacked tryptic activity. Three separate stool specimens were recommended for diagnosis.

Fig.4: Stool suspensions on gelatine plate

A photograph in the paper shows gelatin film acted upon by stool suspensions from three individuals. Concentrations of 1/5 on upper row and 1/10 are on the lower row being the more dilute (figure 4).
1- Normal infant’s stool with extensive white areas where there has been extensive digestion of the gelatin by the trypsin in the stool..
2- Infant with cystic fibrosis – no digestion of gelatin.
3- CF infant on treatment with pancreatin now digesting gelatin.


Subsequent papers on the gelatin film tests discussed problems with interference from faecal bacteria digesting the gelatin on the film (e.g. Johnstone DE, Neter E. Pediatrics 1951; 7:483-490. [PubMed]). Bodian (1952) notes that Dorothy Andersen was the first to suggest testing for stool trypsin - an extremely important test prior to the availability of the sweat test for there were difficulties in duodenal intubation in small infants to obtain fluid for trypsin estimation.
Before the sweat test became available during the Fifties, demonstration of a pancreatic enzyme abnormality was mandatory in making a correct diagnosis of CF (which was essentially a death sentence for the infant) and differentiating the condition from coeliac disease from which the infant may recover. Duodenal intubation to measure tryptic activity was the method used and recommended by Dorothy Andersen.  

 

1950 Gibbs GE, Bostick WL, Smith PM. Incomplete pancreatic deficiency in cystic fibrosis of the pancreas. J Pediatr 1950; 37:320-325. [PubMed]
A series of 38 patients included two proved at autopsy to have CF (by having typical pancreatic histological changes), but with sufficient residual pancreatic exocrine function to achieve normal fat absorption in life. Investigation showed the reduction in pancreatic enzymes was not complete. Possibly another four patients were in this category (13% of the series).


Subsequently it was shown that pancreatic enzyme output could fall to only 10% of normal before intestinal malabsorption occurred (DiMagno et al. 1973 below). The authors mention the previous case report of incomplete pancreatic involvement by Sinclair W Jr, 1948 (above). Also there were soon reports of similar patients who were described as “pancreatic sufficient” i.e. they had sufficient residual pancreatic function to achieve normal fat absorption (Shwachman et al 1952 & 1956 below; di Sant’Agnese al, 1956 below).  

 

1951 Johnstone DE, Neter E. Studies on laboratory diagnosis of cystic fibrosis of the pancreas: Positive gelatine film tests due to gelatine-liquefying bacteria in feces and duodenal juice. Pediatrics 1951; 7:483-490.[PubMed]
A report suggesting that gelatine-liquefying bacteria would interfere with the reliability of the gelatine film test for faecal trypsin which had been described recently by Shwachman (Shwachman et al, 1949 above) and had seemed such a welcome alternative to duodenal aspiration. The authors suggested this complication was related to the increasing frequency of organisms, particularly Pseudomonas aeruginosa, in the stools which they suggested was related to the more prolonged use of antibiotics. Johnstone later showed that if a dilution of faeces of 1 in 120 was used the results were reliable (Johnstone, 1952 below).


This test was very important at the time as the sweat abnormality was not described until 1953 and the sweat tests not widely available until the Sixties. So at this time the diagnosis of CF relied heavily on the clinical picture and the absence of tryptic activity either in the duodenal aspirate or in the stools to identify the pancreatic abnormality.  

 

1952 Shwachman H, Leubner H, Patterson P, Weill CC. Mucoviscidosis with partial pancreatic insufficiency. Am J Dis Child 1952; 84:763-765. [PubMed]
Previously some people with CF had escaped diagnosis as the duodenal trypsin test gave normal results. Progressive loss of pancreatic activity could be shown. Partial pancreatic insufficiency could be demonstrated by various techniques – examining duodenal fluid for various enzymes, viscosity and pH. In the discussion the need to differentiate the various causes of coeliac syndrome was stressed as some coeliac children were being falsely labelled as CF as a result of “the present interest in the cystic fibrosis”. For many years a frequent examination question for medical students was to list the differences between coeliac disease and cystic fibrosis; just as they must know the differences between children with hypothyroidism and Down’s syndrome.


By 1952 Shwachman had obviously become cautious about using secretin intravenously for pancreatic function tests as he warns that secretin “may occasionally produce disastrous effects…..usually not a purified substance….some children are allergic to it”. So although 10 years previously Shwachman had presented a study of secretin pancreatic stimulation tests in children (Maddock et al, 1943 above), he now preferred olive oil as a stimulant and believed “The most reliable diagnostic procedure is study of the duodenal fluid” (also Gibbs et al, 1950 above).
One of a number of papers discussing the estimation of faecal tryptic activity as an indication of pancreatic function, suggesting that absence of tryptic activity in the first few days suggested cystic fibrosis.
It is not surprising that there were a number of papers published on this subject as this was an important test in differentiating CF from coeliac disease in the days before the sweat test became more widely available during the late Fifties and Sixties.  

