Rectal forceps biopsy procedure in cystic fibrosis: technical aspects and patients perspective for clinical trials feasibility

Cystic Fibrosis (CF), the most common life shortening autosomal recessive disease of Caucasian populations, is caused by mutations in the CF transmembrane conductance regulator (CFTR) gene encoding a chloride (Cl^-) channel expressed at the apical membrane of epithelial cells, a major regulator of salt and water transport in epithelia [1]. CF is dominated by respiratory disease but other organs are also affected including the pancreas, intestine and sweat gland as well as male reproductive tract [2].

Although the clinical diagnosis of classic (severe) forms of CF is straightforward, for other patients there is wide variability in the clinical presentation and organ involvement, thus making the CF diagnosis more challenging [3–7]. Moreover, increasing numbers of asymptomatic patients are currently identified through extended programs of CF newborn screening [4, 8–11].

One of the most useful and sensitive laboratory parameters used for the diagnosis and prognosis of CF, is ex vivo assessment of CFTR-mediated Cl^- secretion channel in freshly collected rectal biopsies [8, 9, 12–19]. Moreover, ongoing clinical trials of novel therapeutic CFTR-modulators require improved and robust biomarkers to adequately assess their in vivo efficacy on CFTR. Indeed, there is also great potential to exploit this method to pre-clinically assess compound efficacy directly on human tissues ex vivo, as we previously showed [20] or as a biomarker in clinical trials of novel CFTR-modulators [21–23]. Moreover, it may even be used to evaluate patient/CFTR genotype responsiveness to a drug through a personalized-medicine approach.

For such disseminated usage of this method, standardized operational procedures (SOP) for bowel preparation and biopsing are essential to ensure good tissue viability for the quantitative assessment of the bioelectric parameters [13]. Moreover, since the procedure involves biopsing, a somewhat invasive procedure possibly triggering psychological rejection, there should be clear information on how it is perceived from the patients’ perspective (comfort, invasiveness, pain, sedation requirement, etc.) to obtain an overall assessment of the method.

Our two-fold aim of the present study was: a) to evaluate the technical procedure regarding the quality for bioelectrical/biochemical laboratory analyses of 580 rectal specimens from 132 individuals (CF and non-CF), namely by comparing different bowel preparations and different biopsy forceps sizes as well as regarding the safety of the procedure to the patient; and b) to determine individuals’ assessment regarding the rectal biopsy procedure feasibility to be possibly used as an outcome measure in clinical trials, as successfully described in other studies [24, 25].

Our results demonstrate that best tissue viability for Ussing chamber measurements results after bowel preparation with isotonic solution (NaCl 0.9%) and usage of jumbo (vs standard size) biopsy forceps allowing collection of larger specimens without disrupting tissue integrity. Obtaining rectal biopsies with a flexible endoscope and jumbo forceps is also shown here to be a safe procedure for use in both children and adults (age range was 6 months to 52 years). Data collected on patient’s comfort show that the great majority of individuals (76%) did not report major discomfort due to the short procedure time (max 15 min) and considered it relatively painless (79%), regardless of sedation. Moreover, most individuals accept repeating the procedure for at least one more time (88%) and 53% for more than 4 times, supporting the feasibility of the current approach as an outcome measure in clinical trials.

A flow-chart summarizing the technical biopsing aspects assessed in the present study is shown in Figure 1.

Bowel preparation and biopsy forceps

The bowel cleaning procedures, consisting in administration of an oral laxative (for colonoscopy) or by enema (for sigmoidoscopy) allowed equally good visualization of the rectal mucosa and forceps during the sigmoidoscopy/colonoscopy procedure. Expectedly, jumbo biopsy forceps (3.4 mm Ø) generated larger specimens than standard forceps (2.5 mm Ø), thus facilitating the mounting of the tissue in Ussing chamber inserts.

Data referring to bowel preparation (Table 1, Additional file 1: Table S1) show that there are statistically significant differences (p = 0.054, 90% confidence interval) for tissue integrity between the preparations using NaCl 0.9% (1.19 ± 0.12) and glycerol 12% (1.82 ± 0.21). Oral mannitol was not statistically different from isotonic saline, independently of the biopsy forceps used (Table 1).

