Skip to main content

Staplers in Urologic Surgery

Surgical stapling devices — principally the GIA (gastrointestinal anastomosis) and TA (transverse / thoracoabdominal anastomosis) staplers — have broad applications in urology spanning vascular pedicle control, bowel division and reanastomosis, reservoir / neobladder construction, and bladder / ureteral closure. Their use has evolved significantly with the introduction of absorbable staples (1992), endoscopic / laparoscopic platforms, and robotic stapling systems. This article is the reconstructive-urologist's reference for when to use which stapler, what cartridge to load, and where the evidence actually supports stapled vs. hand-sewn reconstruction.

See also: Bowel Anastomosis, Sutures, Urinary Diversion, Bladder Augmentation.

Part I: Stapler types and fundamental mechanics​

GIA (gastrointestinal anastomosis) stapler​

The GIA stapler simultaneously places two parallel double rows of staples and cuts between them, creating a sealed division of tissue in a single firing. Configurations:

  • Standard tissue GIA — staple heights 3.5–4.8 mm; used for bowel division, detubularization, and reservoir construction.
  • Vascular GIA (Endo-GIA Vascular) — staple height 2.0–2.5 mm (shorter legs for thinner vascular tissue); used for renal-hilum ligation and vascular-pedicle control.
  • Lengths — 30, 45, 60, and 80 mm cartridges.[1]
  • Absorbable GIA — uses polyglyconate (Lactomer) absorbable staples; became available in 1992 and was pivotal for urinary-tract reconstruction.[2][3]

TA (transverse anastomosis) stapler​

The TA stapler places parallel rows of staples without cutting — it staples tissue closed but does not divide it. The surgeon cuts separately with scissors or a knife. This two-step mechanism has important safety implications:

  • FDA MAUDE database review (2009–2019) — no deaths or reoperations with TA staplers, compared with 22 deaths (5.7%) overall with all stapler types. The separation of stapling and cutting allows verification of staple-line integrity before division.[4]

Key distinction — the GIA simultaneously staples and cuts in one firing; the TA staples only, requiring a separate cutting step. This makes the TA inherently safer for vascular control but less efficient for bowel work.

Endoscopic and robotic platforms​

Endoscopic stapling platforms (Endo-GIA, Endo-TA) and robotic stapling systems (Intuitive SureForm, Medtronic Signia-Shaft integration) have extended these techniques to minimally invasive and robot-assisted surgery. The underlying mechanics are the same, but the instrument geometry — longer shaft, angled articulating tip, console-controlled firing — changes handling and approach.

Part II: Vascular pedicle control​

Renal-hilum control during nephrectomy​

Endovascular staplers are the standard method for renal-hilar ligation during laparoscopic and robotic nephrectomy. Two approaches:

En-bloc stapling of the renal hilum (EBSH) — fires the vascular stapler across the entire hilum (artery and vein together) without individual vessel dissection.

  • Lai and Rais-Bahrami meta-analysis (2017), 595 patients — zero cases of arteriovenous fistula (AVF) at mean 26.5-month follow-up; significantly shorter operative time vs. individual vessel ligation; no difference in estimated blood loss or complications.[5]
  • Sherer et al. (2017), 428 patients / 433 renal units — no AVF at mean 51-month follow-up; mean EBL 155 mL.[6]
  • Kouba et al. (2007), 90 en-bloc patients — no AVF at 34-month follow-up.[7]

Individual vessel stapling — each renal artery and vein is dissected and stapled separately. Requires more hilar dissection but theoretically reduces AVF risk (though this risk appears negligible with modern titanium staplers even with en-bloc technique).[8]

Stapler complications during nephrectomy​

The FDA MAUDE database review (2009–2019) identified 383 stapler complications during minimally invasive nephrectomy over 10 years:[4]

Stapler typeShare of complicationsDeathsReoperationsConversion to open
Ethicon Endocutters63%PresentPresentHigher rate
Endo-GIA (Medtronic)28%PresentPresentIntermediate
Endo-TA (Medtronic)9%00Lowest rate

Chan et al. (2000) reported a 1.7% malfunction rate (10 / 565 laparoscopic nephrectomies) with the endovascular GIA stapler. The renal vein was involved in 80% of malfunctions. Open conversion was necessary in 20%. Critically, 70% of failures were attributed to preventable causes (improper loading, incomplete closure, tissue bunching).[9]

Staplers vs. clips for renal pedicle​

A meta-analysis comparing Hem-o-Lok clips vs. staplers for laparoscopic live-donor nephrectomy:[10]

  • No difference in device-failure rate, death rate, or severe hemorrhage.
  • Clips produced longer residual vascular length (advantageous for transplant).
  • Clips were ~$400 cheaper per case.
  • Staplers had shorter warm-ischemia time (mean difference 55.6 s shorter).