 

1952 Johnstone DE. Studies on cystic fibrosis of pancreas: role of various diluents and the dilution factor in interpretation of X-ray film test for fecal trypsin. Am J Dis Child 1952; 84:191-198.
An astonishing 9,600 tests on 963 stool specimens from 216 children! Johnstone had already published on false negativity with the stool gelatine test due to bacterial digestion of the gelatine (Johnstone & Neter, 1951 above). However, in this study he showed that false results did not occur if a 1/120 dilution of the stool was used.


This test was still very important for diagnosis in 1952, as the sweat test was not yet available and duodenal intubation and aspiration was invasive, unpleasant for the infant, and difficult for the nursing and medical staff.

 

1955 di Sant’Agnese PA. Fibrocystic disease of the pancreas with normal or partial pancreatic function: current views on pathogenesis and diagnosis. Pediatrics 1955; 15:683-697. [PubMed]
Another early report of “pancreatic sufficient” patients i.e. defined as those who had sufficient remaining pancreatic function (probably around 10%) to achieve normal intestinal fat absorption without enzyme replacement therapy. On testing, six had partial pancreatic deficiency and 3 normal pancreatic function. These patients provide further evidence that the secretory activity of many and perhaps all exocrine glands, mucus producing and others, is affected rather than the clinical manifestations being due primarily to pancreatic disease and secondary malnutrition e.g. vitamin A deficiency. (See also Gibbs et al, 1950 above; Shwachman et al, 1956 below).

 

1956 Dooley RR, Guilmette F, Leubner H. Patterson PR, Shwachman H, Weil C. Cystic fibrosis of the pancreas with varying degrees of pancreatic insufficiency. J Dis Child 1956; 92:347-368. [PubMed]
A further 17 patients and an estimate that some 10-15% of patients were “pancreatic sufficient” i.e. had sufficient residual exocrine pancreatic function to achieve normal fat absorption. Chest disease may develop before malabsorption occurs thus supporting the original hypothesis of Wolbach and Farber of a generalised disease if modified to include disturbance of sweat glands as an almost constant feature - in contrast to the belief that the problems were secondary to pancreatic malabsorption and malnutrition e.g. vitamin A or other deficiencies. Only one patient had a sweat test performed and that was positive – the sweat abnormality having only been described recently in 1953 (also Gibbs et al. 1950 above; Di Sant’Agnese, 1955 above).

 

 

1957 Fyfe W M. The stool proteolytic activity in the diagnosis of fibrocystic disease of the pancreas. Scot Med J 1957; 2:66 – 68. [PubMed]
Probably the first paper on CF from Scotland – Royal Hospital for Sick Children Glasgow - testing Shwachman’s suggestion that measurement of stool proteolytic activity could replace duodenal intubation in the diagnosis of cystic fibrosis (Shwachman et al, 1949 above). Both duodenal and stool proteolytic activity were examined in 50 infants; 14 infants had absent proteolytic activity and were considered to have CF – confirmed at autopsy in eight. Fyfe concluded that liquefaction of the 1 ml of 7.5% gelatin by a 1:40 or higher dilution of stool was a reliable indication of normal duodenal proteolytic activity.


The author thanks Prof. Stanley Graham and Dr (later Prof) James Hutchinson for help and permission to study their cases - both were leading paediatricians in the UK at the time.

1963 Rick W. Untersuchung zur exokrinen function des pancreas bei zysticher pancreas-fibroses. (Pancreatic exocrine function in CF of the pancreas). Medizinische Welt 1963; 42:2158-9.
An early report of pancreatic function in CF showing reduced volume and reduced bicarbonate secretion. (Quoted by Wong LTK et al, Gut 1982; 23:744-750). Also pancreatic function was reported by Maddock et al, 1943 (above), Kopleman et al, 1985 (below) and Hadorn et al, 1968 (below). Hadorn’s work seemed to have had more impact (possibly as published in English) and is generally regarded as making the major contribution to pancreatic function testing in people with cystic fibrosis.

 

1964 Shwachman H, Diamond LK, Oski FA, Khaw Kon-T. The syndrome of pancreatic insufficiency and bone marrow dysfunction. J Pediatr 1964; 65:645-656. [PubMed]
This is the second most frequent cause of chronic pancreatic disease in children. Six children are described with what became known as the Shwachman–Diamond Syndrome. It is one of a number of conditions eventually recognised as separate entities from cystic fibrosis. The authors advised that the condition should be considered in children at first suspected of having CF but who had repeatedly normal sweat electrolytes. In 1967 Burke and colleagues (Burke V et al. Arch Dis Child 1967; 42:147-157.[PubMed]) reported 19 children with the condition from Royal Children’s Hospital Melbourne having published a preliminary report in 1965 (Colebatch JH et al Lancet 1965; ii: 496). Burke et al commented that in 1964 Bodian et al (Acta Paediatr Uppsala 1964; 53:282-93.[PubMed]) from Great Ormond Street, London had described two children with congenital hypoplasia of the exocrine pancreas both of whom had variable neutropenia and thrombocytopenia but which “did not receive special comment” – which is not entirely true. Bodian et al did however provide a useful review of a previous 18 histologically proven cases in the literature.

