Increased work of breathing and its relationship to dyspnea in malignant pleural effusion

Objectives: Although dyspnea is the most common symptom of pleural effusion (PE), its physiological basis has not yet been fully elucidated. The aim of this work is to investigate the cause of dyspnea before therapeutic thoracentesis (TT) by analyzing the lack of dyspnea relief after TT.

Methods: We retrospectively analyzed data gathered during TT. Among others, our database includes measurements of instantaneous pleural pressure (Ppl) in the ipsilateral hemithorax and airflow in the mouth (during TT), as well as arterial gas tensions (AGT) and dyspnea characterized quantitatively via the Modified Borg Scale (before and after TT). As the Borg scale is a subjective measure, the change in dyspnea (dB) was used in reliable quantitative analyses. Differences in various parameters and their changes between patients who reported dyspnea relief and the other patients (the YES and NO groups, respectively) were studied. Additionally, correlations between dB and these parameters (and their changes) were studied.

Results: Only the amplitude of Ppl changes related to breathing after TT was significantly different (higher) in group NO than in group YES (p < 0.003; the large effect size). dB correlated with this amplitude and the volume of withdrawn fluid (r = −0.51 and 0.51, respectively), but it correlated with neither changes in AGT nor minute ventilation.

Conclusion: The results suggest that the key mechanism of dyspnea in patients with malignant PE is related to reduced total lung compliance due to collapse of a lung part, leading to an increase in the work of breathing required to maintain adequate minute ventilation.

1 Introduction

Dyspnea is reported by approximately 50%–65% of patients with malignant pleural effusion (PE), compromising their quality of life and often being the primary reason for seeking emergency care (Antony et al., 2001; Piggott et al., 2023). Although various physiological mechanisms of dyspnea in patients with PE have been considered for decades, the key mechanism seems to remain unclear (Parshall et al., 2012; Thomas et al., 2015).

The accumulation of pleural fluid is associated with compression of the ipsilateral lung, which may result in increased ventilation‒perfusion mismatch and impaired gas exchange. However, neither low arterial O₂ (PaO₂) nor elevated CO₂ (PaCO₂) partial pressures can be responsible for dyspnea in patients with PE, as both may either decrease or increase after therapeutic thoracentesis (TT) despite dyspnea relief, and the changes in PaO₂ and PaCO₂ do not correlate (Karetzky et al., 1978; Brandstetter and Cohen, 1979; Agustí et al., 1997; Stecka et al., 2018; Muruganandan et al., 2020; Taylor et al., 2021; Zielinska-Krawczyk et al., 2022). Therefore, another cause of dyspnea in PE patients should be considered.

The hydrostatic pressure of the pleural fluid exerted on the ipsilateral hemidiaphragm may lead to flattening or even reversal, significantly impairing diaphragm function; likewise, a large volume of pleural fluid puts pressure on the chest wall (overstretching intercostal muscles) and mediastinum, resulting in its contralateral shift and changes in respiratory mechanics (Thomas et al., 2015; Skaarup et al., 2020). Recent studies have posited that dyspnea in PE may stem primarily from abnormal ipsilateral hemidiaphragm function. Consequently, the reduction in dyspnea following TT may be related to the restoration of the ipsilateral hemidiaphragm shape and function (Muruganandan et al., 2020; Skaarup et al., 2020; Muruganandan et al., 2023). This, however, has not been supported by a more recent study (Fjaellegaard et al., 2024). On the other hand, although dyspnea relief can be statistically correlated with improvement in the ipsilateral hemidiaphragm shape and function, these changes may not be causally linked, i.e., TT could influence both dyspnea and hemidiaphragm function independently. This could explain why dyspnea does not resolve in some patients despite the hemidiaphragm returning to its normal upward curve after TT (Muruganandan et al., 2020; Psallidas et al., 2017).

Given the difficulties in determining the physiological causes of dyspnea before TT, we propose a kind of reversal of the problem, i.e., investigation of the reasons for the lack of dyspnea relief after TT. We attempted to use our own database to study this problem. This database contains various data gathered during TT, including records of the pleural pressure (P_pl) enabling the determination of the amplitude of P_pl changes related to breathing (P_{pl_ampl}).

