- Email: [email protected]
Suction Rectal Biopsy is Accurate in Late Preterm Infants with Suspected Hirschsprung Disease
YJPSU-59408; No of Pages 4 Journal of Pediatric Surgery xxx (xxxx) xxx
Contents lists available at ScienceDirect
Journal of Pediatric Surgery journal homepage: www.elsevier.com/locate/jpedsurg
Suction Rectal Biopsy Is Accurate in Late Preterm Infants with Suspected Hirschsprung Disease☆ Devin R. Halleran a,b,c, Hira Ahmad b,c, Haley Lehmkuhl c, Peter Baker d, Richard J. Wood c, Marc A. Levitt c, Jeremy G. Fisher a,b,⁎ a
Center for Surgical Outcomes Research, The Research Institution at Nationwide Children's Hospital, Columbus, OH Department of Pediatric Surgery, Nationwide Children's Hospital, The Ohio State University, Columbus, OH Center for Colorectal and Pelvic Reconstruction, Nationwide Children's Hospital, Columbus, OH d Department of Pathology, Nationwide Children's Hospital, Columbus, OH b c
a r t i c l e
i n f o
Article history: Received 6 September 2019 Accepted 29 September 2019 Available online xxxx Key words: Hirschsprung disease Suction rectal biopsy Premature Preterm Neonate
a b s t r a c t Background: Few data are available to substantiate the dogma commonly held by pediatric surgeons and pediatric pathologists that rectal biopsies may be inaccurate or risky in preterm neonates. We aimed to address these concerns. Methods: We performed a single-institution retrospective review of patients who underwent suction rectal biopsies at b 37 weeks corrected age from 2009 to 2018. The primary outcomes were accuracy of biopsy results and procedural complications. Results: Forty-nine patients underwent suction rectal biopsy at b 37 weeks corrected age. Mean gestational age at biopsy was 35.2 weeks (range 32.1–36.9) and mean weight was 2126 g (range 1590–3100). Five (10%) infants had biopsies positive for Hirschsprung Disease. All 5 later underwent pull-through operations and had pathologic confirmation of an aganglionic segment. The remaining 44 (90%) had biopsies showing ganglion cells. None were later found to have HD during the follow up period. Those who underwent biopsy at b2000 g (16/44) had 100% sensitivity [95% CI 48, 100] and specificity [95% CI 92, 100] (2 true positives, 14 true negatives). There were no complications identified. Conclusion: Suction rectal biopsy can be performed safely in preterm infants as small as 1590–2000 g with high accuracy. Clinicians should not hesitate to perform a biopsy for a premature infant when clinically appropriate. Type of study: Study of a diagnostic test. Level of evidence: Level IV. © 2019 Elsevier Inc. All rights reserved.
Delayed passage of meconium is common in premature infants. [1–3] The incidence appears to be inversely correlated with gestational age and as many as 80% of infants b 34 weeks estimated gestational age (EGA) fail to pass meconium within 48 h of birth. [4] Hirschsprung Disease (HD) is typically considered after persistent absence of a bowel movement in the setting of abdominal distention. Once thought to be uncommon in premature infants, recent studies confirm that the incidence is the same in preterm and term neonates at approximately 1 in 5000. [5] ☆ How this paper will improve care: This retrospective review found that suction rectal biopsy has a high degree of accuracy in late preterm infants with suspected Hirschsprung disease, including a subset 1590–2000 g. These findings suggest that biopsy can be considered in this group when clinically indicated. ⁎ Corresponding author at: Center for Surgical Outcomes Research, Department of Pediatric Surgery, Nationwide Children's Hospital, 611 E. Livingston Avenue, Faculty Office Building, 6th Floor, Columbus, Ohio, 43205. Tel.: +1 614 722 3911, fax: +1 614 355 6229. E-mail address: jeremy.fi[email protected] (J.G. Fisher).
Depending on the size and clinical stability of the infant, contrast enema (CE) is the usual next step in evaluation. Though helpful in making the diagnosis and guiding surgical planning, CE has only moderate accuracy (sensitivity 70%, specificity 83%). [6] Suction rectal biopsy (SRB), most often performed with a specialized suction device, is considered the gold standard for diagnosis of HD. However, suction rectal biopsy is often delayed until the child reaches term adjusted gestational age owing to a belief that suction rectal biopsy is unreliable in preterm infants. Neonates awaiting biopsy therefore often go untreated, causing a delay in enteral tolerance and prolonged parenteral nutrition. Alternatively, those with suspected HD may be treated with rectal irrigations and thus a number of infants without the disease may receive unnecessary instrumentation and invasive therapy. Despite the concerns for misleading histology and complications from SRB in preterm neonates, little data exist regarding its accuracy and safety of this technique in such patients. The objective of this study was to quantify the sensitivity,
https://doi.org/10.1016/j.jpedsurg.2019.09.055 0022-3468/© 2019 Elsevier Inc. All rights reserved.
Please cite this article as: D.R. Halleran, H. Ahmad, H. Lehmkuhl, et al., Suction Rectal Biopsy Is Accurate in Late Preterm Infants with Suspected Hirschsprung Disease, Journal of Pediatric Surgery, https://doi.org/10.1016/j.jpedsurg.2019.09.055
2
D.R. Halleran et al. / Journal of Pediatric Surgery xxx (xxxx) xxx
specificity, and complication rate of suction rectal biopsy in preterm infants.
