Failure to improve parameters of lactose maldigestion using the multiprobiotic product VSL 3 in lactose maldigesters : A pilot study

Division of Gastroenterology, Department of Medicine, McGill University School of Medicine, Montreal, Quebec Correspondence and reprints: Dr A Szilagyi, The Sir Mortimer B Davis Jewish General Hospital, 3755 Cote St Catherine Road, Room G-327, Montreal, Quebec H3T 1E2. Telephone 514-340-8144, fax 514-340-8282, e-mail aszilagy@gas.jgh.mcgill.ca Received for publication October 13, 2003. Accepted December 17, 2003 R Yesovitch, A Cohen, A Szilagyi. Failure to improve parameters of lactose maldigestion using the multiprobiotic product VSL3 in lactose maldigesters: A pilot study. Can J Gastroenterol 2004;18(2):83-86.

T he concept of prebiotics was suggested by Gibson and   Roberfroid (1).It is based on the theory that some undigested nutrients (usually polysaccharides) which bypass the upper gastrointestinal tract specifically alter the metabolism and expand bacterial populations (usually lactic acid-producing bacteria) thought to exert beneficial effects on the host.There are a number of known natural (those found in leeks, hickory, onions [1] and breast milk [2]) and manufactured (eg, transgalactoliposaccharides [3] and lactulose [4]) prebiotics.Lactose in lactose maldigesters (lactase nonpersistent [LNP] subjects) may be a conditional natural prebiotic because approximately 67% to 70% of the world's population lose the ability to digest lactose (5).In these LNP subjects it has been reported that continued lactose consumption leads to improved handling of the disaccharide (6,7).The phenomenon of improved lactose tolerance on rechallenge after consuming lactose for a defined period of consumption is putatively due to colonic bacterial alteration of metabolism and expansion of targeted microbial flora (7,8).This process is termed colonic adaptation and is defined clinically by reduced measured breath hydrogen (H 2 ) after rechallenge with the target sugar (in this case lactose) (7).Improved symptoms of intolerance and increased fecal β-galactosidase also constitute part of adaptation.In vitro data suggest that adding Lactobacillus or Bifidobacteria species to an adapting system facilitates lactose digestion (9,10).Furthermore, there is a single study suggesting that H 2 production is reduced on lactulose rechallenge after consuming lactobacillus parvum 299V (11).Lactulose and lactose may exert similar effects on colonic bacteria (12).
Although beneficial in its own right (improved nutritional assimilation) the process of adaptation and measurement of resulting breath H 2 may be used as a functional test of colonic bacterial status.For example, we recently showed that patients with inflammatory bowel disease (IBD) in remission are slow or unable to adapt to lactulose, as are healthy controls based on the above outlined concept (13).This adaptive test supports the notion that flora in IBD is abnormal (14,15).
In an effort to further evaluate the potential of colonic adaptation as a biological test of colonic floral function we attempted to link the test with a probiotic of suggested benefit in IBD.The eight-probiotic-containing product VSL3 (Seaford Pharmaceucticals Inc, Ontario) has been shown to be therapeutically useful in pouchitis (16,17) and ulcerative colitis in remission (18,19), and preliminary data show benefit in active ulcerative colitis (20).We therefore undertook a small pilot study and fed two doses of VSL3 to subjects with known lactose maldigestion.However, no improvement was found.

SUBJECTS AND STUDY DESIGN
Ten healthy paid volunteers of multiethnic origin who were previously established to be lactose maldigesters (four men, six women; mean age of 38.6±14.6 years) were recruited.The subjects underwent two lactose (50 g) challenge tests.The first was followed by ingestion of VSL3 bacteria and then a second similar challenge test was carried out at the end of the adapting period after 17 days.The 17-day period was arbitrarily chosen based on the previously published observation that adaptation is associated rapidly with altered floral microorganisms with lactose (8).Secondly, two other studies using naïve probiotics (21,22) showed that bacteria were readily cultured from stool in a short period of time (one week or less).Five subjects consumed 1 g (450×10 9 ) of the bacteria and the other five consumed 4 g (4×450×10 9 ) bacteria.Bacteria were usually mixed with water and taken with meals.The side effects were minor and consisted mainly of nausea, mostly in the high dose group.
The VSL3 probiotic contains four strains of Lactobacillus (L casei, L plantarum, L acidophilus and L bulgaricus), three of Bifidobacterium (B longum, B breve and B infantis), and one strain of Streptococcus salivarius subspecies thermophilus.The probiotics were a gift of Dr C De Simone (Italy) via Seaford Pharmaceuticals Inc (Ontario).Three random batches were checked for viability by plating the contents of a packet on standard blood agar plates.The microbiological evaluation was carried out in the Mortimer B Davis Jewish General Hospital hospital laboratory under the supervision of Dr M Miller.
Informed consent was obtained from all participants and this study was approved by the Research Ethics Committee of the Sir Mortimer B Davis Jewish General Hospital.

