Serial evaluation of clinical, functional, and structural pulmonary changes in 10 dogs with leptospirosis

Authors: Carole Bringold, Ariane Schweighauser, Sandro Walther, Bianca Lassen-Schmidt, Pascale K Gendron, Thierry Francey

Abstract

Background

Leptospirosis is a worldwide zoonosis with multisystemic consequences, including the pulmonary hemorrhage syndrome associated with a high mortality rate in dogs.

Hypothesis/Objectives

Serial assessment of clinical, functional, and structural pulmonary changes in dogs with acute leptospirosis during the first 8 days of hospitalization.

Animals

Ten client-owned dogs diagnosed with leptospirosis.

Methods

Prospective, descriptive case series with daily clinical examination, arterial blood gas analyses on d1-4, d6, and d8 and thoracic computed tomography (CT) with pulmonary segmentation and lung densitometry analysis on d1, d4, and d8.

Results

Clinical respiratory impairment was observed in most dogs during the measurement period, including 1 dog with fatal pulmonary hemorrhage. Half of the dogs showed hypoxemia (arterial partial pressure of median, 56.6 mmHg; range, 49.6-67.8 mmHg) with increased alveolar-arterial PO2 gradient (A-a gradient) (37.9 mmHg [21.9-405.4 mmHg]) over the study period. Alveolar ventilation was only minimally affected (arterial partial pressure of carbon dioxide 35.2 mmHg [25.9-42.1 mmHg]). Quantitative CT analyses identified increased mean lung attenuation (MLA) in all dogs initially (−612 HU [−475 to −681 HU]) with progressive improvement in most dogs (d4: −646 HU [−583 to −746 HU]; −709 HU [−624 to −811 HU]). Whereas clinical score and CT changes improved progressively in most dogs, the blood gas results showed a fluctuating pattern.

Conclusions and clinical importance

The clinical pulmonary manifestations of these dogs with leptospirosis were characterized by hypoxemia, increased A-a gradient, and increased lung density. The clinical course seemed to more closely parallel the structural changes observed on CT rather than the functional alterations assessed with arterial blood gas analyses.

Introduction

Leptospirosis is a zoonosis of worldwide importance caused by pathogenic spirochetal bacteria of the genus Leptospira. The most common clinical manifestations are acute kidney injury (AKI), cholestatic hepatopathy, pulmonary hemorrhage, and hemostatic disorders.^1^^,^^2^ Leptospiral pulmonary hemorrhage syndrome (LPHS), also known as the severe pulmonary form of leptospirosis, is associated with high morbidity and mortality rates in dogs and humans.^3–5^ Its occurrence has been reported increasingly in recent years.^4^^,^^6^ The exact pathogenesis of LPHS remains unclear and seems to be multifactorial,^2^ including endothelial damage caused by leptospiral toxin,^7^^,^^8^ immunologic mechanisms with deposition of immunoglobulins and complement on the alveolar surface,^9^ or disseminated intravascular coagulopathy.^10–12^

Little is known about the progression and regression patterns of suspected pulmonary bleeding caused by LPHS, and functional consequences beyond clinical observation have not yet been described for dogs. Arterial blood gas (ABG) analysis is the most appropriate diagnostic tool to assess ventilation and oxygenation in the clinical setting and thus to diagnose and quantify alterations in pulmonary function.^13^ In a study including 24 humans with acute leptospirosis, 17% showed clinical signs of respiratory difficulties, 33% radiographic changes of the lung, 70% alveolar hyperventilation with hypocapnia, and 75% hypoxemia. The ABG analysis suggested pulmonary venoarterial shunts in affected pulmonary areas and the most severe hypoxemia was observed in patients with combined oliguric renal failure and pulmonary impairment.^14^ The value of ABG analysis as a real-time indicator of disease severity in dogs with LPHS, however, has not been evaluated. The qualitative pulmonary changes observed on computed tomography (CT) of the same study population have been previously reported.^15^ Although traditional visual assessment documents disease distribution and patterns, quantitative assessment objectively quantifies the extent of airspace filling, and thus provides objective, reproducible data regarding disease severity and progression.^16^

With this case series, we aimed to prospectively and serially describe the clinical, functional, and quantitative structural pulmonary changes observed in a small number of dogs with well-characterized acute leptospirosis.

Materials and methods

Detailed information about materials and methods is presented as the Supplementary Appendix. All data are presented as median [range] as summary statistics.