Table 1 Patient demographics. Characteristic
1. Methods 1.1. Cohort identification We performed a single institution retrospective review of all patients who underwent suction rectal biopsy from January 1, 2009 to June 3, 2018. Patients with a gestational age less than 37 weeks at the time of biopsy were considered preterm and included in the study. Medical charts were reviewed to extract the following data: sex, gestational age at biopsy, weight at biopsy, biopsy results, procedural complications, occurrence of repeat biopsy, surgical procedures performed (as applicable), and clinical outcome. This study was approved by the institutional review board (IRB18-00677). 1.2. Suction rectal biopsy technique All SRBs were performed by the pediatric surgical service, typically by a fellow with a board certified pediatric surgeon immediately available if not present at the bedside in the neonatal intensive care unit. At our institution, the RBi2 Suction Rectal Biopsy System (Aus Systems, Australia) has been preferentially used; however, an older model is still available and was likely employed to perform some of the included biopsies. We used the standard technique described on the RBi2 website. [7] In brief, following a rectal irrigation, the device with the cartridge was inserted into the anus to approximately 2–2.5 cm from the anal verge. Suction was applied and the blade on the cartridge is deployed by pulling the device trigger. At least three biopsies were performed posteriorly and laterally (in the 4, 6, and 8 o'clock positions). Specimens were immediately hand delivered to pathology in saline. 1.3. Histologic examination The biopsies were immediately evaluated under a dissecting microscopic for the presence of submucosa. For some biopsies with adequate tissue, one piece was snap frozen. Following fixation in 10% buffered formalin, the biopsies were processed and embedded by routine procedures. Nine slides were prepared; eight slides with 15 sections on each slide were stained with hematoxylin and eosin and one slide with two sections was held unstained. In addition to evaluation for submucosal ganglion cells, the maximum submucosal nerve diameter was determined using a calibrated ocular micrometer. If indicated, the pattern of mucosal small nerve branches was assessed by either calretinin immunohistochemical stain or acetylcholinesterase enzyme histochemical stain. The latter was used in the earlier years of the study. All pathology slides were reviewed by a pathologist with subspecialty training in pediatric pathology. 1.4. Statistics Patient demographic and clinical characteristics were reported using medians for continuous variables and frequencies and percentages for categorical variables. Sensitivity, specificity, positive predictive value, and negative predictive value were calculated. All statistical analyses were performed using STATA 14.0 (StataCorp LP, College Station, Texas, USA). 2. Results Of 364 infants who underwent SRB over the study period, 49 (13%) were younger than 37 weeks gestational age at the time of biopsy. Table 1 describes the demographic information of patients in the
Male sex, n (%) Median age at biopsy (range) Median weight at biopsy (range) EGA at biopsy –33 weeks 33–34 weeks 34–35 weeks 35–36 weeks 36–37 weeks Weight at biopsy b1750 g 1750–1999 g 2000–2249 g 2250–2499 g 2500–2999 g N3000 g Biopsy complications Perforation Hemorrhage Stricture
28 (57) 35.4 weeks (32.1–36.9) 2059 g (1590–3100) n (%) 2 (4) 6 (12) 11 (22) 12 (24) 18 (37) 3 (6) 16 (33) 16 (33) 7 (14) 5 (10) 2 (4) 0 (0) 0 (0) 0 (0)
study cohort. Males comprised 57% of the cohort. The median age at the time of biopsy was 35.4 weeks (range 31.1 weeks–36.9 weeks), and the median weight at biopsy was 2059 g (range 1590–3100 g). There were no complications related to SRB recorded in the medical records for this cohort. Histologic examination revealed normal ganglion cells in 44/49 (90%) patients, and absent ganglion cells with hypertrophic nerves N 40 μm along with abnormal pattern of small mucosal nerve branches in five (10%) patients. All five patients with absent ganglion cells on biopsy underwent a pull through operation [Swenson (2), Soave (2), or Duhamel (1)]. The diagnosis of Hirschsprung disease was confirmed through demonstration of an aganglionic segment on final pathology in all five cases. Of the 44 patients with ganglion cells on initial biopsy, no patients were later found to have Hirschsprung disease. Alternative pathologic causes for delayed passage of meconium were identified in four patients and included cystic fibrosis (n = 1), jejunal atresia (1), allergic colitis (1), and small left colon syndrome (1). Symptoms resolved with appropriate treatment in each case. The remaining 40 patients in whom no pathologic explanation was identified demonstrated a spontaneous resolution of symptoms. The clinical outcomes of patients with a diagnosis of Hirschsprung on SRB are described in Table 2. The median adjusted gestational age of patients with absent ganglion cells on SRB was 36 weeks and 3 days (range 35 weeks 2 days–36 weeks 6 days) and median weight was 2560 (range 1935–2834) g. Three patients underwent single stage pull through at a median age of 36 weeks 3 days (range 36 weeks 2 days– 37 weeks 2 days). The remaining two patients underwent leveling sigmoid colostomy in the newborn period followed by definitive pull through within the first year of life. There were no surgical complications related to pull through or stoma creation in this cohort. In the cohort overall, SRB was 100% sensitive [95% CI 48, 100] and specific [95% CI 92, 100] for the diagnosis of HD (5 true positives, 44 true negatives), yielding positive and negative predictive values of 100%. The subset who underwent biopsy at less than 2000 g (16/44) had 100% sensitivity [95% CI 16, 100] and specificity [95% CI 80, 100] (2 true positives, 14 true negatives). 3. Discussion With 100% sensitivity and specificity of suction rectal biopsy (SRB) for Hirschsprung disease in this cohort of late preterm neonates, this study demonstrates that such infants can safely undergo biopsies that yield accurate results. Further, these findings held for the subset of small neonates who weighed less than 2000 g at the time of biopsy
Please cite this article as: D.R. Halleran, H. Ahmad, H. Lehmkuhl, et al., Suction Rectal Biopsy Is Accurate in Late Preterm Infants with Suspected Hirschsprung Disease, Journal of Pediatric Surgery, https://doi.org/10.1016/j.jpedsurg.2019.09.055
D.R. Halleran et al. / Journal of Pediatric Surgery xxx (xxxx) xxx
3
Table 2 Clinical details of patients with SRB consistent with Hirschsprung disease. Pt #
Age at birth (weeks)
Age at biopsy (weeks)
Weight at biopsy (g)
Nerve size (μm)
LC or PPT
Age at LC (weeks)
Weight at LC (g)
Age at pull-through
Weight at pull-(g)
1 2 3 4 5
34.1 31.1 35.0 33.0 36.0
35.3 35.7 36.4 36.7 36.9
2560 1935 2600 1995 2834
50 60 75 50 98
PPT PPT LC LC PPT
NA NA 36.6 39.0 NA
NA NA 2615 g 2170 g NA
36.3 weeks 36.4 weeks 6 months 8 months 37.3 weeks
2665 3680 7150 6900 2854