Vascular-pedicle control during radical cystectomy​

  • Chang et al. (2003) — prospective RCT, 70 patients. The Compact Flex Articulating Linear Cutter significantly decreased blood loss (p = 0.007), transfusion requirement (p = 0.006), and mean units transfused (p = 0.003) vs. traditional suture ligation, with no stapler-related complications.[11]
  • Hanash et al. (2000) — 16 patients, Endo-GIA-30-3.5 for pedicle control during radical cystectomy, significant reduction in blood loss and shorter operative time vs. conventional technique.[12]
  • Thompson et al. (2014) — prospective RCT, 80 patients, Endo-GIA Stapler vs. LigaSure: no difference in blood loss (708 vs. 687 mL, p = 0.850), operative time, or transfusion requirement. LigaSure was significantly less costly ($625 vs. $1,490, p<0.001).[13]
  • Gould and Borer (1996) — GIA vascular stapler for ligation of the dorsal-vein complex and lateral pedicles during radical retropubic prostatectomy; mean OR time 2 h, EBL 400 mL, 81% continence, 29% potency.[14]

Part III: Bowel division and reanastomosis​

During any urologic procedure requiring bowel harvest (ileal conduit, neobladder, augmentation cystoplasty), the GIA stapler is used for bowel division (isolating the ileal or colonic segment) and side-to-side functional end-to-end bowel reanastomosis (restoring intestinal continuity after segment harvest).

Stapled ileo-ileal anastomosis during RARC​

Saxena et al. (2025), robotic stapler technique for ileo-ileal anastomosis during RARC in 170 patients:[15]

  • Postoperative ileus rate: 7.0%.
  • Small-bowel obstruction: 4.7%.
  • Zero intraoperative bowel injuries.
  • Applicable to ileal conduit (75.8%), neobladder (21%), and continent pouch (2.3%).

GIA-60 vs. GIA-80 for bowel reanastomosis​

Ghanaat et al. (2018) compared GIA-60 vs. GIA-80 for bowel reanastomosis during radical cystectomy in 511 patients. Stapler size was not independently associated with postoperative ileus (OR 1.11, p = 0.6). Positive fluid balance was the strongest predictor of ileus (p = 0.019).[1]

Historical — autosuture ileal conduit​

Karamcheti et al. (1978) reported one of the earliest large series of 110 ileal conduits constructed with autosuture stapling devices, demonstrating significant reduction in operative time and postoperative morbidity vs. conventional suture technique.[16]

Part IV: Neobladder and reservoir construction​

The most complex and controversial application of staplers in urology. The key question is whether stapled reservoirs provide equivalent functional outcomes to hand-sewn reservoirs.

A. Absorbable-stapled reservoirs​

W-stapled ileal neobladder (Montie et al., 1994). Landmark technique using absorbable GIA and TA staples to construct a W-configured ileal neobladder:[2]

  • 15 patients after cystoprostatectomy.
  • Reservoir construction time: 12–21 minutes (dramatically faster than hand-sewn).
  • Urodynamic function and continence comparable to other ileal reservoirs.

However, the expanded comparative series (Montie et al., 1996) revealed a critical limitation: 6 of 19 evaluable patients (32%) had unsatisfactory reservoir characteristics — 3 required augmentation cystoplasty and 3 had higher-pressure, smaller-volume reservoirs. The authors concluded that the W-stapled configuration was inferior to hand-sewn ileal or ileocolic neobladders, possibly due to reservoir design, ischemia, or staple-material reaction.[17]

Pediatric continent reservoirs (Kirsch et al., 1996). Absorbable staples for right-colon pouch construction in 18 children:[3]

  • Decreased operative time by ~1 hour.
  • Saved ~$350 per case.
  • Only 1 significant complication (delayed pouch rupture, not staple-related).
  • No reservoir stones at up to 3 years.