As it is shown in Figure 1A, the work done against elastic forces during one breath is approximately equal to 0.5⋅P_{pl_ampl}⋅V_T. As the minute work, commonly called work of breathing (WOB), is equal to the work done in one breath multiplied by the respiratory rate (RR), the elastic WOB (WOB_e) is equal to 0.5⋅P_{pl_ampl}⋅V_T⋅RR. Since the minute ventilation is equal to V_T⋅RR, we have:

${\text{WOB}}_{\mathrm{e}}=0.5·{\mathrm{P}}_{\text{pl}\_\text{ampl}}·\left({\mathrm{V}}_{\mathrm{T}}·\text{RR}\right)=0.5·{\mathrm{P}}_{\text{pl}\_\text{ampl}}·{\mathrm{V}}_{\mathrm{E}}$

Figure 1

Figure 1. Influence of pleural effusion on work of breathing – schematic diagrams. Bold red and blue curves – schematic nonlinear pressure-volume curves for the chest wall and lungs, respectively, characterizing their static, nonlinear compliances; V – the current volume of the chest cavity/lungs; FRC – functional residual capacity; V_T – a fixed tidal volume; P_pl – pleural pressure, P_{pl_FRC} – P_pl at FRC; P – P_atm-P_pl in the case of the chest wall and P_pl-P_atm in the case of lungs, P_{pl_ampl} – the change in P_pl required to inspire V_T; W – work done during one breath; V_PE – the pleural fluid volume; C_R and C_L – volume-dependent compliance of the right and left lung, respectively; C_collapsed – decrease in the (left) lung compliance caused by ‘elimination’ of a lung part due to its collapse. (A) To increase the lung volume by ΔV, respiratory muscles have to develop the pressure P1 that decreases P_pl to the value that corresponds to FRC+ΔV. According to the laws of physics, the work ΔW1 done by these muscles is equal to P1⋅ΔV. Further increase of V by next ΔV requires pressure P2 and the work ΔW2 is equal to P2⋅ΔV, etc. The total work done by the muscles to inspire V_T is the sum of W1, W2, W3 … Wn, i.e., it is equal to the area of the grey triangle that is equal to 0.5⋅V_T⋅P_{pl_ampl}. (B) Pleural fluid takes up space in the chest decreasing the chest cavity volume, which leads to a downward shift of the chest wall characteristics. Collapse of a lung part leads to compression of the lung characteristics. Although the above leads to a decrease in FRC, its value corresponds to smaller differential lung compliance. As a result, P_{pl_ampl} necessary for the required V_T and WOB (grey triangle area) are greater than in the healthy respiratory system. (C) Pleural fluid withdrawal causes return of the chest wall characteristics to the original position; however, if the collapsed lung part is not recruited, the characteristics of the lungs does not change. FRC increases to the value for which this characteristics is much more horizontal, i.e., the differential compliance is smaller. As a result, P_{pl_ampl} necessary for the required V_T and W are greater than before the fluid withdrawal. (D–F) present schematically contributions of individual lungs to the total lung compliance presented in (A), (B), (C), respectively. The magnitude of arrows indicates contributions of individual hemidiaphragms to W. (D) The pressure-volume curve for lungs illustrates the sum of volume-dependent C_R and C_L. (E) C_L(B) is smaller than C_L(A) by C_collapsed due to collapse of a lung part; therefore, W must increase to maintain V_T. (F) If the atelectatic part of the left lung is not recruited, the ventilated part must overexpand to take place of withdrawn pleural fluid; therefore, since C_L is volume-dependent, C_L(C) is smaller than C_L(B). As a consequence, W is greater, not smaller, than before therapeutic thoracentesis.

Thus, if V_E remains relatively stable, P_{pl_ampl} can be used to characterize WOB_e. Consequently, this study is particularly focused on analyzing the relationship between dyspnea and WOB_e.

2 Materials and methods

2.1 Patients

This study involved a retrospective analysis of data obtained from a comprehensive prospective project in which multiple physiological parameters were measured before, during, and after TT. This project received approval from the Institutional Review Board (KB 105/2012) and was registered on ClinicalTrials.gov (NCT02192138). Medical procedures were conducted at the Department of Internal Medicine, Pulmonary Diseases and Allergy of the Medical University of Warsaw.

The data used for analysis were obtained from patients with malignant PE referred to the hospital for TT. The inclusion criteria were as follows: (1) age between 18 and 85 years, (2) symptomatic pleural effusion occupying at least one-third of the hemithorax determined by posteroanterior chest radiogram, (3) the severity of symptoms (including dyspnea) warranting TT, (4) absence of contraindication for TT, and (5) signed consent to participate in the study. The following exclusion criteria were applied: (1) poor general health condition warranting non-extension of the procedure, (2) mechanical ventilation due to respiratory failure, and (3) unstable hemodynamic or respiratory status not related to pleural effusion.

2.2 Measurements and parameters

TT was performed with the patient in a sitting position. The pleural fluid was withdrawn intermittently, with 1-min breaks for measurement purposes.