This was an important paper for paediatricians dealing with children who had CF as it identified a specific condition (exocrine pancreatic insufficiency, bone marrow dysfunction, skeletal abnormalites and short stature with normal sweat tests) likely to be encountered in specialist CF centres. In fact we identified five such children with the Shwachman-Diamond syndrome in Leeds over 20 years in the course of investigation of children considered initially to have cystic fibrosis.

 

1966 Barbero GJ, Siblinga MS, Marino JM, Seibel R. Stool trypsin and chymotrypsin. Value in the diagnosis of pancreatic insufficiency in cystic fibrosis. Am J Dis Child 1966; 112:536-540. [PubMed]
These are useful non-invasive tests for confirming pancreatic insufficiency used for many years – chymotrypsin being preferred to trypsin as it is the more stable. Their value was confirmed in subsequent reports (Brown et al, 1988 below). This present study used the method of Haverbach BJ, et al, (Gastroenterol 1963; 44:588-597). There was clear separation of CF and non-CF and the authors concluded it to be reliable in identifying those children with cystic fibrosis. Chymotrypsin values were : in CF < 58 ug/g and in non-CF > 75 ug/g stool and trypsin: in CF < 30 ug/g and in non-CF >80 ug/g stool  

 

1968 Hadorn B, Zoppi G, Shmerling DH, Prader A, McIntyre I, Anderson CM. Quantitative assessment of exocrine pancreatic function in infants and children. J Pediatr 1968; 73:39-50. [PubMed]
One of a number of papers by Hadorn et al on stimulated pancreatic function in CF using a triple lumen tube. The method was used in a few CF Centres, including our own in Leeds, during the Seventies, but it was a difficult test to perform and very invasive for the children. The findings in the present group of 10 children with CF were a low volume of fluid and low bicarbonate output even when the enzyme concentrations were not markedly reduced.
After intravenous injection of pancreozymin (P) bile stained juice high in enzymes is secreted. After secretin (S) injection the volume and bicarbonate concentration is increased but the enzyme concentrations are reduced. In CF the secretion of both is markedly reduced (figure 5).
Also from the same group Zoppi G et al. (Helvet Paediatr Acta 1968; 6:577-590. [PubMed]) found a higher protein content in the pancreatic juice of both CF and non-CF pancreatic insufficiency using electrophoresis.

Fig. 5: Pancreozymin-secretin pancreatic function test in a normal person as described by Hadorn. From Paediatric Gastroenterology, C M Anderson, V Burke (eds). Blackwell Scientific Publications, Oxford. 1973: 302.

 

1968 Hadorn B, Johansen PG, Anderson CM. Pancreozymin secretin tests of exocrine pancreatic function in cystic fibrosis and the significance of the result for the pathogenesis of the disease. Can Med Ass J 1968; 98:377-85. [PubMed]
Thirteen of 176 children with CF in the Melbourne clinic had normal fat excretion. Ten of these pancreatic sufficient patients (PS) had pancreozymin-secretin tests confirming that the water and bicarbonate secretion from the pancreas was more severely affected than the enzyme secretion. They secreted small volumes of pancreatic juice with very low bicarbonate content and but abnormally high enzyme concentrations.

Evidence was presented that the pathological features in the pancreas could result from a failure to produce an adequate amount of electrolyte containing fluid. In text the authors make the following important comments “it is possible that this (the extensive changes in the pancreas) is at least partially caused by the “organic fraction” which is abnormally concentrated in enzymes. Although proteolytic enzymes are originally present in an inactive form in pancreatic juice it is not unlikely that they may become activated if the flow of secretion is markedly reduced or stagnates. Furthermore, because the secretion is more concentrated and more viscous, blockage of tubules followed by dilatation of the acini may lead to their rupture, and the intraluminal material containing activated enzymes could spread into the surrounding tissue causing irritation and destruction followed by atrophy and fibrosis of parts of the pancreas or entire organ”

This is a very attractive suggestion and likely to be the correct explanation for the early onset of the pancreatic damage starting with the electrolyte/fluid abnormality, now (but not then) known to be related to abnormalities of fluid and electrolyte transport across the cell membranes and analogous to the changes occurring in other organs – but earlier and more severe as the other organs do not have to contend with the damaging effects of the pancreatic enzymes. Of course, the abnormality of membrane transport had not been described at this stage.


Pancreatic function had been studied by Correia JP & Barros F. (Study of the exocrine function of the pancreas with secretin and pancreozymin. Journal do Medico 1963; 52:581-92.[PubMed]). Rick (1963, above) produced an early report of pancreatic function in CF showing reduced volume and reduced bicarbonate secretion (quoted by Wong LTK et al, Gut 1982; 23:744-750.[PubMed]). Also pancreatic function in CF had been reported first by Maddock et al, 1943 (above), and subsequently by Kopleman et al, 1985 (below). Hadorn’s work seemed to have had more impact (possibly as published in English) and is generally regarded as making the first major contribution to pancreatic function testing in CF.  