Lactose challenge tests
After a specified supper (with avoidance of lactose gas-forming foods; rice and hamburger were suggested) volunteers were asked to fast (water ad libitum).None of the volunteers took antibiotics, analgesics or any anti-or prokinetic medications.They were asked not to chew gum or smoke but were allowed quiet movement throughout a 4.5 h measurement of exhaled H 2 .A 50 g lactose load mixed in water was consumed and after a baseline measurement, breath H 2 was measured every 30 min using a validated handheld H 2 electrochemical sensor (EC60 Vitalograph hydrogen breath monitor, Bedfont Scientific Ltd, United Kingdom) (23,24).The model uses a sealed electrochemical sensor which can detect H 2 in parts per million (ppm) v/v in a range of 0 ppm to 2000 ppm.The average of three exhaled breaths was used at each time interval.Results at each interval were corrected by subtracting the baseline H 2 ppm.In addition, symptoms were recorded by subjects each half hour and graded.The targeted symptoms were bloating, gas, cramps and diarrhea.For each, a score of 1 was assigned for mild (subject aware), 2 for definite presence and 3 for severe symptoms.Although theoretically the maximum score was 108 (9×12 periods) it is virtually impossible to score 12 at each interval.Therefore, for practical purposes, the score system was an open-ended scale.In the case of diarrhea, each loose bowel movement was scored every 30 min.This ordinal scale system was used by the authors previously (12).At the end of 17 days, subjects were asked to repeat the 50 g lactose challenge and record symptoms (test 2).

Statistical analysis
Student's paired t tests were used to analyze differences between test 1 and test 2 in both groups.Based on the authors' previous experience of inducing colonic adaptation with lactulose at low dose, it was reasoned that eight to 10 patients per group would suffice to demonstrate a significant difference of a 45% reduction in the sum of breath H 2 on test 2.This difference was chosen because we previously showed (12) that using lactulose as a prebiotic a reduction of 43% between two challenge doses of lactose was statistically significant (12).While this difference may or may not be clinically significant at an individual level for lactose tolerance, it clearly allows demonstration of a population test effect on colonic microbial function.In fact, in our original abstract, six subjects already demonstrated a statistically significant trend at P=0.054 (25).Pearson correlation was used to compare symptoms with sum of breath H 2 .Statistical significance was accepted at P<0.05.

RESULTS
Four men and six women with a mean age of 38.6±14.6 years (range 24 to 66) undertook the studies.The average baseline H 2 ppm in test 1 was 0.8 ppm (range 0 to 2) in test 1 (n=10) and 4.5 ppm (range 0 to 16) in test 2 (n=10).Changes in Σ4 h breath H 2 and symptom scores for the individuals in group I (one packet/day VSL3) and group II (four packets/day VSL3) are shown in Table 1.Mean ± SD for Σ4 h breath H 2 for test 1 was 279.8±16.4(CI 256 to 297) for group I, and 323±158.9(CI 192 to 597) for group II.In test 2 Σ4.5 h breath H 2 for group I was 302±157.8(CI 152 to 480) and group II 401.0±198.8(CI 139 to 682).While these differences were statistically insignificant, there was a trend for higher rather than lower Σ4.5 h breath H 2 on test 2 in group II.Similarly, there were no statistically significant changes in symptom score for either group I or group II.While there was a weak significant correlation between global symptom scores for both groups in test 1 (r=0.65),this possible relationship was lost in test 2 (r=0.131).Random cultures of probiotics showed that the ingested products were viable.