Dyspnea was assessed just before and after TT via the Modified 10-point Borg Scale. Dyspnea relief associated with TT was quantified by the difference (ΔB) between the Borg Scores before (B_pre) and after TT (B_post), i.e., ΔB = B_pre− B_post. The instantaneous values of P_pl in the ipsilateral hemithorax (digital pleural manometer, IBBE, Poland) and airflow through the mouth (modified LungTest 1,000 spirometer, MES, Poland) were recorded and synchronized for further analyses. V_E was calculated as the sum of tidal volumes during the period of the spirometric measurement divided by the duration of this period. The median value of P_{pl_ampl} after TT (P_{pl_ampl post}) was construed as an index of WOB_e at the end of the procedure.

Arterial blood gases were measured 1 hour before and 1 hour after TT. Arterial blood samples were collected from the radial or ulnar artery of patients breathing ambient air. Blood samples were analyzed within 15 minutes after collection using the Blood Gas Analyzer ABL 800 FLEX (Medical ApS, Brønshøj, Denmark).

More details about patients and methods can be found in previous articles (e.g., Zielinska-Krawczyk et al., 2018; Zielinska-Krawczyk et al., 2022).

2.3 Data analyses

Since dyspnea is a subjective sensation and may be perceived and reported differently by each patient, the absolute value of dyspnea score might not be directly comparable between patients. The relative dyspnea index that are ΔB and ΔB/B_pre seem to be better parameters for objective quantitative analysis; therefore, they were compared with other parameters and their changes.

Since dyspnea relief after TT was observed in only some patients, we divided our patients into two groups: the YES group consisted of patients who reported dyspnea relief (i.e., ΔB>0) and the NO group consisted of the other patients (ΔB≤0). Then, we analyzed which of the following parameters (or their changes) differentiated these groups: PaO₂, PaCO₂, V_E, the volume of withdrawn pleural fluid (V_w) and P_{pl_ampl post}.

If a patient reported insignificant dyspnea before TT, i.e., if B_pre≤2, he/she could not report significant ΔB regardless of the true degree of dyspnea relief. This might distort possible correlations between ΔB and other parameters or their changes; therefore, the main analyses were performed for patients who reported B_pre>2, however correlations were also calculated for all patients.

2.4 Statistical methods

Statistical analyses were performed via the Statistica 10 package (StatSoft Inc.). As some of the analyzed data had distributions different from the normal distribution, nonparametric statistical methods were used. Statistical significance was considered when p < 0.05. The Spearman correlation coefficient was used to estimate associations between the analyzed parameters. The difference in parameters between the NO and YES groups was assessed via the Mann–Whitney U test, and the Glass rank-biserial correlation (rg) was used to estimate the effect size; if rg > 0.45, then the effect size was interpreted as large.

3 Results

In general, data for 41 patients were analyzed; however, due to technical issues, arterial gasometry could not be done before or after TT in 14 patients. Table 1 presents the characteristics of the whole sample, and the YES and NO groups separately. Neither arterial blood gases (Figure 2B) nor V_E nor their changes differed among these groups (Table 1). B _pre and V_w were greater in the YES group; nevertheless, these differences were statistically insignificant. Moreover, although dispersion of PaO₂ values was smaller in patients who reported not severe dyspnea before TT, neither the median values nor PaO₂ changes were different (Figure 2A).

Table 1

Table 1. Characteristics of the groups.

Figure 2

Figure 2. Changes in arterial oxygen tension. PaO_{2 pre} and PaO_{2 post} – the arterial oxygen tensions measured before (Pre) and after (Post) thoracentesis, respectively. (A) B_pre < 4: patients who reported no to moderate dyspnea before thoracentesis (Borg score <4); B_pre ≥ 4: patients with more severe dyspnea. (B) Group NO: patients who reported no dyspnea relief or worsening after therapeutic thoracentesis, Group YES – patients who reported dyspnea relief.

P_{pl_ampl post} was the only parameter that significantly differentiated these groups: it was much greater in the NO group than in the YES group (Table 1). The association between P_{pl_ampl post} and dyspnea is additionally confirmed by the statistically significant inverse correlation between P_{pl_ampl post} and ΔB (Table 2; Figure 3A) and the significant positive correlation between P_{pl_ampl post} and B_post in patients reporting more considerable dyspnea before TT, i.e., in 33 patients with B_pre>2 (Table 3; Figure 3B).

Table 2

Table 2. Spearman’s correlations between changes in dyspnea (ΔB) and other parameters.