 

1975 Shwachman H, Lebenthal E, Khaw P-T. Recurrent acute pancreatitis in patients with cystic fibrosis with normal pancreatic enzymes. Pediatrics 1975; 55:86-94. [PubMed]
Ten adolescents and young adults from over 2000 patients with CF, who were all pancreatic sufficient, developed recurrent pancreatitis – two before the diagnosis of CF was made. There were typical abdominal symptoms of acute pancreatitis and elevated serum and urinary amylase and serum lipase. Three patients eventually died and the autopsies showed characteristic lesions of cystic fibrosis. The authors mention only one previously recorded patient with pancreatitis (di Sant’Agnese PA, Lepore MJ. Involvement of abdominal organs in cystic fibrosis of the pancreas. Gastroenterology 1961; 40:64).
Pancreatitis came to be recognised as a complication in people with CF who were “pancreatic sufficient” i.e. had at least 10% residual pancreatic function and did not require pancreatic enzyme treatment to prevent intestinal malabsorption. Subsequently, after the gene had been identified, a higher than expected number of non-CF people with pancreatitis were found to have one CF mutation (Sharer et al, 1998 below; Cohn et al, 1998 below).

 

Fig. 6: Paul Quinton. From www.biomed.ucr.edu website.

1983 Quinton PM. Chloride impermeability in cystic fibrosis. Nature 1983; 301:421-422. [PubMed]
Another landmark paper in the understanding the CF defect by Paul Quinton (figure 6) , who himself has CF. He later recalls (Quinton, 1999 below) that Michael Knowles (1981 above) reported a significantly larger than normal electronegative potential across the nasal epithelium, along with the fact that NaCl absorption was inhibited in the CF sweat ducts and also that sodium was relatively more absorbed than chloride. This gave Quinton the idea that the basic defect in the CF duct had to be due to an anion impermeability and not defective anion exchange.
Using a series of microperfusion experiments of sweat glands (it is said from his own forearms) he measured the electrolytes by “energy dispersive X-ray analysis”. The chloride impermeability he had shown in sweat glands was the basis for the raised sweat electrolytes and provided a physiological explanation for the high salt content of CF sweat and also “provided the first description of a basic cellular defect that has since proven to be uniformly inherent in all CF affected cells”.
Paul Quinton holds the Nancy Olmsted Chair of Pediatric Pulmonology UC San Diego and is Professor of Biomedical Sciences at UC, Riverside.  

 

1984 Sato K, Sato F. Defective beta adrenergic response of cystic fibrosis sweat glands in vivo and in vitro. J Clin Invest 1984; 73:1763-1771. [PubMed]
Abnormal sodium chloride absorption from the ducts of the sweat glands had been known as the only defect in CF sweat glands. These authors fortuitously found that the secretory portion of CF sweat glands is also abnormal in that it failed to show a sweating response to beta adrenergic stimulation (isoproterenol) both in vivo and in vitro. Their data, after detailed study of the sweat glands, suggested that beta adrenergic regulation was abnormal in CF sweat glands and justified further investigations into the mechanism of beta adrenergic regulation of the eccrine sweat gland in both normal and CF subjects. Subsequently it was shown that the insensitivity was due to the fact that in the sweat gland b-adrenergically stimulated fluid secretion was coupled with the same chloride channel as was altered in salt absorption.  

 

1984 Sturgess JM. Structural and developmental abnormalities of the exocrine pancreas in cystic fibrosis. J Pediatr Gastroenterol Nutr 1984; 3 Suppl 1:S55-66.[PubMed]
Early signs of a deficiency in exocrine pancreatic tissue at 32-38 weeks post-conceptional age suggest that there is a lack of normal maturation of pancreatic exocrine tissue that occurs in utero, with a degenerative process supervening after birth. The volumes of the acinar and duct lumen are markedly increased, up to 10 fold normal volume in CF subjects. The diagnosis of CF within the first few months of life is difficult when based on conventional or subjective pancreatic histology. But by quantitative microscopy of the pancreas as describe by Jennifer Sturgess, the pathologist in Toronto, an objective approach is available that clearly distinguishes CF from control subjects. In this retrospective survey, 93% of CF infants were discriminated from normals; only 2 of 30 cases (7%) were not clearly differentiated from controls.


Earlier studies had reported some CF infants had normal pancreatic histology (Oppenheimer & Esterly, 1973 above).  

 

1985 Kopleman H, Durie P, Gaskin K, Weizman Z, Forstner G. Pancreatic fluid secretion and protein hyperconcentration in cystic fibrosis. N Eng J Med 1985; 312:329-334. [PubMed]
Using the expertise in pancreatic function testing developed in Toronto, the secretion of pancreatic protein and water was studied in 28 patients with CF and 21 controls matched for pancreatic acinar function as defined by trypsin secretion, using a quantitative-marker perfusion technique and continuous intravenous secretin-pancreozymin stimulation. Secretions from the patients with CF contained significantly higher concentrations of protein than those from the controls and their fluid secretion was significantly decreased.
Somewhat analogous to the situation in the airways, the authors concluded that fluid secretion in patients with CF may be a rate-limiting factor in protein output and that a limited flow of hyperconcentrated protein secretions may predispose to protein precipitation and ductal obstruction in the pancreas. (also Kopleman et al, 1988 below).  

