Advertisement

Clinical characteristics of patients with pulmonary infarction – A retrospective review

Open ArchivePublished:April 17, 2018DOI:https://doi.org/10.1016/j.rmed.2018.04.008

      Highlights

      • Pulmonary infarction mortality is comparable to uncomplicated pulmonary embolism.
      • Pulmonary infarction disproportionately occurs in the lower lobes.
      • Hemoptysis was rare in patients who developed pulmonary infarcts.

      Abstract

      Background

      Pulmonary infarction is an infrequent complication of pulmonary embolism. Traditionally, it has been regarded as a sign of worse outcome because ischemia can only occur by the simultaneous failure of all oxygenation sources to the area of infarct, but supporting evidence is limited.

      Methods

      We identified 74 cases of pulmonary infarction over 5 years at a single academic center via review of radiographic reports. Contrast-enhanced chest CT scans were examined to confirm evidence of pulmonary infarction, and patient clinical characteristics and imaging results were studied.

      Results

      Survival to discharge was high (97%). Patients most commonly presented with dyspnea (69%), chest pain (46%), and swelling or pain in the lower extremities (31%), while underlying risk factors included history of malignancy (41%) and surgery within 30 days (24%). Many patients had concurrent cardiovascular (59%) and pulmonary disease (22%). Infarction disproportionately affected the lower lobes.

      Conclusions

      Survival after diagnosis of pulmonary infarction is comparable to uncomplicated pulmonary embolism, suggesting that outcome is not worse. While emboli occurred in multiple lobar sites, pulmonary infarction occurred most commonly in the lower lobes, suggesting unique underlying physiological mechanisms in pulmonary infarction development.

      Keywords

      1. Introduction

      Pulmonary infarction is an uncommon and poorly understood complication of pulmonary embolism (PE) [
      • Miniati M.
      Pulmonary infarction: an often unrecognized clinical entity.
      ]. While pulmonary embolism is common in the United States (incidence of ∼110 per 100,000) [
      • Wiener R.S.
      • Schwartz L.M.
      • Woloshin S.
      Time trends in pulmonary embolism in the United States: evidence of overdiagnosis.
      ], pulmonary infarction is not frequently identified [
      • Miniati M.
      Pulmonary infarction: an often unrecognized clinical entity.
      ]. Pulmonary tissue necrosis as a result of pulmonary embolism does not occur frequently, because of three non-redundant pulmonary oxygenation sources: the pulmonary arterial circulation (venous blood), the bronchial circulation (arterial blood), and direct alveolar ventilation via alveolar pores and/or airway anastomoses [
      • Tsao M.S.
      • Schraufnagel D.
      • Wang N.S.
      Pathogenesis of pulmonary infarction.
      ]. Pulmonary infarction presumably represents ischemia by simultaneous failure of all three oxygenation sources, suggesting significant pulmonary arterial occlusion extent or the occurrence of pulmonary embolism in an already compromised patient [
      • Parambil J.G.
      • Savci C.D.
      • Tazelaar H.D.
      • Ryu J.H.
      Causes and presenting features of pulmonary infarctions in 43 cases identified by surgical lung biopsy.
      ]. It is traditionally associated with worse outcomes due to poorer functional status [
      • Schraufnagel D.E.
      • Tsao M.S.
      • Yao Y.T.
      • Wang N.S.
      Factors associated with pulmonary infarction. A discriminant analysis study.
      ], but specific data addressing this assumption is limited.
      Dalen and colleagues proposed that pulmonary infarction was not related to the severity of pulmonary embolism (as judged by the size and extent of central clot), but rather, the degree of distal vessel occlusion [
      • Dalen J.E.
      • Haffajee C.I.
      • Alpert 3rd, J.S.
      • Howe J.P.
      • Ockene I.S.
      • Paraskos J.A.
      Pulmonary embolism, pulmonary hemorrhage and pulmonary infarction.
      ]. This would then suggest that pulmonary infarction may not correlate with mortality or clinical hemodynamic instability [
      • Cha S.I.
      • Shin K.M.
      • Lee J.
      • et al.
      Clinical relevance of pulmonary infarction in patients with pulmonary embolism.
      ]. Others have further postulated that some historically accepted risk factors (such as cardiovascular disease) may occur with less frequency in those that develop pulmonary infarction, because of a lack of collateral circulation development in peripheral lung regions when cardiovascular function is adequate [
      • Miniati M.
      • Bottai M.
      • Ciccotosto C.
      • Roberto L.
      • Monti S.
      Predictors of pulmonary infarction.
      ]. Few studies have actively excluded pulmonary infarction from statistics in acute pulmonary embolism outcomes, although isolated reports have been published [
      • Kirchner J.
      • Obermann A.
      • Stuckradt S.
      • et al.
      Lung infarction following pulmonary embolism: a comparative study on clinical conditions and CT findings to identify predisposing factors.
      ].
      We performed a retrospective analysis of pulmonary infarctions occurring at a large, urban, tertiary academic medical center in the United States, to describe the clinical presentation, risk factors, and features of this disorder, as well as describe the involved clot burden and location, that may be unique to pulmonary infarction. We hypothesized that patients with pulmonary infarction may not have a worse outcome than patients with uncomplicated pulmonary embolism in general.