Figure 3

Figure 3. Relationships between selected parameters in patients with B_pre>2. Subscripts “pre”and “post”indicate the value of a parameter before and after pleural fluid withdrawal, respectively; P_{pl_ampl post} – the amplitude of pleural pressure changes related to breathing (after the fluid withdrawal); (B) Borg score characterizing dyspnea reported by a patient; ΔB/B_pre – relative dyspnea relief, i.e., the decrease in B (=B_pre-B_post) caused by pleural fluid withdrawal expressed as percentage of B_pre; ΔV_E/V_{E pre} – relative minute ventilation (V_E) increase (=V_{E post}-V_{E pre}) expressed as percentage of V_{E pre}; ΔPaO₂/PaO_{2 pre} – PaO₂ increase expressed as percentage of PaO_{2 pre}. (A) there is an inverse correlation between dyspnea relief and the amplitude; (B), there is a positive correlation between dyspnea after pleural fluid withdrawal and the amplitude; (C,D) – dyspnea relief is associated with neither changes in V_E nor in PaO₂.

Table 3

Table 3. Other statistically significant Spearman’s correlations for the B_pre>2 subsample.

Although the minute ventilation (both V_{E pre}, V_{E post} and ΔV_E) did not differ significantly between the YES and NO groups (Table 1), V_{E pre} was statistically significantly correlated with ΔB (Table 2) and B _pre (Table 3) in the subsample B_pre>2. Table 3 shows other statistically significant correlations.

4 Discussion

There are a number of underlying causes of dyspnea, including various cardiovascular, neuromuscular or respiratory disorders, pain, psychiatric disorders and others (Coccia et al., 2016; Beaudry et al., 2022; Ritter et al., 2024). Some of them may also be present in patients with PE. Nevertheless, since pleural fluid withdrawal usually alleviates dyspnea (the high correlation between ΔB and V_w, Table 2), the presence of this fluid should be suspected as the main cause in patients with PE not related to congestive heart failure (dyspnea does not correlate with the pleural fluid volume in patients with cardiac-related PE (Wijayaratne et al., 2024)); however, this explains only the reason for dyspnea but does not explain its pathophysiological mechanism, i.e., the direct cause.

Changes of PaO₂ and PaCO₂ during TT appeared to be not associated with changes in dyspnea (Table 2); moreover, these changes were insignificant and could be both positive and negative (Table 1), even in the group YES (Figure 2B), which agrees with results of in silico studies (Gólczewski et al., 2025). If low PaO_{2 pre} was indeed a significant cause of considerable dyspnea before TT, more significant dyspnea relief (i.e., greater ΔB) after TT should be associated with more significant improvement of PaO₂ (i.e., greater ΔPaO₂) giving a significant positive correlation, which was not observed (Table 2; Figure 3D). Additionally, the higher the level of dyspnea was before TT, the less significant the PaO₂ improvement was observed (inverse correlation between B_pre and ΔPaO₂, Table 3), whereas if low PaO₂ would be the reason for dyspnea, one could expect more significant improvement in patients with higher B_pre. Thus, our results confirm previous suggestions that low PaO₂ and/or elevated PaCO₂ are not directly responsible for dyspnea in patients with PE not related to cardiovascular diseases.

In our patients with considerable dyspnea at baseline (i.e., B_pre>2), V_{E pre} demonstrated a statistically significant positive correlation with both B_pre (Table 3) and ΔB (Table 2). This might suggest that increased V_E could be perceived by patients as dyspnea. However, if increased V_E were indeed a substantial cause of dyspnea, relief of dyspnea would be associated with a decrease in V_E. In, fact, changes in V_E were generally small and not associated with ΔB (Figure 3C). Furthermore, they could be both positive and negative, even in the group YES (Table 1). Additionally, ΔV_E% exhibited a negative correlation with B_pre (Table 3). The above proves that more intensive ventilation could not be a substantial cause of the feeling of dyspnea in our patients.

Thus, the observed correlations (Table 2 and 3) between B _pre and ΔB on the one hand, and V_{E pre}, ΔV_E% and ΔPaO₂ on the other hand were likely related to secondary associations only.

P_{pl_ampl post} was the only parameter that was significantly different between the YES and NO groups (Table 1) and was correlated with dyspnea change (ΔB in Table 2) as well as with dyspnea after TT (B_post in Table 3). The observed inverse correlation between ΔB and P_{pl_ampl post} (Table 2) suggests that the more intensive the work of the respiratory muscles was after TT, the smaller the relief of dyspnea was reported. Moreover, high or even very high P_{pl_ampl post} in patients in the group NO suggests that inspiratory muscles worked very effectively in these patients. This phenomenon is most likely related to the ipsilateral hemidiaphragm, as it is the main inspiratory muscle producing P_pl changes in this cavity (Figure 1F). Thus, on the one hand, patients from this group reported either no dyspnea relief (10 patients) or even dyspnea increase (5 patients); on the other hand, the respiratory muscles worked very effectively since they could produce such a high P_{pl_ampl post}. This seems to confirm the most recent results reported by Fjaellegaard et al. (Fjaellegaard et al., 2024) that neither the normal shape nor movement of the diaphragm after TT is associated with potential dyspnea relief, as it has been suggested lately by other authors (Muruganandan et al., 2020; Skaarup et al., 2020; Muruganandan et al., 2023).