 

1988 Kopelman H, Corey M, Gaskin K, Durie P, Weizman Z, Forstner G. Impaired chloride secretion, as well as bicarbonate secretion, underlies the fluid secretory defect in the cystic fibrosis pancreas. Gastroenterology 1988; 95:349-355. [PubMed]
Pancreatic fluid and electrolyte secretion was assessed in 56 patients with CF and 56 non-CF control subjects undergoing stimulated pancreatic function tests. The CF subjects secreted significantly less fluid than control subjects caused by impaired chloride, as well as bicarbonate, secretion. Paul Quinton, commenting on this and the 1985 paper (Kopleman et al, 1985 above) from Toronto, noted that both studies showed that bicarbonate and volume outputs were abnormally low in CF and that chloride impermeability had a pronounced negative effect on exocrine bicarbonate transport…. Without sufficient fluid and bicarbonate digestive proenzymes stagnated and activated prematurely in the pancreatic ducts… resulting in micro autolysis, focal injury, inflammation, infiltration, calcification, plugged ducts, fibrosis, and scarring until the entire parenchyma of the pancreas like the lung, but from a seemingly different cause, was destroyed. This seemed a very reasonable explanation as to why the lungs are virtually normal at birth and the pancreas was already severely damaged in many infants (also Kopleman et al, 1985 above).

 

1988 Brown GA, Sule D, Williams J, Puntis JWIL, Booth IW, McNeish AS. Faecal chymotrypsin: a reliable index of exocrine pancreatic function. Arch Dis Child 1988; 63:785-789. [PubMed]
Specimens of meconium and random stools were collected sequentially from 25 healthy newborn babies over the first eight to 14 days of life. The stool chymotrypsin concentrations increased from birth to a maximum at four days of age and then fell again over the next four days. In 22 newborn babies suspected of meconium ileus and later confirmed to have CF, faecal chymotrypsin concentrations were all appreciably reduced. Measuring faecal chymotrypsin concentrations was a reliable procedure for identifying pancreatic exocrine insufficiency in the newborn. (also from Birmingham, Brown GA et al. Arch Dis Child 1988; 63:1229-1233. [PubMed]).


This test was very useful in management of CF patients and we have discussed the use of the test recently as follows – “occasional measurements of faecal chymotrypsin are useful for monitoring pancreatic enzyme replacement therapy. Values that remain low on treatment indicate either inadequate prescription or non adherence with treatment although a normal value does not exclude persisting significant steatorrhoea. Very high values can indicate that the dose is excessive. High values also occur if the patient has rapid intestinal transit. Conversely constipation may be associated with low values. If low values persist alongside symptoms of malabsorption treatments to reduce gastric acid secretion may help improve enzyme efficacy and enable reduction in enzyme intake (Littlewood et al. Pediatr Pulmonol 2006; 41:35-49. [PubMed]). Unfortunately the test has been withdrawn by some UK laboratories since faecal elastase 1 has been available as a marker of pancreatic function.  

 

1990 Waters DL, Dorney SF, Gaskin KJ, Gruca MA, O'Halloran M, Wilcken B. Pancreatic function in infants identified as having cystic fibrosis in a neonatal screening program. N Eng J Med 1990; 322:303-308. [PubMed]
Assessment of pancreatic function in 78 children identified in the New South Wales neonatal screening program as having cystic fibrosis. Measurements of faecal fat excretion, pancreatic-stimulation tests, and estimations of the serum level of pancreatic isoamylase indicated that 29 of the 78 children (37 percent) had substantial preservation of pancreatic function. These 29 children (median age, four years) had growth that was close to normal and comparable to growth in children with severe pancreatic insufficiency who received oral enzyme therapy. Pancreatic insufficiency subsequently developed in 6 of the 29 patients, at 3 to 36 months of age. The authors concluded that the serum immunoreactive-trypsin assay used in neonatal screening programs identifies patients with CF who have sufficient pancreatic function to achieve normal fat absorption and that a substantial proportion of infants identified as having cystic fibrosis are in this category (also Gaskin et al, 1991 below)
Thirty four of these 76 children had their pancreatic function assessed an average of 2.3 years after diagnosis – so subsequently a further 20 infants were studied at the time of diagnosis (Gaskin et al, 1991 below).


The number of children who were pancreatic sufficient in this series is considerably higher than expected but would depend on the proportion carrying a mild mutation.

 

1991 Gaskin K, Waters D, Dorney S, Gruca M, O'Halloran M, Wilcken B. Assessment of pancreatic function in screened infants with cystic fibrosis. Pediatr Pulmonol 1991; Suppl 7:69-71. [PubMed]
Previously these authors reported that 37% of infants with CF diagnosed by neonatal screening with the dried blood spot immunoreactive trypsin assay were pancreatic sufficient (Waters et al, 1990 above). However, 34 of the 78 infants had pancreatic function tests an average 2.3 years after diagnosis, thus it was possible that the percentage with neonatal pancreatic sufficiency was even underestimated, due to the loss of pancreatic function with time in some infants. To assess this hypothesis the authors assessed pancreatic function at the time of diagnosis in a further 20 infants since the completion of the previous study. Results of fecal fat determinations and/or pancreatic stimulation tests indicated that no less than 10 (50%) of these infants have pancreatic sufficiency. Combining these results with those of the previous study, 31 of 64 patients (48%) have pancreatic sufficiency at this early age. The authors monitored the progression of pancreatic disease in the 39 children with pancreatic sufficiency recognized to date. Eleven have developed pancreatic insufficiency and require enzyme replacement therapy. Five others have shown further improvement of colipase secretion with age.
So the authors confirmed their previous conclusion that the dried blood immunoreactive trypsin screening program for cystic fibrosis does recognize patients with pancreatic sufficiency and at diagnosis nearly half their patients were in this category. To date, 28% of patients with pancreatic sufficiency have demonstrated a variable decline in pancreatic function with age.

