Determination of patient-specific airway stent fit using novel 3D reconstruction measurement techniques: a 4-year follow-up of a patient

Abstract

Managing complex benign airway disease is a major challenge in interventional pulmonology. With the introduction of additive manufacturing in the medical field, patient-specific (PS) implants are an innovate prospect for airway management. Historically, stents were oversized to resist migration. However, the optimal degree and impact of stent oversizing remains unclear. The ability to design stents based on computed tomography (CT) invites opportunity to understand sizing. Here, we report a novel three-dimensional (3D) image reconstruction tool to quantify fit repeatedly over time. Analysis of CT imaging before and after successive stent implants in a single patient with different areas of stenosis and malacia was done. Nine PS airway stents over 4 years (five left mainstem and four right mainstem) were studied. The distance between the airway model and stent was calculated. The CT images were correlated to stent designs in CloudCompare software (v2.10-alpha) for novel analysis. Heat map was exported depicting the distances between the airway and the stent to the clinician’s prescribed stent model. Corresponding histograms containing distances, mean, and standard deviation were reported. It is possible to measure stent fit based on heat map quantification on patient imaging. Observation of the airway over time and stent change suggests that the airway became more open over time requiring increased stent diameters. The ability to design and measure stent fit over time can help quantify the utility and impact of PS silicone airway stent. The airway appears to display plasticity such that there is notable change in stent prescription over time.

Introduction

The first report of airway stent insertion date back as early as the 19th century, but it was not until the 1960s when a dedicated airway stent became available and was widely used. ¹

Managing complex nonmalignant airway disease is a major challenge in interventional pulmonology.^2,3 Over the past three decades, airway stenting has become an integral part of the therapeutic bronchoscopic management of benign and malignant central airway diseases. Airway stent insertions often achieve immediate relief of dyspnea. ⁴ Moreover, airway stenting has consistently improved airflow obstruction parameters on pulmonary function testing. ⁵

Silicone stents are the gold standard for inoperable, nonmalignant airway disease due to the ease of placement and removal with rigid bronchoscopy. ⁶ These have proven to be modifiable prior to implantation, which addresses some of the mechanical issues with ‘fit’. ⁷ However, challenges include infection, migration, mucus plugging, and granulation, necessitation of lifelong management, and recurrent stent replacement^8–17 Many variables in stent design have been altered over the years, ranging from basic materials, coverings, size, shape, to deployment and other properties.

The US market has very few stent options and precious little comparative data. Choosing a stent is often left to the experience and skill of the operator. Sizing a stent is also a matter of dogma within the constraints of availability and post-production modifications. We have been working on patient-specific (PS) airway stents for a number of years.^3,13 An airway stent remains an option of last resort because of its many challenges. The purpose of this article was to evaluate a methodology using software to look at the stents’ fit within the airway by visual assessment.

Procedure

Analysis of CT imaging before and after successive stent implants in a single patient with different areas of stenosis and malacia was done. Nine PS airway stents over 4 years (5 left mainstem and 4 right mainstem) were studied.

The PS airway stents were designed using VisionAir software, which utilized the patient’s computed tomography (CT) scan to render an airway model. His airway model and the designed stent were uploaded into CloudCompare software (v2.10-alpha) for novel analysis. The treatment area was cropped at the proximal and distal ends so that only the stented region of the airway would be analyzed.

A heat map was exported, depicting the distances between the airway and the stent onto the clinician’s prescribed stent model. A corresponding histogram showing the distance values, mean deviation, and standard deviation were also reported. Regions in which the stent was oversized was observed by the red variation on the stent surface, which meant a larger deviation was detected between the stent and airway. Where the stent and airway distance were minimal or just contacting one another was depicted by the green variation. Areas in which the airway was larger than the stent was depicted by the blue variation. For example, see Figure 1(ic). This was done to observe the interface between the stent and airway anatomy to help determine the clinical effects of the stents and physician-directed design.

Figure 1.

