Does the Reversal and Normalization of Airway and Pulmonary Vascular Remodeling Accompany Asthma Remission?

To the Editor:

In patients with asthma who achieve clinical remission, ascertaining whether and when this is accompanied by the reversal of remodeling and the normalization of asthma lung pathologic conditions is difficult. Although ample evidence supports the achievement of clinical remission based on exacerbations, oral corticosteroid (OCS) use, stability of lung function, and asthma control,^1,2 whether such patients also experience some reversal of lung remodeling, which was previously proposed¹ as a critical component of “complete remission,” is unknown.

To address this knowledge gap, we generated a pulmonary imaging 4-domain index of normal airway and pulmonary vascular morphology to objectively evaluate the reversal of lung remodeling in patients with asthma. We evaluated 12 participants (10 female; 61 ± 16 years of age) with severe eosinophilic asthma (n = 12; 61 ± 16 years) before and after 2.5 years of anti-IL5Rα therapy; participants provided written informed consent to a protocol to Western University Health Sciences Research Board (IRB#00000940),³ and received anti-IL5Rα (30 mg per dosing regimen),⁴ and returned for a 2.5-year-follow-up visit.^5,6

Pulmonary function tests, Asthma Control Questionnaire (ACQ) scores, Asthma Quality of Life Questionnaire scores, and St. George’s Respiratory Questionnaire scores were acquired at baseline and 4-, 8-, 16-, 48- and 130-week (2.5-year) visits. ¹²⁹Xe MRI was acquired at all study visits, and CT was performed at baseline and 2.5 years.⁴ Exacerbations and OCS use were documented directly with participants and via their electronic health record.

CT images were analyzed to quantify total airway count (TAC), airway wall thickness percent (WT%) and airway lumen area (LA), using VIDAVision software (VIDA Diagnostics Inc).⁴ The Chest Imaging Platform (Brigham and Women’s Hospital) was used to quantify CT pulmonary vascular measurements.⁷ CT mucus-score was quantified⁶ using published methods.⁸ Anatomic proton (¹H) and ¹²⁹Xe MRI were acquired, and MRI ventilation defect percentage (VDP) was quantified.⁴ Clinical remission was classified² as previously described using relaxed (no exacerbations/OCS 12 months, ACQ-5 ≤ 1.5, stable lung function) and strict criteria (no exacerbations/OCS 12 months, ACQ-5 ≤ 0.75, lung function = FEV₁ ≥ 80%_pred or ΔFEV₁ ≥ 200 mL). We used Holm-Bonferroni paired t tests and linear mixed models (SPSS Statistics 29.0; IBM) to determine significant differences over time. Figure 1A provides the markers used to define clinical remission² and pathologic normalization.

Figure 1.

Schematic diagram of imaging and clinical remission measurements, measurement changes over time and overall summary of clinical and imaging/pathologic remission findings. A, Schematic pulmonary ¹²⁹Xe ventilation MRI and CT providing structural and functional information used to evaluate pathologic normalization. Measurements include CT TAC, LA, WT%, and mucus score as well as MRI VDP and CT BV₅/TBV. Clinical measurements of remission included exacerbations, LTOCS use, ACQ-5 and FEV₁ for both relaxed and strict definitions (2) of clinical remission. B, Spaghetti plot for FEV₁ %_pred, ACQ-5, number of exacerbations, and oral corticosteroid use for all (n = 12) participants at baseline, 4, 8, 16 weeks, 1 year, and 2.5 years. C, Spaghetti plot for Mucus-score, VDP, TAC, and BV₅/TBV for all (n = 12) participants at baseline, 4, 8, 16 weeks, 1 year, and 2.5 years. D, Number of participants achieving clinical remission and pathologic normalization across multiple domains (2-, 3-, and 4-domain criteria) at both 1-year and 2.5-year follow-up. Clinical remission is shown in gray, and imaging/pathologic normalization in blue. E, Spaghetti plot for ACQ-5, FEV₁ %_pred, VDP and mucus-score results for n = 5 participants with both clinical and imaging/pathologic remission. (n = 2 had a mucus score of 0 at baseline and 2.5-year follow-up). %_pred = percent predicted; ACQ-5 = Asthma Control Questionnaire - 5; BV₅ = volume of small blood vessels with cross-sectional area < 5mm²; LA = lumen area; LTOCS = long-term oral corticosteroid; OCS = oral corticosteroid; TAC = total airway count; TBV = total blood volume; VDP = ventilation defect percent; WA = wall area; WT% = wall thickness percent.