including as low as 1590 g. This is a good example of the adage that it is not the unanswered questions that matter most but sometimes the unquestioned answers. The study presented here questions the dogma of avoiding biopsy in the premature infant. SRB has proven to be a valuable diagnostic technique in the work up of HD since its description in 1965 and offers clinicians an accurate and minimally invasive technique that can be performed at the bedside and without the need for general anesthesia. [8] The technique involves sampling mucosa and submucosa to detect the presence or absence of ganglion cells and hypertrophic nerve trunks in Meissner's plexus. [9] A recent systematic review demonstrated an average sensitivity and specificity of SRB of 96.8% and 99.4%, respectively, when compared against full thickness biopsy. [10] Very few studies detailing the accuracy of SRB in preterm infants exist at present. Collectively, these studies noted a high degree of sensitivity and specificity of SRB for HD, but many of the patients underwent their biopsies close or after 40 week gestational age. In these prior studies the sensitivity ranged from 46%–100% and the specificity was between 97% and 100%. [11–13] Our study differs from those previously published in that we focus exclusively on a preterm cohort of patients less than 37 gestational weeks at the time of biopsy. Taken together, the available data suggest that the accuracy of SRB in preterm infants is comparable to that of older children. In the case of the premature infant, clinicians have traditionally waited until an infant approaches a term corrected age or weight to perform a suction biopsy. [14,15] This practice has been driven in large part by early studies of the developing enteric nervous system which demonstrated the process of maturation of ganglion cells in the bowel of neonates that continues into early childhood. [16–18] Of particular relevance to the practice of suction biopsy, the development of Meissner's plexus appears delayed relative to Auerbach's plexus. [16] The implication of this developmental understanding therefore is that immature ganglion cells might be misinterpreted on histologic examination, and thus lead the clinician to an incorrect diagnosis in a neonate. Although it is not fully known how the presence of immature ganglion cells affects the accuracy of suction biopsies in a preterm cohort, this study demonstrates that the diagnosis of HD can be made reliably. Complications after suction rectal biopsy are rare and include bleeding, perforation, and pelvic sepsis. [10,19,20] These occur in approximately 1 in 150 procedures, although they appear to happen more commonly in younger children, likely as a result of the thinner rectal wall as demonstrated by the increased presence of circular muscle layer in biopsy specimens compared to older children. [10] As the biopsy specimen represents a greater portion of the circumference of the neonatal bowel, a theoretical risk of rectal stricture exists, although this was not seen in our cohort of patients. Furthermore, no complications were reported in the prior studies of preterm infants, which when taken in aggregate with our results suggest that the safety profile of this practice in late preterm infants is similar to that of older children. [11–13] The ideal lower weight limit of feasibility and safety has yet to be clearly identified, though this series suggest that number is likely less than 2000 g. From a clinical standpoint, earlier diagnosis or exclusion of HD would lead to more appropriate delivery or cessation of invasive therapy. While an early confirmation of HD does not necessarily change a patient's care when there was a strong suspicion for the diagnosis, an
early biopsy may significantly change management in neonates with either unexpected negative or positive biopsy results. Patients with a strong suspicion of HD, but for whom a biopsy is ultimately negative, may avoid repeated rectal irrigations and a period of prolonged suboptimal feeding as the search for the correct diagnosis is delayed. Conversely, infants with a low pretest probability of HD who turn out to have it can begin treatment, such as irrigations, earlier, avoiding potential delays in feeding. Overall, early diagnosis (or exclusion) of HD is likely to result in an overall reduction in the use of healthcare resources and may reduce the overall morbidity experienced by this population. However, further study is needed to elucidate how earlier diagnosis in a preterm cohort impacts these outcomes. A preterm diagnosis of HD was verified at the time of surgical intervention in all five patients with absent ganglion cells in our cohort. An early diagnosis allowed for appropriate therapies to be initiated and allowed for shared decision making between the involved clinical teams and family. Three were managed with definitive, single-stage pull through in the early newborn period, while the remaining two patients were managed with a two-stage procedure. [Table 2] There were no surgical complications in these patients, and early surgical intervention obviated the need for an unnecessarily lengthy period of irrigations or parenteral nutrition. Common pitfalls in the analysis of specimens for Hirschsprung disease in patients of all ages include inadequate sampling of the submucosal plexus above the physiologic aganglionic segment of the distal rectum. This requires an understanding of the normal histology of the distal rectum and anal canal where the wall is expectedly aganglionic but, importantly, lacks the hypertrophic nerves characteristic of HD. In the present study and those previously discussed, no inadequate samples were reported. [11–13] It should be emphasized that accuracy noted in this study reflects use of biopsy in a highly select group of patients with high pretest probability for HD based on clinical and radiographic findings. We would also note that the study population consisted of late preterm infants (all N 32 weeks), and thus these findings may not apply to smaller neonates. The diagnostic strategy to distinguish dysmotility of prematurity was not investigated here. In particular, it is worth noting that the safety of biopsy in the very low birth weight cohort (b1500 g) was not studied. There are a number of limitations to this small, retrospective study that limit its use in a broader population. While the results are encouraging that SRB can be performed early, the complications of concern occur very infrequently and thus a study of this size is underpowered to identify their true incidence. Further, the study did not control for important clinical characteristics, such as physiologic state, that may influence timing of biopsy. These data reflect the practice of a single institution with high volume and highly subspecialized pathologists, which may also reduce their generalizability. Further work in the preterm cohort, including prospective study based on clear protocols for biopsy timing, is needed to establish guidelines for the optimal timing HD workup in this group. 4. Conclusion These results suggest that suction biopsy can be performed with minimal risks in late preterm infants as small as 1590–2000 g with a high degree of accuracy. Clinicians should have a low threshold to biopsy a premature infant when clinically warranted.
Please cite this article as: D.R. Halleran, H. Ahmad, H. Lehmkuhl, et al., Suction Rectal Biopsy Is Accurate in Late Preterm Infants with Suspected Hirschsprung Disease, Journal of Pediatric Surgery, https://doi.org/10.1016/j.jpedsurg.2019.09.055
4
D.R. Halleran et al. / Journal of Pediatric Surgery xxx (xxxx) xxx
Acknowledgements
patients who are pre-term so you can shift your focus to other diagnoses.