(ia) Airway model (A1) rendered from patient’s CT scan data used to design L1. (ib) Patient-specific stent design, L1, overlaid with airway model, A1. (ic) Patient-specific airway stent, L1, containing heat map. (id) Corresponding histogram containing distance values calculated between L1 and airway model. (iia) Airway model (A2) rendered from patient’s CT scan data used to design L2. (iib) Patient-specific stent design, L2, overlaid with airway model, A2. (iic) Patient-specific stent, L2, containing heat map. (iid) Corresponding histogram containing distance values calculated between L2 and airway model. (iiia) Airway model (A3) rendered from patient’s CT scan data used to design L3. (iiib) Patient-specific stent design, L3, overlaid with airway model, A3. (iiic) Patient-specific stent, L3, containing heat map. (iiid) Corresponding histogram containing distance values calculated between L3 and airway model. (iva) Airway model (A4) rendered from patient’s CT scan data used to design L4. (ivb) Patient-specific stent design, L4, overlaid with airway model, A4. (ivc) Patient-specific stent, L4, containing heat map. (ivd) Corresponding histogram containing distance values calculated between L4 and airway model. (va) Airway model (A5) rendered from patient’s CT scan data used to design L5. Note that one can see L4 stent in the image. (vb) Patient-specific stent design, L5, overlaid with airway model, A5. (vc) Patient-specific stent, L5, containing heat map. (vd) Corresponding histogram containing distance values calculated between L5 and airway model. (via) Airway model (A2) rendered from patient’s CT scan data used to design R1. (vib) Patient-specific stent design, R1, overlaid with airway model, A2. (vic) Patient-specific stent, R1, containing heat map. (vid) Corresponding histogram containing distance values calculated between R1 and airway model. (viia) Airway model (A3) rendered from patient’s CT scan data used to design R2. (viib) Patient-specific stent design, R2, overlaid with airway model, A3. (viic) Patient-specific stent, R2, containing heat map. (viid) Corresponding histogram containing distance values calculated between R2 and airway model. (viiia) Airway model (A4) rendered from patient’s CT scan data used to design R3. (viiib) Patient-specific stent design, R3, overlaid with airway model, A4. (viiic) Patient-specific stent, R3, containing heat map. (viiid) Corresponding histogram containing distance values calculated between R3 and airway model. (ixa) Airway model (A5) rendered from patient’s CT scan data used to design R4. (ixb) Patient-specific stent design, R4, overlaid with airway model, A5. (ixc) Patient-specific stent, R4, containing heat map. (ixd) Corresponding histogram containing distance values calculated between R4 and airway model. At the 4-month follow-up, a CT scan showed regression of the airway inflammation around the stent in the left mainstem bronchus, see Figure 2. The patient had procedures every 135 days (versus 39 days on average prior to implantation of the patient-specific stent). L1 remained implanted for 401 days (versus average stent life of 52 days prior to implantation of the patient-specific stent). After 401 days, L1 was removed per recommendation of stent life (per commercially available silicone stents) and replaced by L2.

Case report

A 57-year-old male with airway complications of granulomatosis with polyangiitis (Wegener’s) had been managed in our practice since 2006. His left mainstem bronchus, bronchus intermedius, and the right upper lobe orifice were all affected. The patient had suffered from debilitating respiratory obstructive symptoms and recurrent pneumonia, requiring bilateral stents for multifocal airway disease.

Since 2006, he had undergone 30 bronchoscopic procedures to maintain airway patency, as well as to maintain and manage stent complications. Previous procedures involved systemic therapies and many variations of currently commercially available stents with immediate relief, yet he consistently suffered from stent-related complications with a few weeks or months of implantation. The FDA-granted compassionate use exemption for a PS airway stent to be implanted in the left mainstem bronchus. Subsequent compassionate use allowed an additional eight PS stents to be implanted in both main stem airways over the next several years (a total of nine PS stents, five in the left mainstem, and four in the right mainstem). Due to the patient’s complex airway disease, the PS airway stents were designed to manage the left and right airways differently. The left side of the airway was malacic and therefore the stents were designed to maintain the volume of the airway. The right side of the airway had been stenotic due to recurrent airway strictures, so these stents were designed to open the airway at the regions of stenosis.

Results

Left mainstem

L1 was implanted in the left mainstem to replace a silicone stent that comprised of three different stents (Lymol 16-13-13 Y with Ø13 mm branch cut to 22 mm; Novatech Ø12 mm × 40 mm cut to 20 mm; Novatech Ø9 mm × 20 mm cut to 5 mm). There was no significant oversizing observed for L1 (11-6-6), 3.441% of distances between the stent and airway model were greater than 2.0 mm, see Figure 1(i).

L2 had increased the diameter of all the branches (14-10-10) due to regression of inflammation in the left mainstem bronchus. Moderate oversizing was observed in the carina region and left upper lobe, 24.231% of distances between the stent and airway model were greater than 2.0 mm, see Figure 1(ii). During follow-up, mild granulation tissue formation and biofilm was noted on the stent. L2 remained implanted for 223 days and was then replaced by L3 due to possibly related, recurrent infections.