Study population and inclusion criteria

Client-owned dogs > 10 kg body weight diagnosed with leptospirosis at the Small Animal Clinic of the Vetsuisse Faculty, University of Bern, Switzerland, between June and September 2012 were evaluated. The same criteria for the definition of a case of leptospirosis were used as later defined in the 2023 Consensus Statement on Leptospirosis in Dogs from the American College of Veterinary Internal Medicine (ACVIM).^2^

Clinical examination, standard laboratory examinations, and treatment

A full physical examination was performed daily including a standardized assessment of the respiratory system. A subjective respiratory score was defined, assigning a score of 0 for dogs without evidence of increased respiratory effort, 1 for dogs with mildly (barely visible) increased effort, 2 for dogs with moderately (clearly visible) increased effort, and 3 for dogs with severely increased respiratory effort, in which oxygen supplementation or mechanical ventilation was indicated. Arterial hemoglobin oxygen saturation was assessed using pulse oximetry (SpO2). Initial laboratory evaluation (d0) included a CBC, serum biochemistry profile, coagulation profile, and urinalysis for non-anuric dogs. Treatment during hospitalization was based on the specific needs of the individual dogs and consistent with the later published Guidelines of the ACVIM Small Animal Consensus Statement on Leptospirosis.^17^

Arterial blood gas analysis

Arterial blood gas analyses were performed on d1-4, d6, and d8 of hospitalization. The specific methodology for ABG sampling is presented in the Supplementary File. Blood oxygenation was assessed using the arterial partial pressure of oxygen (PaO2) and PaO2/FiO2 ratio (with FiO2 = fraction of inspired oxygen) when data included dogs treated with supplemental oxygen. The alveolar-arterial PO2 gradient (A-a gradient; P(A-a)O2) was calculated to further differentiate among possible causes of hypoxemia. Alveolar ventilation was assessed using the arterial partial pressure of carbon dioxide (PaCO2).

Diagnostic imaging acquisition and analysis

Thoracic radiographs were performed within the first 24 h of presentation (d0) and thoracic CTs were performed under mild sedation and spontaneous breathing on d1, d4, and d8. Pulmonary tissue segmentation was performed using an automatic algorithm of the software application MeVisPULMO.^18^ For cases with severe lesions, manual correction using a 2-dimensional (2D) drawing tool was required. The software application MeVisPULMO provided a 3-dimensional (3D) reconstruction of the entire lung to assess total lung volume and mean lung attenuation (MLA), based on a density histogram analysis. Mean lung attenuation values between −700 and −850 HU were considered normal.^19^

Results

Study population, disease manifestations, and treatment

Ten dogs of 6 different breeds (Labrador Retriever [n = 3], Airedale Terrier [n = 1], Barbet [n = 1], Mastiff [n = 1], English Setter [n = 1], White Swiss Shepherd dog [n = 1], and 2 mixed breed dogs) were described. Nine dogs were male (8 intact, 1 neutered) and 1 dog was a spayed female. The median age at presentation was 4.1 years [6 months to 13 years] and median body weight 28.8 kg [12.1-74.4 kg]. These 10 dogs represent all cases presented and diagnosed with leptospirosis at the Small Animal Clinic of the Vetsuisse Faculty, University of Bern, Switzerland between June and September 2012 because all dogs met the inclusion criteria (described in the Supplementary Appendix).

The diagnosis of leptospirosis was confirmed in 9 dogs. Confirmation of the diagnosis could not be obtained because of early death from suspected pulmonary hemorrhage in 1 dog and this case was defined as probable. The diagnostic test results are provided as the Supplementary Table. The first clinical signs were observed by the owners a median of 3 days (2-7 days) before presentation. Most relevant signs at admission included lethargy (n = 10), inappetence to anorexia (n = 8), vomiting (n = 8), diarrhea (n = 9), oligoanuria (n = 8), and dyspnea (n = 7). During the study period, 6 dogs showed 3 (AKI, suspected pulmonary hemorrhage, and hemostatic disorders) and 4 dogs showed 2 (AKI and suspected pulmonary hemorrhage) of the 4 typical organ manifestations of leptospirosis.

Treatment during hospitalization included antimicrobial treatment, IV fluids, analgesics, antiemetics, gastric protectants and, if indicated, mild sedatives to minimize the risk of pulmonary bleeding caused by excessive activity and arousal level. Further treatment included oxygen supplementation and mechanical ventilation (n = 1), esophageal tube feeding (n = 6), hemodialysis (n = 8), and therapeutic plasma exchange (n = 1).

One dog was euthanized during the study period (d2) because of respiratory failure. Two dogs were euthanized later during hospitalization (d9 and d30) because of respiratory failure (n = 1) or persistent severe lethargy and general weakness without signs of respiratory impairment (n = 1). Two additional dogs were euthanized after discharge (d36 and d37) because of persistent severe azotemia (n = 1) or severe lethargy and weakness (n = 1).

Standard laboratory results

The most relevant laboratory findings are shown in Table 1. At the time of presentation, 1 dog was classified as AKI International Renal Interest Society (IRIS) grade III, 6 as grade IV, and 3 as grade V. Urinalysis was performed in 4 dogs on the day of admission (d0).