The authors have no relevant financial conflicts of interest to disclose. References Appendix A. Discussion: Suction Rectal Biopsy Is Accurate in Late Preterm Infants with Suspected Hirschsprung Disease Presenter: Devin Halleran. Q. Andrea Bishop, Children's Hospital, Colorado I was just wondering how do you put in the equation the experience of the pathologist? Because yes, it can be done. But I think I am mainly concerned about the patient that has received an operation, and then later we find that they never had Hirschsprung. And that's irreversible. Dr. Halleran That's correct, and thank you for your question, Dr. Bishop. You're absolutely right. In Columbus we have a very experienced staff of pathologists with a very high volume experience with Hirschsprung disease. We didn't specifically look at that in our study, but there is literature suggesting that a pediatric subspecialty in pathology adds to the diagnostic field of diagnosing Hirschsprung disease. So I agree that you would expect a higher accuracy with more pediatric experience, but that's not something that we looked at. Q. Karen Perone, Ann Arbor What gun did you use? Dr. Halleran Yes, we used the RB-1. Q. And so Did you use the monometer? Did you use a lower threshold for the lower weight? Dr. Halleran No, we used the standard technique that we use in the children. Approximately 1 cc above the dente line. Q. Right but did you use a monometer when you used the RB-1 to know the pressure you were pulling? Dr. Halleran Not typically. But again, that's not our standard practice. Q. … Italy Thank you very much for demonstrating the possibility to have a very early diagnosis. But my question is, what is the importance of the clinical point of view? You think that is not dangerous? You can achieve a diagnosis if you have enough pathological expertise. But do you think it's really useful on the clinical point of view for these children? Dr. Halleran Thank you for your question. I don't think that the real clinical value is in diagnosing Hirschsprung disease. I think that the real benefit is in excluding Hirschsprung disease in
[1] Weaver LT, Lucas A. Development of bowel habit in preterm infants. Arch Dis Child 1993;68(3 Spec No):317–20. [2] Verma A, Dhanireddy R. Time of first stool in extremely low birth weight (b or = 1000 grams) infants. J Pediatr 1993;122(4):626–9. [3] Bekkali N, Hamers SL, Schipperus MR, et al. Duration of meconium passage in preterm and term infants. Arch Dis Child Fetal Neonatal Ed 2008;93(5):F376–9. [4] Arnoldi R, Leva E, Macchini F, et al. Delayed meconium passage in very low birth weight infants. Eur J Pediatr Surg 2011;21(6):395–8. [5] Baxter KJ, Bhatia AM. Hirschsprung's disease in the preterm infant: implications for diagnosis and outcome. Am Surg 2013;79(7):734–8. [6] de Lorijn F, Kremer LC, Reitsma JB, et al. Diagnostic tests in Hirschsprung disease: a systematic review. J Pediatr Gastroenterol Nutr 2006;42(5):496–505. [7] Specialty Surgical Products Inc. Rbi2 suction rectal biopsy system data sheet. Retrieved March 9, 2019, from https://ssp-inccom/wordpress/wp-content/uploads/ 2015/07/rbi2-brochure-electronic-24Aug2015pdf. [8] Dobbins 3rd WO, Bill Jr AH. Diagnosis of Hirschsprung's disease excluded by rectal suction biopsy. N Engl J Med 1965;272:990–3. [9] Andrassy RJ, Isaacs H, Weitzman JJ. Rectal suction biopsy for the diagnosis of Hirschsprung's disease. Ann Surg 1981;193(4):419–24. [10] Friedmacher F, Puri P. Rectal suction biopsy for the diagnosis of Hirschsprung's disease: a systematic review of diagnostic accuracy and complications. Pediatr Surg Int 2015;31(9):821–30. [11] Allen AR, Putnam AR, Presson AP, et al. Accuracy of suction rectal biopsy for diagnosis of Hirschsprung's disease in neonates. Eur J Pediatr Surg 2019;29(5):425–30. https://doi.org/10.1055/s-0038-1667040. [12] Keyzer-Dekker CM, Sloots CE, Schokker-van Linschoten IK, et al. Effectiveness of rectal suction biopsy in diagnosing Hirschsprung disease. Eur J Pediatr Surg 2016;26 (1):100–5. [13] Meinds RJ, Kuiper GA, Parry K, et al. Infant's age influences the accuracy of rectal suction biopsies for diagnosing of Hirschsprung's disease. Clin Gastroenterol Hepatol 2015;13(10):1801–7. [14] Niramis R, Tongsin A, Lertsatit A, et al. How to manage low gut obstruction in neonates with immature ganglion cells in the colonic wall? J Med Assoc Thai 2014;97 (Suppl. 6):S66–73. [15] Sharp NE, Pettiford-Cunningham J, Shah SR, et al. The prevalence of Hirschsprung disease in premature infants after suction rectal biopsy. J Surg Res 2013;184(1): 374–7. [16] Smith B. Pre- and postnatal development of the ganglion cells of the rectum and its surgical implications. J Pediatr Surg 1968;3(3):386–91. [17] Burki T, Kiho L, Scheimberg I, et al. Neonatal functional intestinal obstruction and the presence of severely immature ganglion cells on rectal biopsy: 6 year experience. Pediatr Surg Int 2011;27(5):487–90. [18] AG B. JL E. Functional obstruction of the intestine due to neurological immaturity. Prog Pediatr Surg 1971;3:37–52. [19] Rees BI, Azmy A, Nigam M, et al. Complications of rectal suction biopsy. J Pediatr Surg 1983;18(3):273–5. [20] Alizai NK, Batcup G, Dixon MF, et al. Rectal biopsy for Hirschsprung's disease: what is the optimum method? Pediatr Surg Int 1998;13(2–3):121–4.