Bowel Preparation		NaCl 0.9%		Glycerol 12%		Mannitol 20%
Biopsy Forceps		Jumbo	Standard	Jumbo	Standard	Jumbo	Standard
Tissue integrity		1.00 ± 0.11	2.00 ± 0.35	1.48 ± 0.31	2.21 ± 0.23	1.02 ± 0.02	1.97 ± 0.30
Friability		1.23 ± 0.12	1.36 ± 0.24	1.50 ± 0.31	1.81 ± 0.22	1.13 ± 0.33	1.33 ± 0.25
Bleeding		1.01 ± 0.08	1.21 ± 0.18	1.27 ± 0.21	1.16 ± 0.29	1.05 ± 0.22	0.83 ± 0.32
Mucus		1.12 ± 0.11	1.18 ± 0.17	0.91 ± 0.24	0.88 ± 0.13	0.74 ± 0.13	1.06 ± 0.13
Sub-mucosa (n)	Yes	16	3	3	1	2	1
	No	42	9	4	4	12	10
R te (Ω.cm 2 )		21.82 ± 1.03	14.37 ± 1.21	18.75 ± 2.66	13.57 ± 2.70	18.69 ± 0.98	13.32 ± 1.11

NOTE: Results are mean ± SEM; n = 107.

Regarding biopsy forceps, mean values for tissue integrity (Jumbo = 1.07 ± 0.09 vs Standard = 2.04 ± 0.18) and also for R_te (Jumbo = 20.97 ± 0.81 Ω.cm² vs Standard = 14.01 ± 0.86 Ω.cm²) are statistically different between the forceps used (p = 5.51x10^-7 for tissue integrity and p = 8.42x10^-8 for R_te, Table 1, Additional file 1: Table S2). Friability (i.e. tissue susceptibility to breakdown with manipulation) is the only parameter significantly affected by the presence of sub-mucosa (Additional file 1: Table S3, p = 0.008), but this may result from more tissue manipulation being required to remove the sub-mucosa. Regarding bleeding and mucus, their presence/abundance were not influenced by bowel preparation nor biopsy forceps (Table 1).

In addition, we also found statistically significant correlations (Table 2, Figure 2) between R_te and tissue integrity (r = −0.438, p = 8.51x10^-5) and between R_te and friability (r = −0.305, p = 6.65x10^-3), which is supported by partial correlations with bowel preparation and biopsy forceps. Data concerning presence/abundance of blood and mucus do not correlate with tissue viability-R_te (Table 2, Figure 2). We found no influence by usage of sedation in the procedure regarding tissue viability (Additional file 1: Table S4).

	Transepithelial resistance (Rte)
	Pearson ( r )	p-value
Tissue integrity	−0.438	8.51 x 10 -5
Bowel preparation	−0.403	3.74 x 10 -4
Biopsy forceps	−0.244	0.036
Friability	−0.305	6.65 x 10 -3
Bowel preparation	−0.301	7.88 x 10 -3
Biopsy forceps	−0.277	0.015
Bleeding	−0.176	0.110
Bowel preparation	−0.185	0.093
Biopsy forceps	−0.182	0.100
Mucus	+0.078	0.507
Bowel preparation	+0.032	0.787
Biopsy forceps	+0.115	0.330

NOTE: Pearson (r) and p-values indicating significant correlations are highlighted (n = 107).

Histological and macroscopic evaluation of rectal tissues and biochemical analysis

Histological examination of rectal specimens revealed no obvious abnormalities and only some inflammatory cells were found, independently of both tissue origin (CF or non-CF) or bowel preparation (Figure 3A).

Tissues were evaluated for integrity, friability, bleeding, mucus, and thickness of biopsy (presence of sub-mucosa or not) under a dissecting microscope by two different technicians in blinded way (Table 1). As shown above, the scores obtained from this evaluation were correlated with bowel preparation, biopsy forceps and measurements of transepithelial resistance (R_te) to assess tissue viability (Table 2).

We have also used these native colonic tissues to further look at maturation pattern and localization of wt- and mutant-CFTR protein, which can serve as a valuable tool to evaluate the effects of CFTR-modulators together with Ussing chamber measurements of CFTR-mediated Cl^- secretion. As shown in Figure 3B, we were able to detect CFTR protein at the plasma membrane for both wt-CFTR and F508del/P205S-CFTR (P205S is a class IV mutation), but F508del/F508del-CFTR was retained into cytoplasm, close to the basolateral membrane. These findings were in accordance with the maturation pattern of CFTR protein (Figure 3C), which was found to be fully-glycosylated (presence of band C) for wt-CFTR and another class IV CFTR-mutant (R334W), but failed to mature in the biopsy from a F508del/F508del patient.

Measurements of CFTR function in rectal biopsy specimens have proven its value in the fine-diagnosis of patients with milder or “non-classical” forms of CF, in particular when sweat test results are equivocal or borderline and/or if CFTR-disease causing mutations are not readily identified by DNA mutation analysis [12, 14, 16]. This approach also serves as a sensitive test to predict the prognosis when rare CFTR mutations are not identified by standard screening tests [12, 14, 16]. Compared to the airways, the rectal epithelium is easily accessible at any age, expresses higher levels of CFTR, thus increasing robustness of the measurements and, as shown here (Figure 3-A), does not undergo major secondary tissue destruction/remodelling as those occurring in CF airways [13].