DISCUSSION
This small pilot study did not show any benefit for improving lactose intolerance under laboratory conditions.Nevertheless, we felt it is still valuable to report it because of the principle involved in the logic of such studies.
To date, the only successful studies showing improvement of lactose maldigestion with probiotics are those using yogurt (26), other fermented products (27,28) or the addition of live probiotics to milk (29,30).The mechanism in improvement is thought to be due to three factors.First, fermented products contain 30% less lactose.Second, the enzyme β-galactosidase is provided exogenously and continues to digest lactose in the small bowel of LNP subjects.Finally, yogurt may delay gastric emptying and intestinal transit time, thereby reducing the fractional quantity of lactose at the cecum (26).Delayed rate of delivery is putatively associated with reduced symptoms (31,32).
Providing probiotics directly removes the physiological attributes of fermented products.The results obtained here and noted in the literature (21,22) stand in stark contrast to the ability to cause colonic adaptation and improvement in lactose maldigestion with continued lactose consumption (6,7).The process of adaptation with lactose alters fecal flora and favours the emergence of Lactobacillus and Bifidobacteria species (8).
The specific rationale of using VSL3 is that this probiotic has been successfully used therapeutically in IBD (16)(17)(18)(19)(20).In the paradigm of providing naïve probiotics, the following assumptions are made: in lactose maldigesters minimal or zero consumption of lactose may lead to decreased populations of putatively beneficial bacteria; with regular consumption of lactose such bacteria are now numerically expanded and metabolically activated; and if such bacteria (Lactobacillus and Bifidobacteria) are now given with fermented products or even unfermented milk they may become metabolically activated, facilitating lactose consumption before colonic embedding.If we now wish to establish that unique expansion of an important disease-beneficial bacteria is necessary for enhanced metabolism of lactose on challenge, prestimulation would abrogate such a conclusion.In vitro evidence suggests that infusion of naïve Lactobacillus or Bifidobacteria aids in consumption of lactose in fecal slurries (9,10).It is not unreasonable to try to reproduce this effect in vivo.
There are indeed few studies that have examined the possible impact of direct probiotics on lactose maldigestion.Hove et al (21) used simultaneous addition of 12 capsules of mixed Lactobacillus and Bifidobacteria species to a 50 g lactose challenge test.They did not find any improvement in breath H 2 despite previously demonstrating the ability to recover bacteria from stool.More recently, Saltzman et al (22), in a similar elaborate study used a special L acidophilus BG2FO4 species (possessing a high β-galactosidase content) which too was cultured from stool after seven days in lactose maldigesters.However, breath H 2 tests at the end did not differ significantly from the beginning, although the sum did increase on the second challenge test (22).
In our own study we also were unable to confirm adaptation by directly providing probiotics.While we were not able to verify colonic bacterial embedding with VSL3 it has been published that such bacteria are cultured from stool after 20 days (33).We feel that some bacterial embedding did occur in the present study because our 17-day adaptation period is close to 20 days and stools were not tested between 10 and 20 days after commencement of feeding, leaving open the possibility of earlier embedding (32).Furthermore, a dose response is suggested to have occurred in the current study because of a noted trend toward increased breath H 2 production with a higher dose of bacteria.This finding was similar to that in the report by Saltzman et al (22).
A longer feeding period and achievement of our intended group sizes may have strengthened our results.It is important to note, however, that the observed trend showed an outcome opposite to what we were expecting.Therefore, achieving a significant difference with the currently observed trend would not have helped our goal.As a result, based on the reports from the literature and this study, we can cautiously conclude that providing metabolically naïve probiotics will not improve lactose maldigestion.

CONCLUSION
This pilot study does not provide a quick link test connecting probiotic-induced colonic adaptation and bacteria proven to be beneficial in therapy of IBD.Despite the limitations of this small pilot study, the results mimic those reported for other efforts to induce improvement in lactose handling by providing naïve probiotics directly.The uniformity of failure suggests that targeted bacteria to date need to be metabolically turned on before demonstrating efficacy.However future studies should include a longer colonic embedding period to determine whether metabolism of lactose might be altered over the extended time.

TABLE 1
Demographics and distribution of the Σ Σ of 4.5 h breath hydrogen (BH 2 ppm) and Σ Σ of each 30 min (for 4.5 h) symptom scores (ss) for group I (low dose VSL3) and group II (high dose VSL3) Patient Sex Age Test 1 BH 2 Test 1 ss Test 2 BH 2 Test 2 ss None of the differences were statistically significant.However, the P value was 0.76 for Σ 4.5 h BH 2 in group I and 0.18 in group II comparing mean BH 2 for test 1 versus test 2, respectively.F Female; M Male