Prognosis of primary pulmonary adenocarcinoma after surgical resection in small‐breed dogs: 52 cases (2005‐2021)

Abstract Background Tumor size is an important prognostic factor in lung cancer in dogs, and the canine lung carcinoma stage classification (CLCSC) recently has been proposed to subdivide tumor sizes. It is unclear if the same classification scheme can be used for small‐breed dogs. Objectives To investigate whether the tumor size classification of CLCS is prognostic for survival and progression outcomes in small‐breed dogs with surgically resected pulmonary adenocarcinomas (PACs). Animals Fifty‐two client‐owned small‐breed dogs with PAC. Methods Single‐center retrospective cohort study conducted between 2005 and 2021. Medical records of dogs weighing <15 kg with surgically resected lung masses histologically diagnosed as PAC were examined. Results The numbers of dogs with tumor size ≤3 cm, >3 cm to ≤5 cm, >5 cm to ≤7 cm, or >7 cm were 15, 18, 14, and 5, respectively. The median progression‐free interval (PFI) and overall survival time (OST) were 754 and 716 days, respectively. In univariable analysis, clinical signs, lymph node metastasis, margin, and histologic grade were associated with PFI, and age, clinical signs, margin, and lymph node metastasis were associated with OST. Tumor size classification of CLCS was associated with PFI in all categories, and tumor size >7 cm was associated with OST. In multivariable analysis, tumor size >5 cm to ≤7 cm and margin were associated with PFI, and age was associated with OST. Conclusions and Clinical Importance The tumor size classification of CLCS would be an important prognostic factor in small‐breed dogs with surgically resected PACs.


| INTRODUCTION
Primary lung cancer is a common malignancy in people 1 but is relatively rare in dogs. 2 Among lung tumors in dogs, primary pulmonary carcinomas (PPCs) are most common, and among PPC cases, pulmonary adenocarcinomas (PACs) are reported to account for 60%-80%. [3][4][5][6] Prognostic factors for lung cancer in people include histologic classification, 7-9 performance status, 9,10 age, 10 sex, 10,11 smoking, 9,12 molecular biology signatures, 13 and stage classification. The International Association for the Study of Lung Cancer recently has published a revised version of the tumor-node-metastasis stage groupings used for lung cancer in human patients. 14 Tumor size was further subdivided in the revised version, indicating that it is an important prognostic factor in lung cancer patients. Although a staging system for lung tumors in dogs was proposed by the World Health Organization in 1980, this staging system did not include an assessment of tumor size. However, tumor volume [15][16][17] and longest tumor diameter 6,17 have been reported to be associated with survival in dogs with PPC, suggesting that tumor size might be an important prognostic factor in dogs. In 2020, the canine lung carcinoma stage classification (CLCSC) based on stage classification in humans was proposed. 18 In the CLCSC, tumor size was subdivided into 4 categories: ≤3 cm, >3 cm to ≤5 cm, >5 cm to ≤7 cm, and >7 cm.
Each CLCSC component, tumor size, lymph node metastasis, and distant metastasis was associated with survival time in PPC of dogs, suggesting that the CLCSC is an important prognostic factor for dogs with PPC. Subsequently, a relationship between CLCSC and survival also was observed in a larger population of dogs. 6 However, unlike humans, dog breeds have large differences in body size. In the 2 reports that evaluated CLCSC, the median weight of the dogs was >20 kg. 6,18 To our knowledge, the prognosis of surgically resected PACs in small-breed dogs has not been clarified, and whether the tumor size classification of CLCS could be applied to PAC of small-breed dogs has not been examined. We hypothesized that the tumor size classification of CLCS would have a prognostic impact on PAC in dogs weighing <15 kg. Therefore, we aimed to investigate whether the tumor size classification of CLCS is prognostic for survival and progression outcomes in small-breed dogs with surgically resected PACs. We also explored other potential risk factors related to progression outcomes and survival times.