L3 design had slightly reduced the proximal diameter and significantly reduced the left upper lobe diameter (13-6-10) to decrease the oversizing observed in the previous design. Moderate oversizing was still observed in the carina and left lower lobe but reduced from the previous L2 design, see Figure 1(iii); 11.488% of distances between the stent and airway model were greater than 2.0 mm. Mild granulation tissue and biofilm were observed during follow-up, but overall, there was an improvement in patient’s symptoms Figure 5. After 313 days, L3 was removed to see if he could become stent-free.

Figure 5.

(a) An eccentric scar band and reactive granulation tissue completely obstructing the bronchus intermedius of the lung and (b) Granulation tissue partially obstructing the Left upper lobe.

A clinical decision to implant a covered-metal stent (Bonastent Ø12×30) was made intraoperatively following the removal of L3 due to severe malacia. The Bonastent remained implanted for 57 days and was then removed due to worsening symptoms, granulation tissue formation, and mucus plugging. A new covered-metal stent (Aero Ø12×30) was implanted, which quickly resulted in granulation tissue formation and mucus plugging. The Aero stent was removed after 36 days and replaced by L4 after obtaining FDA approval for compassionate use. It was observed that using commercial stents was associated with rapid regression.

Prior to implantation of L4, the patient was experiencing significant obstructive symptoms. L4 had reduced the left lower lobe diameter (13-6-6) compared with the previous PS stent design, L3 (13-6-10). Significant oversizing was observed throughout the entire stented region, 69.349% of distances between the stent and airway model were greater than 2.0 mm, see Figure 1(iv). This suggests that the airway was regressing with wall inflammation and narrowing despite the use of commercial stents, requiring a smaller replacement implant compared with L3. There was a small amount of granulation tissue formation in the left upper division, but this was noted prior to implantation of L4. Over time, the patient had mild obstructive symptoms and infrequent infections. Note the space around the stent on the CT scan, see Figure 1(iv). This influenced the clinical decision to increase the diameters of a new PS stent, L5. After 299 days, L4 was removed and replaced by L5.

All diameters in L5 design were increased (16-7-10) from the previous L4 (13-6-6) design due to a reduction in inflammation in the left mainstem bronchus. Minimal oversizing was observed, 8.045% of distances between the stent and airway model were greater than 2.0 mm, see Figure 1(v). L5 was implanted for 537 days and then replaced.

Right mainstem

Prior to implantation of R1, a straight silicone stent was implanted into the right main stem (Novatech Ø9 mm × 20 mm). The Novatech stent had significant stenosis and mucus impaction in the right bronchus intermedius and some progression in the right upper lobe stenosis. The Novatech stent was removed and replaced by R1 (12-8-10). During follow-up, moderate granulation tissue and biofilm formation were noted near the carina of the right mainstem extending into the right upper lobe, where moderate oversizing was observed; 23.242% of distances between the stent and airway model were greater than 2.0 mm, see Figure 1(vi). R1 was implanted for 223 days and then replaced by R2.

R2 design slightly increased the proximal branch diameter (13-7-10). The right upper lobe diameter was slightly reduced (by 1.0 mm) due to granulation tissue formation observed with the prior stents. Minimal oversizing was observed at the carina, 4.291% of distances between the stent and airway model were greater than 2.0 mm, see Figure 1(vii). R2 was removed after 313 days and the right mainstem of the airway was left stent-free.

Due to a worsening of his symptoms, R3 (10-6-9) was implanted 93 days after the removal of R2 (13-7-10). All branch diameters of R3 design were reduced (compared with R2) due to the narrowing of the stent-free right mainstem bronchus. Significant oversizing was observed along the main branch extending into the bronchus intermedius, 54.368% of distances between the stent and airway model were greater than 2.0 mm, see Figure 1(viii), Figure 5. R3 was removed after 299 days and replaced by R4.

R4 had increased the diameters of all the branches (13-8-11) compared with R3 (10-6-9), creating a similar design to R2 (13-7-10). The increase in diameters, however, did not lead to an observation of an increase in oversizing, 7.466% of distances between the stent and airway were greater than 2.0 mm, see Figure 1(ix). R4 was implanted for 537 days and then replaced.

The stent data of each PS stent are depicted in Table 1.

Table 1.

Stent data of each patient-specific stent.

