Digestive Diseases and Sciences, I, bL 41. No. I0 (October 1996), pp. 2095-2099

Jejunal Water Electrolyte Transport Human Cryptosporidiosis

in

PAUL KELLY, MB, ANDREW V. THILLAINAYAGAM, MB, JACQUELINE SMITHSON, MB, JOHN B. HUNT, MD, ALASTAIR FORBES, MD, BRIAN G. GAZZARD, MD,

and MICHAEL J.G. FARTHING, MD

Cryptosporidiosis may have severe clinical consequences in both immunocompromised and immunocompetent individuals. However, pathophysiological mechanisms that are responsi- ble for diarrhea are poorly understood. We performed jejunal perfusion studies in patients with human immunodeficiency virus-related cryptosporidial diarrhea to measure water and electrolyte transport in vivo. Five patients with human immunodeficiency virus-related cryptosporidiosis and nine healthy volunteers were studied using a triple-lumen steady-state jejunal perfusion technique. Stool volume measurement and distal duodenal biopsy showed that the patients had diarrhea (600-1500 ml/24 hr) and morphological abnormalities of small intestinal mucosa. Net water, sodium, and chloride movement in the jejunum was not significantly different from healthy controls. In these patients with watery diarrhea and morphological mucosal abnormalities, we found no evidence that cryptosporidial diarrhea was due to a secretory state in the proximal small intestine. We conclude that diarrhea may be due to secretion of electrolytes and water eIttux more distally or to other abnormalities of gastrointestinal function.

KEY WORDS: cryptosporidium; diarrhea; acquired immune deficiency syndrome; intestinal transport; diarrhea pathogenesis.

Cryptosporidial infection is an important contributor to the gastrointestinal problems that characterize the acquired immune deficiency syndrome (AIDS), being responsible for chronic diarrhea (1-3) and for some cases of sclerosing cholangitis (4). Cryptosporidiosis is also responsible for diarrheal illness in children, for outbreaks associated with contamination of the water supply, and for diarrhea in travellers (5, 6). A water- borne outbreak in 1993 in Wisconsin, USA, involved

Manuscript received May 5, 1995; revised manuscript received October 30, 1995; accepted July 10. 1996.

From the Digestive Diseases Research Centre. Medical College of St. Bartholomew's Hospital, London EI2AT, UK; Chelsea and Westminster Hospital, 369 Fulham Road, London SWI0 9NH, UK; and St. Mark's Hospital, City Road, London ECIV 2PS, UK.

This work was supported by the Smith and Nephew Foundation. Address for reprint requests: Dr. Paul Kelly, Digestive Diseases

Research Centre, Medical College of St. Bartholomew's Hospital, 2 Newark Street, London El 2 A T . UK.

an estimated 403,000 individuals (7). In endemic childhood diarrhea in developing countries, this pro- tozoan infection is associated with a substantially higher mortality than other diarrheas (8).

The pathophysiology of human cryptosporidiosis is uncertain and currently the focus of debate (9). It has been frequently assumed that there must be a net secretory state in the small intestine, on the grounds that there are individual cases in which the stool volume is very high, and because of the case reported by Modigliani et al (10), in which a secretory state was demonstrated. Recently, Guarino et al (11) have de- scribed an enterotoxic factor in stool supernatant from infected calves that increased short-circuit cur- rent in human jejunum mounted in Ussing chambers, suggesting that C. parvum or the infected bowel lib- erates a secretory enterotoxin.