      2. Materials & methods

      This study was conducted in accordance with the amended Declaration of Helsinki and approved by the Thomas Jefferson University Institutional Review Board under Control #17D.102 without explicit consent from the participants.
      We searched for the terms “pulmonary infarct” or “pulmonary infarction” on intravenous contrast-enhanced chest CT scans obtained within a five-year interval (between January 1, 2012 and December 31, 2016). Reports containing those terms separated by 3 or fewer words (e.g., “pulmonary ischemia or infarction”) were included to maximize the chance that all potential cases were identified.
      Of the 572 reports obtained, we automatically discarded cases using computerized text searching of the complete radiographic reports based upon the terms “no pulmonary infarct”, “no evidence of pulmonary infarct”, “no findings of pulmonary infarct”, and “no evidence of right heart strain or pulmonary infarct”. Of the 411 remaining reports, duplicate patients were manually removed. The resulting 269 reports were individually, manually examined to eliminate those with negative results (e.g., “no evidence of consolidation or opacity to correspond to a pulmonary infarct”), insufficient clinical documentation, or irrelevancy (e.g., “splenic infarct”).
      The radiographic CT images of 116 cases were then examined by a board-certified cardiothoracic radiologist (BS) who was blinded to the clinical characteristics of each patient to confirm the presence of infarction based the following criteria (adapted from criteria suggested by Revel and colleagues) [
      • Revel M.P.
      • Triki R.
      • Chatellier G.
      • et al.
      Is it possible to recognize pulmonary infarction on multisection CT images?.
      ]:
      The CT results were evaluated for location of the embolism, including the specific location in the main pulmonary artery trunk, main pulmonary arteries, lobar pulmonary arteries, segmental pulmonary arteries, and sub-segmental pulmonary arteries. The PE protocol at our center does not include CT scans of the lower extremities and the great veins of the abdomen due to concerns of increased radiation and contrast agent exposure, but ultrasonography of the lower extremities was obtained if it was performed within 3 days of or during admission.
      Patient charts were also examined for possible alternative explanation for the parenchymal opacities. The white blood cell count, neutrophil count, and blood/sputum cultures were analyzed as close to admission as possible, with a margin of 3 days before and after admission. Four patients were excluded as part of this analysis, because of positive blood cultures and suspected diagnosis of septic emboli. All four had either bacterial endocarditis or an indwelling vascular device. As definitive tissue biopsy is not considered the standard of care and occurs quite rarely, infarction was confirmed for every patient only through review of characteristic CT imaging findings in the absence of other possible explanation provided in clinical documentation.
      Clinical records from this cohort of 74 patients were then examined for data related to demographics, survival, length of clinical stay, symptoms at presentation, comorbidities, treatments, laboratory findings, and radiologic imaging characteristics of noted pulmonary embolism(s) and infarction(s).

      3. Statistical methods

      Frequency counts and percentages were reported to summarize categorical variables. Means ± standard deviations were reported to summarize continuous variables. Generalized linear logistic regression modeling, accounting for the correlation among pulmonary infarction patients having emboli in multiple lobar locations, was used to estimate odds ratios for characterizing and comparing the associations between the lobar location(s) of their emboli and their pulmonary infarction(s).