In this study there were a surprisingly large number of infants who were pancreatic sufficient (48%) and this is quite different from our experience of screened infants with CF over the last 30 years in Leeds, although, of course, the frequency will depend on the mutations which the infants have. For example of the last 15 screened infants with CF in Leeds only 2 were pancreatic sufficient (Wolfe et al. J Cyst Fibros 2005; 4(S1):S94).
The faecal pancreatic elastase is now a convenient and reliable way of determining pancreatic function and following the progress of pancreatic function in these infants. The lesson here is that not all newborns with CF require enzyme replacement therapy so it is important to make sure there is evidence of pancreatic insufficiency before commencing enzymes - in practice easily done with a faecal elastase measurement and a small specimen of stool for fat microscopy (Walters et al, 1990 above).

The faecal pancreatic elastase test is also reliable indicator of pancreatic function even when the infant is taking enzyme replacement therapy - in this respect it differs from faecal trypsin and chymotrypsin. 

 

1992 Kristidis P, Bozon D, Corey M, Markiewicz D, Rommens J, Tsui LC. Genetic determination of exocrine pancreatic function in cystic fibrosis. Am J Hum Genet 1992; 50:1178-1184. [PubMed]
A review of the association of so-called "mild" mutations with pancreatic sufficiency. Although the majority of CF mutations - including the most common, delta F508 - are strongly correlated with pancreatic insufficiency (PI), approximately 10% of the mutant alleles may confer pancreatic sufficiency (PS). To extend this observation, genomic DNA of 538 CF patients with well-documented pancreatic function status were analyzed for a series of known mutations in their CFTR genes. A total of 30 different, complete genotypes could be determined in 394 (73%) of the patients. The data showed that each genotype was associated only with PI or only with PS, but not with both. This result is thus consistent with the hypothesis that PI and PS in CF are predisposed by the genotype at the CFTR locus; the PS phenotype occurs in patients who have one or two so-called mild CFTR mutations, such as R117H, R334W, R347P, A455E, and P574H, whereas the PI phenotype occurs in patients with two severe alleles, such as delta F508, delta I507, Q493X, G542X, R553X, W1282X, 621 + 1G---- T, 1717-1G----A, 556delA, 3659delC, I148T, G480C, V520F, G551D, and R560T
This large study from Toronto strengthens the view that pancreatic function status in CF is genetically determined by specific mutations at the CF locus. Subsequent studies established this beyond any doubt.  

 

1997 Wallis C, Leung T, Cubbitt D, Reynolds A. Stool elastase as a diagnostic test for pancreatic function in children with cystic fibrosis. Lancet 1997; 350:1001.[PubMed]
The first published report of faecal pancreatic elastase 1 in children with CF from Colin Wallis, (figure 42) paediatrician at Great Ormond Street, London. All non-CF children had normal faecal elastase levels but 26 of 30 children with CF had undetectable levels. Three of the others were pancreatic sufficient. Normal levels were reached by normal infants within 2 weeks.


The test was further evaluated in a larger series of children with CF in Leeds (Cade et al, 2000 below) and became the accepted method of determining pancreatic function. The test had a major advantage over faecal chymotrypsin as it still remained low (negative) in pancreatic insufficient patients whilst the patient was taking pancreatic enzyme supplements. This test was a major advance - a really useful standardised non-invasive test for accurately confirming the presence of pancreatic insufficiency without having to stop the pancreatic enzyme treatment of patients referred to a CF Centre for full investigation - a far cry from duodenal intubation for tryptic activity which was essential before the sweat test was introduced during the Fifties; also it was a major advance on faecal chymotrypsin. However, faecal chymotrypsin did provide a useful indication as to whether the patient was taking their pancreatic enzyme supplements – a low chymotrypsin when thought to be taking pancreatic enzymes being suggestive of non-adherence.  

 

1998 Sharer N, Schwarz M, Malone G, Howarth A, Painter J, Super M, Braganza J. Mutations of the cystic fibrosis gene in patients with chronic pancreatitis. N Eng J Med 1998; 339:645-652. [PubMed]
The pancreatic lesions of cystic fibrosis develop in utero and closely resemble those of chronic pancreatitis. These authors hypothesized that mutations of the cystic fibrosis transmembrane conductance regulator (CFTR) gene may be more common than expected among patients with chronic pancreatitis. Therefore 134 consecutive patients with chronic pancreatitis were examined for 22 mutations of the CFTR gene that together accounting for 95 percent of all mutations in patients with CF in North West England. The 94 male and 40 female patients ranged in age from 16 to 86 years. None had a mutation on both copies of the CFTR gene but 18 patients (13.4 percent) had a CFTR mutation on one chromosome, as compared with a frequency of 5.3 percent among 600 local unrelated partners of persons with a family history of cystic fibrosis (P<0.001). A total of 10.4 percent of the patients had the 5T allele in intron 8 (14 of 134), which is twice the expected frequency (P=0.008) (error - see below). Four patients were heterozygous for both a CFTR mutation and the 5T allele. Patients with a CFTR mutation were younger than those with no mutations. None had the clinical or laboratory features of cystic fibrosis. The authors concluded that mutations of the CFTR gene and the 5T genotype are associated with chronic pancreatitis.