All-stapled colonic reservoir (Parra, 1991). 27 patients with continent colonic reservoirs constructed entirely with stapling techniques:[18]

  • Mean reservoir construction time 95.4 minutes.
  • Continence achieved in all but 1 patient.
  • Urodynamic capacity mean 750 cc with low filling pressures (8.1 cm H₂O).
  • No ureteral reflux or obstruction at mean 22.8 months.

B. Titanium (nonabsorbable) stapled reservoirs​

With the rise of laparoscopic and robotic surgery, titanium staples have been reintroduced for intracorporeal neobladder construction because they are compatible with endoscopic stapling platforms:

  • Muto et al. (2016) — largest series: 606 stapled orthotopic neobladders over 20 years.[19] Median OR time 205 min; 60-month daytime continence 96%, nighttime 72%; urodynamic parameters stable 12–60 months; early complications 24%, late 23%.
  • Muto et al. (2005) — GIA-stapled Camey II neobladder with serosal-lined antireflux ureteroileal implantation in 43 patients; reduced OR time (~1 h 45 min for neobladder + ureteroileal anastomosis); low complication rate (6.9%) at median 38 months.[20]
  • Abreu et al. (2006) — laparoscopic U-shaped neobladder constructed entirely with titanium staples; technically feasible with the caveat that "continued surveillance is mandatory to determine the lithiasis-inducing potential."[21]

C. Stapled vs. hand-sewn intracorporeal neobladder — the most current evidence​

Mastroianni et al. (2025), the most recent comparative study, 116 patients (56 stapled vs. 60 hand-sewn intracorporeal orthotopic neobladder during RARC):[22]

OutcomeStapled (ST-iON)Hand-sewn (HS-iON)p-value
Perioperative complicationsNo differenceNo differenceNS
30-day complicationsNo differenceNo difference0.17
90-day complicationsNo differenceNo difference0.69
Hospital length of stayLongerShorter0.001
Stone formationHigherLower0.01
Daytime continence recoveryLowerHigher0.001
Nighttime continence recoveryLowerHigher0.001

The authors concluded that hand-sewing may represent a preferable option for intracorporeal neobladder configuration. Stapled neobladders did not reduce operative times or perioperative complications, while significantly increasing stone-formation risk and impairing continence recovery. The only independent predictor of continence recovery was a sex-sparing approach; excluding sex-sparing patients, the hand-sewn approach was the only predictor of nighttime continence recovery (HR 2.6, p = 0.009).[22]

Part V: The stone-formation problem​

The central concern with staples in the urinary tract is their potential to serve as a nidus for calculus formation. This risk differs fundamentally between absorbable and nonabsorbable staples.

Nonabsorbable (titanium / stainless steel) staples​

  • Dangman and Lebowitz (1991) reviewed 30 patients with bowel interposed in the urinary tract using metallic staples and found stones in 8 patients (27%). The stones formed on exposed staple portions and had a characteristic radiologic appearance — the staple was eccentric (not central) within the calculus.[23]
  • Edlich et al. (1987) stated definitively: "Nonabsorbable sutures or staples should not be used in the urinary tract because they predictably promote urolithiasis."[24]
  • All alloplastic materials in the urinary tract are susceptible to encrustation, affected by surface roughness, hydrophobicity, urinary pH, and infection with urease-producing bacteria.[25]

Absorbable staples​

Arif et al. (1999) — prospective RCT comparing absorbable vs. nonabsorbable staples for antireflux-valve construction in hemi-Kock pouches (50 patients):[26]

  • Pouch stones: 1/21 (4.8%) absorbable vs. 6/20 (30%) nonabsorbable (significant).
  • Valve extussusception: 1/21 vs. 3/20.
  • Absorbable staples significantly decreased stone formation and improved valve stability.