| Case selection
A retrospective cohort study was performed at the Japan Small Ani- of death was unknown) also were collected. When follow-up data were limited, medical records were obtained from the primary veterinarian, or updates were obtained by telephone from the owner. The presence of clinical signs was defined as the initial presentation with any potential clinical sign associated with a lung tumor, including coughing, tachypnea, dyspnea, lethargy, exercise intolerance, anorexia, or hemoptysis, whereas no clinical signs at diagnosis were defined as incidental identification of a lung tumor on imaging with no clinical signs associated the lung tumor. 18 The CLCSC system was used for staging and tumor size classification. 15 The anatomic location and longest tumor diameter of PAC and lymph node status were recorded according to the CT data that was reassessed by a single diagnostic radiologist (A.T). If the CT data were incomplete, the data were supplemented with the original diagnostic imaging reports. If heterogeneous contrast-enhanced lymph node patterns were seen, the longest lymph node diameter was >12 mm, or a lymph node/ vertebral body ratio was >1.05 on CT, lymph node metastasis was suspected. 19 Diagnosis, lymph node status, presence, and location of intrapulmonary nodules separate from the primary tumor, tumor invasion into specific organs, and surgical margins were recorded from the original histopathologic reports. Because the histologic pattern and grade of PAC were not evaluated in the original histopathology report, those were reassessed by 2 pathologists (Y.K and K.N). The PAC histologic grade was based on the PPC grading system. 20 Complete resection was defined as a lack of tumor cells at the margins. The CT scanning and reconstruction parameters used at the JSACC for the 4MDCT scanner were 120 kVp, 120 mA, rotation time = 0.75 seconds, slice thickness = 1-2 mm, beam pitch = 1.374, reconstruction interval = 1-2 mm, and a standard algorithm, whereas for the 80MDCT scanner these were 120 kVp, 350 mA, rotation time = 0.5 seconds, slice thickness = 0.5 mm, beam pitch = 1.388, reconstruction interval = 0.5 mm, and a standard algorithm. Iopamidol (Oypalomin; Fuji Pharma) was used as the contrast agent for CT. This contrast agent was administered at a dosage of 2-2.5 mL/kg (600-750 mg/kg) for >10 seconds via the cephalic vein, and images were taken 120 seconds after administration.

| Statistical analysis
Progression-free interval (PFI) was defined as the time from surgery to progressive disease (PD). Progressive disease was defined as recurrence, dissemination, metastasis, or malignant pleural effusion noted postoperatively and was confirmed when the diagnosis was based on imaging, cytologic findings or histologic findings. Dogs were censored from PFI analysis if no detectable PD was noted at the last follow-up or death. Overall survival time (OST) was defined as the time from surgery to death from any cause. Dogs that were lost to follow-up or remained alive were censored from OST analysis at the date of the last contact.
The Kaplan-Meier method was used to estimate and depict PFI and OST. Interval or survival between groups was compared using the log-rank method. Age, weight, clinical signs at the initial examination, lymph node status (N0 vs N1, 2), completeness of tumor resection (margin), histologic grade (grade 1-3), histologic patterns (papillary vs other), tumor size classification of CLSC (≤3 cm, >3 cm to ≤5 cm, >5 cm to ≤7, or >7 cm), and adjuvant chemotherapy were analyzed as variables to identify potential risk factors associated with PFI and OST. These data were reported as median and interquartile range (IQR). Lymph node metastasis was evaluated by histopathologic examination, and as in previous reports, for dogs in which lymph node resection or biopsy was not performed, lymph node status was recorded as N0. 6,18,20 Dogs that died perioperatively were excluded from the analysis of PFI and OST of adjuvant chemotherapy. To determine potential risk factors associated with PFI and OST, univariable and multivariable analyses were performed using the Cox proportional hazards regression model. Multivariable analyses included variables for which P < .05 was obtained in univariable analyses. These results were presented as hazard ratio (HR) and 95% confidence intervals (CI). The potential risk factors between the dogs receiving or not receiving adjuvant CBDCA were compared using Fisher's exact test.
P values <.05 were considered significant. Statistical analysis was performed using EZR software (Saitama Medical Centre, Jichi Medical University, Saitama, Japan).

| Patient population
Fifty-two client-owned small-breed dogs met the criteria for inclusion.

| Diagnostic testing and staging
Contrast-enhanced thoracic CT was performed in all cases. Three patients had CTs performed at other institutions. The CT data were partially or totally lost for 2 dogs and could not be reviewed, and thus the description from the original diagnostic imaging report was used.
Those reports described the anatomic location and the longest diameter of the mass but did not describe the details of the lymph nodes.   Table 1).
Six dogs showed no improvement in coughing after surgery, and the suspected causes of coughing (including causes affecting >1 dog) were tracheal or bronchial-related disease (n = 3), myxomatous degenerative mitral valve disease (n = 3), carcinomatous pleuritis (n = 1), pneumonia (n = 1), and pneumothorax (n = 1). The remaining 3 dogs, in which improved coughing could not be confirmed, died during the perioperative period, and therefore it was not possible to assess whether coughing had improved.