In general, WOB_e depends directly on P_{pl_ampl} according to the fundamental equation WOB_e = 0.5⋅P_{pl_ampl}⋅V_E. Thus, regardless of which respiratory muscles are responsible for such high P_{pl_ampl post} in patients in the group NO, these patients may report dyspnea due to high WOB_e. In the YES group, WOB_e was significantly lower after TT, and those patients reported dyspnea relief. The association between post-TT dyspnea and post-TT WOB_e seems to be additionally confirmed by the correlation between B_post and P_{pl_ampl post} (Table 3). Note, however, that although this correlation is statistically significant, it is not very high. This may suggest that either other factors could also contribute to dyspnea or subjective feeling of increased WOB_e was different in individual patients or both.

Unfortunately, WOB_e before TT cannot be assessed on the basis of P_{pl_ampl} measured in the ipsilateral hemithorax because the ipsilateral hemidiaphragm has an insignificant contribution to respiration before TT (Figure 3B), and WOB_e is mostly done by the contralateral hemidiaphragm; in particular, P_{pl_ampl} in the ipsilateral hemithorax may have negative values when the ipsilateral hemidiaphragm is inverted (which would yield a ridiculous result, i.e., a ‘negative WOB_e’). Nevertheless, WOB_e before pleural fluid withdrawal can be assessed on the basis of the laws of physics since, in general:

${\text{WOB}}_{\mathrm{e}}=0.5·\text{RR}·{\mathrm{V}}_{\mathrm{T}}^2/{\mathrm{C}}_{\text{tot}}=0.5·{\mathrm{V}}_{\mathrm{E}}^2/\left(\text{RR}·{\mathrm{C}}_{\text{tot}}\right)$

where C_tot is the total respiratory system compliance. Thus, since neither V_E (Table 1) nor the median RR (Zielinska-Krawczyk et al., 2018) changes significantly during TT, changes in C_tot are responsible for changes in WOB_e. Before TT, the pleural fluid causes collapse of a part of the ipsilateral lung or even the whole ipsilateral lung is not ventilated; this means that C_tot is reduced proportionally. According to the above formula, WOB_e is increased to the same extent. Moreover, lung compliance is nonlinear. Therefore, the same V_T ventilated smaller amount of lungs requires an additional increase of WOB_e (Figure 1C).

The pleural fluid compresses both the lungs and bronchi; therefore, the total WOB = WOB_e + WOB_r (where WOB_r is the resistive work) is additionally increased by raised work against bronchi resistance. Moreover, under physiological conditions, ribcage elastance helps inspiration, whereas additional effort is required if the thoracic cavity is expanded by PE over the volume for which the trans-wall pressure is positive at the FRC. Thus, WOB has to be increased before TT, and patients can interpret this increase as dyspnea regardless of the other possible causes.

For reasons discussed above, before TT, WOB must be increased at least owing to the collapse of a part of the lungs (Figures 1B,E), thoracic cavity expansion and narrowed bronchi. After TT, the collapsed parts can be either recruited or not. If these parts are recruited, then C_tot increases to the normal value (Figures 1A,D). In consequence, P_{pl_ampl} = V_T/C_tot can be small, the required WOB_e decreases and dyspnea is reduced as in the group YES. However, if these parts remain collapsed after TT, C_tot is still small, and thus (Figures 1C,F).

a. WOB_e cannot decrease and dyspnea remains,

b. P_{pl_ampl} in the contralateral hemithorax has to be high because of the small C_tot,

c. as the ipsilateral hemidiaphragm is curved upward at least as much as the contralateral hemidiaphragm, high P_{pl_ampl} in the ipsilateral hemithorax, similar to P_{pl_ampl} in the contralateral hemithorax, may be generated,

d. lung compliance nonlinearity may cause an increase in WOB_e (Figure 1C) leading to dyspnea increase, not relief.

The above can explain the lack of dyspnea relief and intriguing dyspnea increase after TT in the group NO if dyspnea in those patients before TT has been caused just by increased WOB_e.