This paper from Manchester UK is reported in some detail as it was the first report of a strong association between mutations in the CFTR gene and pancreatitis in people who did not have cystic fibrosis. Martin Schwarz, Consultant Clinical Molecular Geneticist in Manchester told me that the total percentage with 5T was only 5% (not 10.4% as stated in the paper) and so not significant. The abnormal CFTR genotypes in these patients with pancreatitis resemble those associated with male infertility. The findings of Cohn et al, 1998 were similar to those in this study (Cohn et al, 1998 below) and appeared in the same issue of the New England Journal of Medicine.  

 

1998 Cohn JA, Friedman KJ, Noone PG, Knowles MR, Silverman LM, Jowell PS. Relation between mutations of the cystic fibrosis gene and idiopathic pancreatitis. N Engl J Med 1998; 339: 653-8. [PubMed]
Twenty seven adult patients with idiopathic pancreatitis were tested for 17 CFTR mutations and for the 5T allele in intron 8 of the CFTR gene. (The 5T allele reduces the level of functional CFTR and is associated with an inherited form of infertility in males). Patients with two abnormal CFTR alleles were further evaluated for unrecognized CF-related lung disease, and both base-line and CFTR- mediated ion transport were measured in the nasal mucosa. Ten patients (37%) with idiopathic chronic pancreatitis had at least one abnormal CFTR allele. Eight CFTR mutations were detected. In three patients both alleles were affected but they did not have lung disease typical of CF on the basis of sweat testing, spirometry, or base-line nasal potential-difference measurements but they did have abnormal nasal cyclic AMP-mediated chloride transport.


Both the above papers independently confirmed a definite association of pancreatitis and CF mutations and appeared on Sept 3rd 1998 in the same issue of the New England Journal of Medicine.  

 

2000 Cade A, Walters MP, McGinley N, Firth J, Brownlee KG, Conway SP, Littlewood JM. Evaluation of fecal pancreatic elastase-1 as a measure of pancreatic exocrine function in children with cystic fibrosis. Pediatr Pulmonol 2000; 29:172-176. [PubMed]
Experience of pancreatic elastase-1 measurements (fecal EL-1) from a large series of 142 patients from the paediatric CF centre in Leeds. The value in 135 pancreatic insufficient (PI) patients was 10 microg/g stool (2.5-33); of the 7 pancreatic sufficient (PS) patients, 698 microg/g stool (400.5 - 824.5), and that of the non-CF controls, 615 microg/g stool (420-773). There is an impressive clear cut separation of the groups.

This is one of a number of papers confirming the value of a relatively new and reliable non-invasive test of pancreatic function using a single small faecal specimen; also the result is still valid even when patient is taking pancreatic enzyme replacement therapy. Fecal EL-1 is now used as evidence of pancreatic insufficiency in screened CF infants and is reliable over the age of 2 weeks; in older CF patients at diagnosis; for confirming the need for pancreatic enzymes in patients referred to the CF centre from other hospitals who are already taking enzymes (experience shows that some are pancreatic sufficient and do not require the enzyme supplements they have been prescribed); for annual monitoring of PS patients to detect the onset of PI; and as supporting evidence when excluding the diagnosis of CF in patients attending the pediatric gastroenterology clinic.
The test was undoubtedly a very valuable addition to the tests for pancreatic insufficiency and one of the most useful of any previously described. The first brief report of EL-1 in CF children was in 1997 from Great Ormond Street, London (Wallis et al. Lancet 1997; 350:1001.above [PubMed]) the test having been described earlier in adults by Loser et al. (Gut 1996; 39:580-586. [PubMed] and others). 

 

2000 Feigelson J, Pecau Y, Poquet M, Terdjman P, Carrere J, Chazalette JP, Ferec C. Imaging changes in the pancreas in cystic fibrosis: a retrospective evaluation of 55 cases seen over a period of 9 years. J Pediatr Gastroenterol Nutr 2000; 30:145-151. [PubMed]
The aim of this study was to follow-up the evolution of the CF pancreatic changes in 55 people with CF to define its successive stages in correlation with the clinical, biochemical, and imaging findings. Imaging using mainly echograms and tomodensitometric scans were regularly performed: echograms every 6 months, and tomodensitometric scans every 1 to 2 years. Five groups of patients were identified: those with a normal pancreas (n = 4), incomplete lipomatosis of the pancreas (n = 9), complete lipomatosis of the pancreas (n = 23), cystic pancreas (n = 5), macrocystic pancreas (n = 1), atrophic pancreas (n = 13). Forty episodes of pancreatitis were observed in seven patients – they had bouts of abdominal pain and elevation of lipase levels.