Please cite this article as: D.R. Halleran, H. Ahmad, H. Lehmkuhl, et al., Suction Rectal Biopsy Is Accurate in Late Preterm Infants with Suspected Hirschsprung Disease, Journal of Pediatric Surgery, https://doi.org/10.1016/j.jpedsurg.2019.09.055
Contents lists available at ScienceDirect
Journal of Pediatric Surgery journal homepage: www.elsevier.com/locate/jpedsurg
Suction Rectal Biopsy Is Accurate in Late Preterm Infants with Suspected Hirschsprung Disease☆ Devin R. Halleran a,b,c, Hira Ahmad b,c, Haley Lehmkuhl c, Peter Baker d, Richard J. Wood c, Marc A. Levitt c, Jeremy G. Fisher a,b,⁎ a
Center for Surgical Outcomes Research, The Research Institution at Nationwide Children's Hospital, Columbus, OH Department of Pediatric Surgery, Nationwide Children's Hospital, The Ohio State University, Columbus, OH Center for Colorectal and Pelvic Reconstruction, Nationwide Children's Hospital, Columbus, OH d Department of Pathology, Nationwide Children's Hospital, Columbus, OH b c
a r t i c l e
i n f o
Article history: Received 6 September 2019 Accepted 29 September 2019 Available online xxxx Key words: Hirschsprung disease Suction rectal biopsy Premature Preterm Neonate
a b s t r a c t Background: Few data are available to substantiate the dogma commonly held by pediatric surgeons and pediatric pathologists that rectal biopsies may be inaccurate or risky in preterm neonates. We aimed to address these concerns. Methods: We performed a single-institution retrospective review of patients who underwent suction rectal biopsies at b 37 weeks corrected age from 2009 to 2018. The primary outcomes were accuracy of biopsy results and procedural complications. Results: Forty-nine patients underwent suction rectal biopsy at b 37 weeks corrected age. Mean gestational age at biopsy was 35.2 weeks (range 32.1–36.9) and mean weight was 2126 g (range 1590–3100). Five (10%) infants had biopsies positive for Hirschsprung Disease. All 5 later underwent pull-through operations and had pathologic confirmation of an aganglionic segment. The remaining 44 (90%) had biopsies showing ganglion cells. None were later found to have HD during the follow up period. Those who underwent biopsy at b2000 g (16/44) had 100% sensitivity [95% CI 48, 100] and specificity [95% CI 92, 100] (2 true positives, 14 true negatives). There were no complications identified. Conclusion: Suction rectal biopsy can be performed safely in preterm infants as small as 1590–2000 g with high accuracy. Clinicians should not hesitate to perform a biopsy for a premature infant when clinically appropriate. Type of study: Study of a diagnostic test. Level of evidence: Level IV. © 2019 Elsevier Inc. All rights reserved.
Delayed passage of meconium is common in premature infants. [1–3] The incidence appears to be inversely correlated with gestational age and as many as 80% of infants b 34 weeks estimated gestational age (EGA) fail to pass meconium within 48 h of birth. [4] Hirschsprung Disease (HD) is typically considered after persistent absence of a bowel movement in the setting of abdominal distention. Once thought to be uncommon in premature infants, recent studies confirm that the incidence is the same in preterm and term neonates at approximately 1 in 5000. [5] ☆ How this paper will improve care: This retrospective review found that suction rectal biopsy has a high degree of accuracy in late preterm infants with suspected Hirschsprung disease, including a subset 1590–2000 g. These findings suggest that biopsy can be considered in this group when clinically indicated. ⁎ Corresponding author at: Center for Surgical Outcomes Research, Department of Pediatric Surgery, Nationwide Children's Hospital, 611 E. Livingston Avenue, Faculty Office Building, 6th Floor, Columbus, Ohio, 43205. Tel.: +1 614 722 3911, fax: +1 614 355 6229. E-mail address: jeremy.fi[email protected] (J.G. Fisher).
Depending on the size and clinical stability of the infant, contrast enema (CE) is the usual next step in evaluation. Though helpful in making the diagnosis and guiding surgical planning, CE has only moderate accuracy (sensitivity 70%, specificity 83%). [6] Suction rectal biopsy (SRB), most often performed with a specialized suction device, is considered the gold standard for diagnosis of HD. However, suction rectal biopsy is often delayed until the child reaches term adjusted gestational age owing to a belief that suction rectal biopsy is unreliable in preterm infants. Neonates awaiting biopsy therefore often go untreated, causing a delay in enteral tolerance and prolonged parenteral nutrition. Alternatively, those with suspected HD may be treated with rectal irrigations and thus a number of infants without the disease may receive unnecessary instrumentation and invasive therapy. Despite the concerns for misleading histology and complications from SRB in preterm neonates, little data exist regarding its accuracy and safety of this technique in such patients. The objective of this study was to quantify the sensitivity,
https://doi.org/10.1016/j.jpedsurg.2019.09.055 0022-3468/© 2019 Elsevier Inc. All rights reserved.
Please cite this article as: D.R. Halleran, H. Ahmad, H. Lehmkuhl, et al., Suction Rectal Biopsy Is Accurate in Late Preterm Infants with Suspected Hirschsprung Disease, Journal of Pediatric Surgery, https://doi.org/10.1016/j.jpedsurg.2019.09.055
2
D.R. Halleran et al. / Journal of Pediatric Surgery xxx (xxxx) xxx
specificity, and complication rate of suction rectal biopsy in preterm infants.