We have investigated water and electrolyte trans-

2095 Digestive Diseases and Sciences. VoL 4l. No. 10 (October 1990) {~163-21161%/11 t[|l t - 2~ }~.~55{19~51 D/l| ,t" I t;~h Plenu m Publislaing (.'orpt~rn lion

KELLY E T A L

TABLE 1. CLINICAL FEATURES, DISTAL DUODENAL HISTOPATHOLOGY, AND PARASITOLOGY IN FIVE PATIENTS WITH CHRONIC HIV-RELATED CRYPTOSPORIDIOSIS UNDERGOING JEJUNAL PERFUS1ON STUDIES*

Patient

1 2 3 4 5

Age (yr) 36 32 35 26 29 Sex M M M M F CD4 count (× 106/liter) 12 11 33 87 19 Duration of infection (months) 4 2 2 1 11 Drug therapy prior to study P,C,L,Z,Pe,M D,L,Ac Z,S,Ac,M,T, Pr Z,A,S,Ac P,S,Cp,M,L, Mb,Cy Stool volume (ml/24 hr) 800 600 700 600 1500 Villus height (~m) PVA 247 345 316 398 Crypt depth (/xm) ND 292 389 226 338 Villus-crypt ratio ND 0.85 0.89 1.40 1.18 Biopsy parasites None C. pan,urn None None C. pan,urn Concurrent infection None None E. bieneusi None None

* PVA, partial villous atrophy. ND, not determined. Normal values, based upon normal biopsies measured in this hospital; villus height: mean 330 (range 250-460) ~m; crypt depth: 103 (80-180)/~m; villus-crypt ratio: 3.2 (2.0-5.0). Codes for drugs: Z = Zidovudine, P = paromomycin, Ac = acyclovir, C = codeine phosphate, L = Ioperamide, S = co-trimoxazole, Pe = pentamidine (nebulized) weekly, T = testosterone, M = megesterol, Pr = propranolol, M = metoclopramide, D = dapsone, Cp = cephalexin, Cy = cyclizine, Mb = mebeverine.

port in human cryptosporidiosis using the triple- lumen steady-state jejunal perfusion technique, in patients with human immunodeficiency virus (HIV)- related chronic C. parvum infection and diarrhea.

MATERIALS AND METHODS

We studied five patients with AIDS and chronic crypto- sporidiosis attending the Chelsea and Westminster HIV Unit. Four were European men and one an African woman. All had stool volumes measured within three days of the study and passed three or more watery stools in the 24 hr before it. All patients underwent thorough diagnostic eval- uation and distal duodenal biopsy for morphometry within three weeks. Drugs were stopped 24 hr before the test. Nine healthy volunteers (male, aged 22-28) without evidence of gastrointestinal disorder were studied as controls.

Diagnostic evaluation comprised a search in stool smears for C. pan,urn by auramine staining, confirmed by a modi- fied Ziehl-Neetsen stain, and staining with calcifluor and strong trichrome for microsporidia. Distal duodenal or je- junal and rectal biopsies were stained with an immunoper- oxidase stain for cytomegalovirus and an immune stain for adenovirus. Small intestinal biopsies were stained with he- matoxylin and eosin to search for parasites, and semithin sections in plastic were stained with toluidine blue.

The triple-lumen perfusion technique was performed as described previously (12, 13). Following an overnight fast, a standard polyvinyl triple-lumen perfusion tube was placed in the proximal jejunum under fluoroscopic control, with the infusion port distal to the ligament of Treitz. A 10-cm mixing segment and a 30-cm test segment were used. Each subject was perfused with a physiological plasma electrolyte solution (PES) containing Na ÷ 140, K + 4.0, CI- 104, HCO3- 40 mmol/liter and polyethylene glycol (PEG) 4000 2.5 g/liter (including [14C]PEG 2/~Ci/liter) as a nonabsorb- able marker. The total osmolality of the perfusate was 289 mosm/kg. The solution was warmed to 37°C and gassed with 95% 02 and 5% CO2 for 30 min before and throughout the

study. PES was infused at 15 ml/min and samples collected at 1.5 ml/min from the 10-cm port. Perfusate was collected by siphon drainage from the distal (40 cm) port. One hour was allowed for achievement of a steady state (equilibration period), and then samples were collected for analysis over the next 30 rain (study period). Assay of the PEG concen- tration was carried out in a LKB Wallac 1210 ultrabeta scintillation counter using 0.5 ml of perfusate in 4.5 ml scintillation fluid (Optiphase Safe, Fisons). Cations were determined in an Instrumentation Laboratory 934 flame photometer, CI- in a Corning 925 chloride meter, and CO2 in a Corning 965 CO2 analyzer. The net electrolyte and water movements were calculated according to standard equations for the triple-lumen technique (14). According to general convention, net water and solute movement have been expressed in milliliters per centimeter per hour or millimoles per centimeter per hour, respectively, with neg- ative values indicating net secretion and positive values indicating net absorption.