Clinical respiratory evaluation

Increased respiratory effort was observed in almost all study a respiratory score ≥ 1 was documented in 7 dogs on d1 and in 9 dogs at least once during the time of observation. One dog with severe dyspnea experienced massive hemorrhage from the airways, which was noted immediately after euthanasia on d2. Altogether, the respiratory manifestations improved after the first 2-3 days in most dogs as shown by decreasing respiratory score over time (Figure 1). Seven dogs had mildly or moderately decreased SpO2 at least once during the study period (Table 2).

Arterial blood gas results

Arterial blood gas results were available for the entire period of assessment (d1-d8) in 8 dogs. The FiO2 was 0.21 for all measurements, except for the ventilated dog on d2 with a FiO2 of 1.0. The ABG results are summarized in Table 2.

Five dogs showed mild hypoxemia and a mildly (n = 4) to markedly (n = 1) decreased PaO2/FiO2 ratio on d1. Three dogs became severely hypoxemic and 4 dogs had a markedly decreased PaO2/FiO2 ratio later during measurement period. The ventilated dog had a normal PaO2 but a markedly decreased PaO2/FiO2 ratio (207 mmHg) on d2. The A-a gradient was increased in all dogs with decreased PaO2 or PaO2/FiO2 ratio and in 1 dog with normal oxygenation indices. The lowest abnormal oxygenation indices, PaO2 and PaO2/FiO2, and the highest abnormal A-a gradients were measured mid-study, with medians on d4, d3.5, and d4, respectively. Severe day-to-day variation in PaO2/FiO2 was observed in all dogs (Figure 2). The SpO2 and PaO2 provided discordant results in the assessment of oxygenation status at several simultaneous measurements (Figure 3). In 21/54 (39%) measurements, either a decreased SpO2 was measured in normoxemic dogs (n = 6) or a normal SpO2 in hypoxemic dogs (n = 15).

Minimal alteration of alveolar ventilation was present in all dogs on least in 1 measurement. Transient hypercapnia (1-2 days) was observed in 3 dogs (41.7 mmHg; [40.7-42.1 mmHg]). Mild to moderate hypocapnia was observed in 24/55 measurements in 8 dogs (31.5 mmHg; [25.9-33.9 mmHg]). In 14 measurements, it was associated with hypoxemia, metabolic acidosis, or both. No association was observed in 10 measurements in which the cause therefore was suspected to be related to pain, stress, or pulmonary disease.

Diagnostic imaging findings

Thoracic radiographs were performed on 9 dogs within the first 24 h of admission (d0). They showed an unstructured interstitial (n = 5), a reticulonodular (n = 2), or a patchy alveolar (n = 2) lung pattern. Distribution of the pulmonary changes was either symmetric (n = 5) or asymmetric (n = 4) and accentuated in the caudodorsal lung fields in all 9 dogs. Minimal (n = 4) to mild (n = 2) pleural effusion was detected in 6 dogs.

Three serial CTs of the thorax were performed on 9 dogs and only the first CT was available in 1 dog. A representative example of serial lung density histograms and corresponding CTs of 1 dog are shown in Figure 4. Lung densitometry measurements of the first CT showed an increased MLA in all dogs (−612 HU; [−475 to −681 HU]). The second and third CTs showed increased MLA in 6/9 and 4/9 dogs, respectively (CT2: −646 HU [−583 to −746 HU]; CT3: −709 HU [−624 to −811 HU]). Decreasing MLA over time was observed in 8 dogs (Figure 5). The total lung volume over the 3 CTs remained within 10% of baseline in 5 dogs, decreased in 2 dogs (−13%, −22%), and increased in 2 dogs (+21%, +36%).

Discussion

We describe the clinical, functional, and structural pulmonary abnormalities in 10 dogs with leptospirosis. Respiratory impairment was observed in most dogs. Although most of them showed clinical respiratory improvement within the first few days, 2 dogs were euthanized as a direct consequence of respiratory failure (1 dog during and 1 dog after the study period).

The main feature of the ABG analyses was hypoxemia in approximately half of the study population, ranging from mild changes to results consistent with acute lung injury (<300 mmHg).^20^ Hypoxemia was consistently associated with an increased A-a gradient. Hypoxemia with an increased A-a gradient reflects venous admixture,^13^ the main cause being low ventilation-perfusion regions, but also right-to-left shunts, small airway and alveolar collapse, or impaired diffusion.^21^^,^^22^ Pulmonary venoarterial shunts in impaired lung areas have been suspected in humans with LPHS, because of an increased pulmonary shunt fraction (Qs/Qt ratio).^14^ The response to mechanical ventilation with 100% oxygen supplementation with increased PaO2 (207 mmHg) and markedly decreased PaO2/FiO2 ratio (207 mmHg) in 1 dog suggests low ventilation-perfusion regions, small airway and alveolar collapse, or impaired diffusion as the primary pathophysiologic mechanism for hypoxemia at least in this dog.^13^ Invasive assessment of the pulmonary shunt fraction was not possible in these client-owned dogs with respiratory impairment.