Concluding, either dyspnea is reduced (ΔB is large) because the WOB_e and P_{pl_ampl post} are small after TT (the YES group in Table 1) or WOB_e is still large, dyspnea persists (ΔB≤0) and P_{pl_ampl post} is high (the NO group). Certainly, some collapsed parts can be recruited, and the others remain collapsed; thus, both ΔB and P_{pl_ampl} can be moderate. As the result, a negative correlation between ΔB and P_{pl_ampl post} (Table 2) and a positive correlation between P_{pl_ampl post} and B_post (Table 3) are observed.

It should be noted that if the mediastinum is very compliant, P_pl need not be very low despite high P_{pl_ampl post} (Gólczewski et al., 2025). Indeed, in such a case, a lack of collapsed lung part recruitment requires high P_{pl_ampl post} to maintain V_E at a necessary level, whereas the space in the ipsilateral hemithorax that has been occupied by pleural fluid may be partly filled with the contralateral lung, which protects against excessive P_pl fall and seemingly suggests an expandable lung. This may partly explain the lack of differences in post-TT dyspnea relief between patients with expandable and non-expandable lung (Petersen et al., 2024). Thus, not only P_pl but also P_{pl_ampl} should be monitored during TT, at least to differentiate expandable and non-expandable lungs.

Although not very high, the statistically significant positive correlation between B_pre and V_{E pre} in the subsample B_pre>2 (Table 3) seems to confirm that WOB_e is responsible for dyspnea in PE, as the greater the V_{E pre} value is, the greater the WOB_e. It is not clear, however, why changes in dyspnea, i.e., ΔB, are not associated with ΔV_E (Table 2), whereas ΔV_E is inversely correlated with B_pre (Table 3). Notably, no correlations between V_E and arterial gas tensions were found. The relatively high inverse correlation between P_{pl_ampl post} and V_{E pre} (Table 3) has not yet been explained.

4.1 Study limitation

The main limitation is related to the fact that neither P_{pl_ampl} in the contralateral hemithorax nor WOB_e is precisely known, particularly before TT; therefore, we have had to rely on laws of physics in the interpretation of results. The fact that a study is retrospective is usually considered a limitation. In our case, however, even a prospective study would not supply better data related directly to WOB_e because additional measurements in patients in such poor condition would be impossible for ethical reasons. Thus, the only true limitation related to the retrospective character of this study is that TT was terminated both because of symptoms (such as too intensive cough, for example,) and when there was no more fluid or when P_pl excessively fell or P_{pl_ampl} increased too much (hence the inverse correlation between P_{pl_ampl post} and V_w in Table 3), which means some lack of sample uniformity. Certainly, the feeling of dyspnea might also be affected by other factors, such as pain, stress or fear, in patients with serious comorbidities undergoing interventional procedure; however, these factors have not been precisely recorded and, therefore, their influence cannot be analyzed here. Nevertheless, the association between WOB and dyspnea was so strong that the influence of the other causes was insufficient to destroy the correlation between P_{pl_ampl post} and dyspnea (both B_post and ΔB).

5 Conclusion

Respiratory chemoreceptors seem to control breathing effectively in patients with malignant PE since arterial blood gases do not change significantly; therefore, changes in dyspnea after pleural fluid withdrawal can be associated with neither blood gas tensions nor V_E changes. Our results suggest that dyspnea in PE not associated with congestive heart failure can be related to a decrease in total lung compliance (due to the collapse of a part or the whole ipsilateral lung). This decrease forces a proportional increase in WOB_e to maintain the ventilation and arterial gas tensions required by respiratory chemoreceptors. The strong association between dyspnea changes and P_{pl_ampl post} (statistical significance, large effect size) confirms that if pleural fluid withdrawal leads to recruitment of the collapsed parts, then both WOB_e and dyspnea decrease; otherwise, WOB_e and dyspnea do not change or can even increase owing to the nonlinearity of lung compliance. Thus, although there can be various other causes of dyspnea in an individual patient with PE, an increase in WOB should also be taken into account. Based on the analyses presented here, we recommend pleural manometry during TT, which will enable to adjust the fluid removal rate to the observed value of P_{pl_ampl}.

Data availability statement

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author/s.

Ethics statement

The studies involving humans were approved by the Institutional Review Board Medical University of Warsaw (KB 105/2012). The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

Author contributions

TG: Conceptualization, Formal analysis, Investigation, Writing – original draft. AS: Formal analysis, Data curation, Writing – original draft. EG: Data curation, Project administration, Writing – review and editing. MZ-K: Data curation, Writing – review and editing. RK: Data curation, Supervision, Writing – review and editing.

Funding

The author(s) declared that financial support was received for this work and/or its publication. This study was supported by the National Science Center, Poland (grant N 2019/35/B/NZ5/02531) and Nalecz Institute of Biocybernetics and Biomedical Engineering, Polish Academy of Sciences.