The study was approved by the Research Ethics Commit- tees of the City and Hackney Health Authority and of the Chelsea and Westminster Hospital. All the subjects gave informed, written consent.

RESULTS

All subjects had diarrhea with increased stool vol- umes and low peripheral blood CD4 cell counts. Small bowel biopsy revealed an active inflammatory cell infiltrate in two, and one patient had concurrent microsporidiosis with Entemcytozoon bieneusi (Table I). Mucosal morphomet ry demonstra ted crypt length- ening.

Figure 1 shows water and electrolyte movement results. Mean water and electrolyte absorption was similar in patients [water 3.4 _+ 1.0 (SEM) ml/cm/hr; sodium 0.39 _+ 0.14 mmol/cm/hr; potassium 0.018 _+

2096 Digestive Diseases and Sciences, Vol, 41, No, 10 (October 1996)

JEJUNAL PERFUSION IN CRYPTOSPORIDIOSIS

A Water absorption (ml.cm-l.h -1)

g,

8,

7

6 5 4

3

2 1

0

-1

:---- 3.4 ---'3.2

i

Cryptosporidiosis Controls

B Sodium absorption (mmol.cm-l.h -1)

1 . 4 .

"L2-

0 . 8 .

0.6'

0.4.

0.2'

0

-0.2

-~ 0.39

I

Cryptosporidiosls Controls

C Chloride movement ( m m o l . c m - l , h -1 )

1

0.8

O.e

0.4

0.2

0

.0.2-

. 0 , 4

*

_0.22 " 0.11 I

Cryptosporidlosls Controls

Fig 1. Net water (a), sodium (b), and chloride (c) transport in five patients with HIV-related cryptosporidiosis and nine healthy vol- unteers. Horizontal bars represent mean values.

0.001; chloride 0.11 ___ 0.11, bicarbonate 0.21 + 0.07] and healthy controls (water 3.2 +-- 1.0 ml/cm/hr; so- dium 0.46 _ 0.16 mmol/cm/hr; chloride 0.22 __+ 0.11). The patient with the tendency toward secretion of water, sodium, or chloride was patient 5 (Table 1) in each case; she was the patient with the highest stool volume.

DISCUSSION

We were unable to demonstrate net secretion of water and electrolytes in the proximal human jejunum during infection with C. parvum, despite clinically

significant chronic diarrhea and the presence of epi- thelial abnormalities in distal duodenal biopsies.

We are aware of only one previous study that suggests that a net secretory state occurs in human cryptosporidiosis (10). The authors reported two cases of cryptosporidial diarrhea; in one of these, with concurrent salmonellosis and cytomegalovirus infec- tion and stool volumes of up to 7 liters/24 hr, the diarrhea was shown to be secretory using a perfusion technique. Biopsies suggested that this individual had prominent jejunal infection. The common belief that cryptosporidial diarrhea has a secretory origin gains support from the observation that fasting does not abolish the diarrhea (15). The studies reported here were conducted in the fasted state; water and elec- trolyte movements could still show abnormalities fol- lowing feeding.

The five subjects reported here had stool volumes of 600-1500 ml/24 hr. The individual with the highest stool volume had marginal evidence of net sodium and water secretion. Those individuals with stool vol- umes of 3-20 liters/24 hr due to cryptosporidiosis, and in whom the diarrhea is not diminished by fasting, may have net small intestinal water and sodium se- cretion. However, these high-output patients are in a minority. In a study from London, only one third of patients had stool volumes greater than 1500 ml/24 hr (16). The diarrhea associated with cryptosporidiosis is variable from day to day and from week to week (17), and so at any one time the proportion of infected individuals with high volume diarrhea may be small. Some variation in stool volume is related to CD4 count, but even in those with CD4 counts below 50/mm 3, there is a wide range of clinical disease expression (3).