We developed a new imaging/pathology-based 4-domain scheme based on previous results,^9,10 which showed that as asthma severity increases, ¹²⁹Xe MRI VDP increases (worsens), CT mucus-occlusions increase,⁸ CT airway wall area and thickness are increased, and both CT total airway count and airway LA diminish or worsen. In addition, the volume of small blood vessels with cross-sectional area < 5 mm² (BV₅) diminishes over time (referred to as pruning),¹¹ perhaps because of blood redistribution from the smaller to the larger blood vessels. Based on these previous findings, we integrated these measurements into an imaging-based index to define pathophysiologic normalization based on healthy reference measurements, as follows: 2 domain imaging normalization = MRI VDP ≤ 4%¹² and CT mucus score ≤ 1; 3 domain imaging normalization = 2 domain + WT% ≤ 18.5% or TAC ≥ 205 or LA ≥ 8.1, and 4 domain imaging normalization = 3 domain + BV₅ normalized to total blood volume (TBV) ≥ 0.56.¹¹

Table 1 shows that the mean ACQ-5 was significantly reduced at 1 year (P = .009) and 2.5 years (P = .009), and this agreed with the significant absolute increase from baseline in mean FEV₁ %_pred of 20% at 1 year (P = .02) and 26% at 2.5 years (P = .009). Mean VDP significantly decreased by 11% (absolute) at 1 year (P = .04) and 13% (absolute) at 2.5 years (P = .004), which coincided with significantly diminished mucus score (P = .03) and significantly increased TAC (P = .01) and BV₅/TBV (P = .04) at 2.5 years. Clinical (Fig 1B) and imaging (Fig 1C) measurements are shown in spaghetti plots for all 12 participants.

Table 1.

Clinical and Imaging Measurements, Changes and Remission Classification, 1 and 2.5 Years After Anti-IL5Rα Initiation

	1 Year (n = 10)			2.5 Years (n = 12)
	Mean (SD)	Δ (SD)	P	Mean (SD)	Δ (SD)	P
Clinical Measurements
FEV1 %pred	82 (17)	20 (22)	.02	89 (19)	26 (23)	.009
ACQ-5	0.4 (0.5)	-1.5 (1.2)	.009	0.5 (0.5)	-1.7 (1.5)	.009
ACQ-6	0.3 (0.5)	-1.4 (1.1)	.005	0.5 (0.5)	-1.6 (1.6)	.004
No EX/OCS ✓	10 (100)	ND	ND	11 (92)	ND	ND
EX/OCS ✕	0 (0)	ND	ND	1 (8)	ND	ND

Clinical Remission	Relaxed	Strict		Relaxed	Strict
2 Domain	10 (100)			11 (92)
3 +Control	9 (90)	8 (80)		11 (92)	7 (58)
3 +Lung function	7 (70)	7 (70)		11 (92)	10 (83)
4 Domain	7 (70)	5 (50)		10 (83)	6 (50)

Imaging/Pathologic Measurements	Mean (SD)	Δ (SD)	P	Mean (SD)	Δ (SD)	P
VDP %	5 (4)	-11 (13)	.04	4 (2)	-13 (11)	.004
Mucus score	ND	ND	ND	1 (1)	-3 (4)	.03
TAC	ND	ND	ND	197 (59)	68 (60)	.01
WT%	ND	ND	ND	18.7 (1.0)	-0.2 (1.0)	.9
LA, mm2	ND	ND	ND	9.4 (3.8)	0.1 (3.7)	.9
BV5/TBV	ND	ND	ND	0.57 (0.03)	0.06 (0.07)	.04
Imaging/Pathologic Remission
2 Domain	ND	ND	ND	8 (67)		ND
3 +TAC or WT% or LA	ND	ND	ND	6 (50)		ND
4 Domain (+BV5/TBV)	ND	ND	ND	5 (42)		ND

Data are presented as No. (%) unless otherwise indicated. Holm-Bonferroni corrected P values for differences between baseline to 1 year and baseline to 2.5 years. Linear mixed model P values were similar to Holm-Bonferroni paired t-test P values. Boldface indicates the values are significantly different.