Titanium staples — intermediate risk​

Contemporary series using titanium staples report stone rates of approximately 6% at median 20-month follow-up.[27] Mastroianni 2025 confirmed significantly higher stone rates with stapled vs. hand-sewn neobladders (p = 0.01).[22] Shao et al. (2015) found neobladder stones in 2 / 30 endoscopic-stapler-sutured vs. 0 / 25 manually sutured patients.[28]

Exposed vs. buried staples​

The critical factor is whether staples are exposed to urine or buried beneath mucosa. Baughman et al. (2003) reported a case of a nearly complete intravesical titanium staple line visible at surveillance cystoscopy after laparoscopic nephroureterectomy — without encrustation at 6 months, but all staples were removed prophylactically.[29]

Julian and Ravitch (1986) evaluated both stainless-steel and absorbable staples in 104 bladder closures in dogs: 4 steel and 29 absorbable closures had exposed staples, but only 3 total developed crystal formation. All closures healed without difficulty.[30]

Part VI: Other urologic applications​

  • Bladder-cuff excision during nephroureterectomy — Kerbl et al. (1993) demonstrated laparoscopic GIA stapler closure of the bladder cuff in 8 pigs and 3 patients; no extravasation, stone, UTI, or abscess at 2–6 months.[31]
  • Laparoscopic pyeloplasty (reduction pelvioplasty) — Grubb et al. (2002) used titanium Endo-GIA for renal-pelvis reduction in 5 patients; shorter OR time (5.5 vs. 6.8 h), no extravasation vs. 2 in the sutured group, no stones at median 27 months.[32]
  • Stapled Boari flap (ureteral substitution) — Gallentine and Harmon (2001) created a stapled neoureter using absorbable GIA in 14 pigs; mean construction time only 15 minutes; mean length 13.4 cm; no residual staples or stones at cystoscopy / necropsy.[33]
  • Continent catheterizable channel construction — Bejany and Politano (1988) used the GIA stapler to create a tapered distal ileal segment for continent catheterizable stomas; full continence in all 10 patients, no reflux or obstruction in 95% of ureters.[34]

Part VII: Practical principles and pearls​

PrincipleDetail
Vascular vs. tissue cartridgesAlways use vascular cartridges (2.0–2.5 mm staple height) for renal hilum and vascular pedicles; tissue cartridges (3.5–4.8 mm) for bowel
Verify complete closure before firing70% of GIA malfunctions during nephrectomy were due to preventable human error — improper loading, incomplete jaw closure, tissue bunching
TA safer than GIA for vascular controlSeparate stapling and cutting steps allow verification before division; zero deaths / reoperations with TA in the MAUDE database
Absorbable staples preferred in the urinary tractSignificantly lower stone-formation rate vs. nonabsorbable; recommended whenever staples will contact urine
Hand-sewn neobladders may be superiorBetter continence recovery and lower stone rates than stapled neobladders in the most recent comparative data
Stapled bowel reanastomosis is standardSide-to-side functional end-to-end stapled anastomosis is the standard technique for restoring bowel continuity
Stapler size (GIA-60 vs. GIA-80) does not affect ileusNo independent association between stapler size and postoperative ileus
Cost considerationsEndo-GIA staplers are significantly more expensive than alternatives (LigaSure $625 vs. $1,490; ~$400 more than Hem-o-Lok clips)
Stone surveillance mandatory with titanium staplesIf nonabsorbable staples are used in the urinary tract, long-term imaging surveillance for stone formation is essential

Summary — stapler applications in urology at a glance​

ApplicationStaplerStaple materialKey evidenceCurrent status
Renal-hilum ligationEndo-GIA Vascular or Endo-TATitanium595 patients, 0% AVF; en-bloc safeStandard of care for laparoscopic / robotic nephrectomy
Radical cystectomy pediclesEndo-GIA VascularTitaniumRCT: decreased blood loss vs. suture; equivalent to LigaSure but more costlyWidely used; LigaSure increasingly preferred for cost
Bowel reanastomosisGIA (tissue)Titanium170 pts RARC: 7% ileus, 4.7% SBO, 0% bowel injuryStandard of care
Neobladder constructionGIA ± TAAbsorbable or titanium606 patients, 96% 60-mo daytime continence; BUT hand-sewn superior in recent comparisonControversial; hand-sewn may be preferable
Augmentation cystoplastyGIA (absorbable)Absorbable18 children: 1-h time savings, no stones at 3 yrViable, especially in pediatrics
Bladder cuff (nephroureterectomy)Endo-GIATitaniumNo early complications; exposed staples possibleAccepted; monitor for exposed staples
Pyeloplasty (reduction)Endo-GIATitaniumShorter OR time; no extravasation; no stones at 27 moLimited adoption; stone risk theoretical
Boari flapGIA (absorbable)Absorbable15-min construction; no stones; mucosa covers staple linePreclinical; promising