Acknowledgements

The authors thank Marcin Michnikowski and Katarzyna Faber for their assistance during therapeutic thoracentesis.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. Generative AI tools were used exclusively for language editing and improving the clarity of the text.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fphys.2025.1664237/full#supplementary-material

Abbreviations

PE, pleural effusion; TT, therapeutic thoracentesis; Ppl, pleural pressure; AGT, arterial gas tensions; dB, ΔB, change in dyspnea, difference between the Borg Scores before and after therapeutic thoracentesis; PaO₂, arterial oxygen partial pressure; PaCO₂, arterial carbon dioxide partial pressure; P_{pl_ampl}, amplitude of pleural pressure changes related to breathing; P_{pl_ampl post}, P_{pl_ampl} after therapeutic thoracentesis; V_E, minute ventilation; WOB, work of breathing (from the point of view of physics it is actually power, not work); WOB_e, work of breathing which is required to overcome elastic forces; WOB_r, work of breathing which is required to is the resistive work; RR, respiratory rate; V_T, tidal volume; B_pre, the Borg Scores before therapeutic thoracentesis; B_post, the Borg Scores after therapeutic thoracentesis; V_w, the volume of withdrawn pleural fluid; rg, the Glass rank-biserial correlation.

References

Agustí A. G., Cardús J., Roca J., Grau J. M., Xaubet A., Rodriguez-Roisin R. (1997). Ventilation-perfusion mismatch in patients with pleural effusion: effects of thoracentesis. Am. J. Respir. Crit. Care Med. 156, 1205–1209. doi:10.1164/ajrccm.156.4.9612113

PubMed Abstract | CrossRef Full Text | Google Scholar

Antony V. B., Loddenkemper R., Astoul P., Boutin C., Goldstraw P., Hott J., et al. (2001). Management of malignant pleural effusions. Eur. Respir. J. 18 (2), 402–419. doi:10.1183/09031936.01.00225601

PubMed Abstract | CrossRef Full Text | Google Scholar

Beaudry R. I., Brotto A. R., Varughese R. A., de Waal S., Fuhr D. P., Damant R. W., et al. (2022). Persistent dyspnea after COVID-19 is not related to cardiopulmonary impairment; a cross-sectional study of persistently dyspneic COVID-19, non-dyspneic COVID-19 and controls. Front. Physiol. 13, 917886. doi:10.3389/fphys.2022.917886

PubMed Abstract | CrossRef Full Text | Google Scholar

Brandstetter R. D., Cohen R. P. (1979). Hypoxemia after thoracentesis: a predictable and treatable condition. JAMA 242, 1060–1061. doi:10.1001/jama.1979.03300100038019

PubMed Abstract | CrossRef Full Text | Google Scholar

Coccia C. B., Palkowski G. H., Schweitzer B., Motsohi T., Ntusi N. A. (2016). Dyspnoea: pathophysiology and a clinical approach. S. Afr. Med. J. 106 (1), 32–36. doi:10.7196/samj.2016.v106i1.10324

PubMed Abstract | CrossRef Full Text | Google Scholar

Fjaellegaard K., Koefod Petersen J., Alstrup G., Skaarup S., Frost Clementsen P., Laursen C. B., et al. (2024). Ultrasound in predicting improvement in dyspnoea after therapeutic thoracentesis in patients with recurrent unilateral pleural effusion. Eur. Clin. Respir. J. 11 (1), 2337446. doi:10.1080/20018525.2024.2337446

PubMed Abstract | CrossRef Full Text | Google Scholar

Gólczewski T., Stecka A. M., Grabczak E. M., Michnikowski M., Zielińska-Krawczyk M., Krenke R. (2025). Hemidiaphragm work in large pleural effusion and its insignificant impact on blood gases: a new insight based on in silico study. Front. Physiol. 16, 1539781. doi:10.3389/fphys.2025.1539781

PubMed Abstract | CrossRef Full Text | Google Scholar

Karetzky M. S., Kothari G. A., Fourre J. A., Khan A. U. (1978). Effect of thoracentesis on arterial oxygen tension. Respiration 36 (2), 96–103. doi:10.1159/000193932

PubMed Abstract | CrossRef Full Text | Google Scholar

Muruganandan S., Azzopardi M., Thomas R., Fitzgerald D. B., Kuok Y. J., Cheah H. M., et al. (2020). The pleural effusion and symptom evaluation (PLEASE) study of breathlessness in patients with a symptomatic pleural effusion. Eur. Respir. J. 55 (5), 1900980. doi:10.1183/13993003.00980-2019

PubMed Abstract | CrossRef Full Text | Google Scholar

Muruganandan S., Mishra E., Singh B. (2023). Breathlessness with pleural effusion: what do we know? Semin. Respir. Crit. Care Med. 44 (4), 502–508. doi:10.1055/s-0043-1769098