Unlike other studies reporting a good correlation between SpO2 and PaO2 in humans and dogs,^23^ both measurement methods did not identify similar degrees of hypoxemia in our study. Therefore, pulse oximetry does not seem to be a reliable surrogate for ABG analyses in dogs with leptospirosis and possible pulmonary involvement. Similar observations reported low correlation between SpO2 and PaO2 and poor sensitivity for the detection of hypoxemia in dogs breathing room air.^24^ Reasons for this discrepancy may include measurement artifacts, altered peripheral perfusion, hypothermia, anemia, variable shifts in the oxygen dissociation curve, or interferences in light transmission and detection because of drugs or unspecified molecules accumulated in these dogs with multiple abnormalities.^24^

Impaired alveolar ventilation was observed in approximately half of all measurements but it was typically very mild and probably of minimal clinical relevance. Hyperventilation was observed in most dogs at least once during the measurement period, similar to the 70% reported for human patients with LPHS.^14^ However, hyperventilation is a nonspecific clinical sign and could be caused by other factors than LPHS, including response to metabolic acidosis or hypoxia, or sympathetic stimulation caused by pain, stress, or anxiety.

Radiographic changes at the time of admission were visible in all dogs for which thoracic radiographs were available and were compatible with the typical pattern reported for dogs with LPHS, characterized by a caudodorsally accentuated interstitial lung pattern. Previous studies describe abnormal lung patterns in 9%-70% of the dogs with leptospirosis.^3^^,^^25^^,^^26^ Quantitative assessment of the pulmonary parenchyma showed increased MLA in all dogs on d1, decreasing in most dogs over time, concurrent with the previously described improvement of most qualitative pulmonary CT findings.^15^ The MLA decreased to normal in half of the dogs until d8 but minimal to mild qualitative structural changes of the pulmonary parenchyma were still visible in all dogs at this time.^15^ Increased lung density generally can be caused by inflammatory, hyperplastic, or neoplastic infiltrates, hyperemia, decreased volume of air in the alveoli (eg, exhalation, atelectasis, or hypoinflation caused by obesity or pleural disorders), or replacement of air by fluid or cells in air spaces.^27^ Based on the qualitative description of the CT findings of our study patients, it can be assumed that the increase in pulmonary density was mostly caused by pulmonary hemorrhage and alveolar edema but, because histology was not available, their relative importance remains unknown.^15^ The MLA does not provide any information on the distribution of the lesions and can be similar in cases with mild but diffuse lesions and in cases with focally severe lesions. Because no contrast study was performed, the possibility of small pulmonary thrombi could not be ruled out. The fact that total lung volume remained stable or decreased in most dogs rules out the possibility that decreasing mean lung density was caused by increasing lung volume.

Limitations of our study include the grading of the respiratory impairment, that was based on a subjective, nonvalidated respiratory score. It is, therefore, difficult to compare with other studies and might have low reproducibility. Furthermore, the use of sedatives and analgesics may have influenced the depth and frequency of respiration. Most other limitations are related to the clinical nature of the study. The CT acquisition technique was without breath-holding and, therefore, breathing artifacts could not be avoided. Furthermore, this technique could have affected lung density because it is influenced by the level of inspiration.^28^ For this reason, results of the quantitative CT analyses might have lower reproducibility than results generated using breathholding acquisition techniques. However, breathholding acquisition techniques can only be performed under general anesthesia in dogs. Another limitation is the lack of histological examination in all dogs.

In conclusion, most dogs in the study had clinical respiratory signs, increased lung density, and oxygenation disturbances suggestive of hypoxemia and venous admixture as the most likely reason for oxygenation impairment. These data further suggest that caution is required when respiratory assessment is limited to a clinical assessment or SpO2 because hypoxemic dogs might not be recognized. An ABG analysis is, therefore, strongly recommended. Furthermore, thoracic CT imaging identified structural abnormalities of the lung parenchyma consistent with LPHS^15^ even in dogs without clinical respiratory impairment or with normal arterial oxygenation. It therefore provides complementary information to the clinical and functional assessments of dogs with leptospirosis. Whereas the quantitative CT analysis seemed to best reflect the progressive clinical improvement in these dogs, the ABG analysis (despite its daily variations), provided valuable information about the mechanism and degree of functional improvement.

Supplementary Material