It has been shown by Guarino et al (11) that human jejunum can respond with a chloride-mediated short- circuit current to a stool supernatant from calves infected with C. parvum. In a neonatal pig model, Argenzio et al (18, 19) found no evidence of a net secretory state for water with in vivo perfusion but did report a reduction in glucose-coupled sodium absorp- tion in the ileum, which was at least partly attribut- able to local prostanoid production. In their co- lostrum-deprived animals, secretion of sodium and chloride was detected in an Ussing chamber prepara- tion.

The trophozoites of C. parvum do not cause en- terocyte destruction, but are found in an intracellular position just beneath, and obliterating, the brush bor- der (20). In immunocompetent children, there is in- flammation and villous atrophy (21), so these

Digestive Diseases and Sciences, Vol. 41, No. 10 (October 1996) 2097

KELLY ET A L

changes, which are also seen in immunocompromised adults, may not simply be due to the presence of HIV. It is possible that alteration in villous enterocyte kinetics leads to an epithelium with reduced absorp- tive capacity as the ratio of mature to immature enterocytes falls (9, 22). In our patients, total crypt cell mass was increased, but net secretion was not observed. Alteration in the villus-crypt ratio does not in itself lead to diarrhea, as individuals can be found with subtotal villous atrophy but without diarrhea, but it may contribute to a multifactorial process.

A possible explanation for the absence of a secre- tory state in our patients may lie in the variable distribution of cryptosporidiai infection in the gut. In an immunocompetent neonatal piglet model, the dis- tribution of infection was found to migrate in a caudal direction over a period of about five days (23). Jejunal or distal duodenal biopsies reveal trophozoites of C. pan,urn in about 50% of cases. Genta et al (24) showed that the intensity of infection as assessed by oocyst counts in stool was reflected in the proportion of duodenal biopsy surface area covered by the tro- phozoites, but no direct relationship was found with the severity of the diarrhea. It may well be that a net secretory state exists only in the most densely infected region of the gastrointestinal tract, and the precise nature of the physiological lesion will therefore de- pend upon the extent of the infection at any one time. In our patients with moderate volume diarrhea, ileal or ileocolonic infection may have been dominant, with a secretory lesion confined to these sites. To confirm or refute this hypothesis would require a study of morphometry and water and electrolyte transport in the ileum and colon, but this would be invasive and probably not easily tolerated. Distur- bances of intestinal motility have not been reported in cryptosporidiosis but may contribute to the produc- tion of diarrhea.

All the patients were taking drugs that could po- tentially alter gastrointestinal transit, but these were all stopped prior to the study and these drugs would not abolish a significant secretory state. Furthermore, the stool volumes were shown to be abnormally high. Our study demonstrates that it is possible to have cryptosporidial diarrhea without proximal small intes- tinal secretion.

REFERENCES

1. Kotler DP, Francisco A, Clayton F, Scholes JV, Orenstein JM: Small intestinal injury and parasitic diseases in AIDS. Ann Intern Med 113:444-449, 1989

2. Gellin J, Soave R: Coccidian infections in AIDS. Med Clin North Am 76:205-234, 1992

3. Blanshard C, Jackson AM, Shanson DC, Francis N, Gazzard B: Cryptosporidiosis in HIV seropositive patients. Q J Med 85:813-823, 1992

4. Forbes A, Blanshard C, Gazzard B: Natural history of AIDS rclated sclcrosing cholangitis: A study of 20 cases. Gut 34:116- 121, 1993

5. Casemore DP: Epidemiological aspects of human cryptospo- ridiosis. Epidemiol Infect 104:1-28, 1990

6. Katelaris P, Farthing MJG: Cryptosporidiosis--an emerging risk to travellers. Travel Med Int 1:10-14, 1992

7. MacKenzie WR, Hoxie N J, Proctor ME, Gradus MS, Blair KA, Pcterson DE, Kazmierczak J J, Addis DG, Fox KR, Rose JB, Davis JP: A massive outbreak in Milwaukee of Co,ptospo- ridium infection transmitted through the public waler supply. N Engl J Med 331:I61-167, 1994