Fig. 7: Abdominal ultrasound showing pancreatic cysts in an adolescent boy with cystic fibrosis

The authors suggest that studies of pancreatic function should be performed routinely in cystic fibrosis, especially in those with pancreatic sufficiency or in patients with normal pancreatic images. Acute pancreatitis was common and should be diagnosed and properly identified to be differentiated from other acute abdominal syndromes occurring in cystic fibrosis – a very important point.
However, it is unlikely that the pancreas will receive such regular and careful attention in many CF centres where more attention is given to the chest than the pancreas and even simple evaluation of intestinal fat malabsorption is unusual!  

 

2004 Borowitz D. Baker SS. Duffy L. Baker RD. Fitzpatrick L. Gyamfi J. Jarembek K. Use of fecal elastase-1 to classify pancreatic status in patients with cystic fibrosis. J Pediatr 2004; 145:322-326. [PubMed]
Intestinal fat absorption and faecal elastase-1 (FE-1) were compared in subjects with CF at 33 CF centers. The authors concluded that FE-1 is an accurate, easily obtained screening test to classify pancreatic status in patients with CF. This information is important for prognostication, treatment, and to avoid misclassification in clinical research. They suggested that measurement of FE-1 should become a standard of care for patients with CF.  

 

2005 Bishop MD, Freedman SD, Zielenski J, Ahmed N, Dupuis A, Martin S, Ellis L, Shea J, Hopper I, Corey M, Kortan P, Haber G, Ross C. Tzountzouris J, Steele L, Ray PN, Tsui LC, Durie PR. The cystic fibrosis transmembrane conductance regulator gene and ion channel function in patients with idiopathic pancreatitis. Hum Genet 2005; 118:372-381. [PubMed]
Twelve (21%) of 56 patients with pancreatitis fulfilled current clinical criteria for the diagnosis of CF, but CFTR genotyping alone confirmed the diagnosis in only two of these patients. The authors concluded that extensive genotyping and ion channel testing (sweat tests and nasal potential differences) are useful to confirm or exclude the diagnosis of CF in the majority of patients with idiopathic pancreatitis.


Obviously CF should be considered in all patients with pancreatitis as it should in all men with congenital absence of the vas deferens.  

 

2005 De Boeck K, Weren M, Proesmans M, Kerem E. Pancreatitis among patients with cystic fibrosis: correlation with pancreatic status and genotype. Pediatrics 2005; 115:e463-9. [PubMed]
This multicentre study of 10071 patients with CF from 29 different countries revealed an estimated overall occurrence of pancreatitis among people with CF of 1.24% (10.3% in pancreatic sufficient and 0.5% in pancreatic insufficient patients).


Further confirmation that pancreatitis is a real risk for people with CF particularly if pancreatic sufficient. The overall incidence of pancreatitis reported by Jan Feigelson (2000 above) was 12.7%. Pancreatitis is identified by typical pain, evidence of inflammation on imaging and raised serum amylase or lipase.  

 

2009 Löhr JM. Hummel FM. Pirilis KT. Steinkamp G. Körner A. Henniges F. Properties of different pancreatin preparations used in pancreatic exocrine insufficiency. Eur J Gastroenterol Hepatol 2009; 21:1024-1031. [PubMed]
Measurements of size, surface, acid resistance, release of enzymes, pharmacokinetics and batch consistency were undertaken. available pancreatin preparations vary widely with respect to investigated parameters, which may have consequences for facilitating optimal digestion.  

 

2009 Gooding I. Bradley E. Puleston J. Gyi KM. Hodson M. Westaby D. Symptomatic pancreatitis in patients with cystic fibrosis. Am J Gastroenterol 2009; 104:1519-1523.[PubMed] The Royal Brompton Hospital database of adult CF patients was searched to identify all patients with symptomatic pancreatitis. Clinical details were taken from the case notes. PS pancreatitis patients were then compared with an age- and sex-matched PS control group drawn from the database. Sixteen patients (9 males) had suffered symptomatic pancreatitis, representing 1.6% of the total database. The mean age at CF diagnosis was 18.7 years, and at presentation with pancreatitis it was 28.8 years. These patients are PS at birth but are more likely to develop late PI status than PS patients without pancreatitis. R117H may be associated with this phenotype.

 

2011 Abu-El-Haija M. Sinkora M. Meyerholz DK. Welsh MJ. McCray PB Jr. Butler< J. Uc A.  An activated immune and inflammatory response targets the pancreas of newborn pigs with cystic fibrosis. Pancreatology 2011; 11:506-515. [PubMed]
In cystic fibrosis pancreatic disease begins in utero and progresses over time to complete destruction of the organ. Although inflammatory cells have been detected in the pancreas of humans and pigs with CF, their subtypes have not been characterized.
Using four-color flow cytometry, the authors analyzed the surface antigens of leukocytes in pancreas, blood, and mesenteric lymph nodes (MLN) of newborn pigs with CF. They discovered an activated immune response that was specific to the pancreas of newborn CF pigs; inflammation was not systemic. The presence of both innate and adaptive immune cells suggests that the disease process is complex and extensive.