| Histology and CLCSC staging
The primary PAC was completely resected in 45 dogs (86.5%). The Lymph node metastases were observed in 9 of 37 dogs (24.3%), of which 6 dogs had suspected lymph node metastases on CT. Thirteen, 28, and 11 dogs were classified as CLCSC stages 1, 2, and 3, respectively. No dogs were classified as stage 4. between dogs that did and did not receive chemotherapy (Table 2).

| Chemotherapy
One dog did not receive adjuvant CBDCA but was given thalidomide postoperatively. The affected dog had multiple lung metastases on Day 193 and died on Day 220.

| Outcomes
Regarding outcomes at the time of analysis, 41 dogs had died (78.8%), 6 were still alive (11.5%), and 5 were lost to follow-up (9.6%).  Acute respiratory distress syndrome because of noncardiogenic pulmonary edema of death in 6 dogs was unknown. Non-neoplastic diseases included congestive heart failure (n = 2), seizures (n = 2), immune-mediated hemolytic anemia (n = 1), obstructive hydrocephalus (n = 1), pneumonia (n = 1), chronic kidney disease (n = 1), and protein-losing enteropathy (n = 1).     17,24,25 Although it is difficult to identify the specific reasons for long-term survival, population characteristics defined by inclusion criteria (ie, restricted to PACs and small-breed dogs) and the absence of cases with distant metastases were considered factors for long-term survival in our study. Additionally, age was associated with OST in both univariable and multivariable analyses, with a 1.28-fold increased risk of death for each additional year. In general, the number of dogs with comorbidities increases with age.

| Outcome after disease progression
The median age at initial presentation in our study was 12 years, and the dogs survived for a relatively long time after surgery, which suggests that the risk of death from diseases other than PAC increases with age and that age may have been associated with OST.
Surgical margins were associated with PFI in both univariable and multivariable analyses in our study. Previous studies have reported that surgical margins of PPCs or primary pulmonary neoplasia (PPN) are associated with OST 6,17,18 and tumor-specific survival 17,18 in univariable analysis but not in multivariable analysis. 6,18 These reports suggested that increased tumor size and invasiveness and location within the lung lobes (hilar and peripheral) may have been associated with surgical margins and influenced the results of multivariable analysis. 6,18 We only investigated some of those previously reported factors, making it difficult to determine the confounding factors between surgical margins and the other factors used in our study. All 7 dogs in our study with incomplete margins were found to have PD postoperatively, and the PFI was significantly shortened. Therefore, when surgical margins are incomplete, postoperative progression should be carefully monitored, and adjuvant therapy should be considered.
The lymph node metastatic rate determined by histology in our study was 24.3%, which was comparable to previously reported lymph node metastatic rates ranging from 22.4% to 32.7% for PPCs. 16,18,24 Previous reports found that N stage affects survival in PPCs or PPNs. 16,18,20,24,25 In our study, N stage was associated with PFI and OST in univariable analysis but not in multivariable analysis. However, T A B L E 3 Univariable Cox regression analysis of potential risk factors for the progression-free interval and overall survival time in 52 small-breed dogs with surgically resected pulmonary adenocarcinomas. Our study had some limitations. First, the study was conducted at a single institution with a relatively small sample size. Second, because of the retrospective study design, some diagnostic methods were incomplete; The CT equipment was modified during the study, and some data was lost; thus, there was incomplete reassessment of CT images. The distance between the surgical margins and the tumor on histologic tissue sections was not evaluated, and therefore the prognostic impact could not be assessed. Third, the PD evaluations were problematic; 9 dogs were diagnosed with PD using only thoracic radiography to identify lung masses. In these dogs, no cytologic or histologic assessments of the lung masses were performed. We also did not perform necropsies on these dogs. It is possible that the lung masses diagnosed as PD were diseases other than PAC. Routine monitoring was recommended, but the follow-up schedule was not standardized. Fourth, because of the long study period, treatment methods were not consistent; surgical techniques for lobectomy were changed (eg, from suture ligation to Endo GIA Stapler) and rescue therapy options after PD were added (eg, TOC).
In conclusion, the tumor size classification of CLCS can be an important prognostic factor in small-breed dogs with surgically resected PACs. Given that postoperative PD was found in 46.2% of the dogs in our study, a randomized trial with a larger sample size is warranted to determine the benefit of adjuvant chemotherapy for dogs with potential risk factors.

ACKNOWLEDGMENT
No funding was received for this study.