Table 1 Patient demographics. Characteristic
1. Methods 1.1. Cohort identification We performed a single institution retrospective review of all patients who underwent suction rectal biopsy from January 1, 2009 to June 3, 2018. Patients with a gestational age less than 37 weeks at the time of biopsy were considered preterm and included in the study. Medical charts were reviewed to extract the following data: sex, gestational age at biopsy, weight at biopsy, biopsy results, procedural complications, occurrence of repeat biopsy, surgical procedures performed (as applicable), and clinical outcome. This study was approved by the institutional review board (IRB18-00677). 1.2. Suction rectal biopsy technique All SRBs were performed by the pediatric surgical service, typically by a fellow with a board certified pediatric surgeon immediately available if not present at the bedside in the neonatal intensive care unit. At our institution, the RBi2 Suction Rectal Biopsy System (Aus Systems, Australia) has been preferentially used; however, an older model is still available and was likely employed to perform some of the included biopsies. We used the standard technique described on the RBi2 website. [7] In brief, following a rectal irrigation, the device with the cartridge was inserted into the anus to approximately 2–2.5 cm from the anal verge. Suction was applied and the blade on the cartridge is deployed by pulling the device trigger. At least three biopsies were performed posteriorly and laterally (in the 4, 6, and 8 o'clock positions). Specimens were immediately hand delivered to pathology in saline. 1.3. Histologic examination The biopsies were immediately evaluated under a dissecting microscopic for the presence of submucosa. For some biopsies with adequate tissue, one piece was snap frozen. Following fixation in 10% buffered formalin, the biopsies were processed and embedded by routine procedures. Nine slides were prepared; eight slides with 15 sections on each slide were stained with hematoxylin and eosin and one slide with two sections was held unstained. In addition to evaluation for submucosal ganglion cells, the maximum submucosal nerve diameter was determined using a calibrated ocular micrometer. If indicated, the pattern of mucosal small nerve branches was assessed by either calretinin immunohistochemical stain or acetylcholinesterase enzyme histochemical stain. The latter was used in the earlier years of the study. All pathology slides were reviewed by a pathologist with subspecialty training in pediatric pathology. 1.4. Statistics Patient demographic and clinical characteristics were reported using medians for continuous variables and frequencies and percentages for categorical variables. Sensitivity, specificity, positive predictive value, and negative predictive value were calculated. All statistical analyses were performed using STATA 14.0 (StataCorp LP, College Station, Texas, USA). 2. Results Of 364 infants who underwent SRB over the study period, 49 (13%) were younger than 37 weeks gestational age at the time of biopsy. Table 1 describes the demographic information of patients in the
Male sex, n (%) Median age at biopsy (range) Median weight at biopsy (range) EGA at biopsy –33 weeks 33–34 weeks 34–35 weeks 35–36 weeks 36–37 weeks Weight at biopsy b1750 g 1750–1999 g 2000–2249 g 2250–2499 g 2500–2999 g N3000 g Biopsy complications Perforation Hemorrhage Stricture
28 (57) 35.4 weeks (32.1–36.9) 2059 g (1590–3100) n (%) 2 (4) 6 (12) 11 (22) 12 (24) 18 (37) 3 (6) 16 (33) 16 (33) 7 (14) 5 (10) 2 (4) 0 (0) 0 (0) 0 (0)
study cohort. Males comprised 57% of the cohort. The median age at the time of biopsy was 35.4 weeks (range 31.1 weeks–36.9 weeks), and the median weight at biopsy was 2059 g (range 1590–3100 g). There were no complications related to SRB recorded in the medical records for this cohort. Histologic examination revealed normal ganglion cells in 44/49 (90%) patients, and absent ganglion cells with hypertrophic nerves N 40 μm along with abnormal pattern of small mucosal nerve branches in five (10%) patients. All five patients with absent ganglion cells on biopsy underwent a pull through operation [Swenson (2), Soave (2), or Duhamel (1)]. The diagnosis of Hirschsprung disease was confirmed through demonstration of an aganglionic segment on final pathology in all five cases. Of the 44 patients with ganglion cells on initial biopsy, no patients were later found to have Hirschsprung disease. Alternative pathologic causes for delayed passage of meconium were identified in four patients and included cystic fibrosis (n = 1), jejunal atresia (1), allergic colitis (1), and small left colon syndrome (1). Symptoms resolved with appropriate treatment in each case. The remaining 40 patients in whom no pathologic explanation was identified demonstrated a spontaneous resolution of symptoms. The clinical outcomes of patients with a diagnosis of Hirschsprung on SRB are described in Table 2. The median adjusted gestational age of patients with absent ganglion cells on SRB was 36 weeks and 3 days (range 35 weeks 2 days–36 weeks 6 days) and median weight was 2560 (range 1935–2834) g. Three patients underwent single stage pull through at a median age of 36 weeks 3 days (range 36 weeks 2 days– 37 weeks 2 days). The remaining two patients underwent leveling sigmoid colostomy in the newborn period followed by definitive pull through within the first year of life. There were no surgical complications related to pull through or stoma creation in this cohort. In the cohort overall, SRB was 100% sensitive [95% CI 48, 100] and specific [95% CI 92, 100] for the diagnosis of HD (5 true positives, 44 true negatives), yielding positive and negative predictive values of 100%. The subset who underwent biopsy at less than 2000 g (16/44) had 100% sensitivity [95% CI 16, 100] and specificity [95% CI 80, 100] (2 true positives, 14 true negatives). 3. Discussion With 100% sensitivity and specificity of suction rectal biopsy (SRB) for Hirschsprung disease in this cohort of late preterm neonates, this study demonstrates that such infants can safely undergo biopsies that yield accurate results. Further, these findings held for the subset of small neonates who weighed less than 2000 g at the time of biopsy
Please cite this article as: D.R. Halleran, H. Ahmad, H. Lehmkuhl, et al., Suction Rectal Biopsy Is Accurate in Late Preterm Infants with Suspected Hirschsprung Disease, Journal of Pediatric Surgery, https://doi.org/10.1016/j.jpedsurg.2019.09.055
D.R. Halleran et al. / Journal of Pediatric Surgery xxx (xxxx) xxx
3
Table 2 Clinical details of patients with SRB consistent with Hirschsprung disease. Pt #
Age at birth (weeks)
Age at biopsy (weeks)
Weight at biopsy (g)
Nerve size (μm)
LC or PPT
Age at LC (weeks)
Weight at LC (g)
Age at pull-through
Weight at pull-(g)
1 2 3 4 5
34.1 31.1 35.0 33.0 36.0
35.3 35.7 36.4 36.7 36.9
2560 1935 2600 1995 2834
50 60 75 50 98
PPT PPT LC LC PPT
NA NA 36.6 39.0 NA
NA NA 2615 g 2170 g NA
36.3 weeks 36.4 weeks 6 months 8 months 37.3 weeks
2665 3680 7150 6900 2854