✓ = achieved; ✕ = not achieved; %_pred = percent predicted; ΔFEV₁ = difference in FEV₁ from baseline; ACQ = Asthma Control Questionnaire; BV₅ = volume of small blood vessels with cross-sectional area < 5 mm²; EX = exacerbation; LA = lumen area; ND = not done; OCS = oral corticosteroid; TAC = total airway count; TBV = total blood volume; VDP = ventilation defect percent; WT% = wall thickness percent.

The number and fraction of participants achieving clinical remission and pathologic normalization are provided in Table 1 and Figure 1D. Using the relaxed definition, 100%, 90%, 70%, and 70% of patients met criteria for 2, 3 (+control), 3 (+lung function), and 4 domain clinical remission at 1 year, and 92%/92%/92%/83% achieving 2, 3 (+control), 3 (+lung function), and 4 domain clinical remission at 2.5 years. Using the strict definition of clinical remission, 100%/80%/70%/50% of patients met criteria for 2, 3 (+control), 3 (+lung function), and 4 domain remission at 1 year, respectively, and 92%/58%/83%/50% achieved this at 2.5 years.

Imaging-index-based normalization was identified in 67%/50%/42% of participants who met 2, 3 (+TAC or +WT% or +LA), and 4 (+BV₅/TBV) domain criteria, respectively, at 2.5 years. There was no evaluation of imaging-index remission at 1 year because CT was not acquired. All 5 participants who achieved 4 domain imaging normalization at 2.5 years also achieved the relaxed 4 domain clinical remission definition. Three participants (P016, P025, and P028) with imaging normalization did not achieve strict 4 domain clinical remission (ACQ-5 = 0.8/1.5/1.0, respectively), perhaps because of the subjective nature of self-reported asthma control. In Figure 1E, the changes over time are shown for the 5 patients who achieved both clinical and imaging remission for ACQ-5, FEV₁, VDP, and mucus score. We did not observe any baseline measurement differences in the 5 patients who achieved imaging remission as compared with the other 7. Hence, we cannot comment on predictive factors that may have influenced clinical or imaging/pathologic remission observed here. What is clear is that the imaging index identified that normalization of CT mucus score, small vessel volume, airway morphology, and MRI ventilation defects was achieved in 5 of 6 patients who achieved clinical remission. The added value of this work is the conceptual introduction of a new composite image-based index of response for longitudinal studies of biologic therapy, which is complementary to clinical indexes of “remission.”

We acknowledge the small sample size (n = 12 vs n = 28 at baseline, of whom 16 were not evaluated), with dropout attributable to a lack of funded anti-IL5Rα access, the influence of pandemic local lockdowns, as well as loss to follow-up and pregnancy. Notwithstanding the small sample size, 5 participants achieved both clinical remission and imaging/pathologic normalization after 2.5 years, which will help guide future studies.

In summary, we developed a novel imaging-based index of normal lung morphology that demonstrated that in 5 of 6 patients who achieved clinical remission, the reversal and normalization of airway and pulmonary vascular remodeling was also achieved after 2.5 years’ anti-IL5Rα. If remission is an emerging goal of asthma treatment, the reversal of lung remodeling alongside airway and vessel normalization can now be measured and considered.

Funding/Support

A. M. and S. T. are currently supported by MSc scholarships from the Natural Sciences and Engineering Research Council (NSERC) Canada and Asthma Canada. M. J. M. is also supported by an NSERC postdoctoral scholarship. G. P. is supported by NSERC, CIHR, and holds a Tier 1 Canada Research Chair. This manuscript describes long-term follow-up results which were funded by the investigator, via an investigator-sponsored study for which baseline data were partially funded by Astra Zeneca.

Financial/Nonfinancial Disclosures

The authors have reported to CHEST the following: G. P. reports financial support was provided by GlaxoSmithKline Inc. G. P. reports a relationship with GlaxoSmithKline Inc that includes: speaking and lecture fees. G. P. reports a relationship with Polarean PLc that includes: consulting or advisory and speaking and lecture fees. A. B. reports relationships with GlaxoSmithKline Inc, Sanofi-Regeneron, and AstraZeneca that includes: speaking, lecture, and consulting fees. None declared (A. M., S. T., M. J. M., H. K. K., H. S., and C. Y.)