Cross-references​

References​

1. Ghanaat M, Winer AG, Sjoberg DD, et al. "Comparison of postradical cystectomy ileus rates using GIA-80 versus GIA-60 intestinal stapler device." Urology. 2018;122:121–126. doi:10.1016/j.urology.2018.09.010

2. Montie JE, Pontes JE, Parulkar BG, Selby T. "W-stapled ileal neo-bladder formed entirely with absorbable staples." J Urol. 1994;151(5):1188–1192. doi:10.1016/s0022-5347(17)35210-2

3. Kirsch AJ, Olsson CA, Hensle TW. "Pediatric continent reservoirs and colocystoplasty created with absorbable staples." J Urol. 1996;156(2 Pt 2):614–617. doi:10.1097/00005392-199608001-00010

4. Gopal N, Long B, Phillips J, Eshghi M. "Endovascular stapler complications during minimally invasive nephrectomy: an updated review of the FDA MAUDE database from 2009–2019." Urology. 2021;153:181–184. doi:10.1016/j.urology.2021.02.010

5. Lai WS, Rais-Bahrami S. "Safety and efficacy of en bloc renal hilar vascular staple ligation: a meta-analysis." J Urol. 2017;197(1):175–181. doi:10.1016/j.juro.2016.07.077

6. Sherer BA, Chow AK, Newsome MJ, et al. "En bloc stapling of the renal hilum during laparoscopic nephrectomy: a double-institutional analysis of safety and efficacy." Urology. 2017;105:69–75. doi:10.1016/j.urology.2017.01.051

7. Kouba E, Smith AM, Derksen JE, et al. "Efficacy and safety of en bloc ligation of renal hilum during laparoscopic nephrectomy." Urology. 2007;69(2):226–229. doi:10.1016/j.urology.2006.09.046

8. Schatloff O, Lindner U, Lindner A. "Current status of en bloc stapling of the renal hilum during laparoscopic nephrectomy." J Laparoendosc Adv Surg Tech A. 2010;20(7):631–633. doi:10.1089/lap.2009.0435

9. Chan D, Bishoff JT, Ratner L, Kavoussi LR, Jarrett TW. "Endovascular gastrointestinal stapler device malfunction during laparoscopic nephrectomy: early recognition and management." J Urol. 2000;164(2):319–321.

10. Liu Y, Huang Z, Chen Y, et al. "Staplers or clips? A systematic review and meta-analysis of vessel controlling devices for renal pedicle ligation in laparoscopic live donor nephrectomy." Medicine (Baltimore). 2018;97(45):e13116. doi:10.1097/MD.0000000000013116

11. Chang SS, Smith JA, Cookson MS. "Decreasing blood loss in patients treated with radical cystectomy: a prospective randomized trial using a new stapling device." J Urol. 2003;169(3):951–954. doi:10.1097/01.ju.0000051372.67213.ca

12. Hanash KA, Peracha AM, Al-Zahrani HM, et al. "Radical cystectomy: minimizing operative blood loss with a 'stapling technique.'" Urology. 2000;56(3):488–491. doi:10.1016/s0090-4295(00)00582-3

13. Thompson IM, Kappa SF, Morgan TM, et al. "Blood loss associated with radical cystectomy: a prospective, randomized study comparing Impact LigaSure vs. stapling device." Urol Oncol. 2014;32(1):45.e11–e15. doi:10.1016/j.urolonc.2013.06.006

14. Gould DL, Borer J. "Applied stapling technique in radical retropubic prostatectomy: efficient, effective and efficacious." J Urol. 1996;155(3):1008–1010.

15. Saxena S, Kim K, Billah MS, et al. "Outcomes of stapled ileo-ileal anastomosis during robot-assisted radical cystectomy with urinary diversion: points of technique." J Endourol. 2025. doi:10.1177/08927790251390881

16. Karamcheti A, O'Donnell WF, Hakala TR, Schwentker FN, Steichen FM. "Autosuture ileal conduit construction: experience in 110 cases." J Urol. 1978;120(5):545–548. doi:10.1016/s0022-5347(17)57268-7

17. Montie JE, Pontes JE, Powell IJ. "A comparison of the W-stapled ileal reservoir with hand-sewn reservoirs for orthotopic bladder replacement." Urology. 1996;47(4):476–481. doi:10.1016/S0090-4295(99)80480-4