including as low as 1590 g. This is a good example of the adage that it is not the unanswered questions that matter most but sometimes the unquestioned answers. The study presented here questions the dogma of avoiding biopsy in the premature infant. SRB has proven to be a valuable diagnostic technique in the work up of HD since its description in 1965 and offers clinicians an accurate and minimally invasive technique that can be performed at the bedside and without the need for general anesthesia. [8] The technique involves sampling mucosa and submucosa to detect the presence or absence of ganglion cells and hypertrophic nerve trunks in Meissner's plexus. [9] A recent systematic review demonstrated an average sensitivity and specificity of SRB of 96.8% and 99.4%, respectively, when compared against full thickness biopsy. [10] Very few studies detailing the accuracy of SRB in preterm infants exist at present. Collectively, these studies noted a high degree of sensitivity and specificity of SRB for HD, but many of the patients underwent their biopsies close or after 40 week gestational age. In these prior studies the sensitivity ranged from 46%–100% and the specificity was between 97% and 100%. [11–13] Our study differs from those previously published in that we focus exclusively on a preterm cohort of patients less than 37 gestational weeks at the time of biopsy. Taken together, the available data suggest that the accuracy of SRB in preterm infants is comparable to that of older children. In the case of the premature infant, clinicians have traditionally waited until an infant approaches a term corrected age or weight to perform a suction biopsy. [14,15] This practice has been driven in large part by early studies of the developing enteric nervous system which demonstrated the process of maturation of ganglion cells in the bowel of neonates that continues into early childhood. [16–18] Of particular relevance to the practice of suction biopsy, the development of Meissner's plexus appears delayed relative to Auerbach's plexus. [16] The implication of this developmental understanding therefore is that immature ganglion cells might be misinterpreted on histologic examination, and thus lead the clinician to an incorrect diagnosis in a neonate. Although it is not fully known how the presence of immature ganglion cells affects the accuracy of suction biopsies in a preterm cohort, this study demonstrates that the diagnosis of HD can be made reliably. Complications after suction rectal biopsy are rare and include bleeding, perforation, and pelvic sepsis. [10,19,20] These occur in approximately 1 in 150 procedures, although they appear to happen more commonly in younger children, likely as a result of the thinner rectal wall as demonstrated by the increased presence of circular muscle layer in biopsy specimens compared to older children. [10] As the biopsy specimen represents a greater portion of the circumference of the neonatal bowel, a theoretical risk of rectal stricture exists, although this was not seen in our cohort of patients. Furthermore, no complications were reported in the prior studies of preterm infants, which when taken in aggregate with our results suggest that the safety profile of this practice in late preterm infants is similar to that of older children. [11–13] The ideal lower weight limit of feasibility and safety has yet to be clearly identified, though this series suggest that number is likely less than 2000 g. From a clinical standpoint, earlier diagnosis or exclusion of HD would lead to more appropriate delivery or cessation of invasive therapy. While an early confirmation of HD does not necessarily change a patient's care when there was a strong suspicion for the diagnosis, an
early biopsy may significantly change management in neonates with either unexpected negative or positive biopsy results. Patients with a strong suspicion of HD, but for whom a biopsy is ultimately negative, may avoid repeated rectal irrigations and a period of prolonged suboptimal feeding as the search for the correct diagnosis is delayed. Conversely, infants with a low pretest probability of HD who turn out to have it can begin treatment, such as irrigations, earlier, avoiding potential delays in feeding. Overall, early diagnosis (or exclusion) of HD is likely to result in an overall reduction in the use of healthcare resources and may reduce the overall morbidity experienced by this population. However, further study is needed to elucidate how earlier diagnosis in a preterm cohort impacts these outcomes. A preterm diagnosis of HD was verified at the time of surgical intervention in all five patients with absent ganglion cells in our cohort. An early diagnosis allowed for appropriate therapies to be initiated and allowed for shared decision making between the involved clinical teams and family. Three were managed with definitive, single-stage pull through in the early newborn period, while the remaining two patients were managed with a two-stage procedure. [Table 2] There were no surgical complications in these patients, and early surgical intervention obviated the need for an unnecessarily lengthy period of irrigations or parenteral nutrition. Common pitfalls in the analysis of specimens for Hirschsprung disease in patients of all ages include inadequate sampling of the submucosal plexus above the physiologic aganglionic segment of the distal rectum. This requires an understanding of the normal histology of the distal rectum and anal canal where the wall is expectedly aganglionic but, importantly, lacks the hypertrophic nerves characteristic of HD. In the present study and those previously discussed, no inadequate samples were reported. [11–13] It should be emphasized that accuracy noted in this study reflects use of biopsy in a highly select group of patients with high pretest probability for HD based on clinical and radiographic findings. We would also note that the study population consisted of late preterm infants (all N 32 weeks), and thus these findings may not apply to smaller neonates. The diagnostic strategy to distinguish dysmotility of prematurity was not investigated here. In particular, it is worth noting that the safety of biopsy in the very low birth weight cohort (b1500 g) was not studied. There are a number of limitations to this small, retrospective study that limit its use in a broader population. While the results are encouraging that SRB can be performed early, the complications of concern occur very infrequently and thus a study of this size is underpowered to identify their true incidence. Further, the study did not control for important clinical characteristics, such as physiologic state, that may influence timing of biopsy. These data reflect the practice of a single institution with high volume and highly subspecialized pathologists, which may also reduce their generalizability. Further work in the preterm cohort, including prospective study based on clear protocols for biopsy timing, is needed to establish guidelines for the optimal timing HD workup in this group. 4. Conclusion These results suggest that suction biopsy can be performed with minimal risks in late preterm infants as small as 1590–2000 g with a high degree of accuracy. Clinicians should have a low threshold to biopsy a premature infant when clinically warranted.
Please cite this article as: D.R. Halleran, H. Ahmad, H. Lehmkuhl, et al., Suction Rectal Biopsy Is Accurate in Late Preterm Infants with Suspected Hirschsprung Disease, Journal of Pediatric Surgery, https://doi.org/10.1016/j.jpedsurg.2019.09.055
4
D.R. Halleran et al. / Journal of Pediatric Surgery xxx (xxxx) xxx
Acknowledgements
patients who are pre-term so you can shift your focus to other diagnoses.