PubMed Abstract | CrossRef Full Text | Google Scholar

Parshall M. B., Schwartzstein R. M., Adams L., Banzett R. B., Manning H. L., Bourbeau J., et al. (2012). An official American thoracic society statement: update on the mechanisms, assessment, and management of dyspnea. Am. J. Respir. Crit. Care Med. 185 (4), 435–452. doi:10.1164/rccm.201111-2042ST

PubMed Abstract | CrossRef Full Text | Google Scholar

Petersen J. K., Fjaellegaard K., Rasmussen D. B., Alstrup G., Høegholm A., Sidhu J. S., et al. (2024). Patient-reported outcome measures in patients with and without non-expandable lung secondary to malignant pleural Effusion—A single-centre observational study. Diagn. (Basel) 14 (11), 1176. doi:10.3390/diagnostics14111176

PubMed Abstract | CrossRef Full Text | Google Scholar

Piggott L. M., Hayes C., Greene J., Fitzgerald D. B. (2023). Malignant pleural disease. Breathe (Sheff) 19 (4), 230145. doi:10.1183/20734735.0145-2023

PubMed Abstract | CrossRef Full Text | Google Scholar

Psallidas I., Yousuf A., Talwar A., Hallifax R. J., Mishra E. K., Corcoran J. P., et al. (2017). Assessment of patient-reported outcome measures in pleural interventions. BMJ Open Respir. Res. 4 (1), e000171. doi:10.1136/bmjresp-2016-000171

PubMed Abstract | CrossRef Full Text | Google Scholar

Ritter O., Noureddine S., Laurent L., Roux P., Westeel V., Barnig C. (2024). Unraveling persistent dyspnea after mild COVID: insights from a case series on hyperventilation provocation tests. Front. Physiol. 15, 1394642. doi:10.3389/fphys.2024.1394642

PubMed Abstract | CrossRef Full Text | Google Scholar

Skaarup S. H., Lonni S., Quadri F., Valsecchi A., Ceruti P., Marchetti G. (2020). Ultrasound evaluation of hemidiaphragm function following thoracentesis: a study on mechanisms of dyspnea related to pleural effusion. J. Bronchology Interv. Pulmonol. 27 (3), 172–178. doi:10.1097/LBR.0000000000000627

PubMed Abstract | CrossRef Full Text | Google Scholar

Stecka A. M., Gólczewski T., Grabczak E. M., Zieliński K., Michnikowski M., Zielińska-Krawczyk M., et al. (2018). The use of a virtual patient to follow changes in arterial blood gases associated with therapeutic thoracentesis. Int. J. Artif. Organs 41 (11), 690–697. doi:10.1177/0391398818793354

PubMed Abstract | CrossRef Full Text | Google Scholar

Taylor T. M., Radchenko C., Sanchez T. M., Shepherd R. W., Kang L., Shojaee S. (2021). The Impact of thoracentesis on postprocedure pulse oximetry. J. Bronchology Interv. Pulmonol. 28 (3), 192–200. doi:10.1097/LBR.0000000000000747

PubMed Abstract | CrossRef Full Text | Google Scholar

Thomas R., Jenkins S., Eastwood P. R., Lee Y. C., Singh B. (2015). Physiology of breathlessness associated with pleural effusions. Curr. Opin. Pulm. Med. 21 (4), 338–345. doi:10.1097/MCP.0000000000000174

PubMed Abstract | CrossRef Full Text | Google Scholar

Wijayaratne T., Yousuf A., Panchal R. (2024). Cardiac related pleural effusions: a narrative review. J. Thorac. Dis. 16 (2), 1674–1686. doi:10.21037/jtd-23-1731

PubMed Abstract | CrossRef Full Text | Google Scholar

Zielinska-Krawczyk M., Grabczak E. M., Michnikowski M., Zielinski K., Korczynski P., Stecka A., et al. (2018). Patterns of pleural pressure amplitude and respiratory rate changes during therapeutic thoracentesis. BMC Pulm. Med. 18 (1), 36. doi:10.1186/s12890-018-0595-7

PubMed Abstract | CrossRef Full Text | Google Scholar

Zielinska-Krawczyk M., Stecka A. M., Grabczak E. M., Michnikowski M., Zieliński K., Korczynski P., et al. (2022). Impact of therapeutic thoracentesis and pleural pressure changes on breathing pattern, dyspnea, lung function, and arterial blood gases. Pol. Arch. Intern. Med. 132 (4), 16185. doi:10.20452/pamw.16185

PubMed Abstract | CrossRef Full Text | Google Scholar