	L1	L2	L3	L4	L5	R1	R2	R3	R4
No. of days implanted	405	223	313	299	537	223	313	299	537
Mean deviation (mm)	0.403	0.902	0.781	2.590	0.709	1.029	0.903	2.208	1.095
Standard deviation	0.923	1.311	1.022	1.189	0.954	1.270	0.642	0.994	0.926
Minimum deviation (mm)	−3.239	−3.143	−1.940	−5.956	−3.162	−2.662	−1.414	−1.971	−3.805
Maximum deviation (mm)	3.297	4.182	4.171	6.159	3.379	3.919	3.146	5.349	2.742
% Of points < 0.0 mm	33.098	29.406	22.892	0.991	19.882	22.194	6.860	0.088	11.388
% Of points > 0.0 mm and < 1.0 mm	40.8	23.982	36.442	3.811	43.618	18.674	46.703	11.773	25.942
% Of points > 1.0 mm and < 2.0 mm	22.661	22.381	29.178	25.849	28.455	35.89	42.146	33.771	55.204
% Of points > 2.0 mm	3.441	24.231	11.488	69.349	8.045	23.242	4.291	54.368	7.466

L5 and R4 remain implanted to date.

Discussion

The purpose of this study was to evaluate a software-based visual assessment of stent fit for PS airway stents and determine whether there were any clinical correlates. The PS stents did not exhibit a pattern of ‘fit’ to indicate that a specific algorithm was used when designing each stent, rather a graphical representation of the clinical decision made by the physician was depicted to address the CT and bronchoscopic findings. It was observed that the PS stents were capable of treating the airway with a more durable response than seen historically in this patient. Notable reduction in the wall thickness and increased lumen was exhibited on the CT images within the first implantation of the PS stent, see Figure 2.

Figure 2.

Inspiratory CT Images. (a) Prior to implantation of the first patient-specific airway stent and (b) 145 days post-implantation of L1 patient-specific airway stent.

An increase in airway volume as well as a reduction of the inflammation of the airway walls can be observed surrounding both main bronchi. A portion of the stent can be seen in b (red arrow).

The regression of inflammation suggested that the subsequent designs needed an increase in their diameters to better represent the changing airway. This is most notably seen in the comparison of the heat map of L4 at implantation [Figure 1(iv)] and the heat map of L4 at removal [Figure 1(v)]. The changes observed in the airway suggest that the airway has plastic properties, which allowed the PS airway stents to re-shape the airway over time, see Figure 3.

Figure 3.

Inspiratory CT image. (a) Prior to implantation of patient-specific airway stent and (b) 132 days post-implantation of L2 and R1 showing a significant increase in airway volume in both the left and right main bronchi.

The removal of the PS airway stents led to an observation of a rapid increase in airway stenosis, which was evidenced by worsening patient symptoms, see Figure 4.

Figure 4.

Inspiratory CT image, 80 days post-removal of L3 and R2. A reduction of the airway volume can be observed in the left and right main bronchi.

The limited observation with commercial stents in both symptoms and imaging suggest that increased wall inflammation may be less with PS stents for this individual. The component of stenosis on the right rapidly recurred when stent-free period was attempted. In a short time, the airway lumen narrowed significantly such that even a smaller replacement stent was associated with increased oversizing. These effects diminished over time, and the subsequent PS stents received were larger, but with less oversizing relative to the lumen.

Observations of the airway made during pre-implantation, implantation, and stent change suggest that the airway became more open over time requiring an increase in stent diameters.

Currently, there are no studies available defining the best way to size airway stents. Previous studies have considered creating a ‘snug’ fit or implanting the largest size possible, but no quantitative measures have been given to define a ‘snug’ fit or to prove the efficacy of using the largest size. In fact, most experts describe a process of measuring the airway and oversizing by 1–2 mm, to prevent migration associated with undersizing and prevent the formation of granulation tissue from high pressure points. There are also clinical decisions to be made about oversizing that differ when treating airway malacia as opposed to stenosis. Based on literature, it appears that a better fitting stent may reduce the current risks associated with airway stents and offer additional benefits, currently unknown.

Conclusion

Based on the results of this study, the PS airway stent may not only be used as a sufficient treatment for airway patency but also used as a correctional treatment over time to re-shape a patient’s airway. These early observations need to be systematically reproduced in rigorous prospective studies. Confounders, operator bias, and additional objective measures of respiratory function will need to be assessed. Furthermore, the use of software applications to define stent fit may have some value in either predicting or better describing best fit.