18. Parra RO. "A simplified technique for continent urinary diversion: an all-stapled colonic reservoir." J Urol. 1991;146(6):1496–1499. doi:10.1016/s0022-5347(17)38149-1

19. Muto G, Collura D, Simone G, et al. "Stapled orthotopic ileal neobladder after radical cystectomy for bladder cancer: functional results and complications over a 20-year period." Eur J Surg Oncol. 2016;42(3):412–418. doi:10.1016/j.ejso.2015.11.010

20. Muto G, Bardari F, D'Urso L. "New serosal lined antireflux ureteroileal implantation technique on a GIA stapler detubularised ileal neobladder: technical considerations and results." Eur Urol. 2005;48(5):826–830. doi:10.1016/j.eururo.2005.03.011

21. Abreu SC, Araújo MB, Silveira RA, et al. "Laparoscopic-assisted radical cystectomy with U-shaped orthotopic ileal neobladder constructed using nonabsorbable titanium staples." Urology. 2006;68(1):193–197. doi:10.1016/j.urology.2006.02.011

22. Mastroianni R, Chiacchio G, Tuderti G, et al. "Stapled vs totally handsewn robotic intracorporeal neobladder: perioperative and functional outcomes." BJU Int. 2025. doi:10.1111/bju.16826

23. Dangman BC, Lebowitz RL. "Urinary tract calculi that form on surgical staples: a characteristic radiologic appearance." AJR Am J Roentgenol. 1991;157(1):115–117. doi:10.2214/ajr.157.1.2048508

24. Edlich RF, Rodeheaver GT, Thacker JG. "Considerations in the choice of sutures for wound closure of the genitourinary tract." J Urol. 1987;137(3):373–379. doi:10.1016/s0022-5347(17)44038-9

25. Choong SK, Whitfield HN. "Urinary encrustation of alloplastic materials." J Endourol. 2000;14(1):19–23. doi:10.1089/end.2000.14.19

26. Arif H, Madbouly K, Mahran MR, Ashamallah A, Ghoneim MA. "A prospective randomized study comparing absorbable and nonabsorbable staples in constructing antireflux valves of urethral hemi-Kock pouches." BJU Int. 1999;84(4):440–443. doi:10.1046/j.1464-410x.1999.00211.x

27. Abreu SC, Fonseca GN, Cerqueira JB, et al. "Laparoscopic radical cystectomy with intracorporeally constructed Y-shaped orthotopic ileal neobladder using nonabsorbable titanium staples exclusively." Urology. 2005;66(3):657. doi:10.1016/j.urology.2005.03.021

28. Shao P, Li P, Ju X, et al. "Laparoscopic radical cystectomy with intracorporeal orthotopic ileal neobladder: technique and clinical outcomes." Urology. 2015;85(2):368–373. doi:10.1016/j.urology.2014.09.059

29. Baughman SM, Sexton W, Bishoff JT. "Multiple intravesical linear staples identified during surveillance cystoscopy after laparoscopic nephroureterectomy." Urology. 2003;62(2):351. doi:10.1016/s0090-4295(03)00335-2

30. Julian TB, Ravitch MM. "Closure of the urinary bladder with stainless steel and absorbable staples." Ann Surg. 1986;204(2):186–192. doi:10.1097/00000658-198608000-00014

31. Kerbl K, Chandhoke P, McDougall E, et al. "Laparoscopic stapled bladder closure: laboratory and clinical experience." J Urol. 1993;149(6):1437–1440. doi:10.1016/s0022-5347(17)36408-x

32. Grubb RL, Sundaram CP, Yan Y, et al. "Use of titanium staples during upper tract laparoscopic reconstructive surgery: initial experience." J Urol. 2002;168(4 Pt 1):1366–1369. doi:10.1016/S0022-5347(05)64450-3

33. Gallentine ML, Harmon WJ. "Ureteral substitution with a stapled neoureter: a simplified Boari flap." J Urol. 2001;166(5):1869–1872.

34. Bejany DE, Politano VA. "Stapled and nonstapled tapered distal ileum for construction of a continent colonic urinary reservoir." J Urol. 1988;140(3):491–494. doi:10.1016/s0022-5347(17)41699-5