Moreover, this technique has even potential for a much wider scope, namely to monitor diseases, affecting epithelial ion transport [29–31]. Although these studies were so far only performed in mice, the results are highly promising for the clinic.

Because functional and biochemical data are not available for most rare mutations, these specimens can also be used to establish functional and biochemical correlations with rare CFTR genotypes (Figure 3B, 3C), as also reported in other studies [12, 14, 16, 28, 32, 33]. Furthermore, this approach may be used as well to validate efficacy of novel CFTR-modulator compound/drugs directly on human native tissues for such rare mutations [21–23]. Indeed, rectal biopsies are already in use as outcome measures in clinical trials for other diseases [24, 25] and for some a high number of biopsies (n = 28) per patient has been reported without complications [25]. The same can thus be translated into the CF field.

Since good tissue viability is critical for quantitative assessment of bioelectric responses our recommendation is that specimens are maintained in appropriate medium on ice until used in functional measurements, which should be performed immediately [13]. To minimize edge damage (and consequent liquid/electric leakage), tissues should be mounted under a dissecting microscope for optimal orientation of biopsy on the insert opening and to prevent tissue damage with excessive instrument manipulation.

The present study aimed to determine whether and how bowel preparation for sigmoidoscopy and choice of biopsy forceps influence tissue viability for subsequent laboratory analyses, in particular, bioelectric measurements. We compared two commonly used solutions for bowel preparation at our endoscopy unit, namely glycerol and isotonic saline enema (for sigmoidoscopy) as well as oral mannitol (for colonoscopy). Our data show that isotonic saline solution, which we speculate is less harmful for the mucosa, is superior to glycerol-based preparation, but there are no significant differences between NaCl-enema and oral mannitol. Indeed, differences were detected between mean values for the tissue integrity parameter between biopsies obtained after NaCl- and glycerol-based bowel preparations, the former showing better results (Tables 1 and Additional file 1: Table S1).

The current study only focuses on the biopsing forceps procedure, as this is the routine procedure at our endoscopy unit. Indeed, this is the approach that we regularly use because it uses an endoscope which allows direct visual inspection of the mucosa area to be biopsied during the procedure to avoid hitting on damaged tissue or any abnormal vascular structures. This reduces the hemorrhage risk for the patient and increases the safety profile of the procedure. Indeed, excessive bleeding, although never occurring in this study, would be rapidly identified and thus immediately managed. It also allows avoiding biopsing the same site twice (Additional file 3 video S1: Rectal biopsy procedure).

However, previous [12, 13, 15, 17–19] and recent reports [14, 16, 34, 35] indicate that biopsies obtained by suction can be similarly applied (albeit without visual contact) and major complications have equally not been reported. In fact, the values for transepithelial resistance found by ourselves [14] (14–21 Ω × cm²) are comparable with those found previously using similar (forceps) [12] - 24 Ω × cm² - or even different (suction) techniques [36, 37]: 27 Ω × cm², and 24–39 Ω × cm² (no mean values given). That approach is thus expected to be equally tolerated by the patients’ while also providing adequate specimens for electrophysiology studies. Here, we did not compare suction and forceps biopsy, but just the forceps size (standard vs jumbo forceps) in terms of how sample size might affect its viability and also regarding safety of the procedure. Overall, mean values for tissue integrity (jumbo = 1.07 ± 0.09 vs standard = 2.04 ± 0.18) and also for R_te (jumbo = 20.97 ± 0.81 Ω.cm² vs standard = 14.01 ± 0.86 Ω.cm²) were shown to be statistically different (p = 5.51x10^-7 and p = 8.42x10^-8, respectively, Tables 1, Additional file 1: Table S2), with jumbo forceps providing the best results.

Importantly, data for both tissue integrity and friability show good correlations (positive and negative, respectively) with tissue viability (assessed by R_te measurements) and are influenced by bowel preparation and biopsy forceps (Table 2, Figure 2). Our data also show that collection of superficial rectal biopsies with (jumbo/standard) forceps constitutes a safe procedure, as we observed no complications, similarly to what others previously reported for studies on Hirschsprung [38] or inflammatory bowel disease [39].

Data concerning presence/abundance of bleeding and mucus do not correlate with tissue viability (R_te), but interestingly there is a trend positive correlation between mucus and R_te (r = +0.078), indicating that presence of mucus (i.e. probably resulting from a less “aggressive” bowel cleaning) could somehow serve to preserve tissue viability (Table 2 and Figure 2).