      4. Results

      Out of the 74 pulmonary infarction patients included in our study, 45% were male and 55% were female (Table 1). The average age was 55 ± 16 years. Most patients who were diagnosed with pulmonary infarction survived to discharge (97%), with a slight reduction in survival rates at 3 months and 6 months (93% and 88%, respectively), comparable to a recent published report for patients with pulmonary embolism with or without infarct (97% to discharge, 92% at 3 months, and 89% at 6 months) [
      • Ng A.C.
      • Chung T.
      • Yong A.S.
      • et al.
      Long-term cardiovascular and noncardiovascular mortality of 1023 patients with confirmed acute pulmonary embolism.
      ] (Table 2). The median length of stay was 8 days. One patient was excluded from calculations involving length of stay because of a 22-month inpatient admission while awaiting heart transplant. Patients most commonly presented with dyspnea (69%), chest pain (46%), and swelling or pain in the lower extremities (31%). Bilateral venous ultrasonography of the lower extremities was available for 51 patients, 29 of which were positive for deep venous thrombosis (DVT) (57%). Of the 29, most were femoropopliteal (83%), but a minority were subpopliteal (17%). At presentation, fever and hemoptysis were less common (11% and 4%, respectively). Syncope was quite rare, affecting only 1% of the cohort (Table 1).
      Table 1Clinical characteristics of patients with pulmonary infarction at presentation (n = 74).
      Total% or SD
      Age (mean ± SD yr)55±16
      Male33(45)
      Length of stay (mean ± SD days)8±11
      Chest pain34(46)
      Dyspnea51(69)
      Syncope1(1)
      Hemoptysis3(4)
      Fever8(11)
      Signs of deep vein thrombosis
      Swelling or pain in the lower extremities.
      23(31)
      a Swelling or pain in the lower extremities.
      Table 2Survival of patients in this study compared to historical controls.
      This study

      (n = 74)
      Kniffen Jr.

      et al. 1994
      Includes mortality from pulmonary embolism regardless of presence of infarction.


      (n = 7174)
      van Beek

      et al. 1997
      Includes mortality from pulmonary embolism regardless of presence of infarction.


      (n = 192)
      Scarvelis

      et al. 2010
      Includes mortality from pulmonary embolism regardless of presence of infarction.


      (n = 498)
      Ng

      et al. 2011
      Includes mortality from pulmonary embolism regardless of presence of infarction.