The authors have no relevant financial conflicts of interest to disclose. References Appendix A. Discussion: Suction Rectal Biopsy Is Accurate in Late Preterm Infants with Suspected Hirschsprung Disease Presenter: Devin Halleran. Q. Andrea Bishop, Children's Hospital, Colorado I was just wondering how do you put in the equation the experience of the pathologist? Because yes, it can be done. But I think I am mainly concerned about the patient that has received an operation, and then later we find that they never had Hirschsprung. And that's irreversible. Dr. Halleran That's correct, and thank you for your question, Dr. Bishop. You're absolutely right. In Columbus we have a very experienced staff of pathologists with a very high volume experience with Hirschsprung disease. We didn't specifically look at that in our study, but there is literature suggesting that a pediatric subspecialty in pathology adds to the diagnostic field of diagnosing Hirschsprung disease. So I agree that you would expect a higher accuracy with more pediatric experience, but that's not something that we looked at. Q. Karen Perone, Ann Arbor What gun did you use? Dr. Halleran Yes, we used the RB-1. Q. And so Did you use the monometer? Did you use a lower threshold for the lower weight? Dr. Halleran No, we used the standard technique that we use in the children. Approximately 1 cc above the dente line. Q. Right but did you use a monometer when you used the RB-1 to know the pressure you were pulling? Dr. Halleran Not typically. But again, that's not our standard practice. Q. … Italy Thank you very much for demonstrating the possibility to have a very early diagnosis. But my question is, what is the importance of the clinical point of view? You think that is not dangerous? You can achieve a diagnosis if you have enough pathological expertise. But do you think it's really useful on the clinical point of view for these children? Dr. Halleran Thank you for your question. I don't think that the real clinical value is in diagnosing Hirschsprung disease. I think that the real benefit is in excluding Hirschsprung disease in
[1] Weaver LT, Lucas A. Development of bowel habit in preterm infants. Arch Dis Child 1993;68(3 Spec No):317–20. [2] Verma A, Dhanireddy R. Time of first stool in extremely low birth weight (b or = 1000 grams) infants. J Pediatr 1993;122(4):626–9. [3] Bekkali N, Hamers SL, Schipperus MR, et al. Duration of meconium passage in preterm and term infants. Arch Dis Child Fetal Neonatal Ed 2008;93(5):F376–9. [4] Arnoldi R, Leva E, Macchini F, et al. Delayed meconium passage in very low birth weight infants. Eur J Pediatr Surg 2011;21(6):395–8. [5] Baxter KJ, Bhatia AM. Hirschsprung's disease in the preterm infant: implications for diagnosis and outcome. Am Surg 2013;79(7):734–8. [6] de Lorijn F, Kremer LC, Reitsma JB, et al. Diagnostic tests in Hirschsprung disease: a systematic review. J Pediatr Gastroenterol Nutr 2006;42(5):496–505. [7] Specialty Surgical Products Inc. Rbi2 suction rectal biopsy system data sheet. Retrieved March 9, 2019, from https://ssp-inccom/wordpress/wp-content/uploads/ 2015/07/rbi2-brochure-electronic-24Aug2015pdf. [8] Dobbins 3rd WO, Bill Jr AH. Diagnosis of Hirschsprung's disease excluded by rectal suction biopsy. N Engl J Med 1965;272:990–3. [9] Andrassy RJ, Isaacs H, Weitzman JJ. Rectal suction biopsy for the diagnosis of Hirschsprung's disease. Ann Surg 1981;193(4):419–24. [10] Friedmacher F, Puri P. Rectal suction biopsy for the diagnosis of Hirschsprung's disease: a systematic review of diagnostic accuracy and complications. Pediatr Surg Int 2015;31(9):821–30. [11] Allen AR, Putnam AR, Presson AP, et al. Accuracy of suction rectal biopsy for diagnosis of Hirschsprung's disease in neonates. Eur J Pediatr Surg 2019;29(5):425–30. https://doi.org/10.1055/s-0038-1667040. [12] Keyzer-Dekker CM, Sloots CE, Schokker-van Linschoten IK, et al. Effectiveness of rectal suction biopsy in diagnosing Hirschsprung disease. Eur J Pediatr Surg 2016;26 (1):100–5. [13] Meinds RJ, Kuiper GA, Parry K, et al. Infant's age influences the accuracy of rectal suction biopsies for diagnosing of Hirschsprung's disease. Clin Gastroenterol Hepatol 2015;13(10):1801–7. [14] Niramis R, Tongsin A, Lertsatit A, et al. How to manage low gut obstruction in neonates with immature ganglion cells in the colonic wall? J Med Assoc Thai 2014;97 (Suppl. 6):S66–73. [15] Sharp NE, Pettiford-Cunningham J, Shah SR, et al. The prevalence of Hirschsprung disease in premature infants after suction rectal biopsy. J Surg Res 2013;184(1): 374–7. [16] Smith B. Pre- and postnatal development of the ganglion cells of the rectum and its surgical implications. J Pediatr Surg 1968;3(3):386–91. [17] Burki T, Kiho L, Scheimberg I, et al. Neonatal functional intestinal obstruction and the presence of severely immature ganglion cells on rectal biopsy: 6 year experience. Pediatr Surg Int 2011;27(5):487–90. [18] AG B. JL E. Functional obstruction of the intestine due to neurological immaturity. Prog Pediatr Surg 1971;3:37–52. [19] Rees BI, Azmy A, Nigam M, et al. Complications of rectal suction biopsy. J Pediatr Surg 1983;18(3):273–5. [20] Alizai NK, Batcup G, Dixon MF, et al. Rectal biopsy for Hirschsprung's disease: what is the optimum method? Pediatr Surg Int 1998;13(2–3):121–4.
Please cite this article as: D.R. Halleran, H. Ahmad, H. Lehmkuhl, et al., Suction Rectal Biopsy Is Accurate in Late Preterm Infants with Suspected Hirschsprung Disease, Journal of Pediatric Surgery, https://doi.org/10.1016/j.jpedsurg.2019.09.055