Therefore, performing bowel preparation with isotonic saline and obtaining the rectal biopsies with jumbo forceps are demonstrated to constitute the best combination for the procedure (Table 1) with the highest mean values for R_te (21.82 ± 1.03 Ω.cm²) and the best for tissue integrity (1.00 ± 0.11). On the other hand, usage of 12% glycerol enema and smaller (standard) forceps produce the less viable tissues, namely (Table 1): worst tissue integrity (2.21 ± 0.23) and R_te (13.57 ± 2.70 Ω.cm²), and highest friability (1.81 ± 0.22), thus rendering quantitative determination of CFTR-mediated Cl^- secretion less reliable. It is likely that usage of the glycerol-based enema, acting as an osmotic laxative and thus increasing the luminal volume, which stretches the mucosa, makes it more susceptible to friability and disruption, compromising tissue integrity. Alternative for bowel preparation in children could be sodium citrate + sodium lauryl sulfoacetate, together with glycerol + sorbitol (“Microlax®”) or dioctyl sulfosuccinic acid sodium salt + sorbitol (“Clyss-go®”), which we had previously experienced to yield viable specimens if done the day before sigmoidoscopy (data not shown), probably allowing rectal mucosa to recover from this enema procedure. However, as these procedures were not rigorously assessed, we cannot compare them we the ones used in the present study.

The easy access to the patients rectum and the low innervation of this area minimizing pain, makes this approach to be expectedly well tolerated [12, 14]. Moreover, modern gastroenterology techniques and instruments currently applied in outcome measures for clinical trials have made this approach increasingly simpler and easier [25]. Indeed, in the present study, concerning only the sigmoidoscopy/rectal biopsy procedure, sedation was used primarily to reduce anxiety and ensure cooperation (only for 24% of the sigmoidoscopy/rectal biopsy cases sedation was preferred, see Methods), especially in small children that generally demonstrate less cooperation in any medical procedure In cooperative, non-anxious patients, our experience recommends that the procedure is performed with no anaesthesia.

Regarding safety of the patients, one CF patient who complained of abdominal pain (see Results) was the only patient with inappropriate bowel cleaning, where more insufflation had to be used in order to carry out the procedure. This patient also had a clinical history of surgical interventions for meconium ileus and colonic adhesions, which might be related to the abdominal pain. We also report minor complications that cannot be fully related to the sigmoidoscopy procedure: one patient who vomited (although his parents informed that this usually happens when he fasts) and another patient who underwent both upper endoscopy and sigmoidoscopy also reported pain, but it is likely that the former procedure was the source of pain.

Thus, the present study also shows that this procedure with jumbo forceps is safe to be applied from young children to adults (age range was 6 months to 52 years), which application is relevant also for other disorders as Hirschsprung or inflammatory bowel disease. Although we have no experience, it also expected that it can be safely applied to newborns, namely those identified in increasing numbers as asymptomatic CF patients by the recently implemented extensive newborn screening programs, merely based on elevated serum concentrations of immunoreactive trypsinogen (IRT). Indeed, these patients, posing new challenges to the CF diagnosis and prognosis are likely candidates to undergo this procedure to find evidence of CFTR (dys)function.

One of the major limitations usually related to patient surveys, is that we can only rely on the responses of the individuals who agree to undergo certain procedures. In this case, we are relying on the perspectives of the patients who agreed to participate on the sigmoidoscopy/rectal biopsy procedure, so there is a selection bias that leads to underscoring the fraction of individuals who do not tolerate this kind of procedure. Nevertheless, we can only rely on the responses of the individuals who agreed to participate, as those who have not undergone the procedure can only have preconceptions and not real experience to assess it. Importantly, patient enquiries demonstrate that for the majority of the individuals (76%) the rectal biopsy procedure is not associated with high levels of discomfort due to the short procedure time (max 15 min, Additional file 3 Video S1: Rectal biopsy procedure), regardless of sedation (Table 3) and no significant differences were found between the “control group” – with an already established CF diagnosis- and the “diagnostic group” (data not shown). Moreover, this shows to be also a relatively painless procedure, as 79% of the individuals did not report pain (Table 3). Nevertheless, “the sigmoidoscopy step” was associated with the highest level of discomfort (Table 4) as possibly expected, probably because of the preconceptions associated with this type of procedure. Accordingly, the individuals interviewed classified the rectal biopsy procedure as more unpleasant than sweat test, spirometry or blood collection (Additional file 1: Table S5). But if these individuals are required to repeat the biopsy procedure, despite some preconception concerns comprising prejudice and discomfort, the great majority (88%) would accept doing it for at least one more time, with 53.3% of the patients accepting to repeat it up to four more times (Table 3). Importantly, we also have to take into account, especially when a research procedure which is not routine is attempted, that adults above 20 years of age are more receptive to repeat the procedure than adolescents or children (as reported by their parents).

In conclusion, results from the present study recommend that in the case of forceps biopsy, rectal biopsies should be obtained with jumbo biopsy forceps after bowel preparation with NaCl isotonic solution to obtain viable specimens for bioelectric measurements for CF studies. The procedure is safe and is well tolerated from the patients’ perspective, demonstrating its feasibility as an outcome measure in clinical trials.