      (n = 1023)
      n%n%n%n%n%
      to discharge73976385793877899297
      3 months579393892
      6 months56881598390989
      a Includes mortality from pulmonary embolism regardless of presence of infarction.
      Comorbidities among patients are summarized in Table 3. Right heart strain (e.g., straightening of the interventricular septum) was noted on CT for 42% of patients. About 15% of patients experienced hypoxemia during their clinical course and required supplemental oxygen. The most common underlying risk factor in our case series was history of malignancy (41%), followed by surgery within 30 days (24%). Trauma or injury prior to presentation was not a common underlying risk factor for pulmonary embolization (3%). Over half of the patients (59%) had underlying cardiovascular disease, while a minority had concurrent pulmonary disease (22%).
      Table 3Comorbidities of patients with pulmonary infarction (n = 74).
      Total%
      Right heart strain31(42)
      Hypoxemia11(15)
      History of cancer30(41)
      Surgery within 30 days18(24)
      Trauma/injury within 30 days2(3)
      Cardiovascular disease44(59)
      Pulmonary disease16(22)
      Therapy data were available for 73 patients at admission and 72 at discharge. At diagnosis of PE, most patients were placed on either low-molecular-weight heparin (55%) or heparin (40%), although one was treated with argatroban and another with apixaban. Two patients received IVC filters instead of anticoagulation therapy due to bleeding concerns. At discharge, patients were split between warfarin (38%), low-molecular-weight heparin (28%), and direct oral anticoagulants (26%). Three patients were kept on fondaparinux and one on heparin. Two patients died before discharge, both of whom were treated with heparin only.
      Arterial and venous blood gas data were available for 22 patients (12 arterial and 10 venous), and two arterial samples were excluded as they were performed over 7 days since the patient's admission. Two patients presented with acidemia (pH < 7.35) and two presented with alkalemia (pH > 7.45). The lowest recorded pH was 7.32 and the highest was 7.53. The arterial PCO2 was, on average, 42 mm Hg at presentation. Patients were often placed on supplemental O2 early in their evaluation, and the average arterial PO2 was 101 mm Hg and O2 saturation was 94% (one patient was reported to have an arterial PO2 < 40 mm Hg and this was discarded as an accidental venous sample).
      Table 4 reveals the distribution of pulmonary emboli by artery location. While many patients had multiple pulmonary emboli, the majority manifested only one infarction (84%). Table 5 demonstrates the infarction characteristics with respect to the lobar locations affected by emboli. A total of 181 lobes were affected by emboli, among which 89 (49%) were associated to infarcts. The highest rates of infarcts per lobe were in the left lower lobe (54.8) and right lower lobe (72.9). The other lobar locations had infarct rates between 26.1 and 30.3. The odds of pulmonary infarction occurrence from emboli in the right lower lobe were more than 7-fold greater compared to the left upper lobe (OR = 7.61; 95% CI 2.48 to 23.41, p < 0.01). There was a significantly increased odds for the occurrence of pulmonary infarction occurring from emboli in the left lower lobe compared to emboli in the left upper lobe (OR = 3.43; 95% CI 1.10 to 10.66, p = 0.03). We also tested this same hypothesis using the right upper lobe as the reference lobe and found comparable results with both the right lower lobe (OR = 6.18; 95% CI 2.34 to 16.30, p < 0.01) and the left lower lobe (OR = 2.78; 95% CI 1.08 to 7.15, p = 0.03). The odds that a patient's pulmonary infarction occurred from emboli affecting the right versus left lung were not statistically significantly different (OR = 1.33; 95% CI 0.74 to 2.40, p = 0.34) (Table 6).
      Table 4Embolism locations in patients with pulmonary infarction (n = 74).
      Pulmonary arteryTotal%
      Main trunk4(5)
      Left main16(21)
      Right main22(28)
      Left upper lobar20(26)
      Left lower lobar33(42)
      Right upper lobar28(36)
      Middle lobar21(27)
      Right lower lobar40(51)
      Presence in segmental artery71(91)
      Presence in subsegmental artery71(91)
      Table 5Infarction characteristics and association statistics by lobar location in patients with pulmonary infarction.
      Lobar LocationAffected
      Number of lobar locations affected with embolism.
      InfarctsInfarct Rate
      Infarcts per embolism affected lobar location.
      Odds Ratio
      Odds of an embolism affected location having an infarct relative to the odds in the left upper lobe.
      95% CI
      95% confidence interval (CI) for the odds ratio.
      p
      Left upper23626.11.00reference
      Left lower422354.83.43(1.10, 10.66)0.03
      Right upper331030.31.23(0.38, 3.95)0.73
      Middle24729.21.17(0.32, 4.21)0.81
      Right lower594372.97.61(2.48, 23.41)<.01
      Total18189
      a Number of lobar locations affected with embolism.
      b Infarcts per embolism affected lobar location.
      c Odds of an embolism affected location having an infarct relative to the odds in the left upper lobe.
      d 95% confidence interval (CI) for the odds ratio.
      Table 6Infarction characteristics and association statistics by lobar location in patients with pulmonary infarction (left vs. right lung).
      LobeAffected
      Number of lobar locations affected with embolism.
      InfarctsInfarct Rate
      Infarcts per embolism affected lobe.
      Odds Ratio
      Odds of an embolism affected lobe having an infarct relative to the odds in the left lobe.
      95% CI
      95% Confidence interval (CI) for the odds ratio.
      p
      Left652944.61.00reference0.34
      Right1166051.71.33(0.74, 2.40)
      Total18189
      a Number of lobar locations affected with embolism.
      b Infarcts per embolism affected lobe.
      c Odds of an embolism affected lobe having an infarct relative to the odds in the left lobe.
      d 95% Confidence interval (CI) for the odds ratio.
      Follow-up CT scans were performed in 32 patients, over a range of 1–69 weeks. It should be noted that because the performance of CT scans was not scheduled according to any discernable interval, the average time to resolution cannot be accurately provided. However, 10 patients had CT scans performed before complete resolution. In these patients, persistence of pulmonary infarction was noted at 10 weeks, on average, and ranged from between 1 and 21 weeks.