8. Molbak K, Hojlyng N, Gonschau A, Sa JCC, Ingholt L, da Silva A, Aaby A: Cryptosporidiosis in infancy and childhood mortality in Guinea Bissau. Br Med J 307:417-420, 1993

9. Sears CL, Guerrant RL: Cryptosporidiosis: The complexity of intestinal pathophysiology. Gastroenterology 106:252-254, 1994 (editorial)

10. Modigliani R, Bories C, Le Charpentier Y, Salmeron M, Messing B, Galian A, Rambaud JC, Lavergne A, Cochand- Priollet B, Desportes I: Diarrhoea and malabsorption in AIDS: A study of four cases with special emphasis on opportunistic protozoan infestations. Gut 26:179-187, 1985

11. Guarino A. Canani RB, Pozio E, Terracciano L, Albano F, Mazzeo M: Enterotoxic effect of stool supernatant of Crypto- spotidium infected calves on human jejunum. Gastroenterology 106:28-34, 1994

12. Hunt JB, Elliott EJ, Fairclough PD, Clark ML, Farthing MJG: Water and solute absorption from hypotonic glucose- electrolyte solutions in human jejunum. Gut 33:479-483, 1992

13. Hunt JB, Thillainayagam AV, Carnaby S, Fairclough PD, Clark ML, Farthing MJG: Absorption of a hypotonic oral rehydration solution in a human model of cholera. Gut 35:211- 214, 1994

14. Cooper H, Levitan R, Fordtran JS, Ingelfinger FJ: A method for studying absorption of water and solute from the human small intestine. Gastroenterology 50:1-7, 1966

15. Gritfin GE: Malabsorption, malnutrition and HIV disease. Baill Clin Gastroenterol 4(2):361-373, 1990

16. Connolly GM, Dryden MS, Shanson DC, Gazzard BG: Cryp- tosporidial diarrhoea in AIDS and its treatment. Gut 29:593- 597, 1988

17. McGowan I, Hawkins A, Weller I: The natural history of cryptosporidial diarrhoea in HIV infected patients. AIDS 7:349-354, 1993

18. Argenzio RA, Liacos JA, Levy ML, Meuten DJ, Lecce JG, Powell DW: Villous atrophy, crypt hyperptasia, cellular infil- tration, and impaired glucose-Na absorption in enteric crypto- sporidiosis of pigs. Gastroenterology 98:1129-1140, 1990

19. Argenzio RA, Lecce J, Powell DW: Prostanoids inhibit intes- tinal NaCI absorption in experimental porcine cryptosporidi- osis. Gastroenterology 104:440-447, 1993

20. Marcial MA, Madara JL: Cryptosporidium: Cellular localiza- tion, structural analysis of absorptive cell-parasite interactions in guinea pigs, and suggestion of protozoan transport by M cells. Gastroenterology 90:583-594, 1986

21. Phillips AD, Thomas AG, Walker Smith JA: Cryptosporidium,

2098 Digestive Diseases and Sciences, Vol. 41, No. 10 (October 1996)

JEJUNAL P E R F U S I O N IN CRYPTOSPORIDIOSIS

chronic diarrhoea, and the proximal small intestinal mucosa. Gut 33:1057-1061, 1992

22. Buret A, Gall DG, Nation PN, Olson ME: Intestinal protozoa and epithelial kinetics, structure and function. Parasitol Today 6:375-380, 1990

23. Vitovec J, Koudela B: Pathogenesis of intestinal cryptosporidi-

osis in conventional and gnotobiotic piglets. Vet Parasitol 43:25-36, 1992

24. Genta RM, Chappell CL, White AC, Kimball KT, Goodgame RW: Duodenal morphology and intensity of infection in AIDS related intestinal cryptosporidiosis. Gastroenterology i05:1769-1775, 1993

Digestive Diseases and Sciences, Vol. 41, No. 10 (October 1996) 2099