      5. Discussion

      Estimates of survival after pulmonary embolism without regard to the presence or absence of infarction vary (78–97% before discharge [
      • Ng A.C.
      • Chung T.
      • Yong A.S.
      • et al.
      Long-term cardiovascular and noncardiovascular mortality of 1023 patients with confirmed acute pulmonary embolism.
      ,
      • Kniffin Jr., W.D.
      • Baron J.A.
      • Barrett J.
      • Birkmeyer J.D.
      • Anderson Jr., F.A.
      The epidemiology of diagnosed pulmonary embolism and deep venous thrombosis in the elderly.
      ,
      • van Beek E.J.
      • Kuijer P.M.
      • Buller H.R.
      • Brandjes D.P.
      • Bossuyt P.M.
      • ten Cate J.W.
      The clinical course of patients with suspected pulmonary embolism.
      ], 92% at 3 months [
      • Ng A.C.
      • Chung T.
      • Yong A.S.
      • et al.
      Long-term cardiovascular and noncardiovascular mortality of 1023 patients with confirmed acute pulmonary embolism.
      ], 83–89% at 6 months [
      • Ng A.C.
      • Chung T.
      • Yong A.S.
      • et al.
      Long-term cardiovascular and noncardiovascular mortality of 1023 patients with confirmed acute pulmonary embolism.
      ,
      • van Beek E.J.
      • Kuijer P.M.
      • Buller H.R.
      • Brandjes D.P.
      • Bossuyt P.M.
      • ten Cate J.W.
      The clinical course of patients with suspected pulmonary embolism.
      ], and 61–81% at 1 year [
      • van Beek E.J.
      • Kuijer P.M.
      • Buller H.R.
      • Brandjes D.P.
      • Bossuyt P.M.
      • ten Cate J.W.
      The clinical course of patients with suspected pulmonary embolism.
      ]). In our cohort of carefully defined patients with pulmonary infarction, survival to discharge was 97%, survival to 3 months was 93%, and survival to 6 months was 88%. This finding is comparable to historical cases, and suggests that patients with pulmonary infarction may not have worse outcomes compared to patients without infarction. Miniati and colleagues propose that the development of infarcts happens in younger and healthier subjects due to the lack of efficient collateral circulation to pulmonary tissues, which presumably develops over many years [
      • Miniati M.
      • Bottai M.
      • Ciccotosto C.
      • Roberto L.
      • Monti S.
      Predictors of pulmonary infarction.
      ].
      Ng and colleagues presented some baseline characteristics in their investigation of the long-term cardiovascular and all-cause mortality in 1023 patients with confirmed pulmonary embolism. Their median length of stay was 7 days, compared to 8 days in our cohort. Documented deep vein thrombosis was noted in 18% of their patients, while we noted the presence of DVT in 59% of those evaluated with ultrasound (26/51 patients). History of malignancy was present in 22% of their patients, compared to our 41%. While we considered hypertension and hyperlipidemia as representing cardiovascular disease (59%), Ng and colleagues separated these two risk factors from their counts, resulting in a 31% rate of hypertension, 14% rate of hyperlipidemia, and 44% rate of cardiovascular disease (e.g., ischemic heart disease, stroke, and valvular heart disease). Chronic pulmonary disease was noted in 14% of these patients, compared to 22% in our series [
      • Ng A.C.
      • Chung T.
      • Yong A.S.
      • et al.
      Long-term cardiovascular and noncardiovascular mortality of 1023 patients with confirmed acute pulmonary embolism.
      ].
      This large series by Ng and colleagues was published before the availability of direct oral anticoagulants. Of their patients who survived to discharge, the vast majority were treated with warfarin (95%) and the remainder with enoxaparin only (5%) [
      • Ng A.C.
      • Chung T.
      • Yong A.S.
      • et al.
      Long-term cardiovascular and noncardiovascular mortality of 1023 patients with confirmed acute pulmonary embolism.
      ]. Our series spanned 5 years and the introduction of direct oral anticoagulants into clinical practice. Nearly all of our patients were treated initially with heparin or low-molecular-weight heparin, and while the largest portion was discharged on long-term warfarin therapy (38%), a significant number were placed on direct oral anticoagulants (26%).
      Because pulmonary infarctions are typically pleural-based, chest pain is expected to be common, and this was the case in our cohort (46%). As infarction leads to necrosis, bleeding or hemoptysis would be anticipated to occur commonly, but we did not find this to be the case (4%). Table 3 lists the underlying clinical causes of emboli; malignancy and surgery are both known to increase the risk of thrombosis [
      • Anderson J.A.
      • Weitz J.I.
      Hypercoagulable states.
      ]. Underlying cardiovascular disease was also common, consistent with the more historically-accepted hypothesis that hemodynamic failure predisposes the development of pulmonary infarction [
      • Tsao M.S.
      • Schraufnagel D.
      • Wang N.S.
      Pathogenesis of pulmonary infarction.
      ]. We also noted that 31% of patients with a follow-up CT continued to show evidence of pulmonary infarction for an average of 10 weeks (range, 1–21 weeks), suggesting that follow-up radiography to document clearing should not occur for at least 2 months.
      We noted a lower lobe predominance of pulmonary infarction in our cohort, in concordance with previously published studies [
      • Miniati M.
      • Bottai M.
      • Ciccotosto C.
      • Roberto L.
      • Monti S.
      Predictors of pulmonary infarction.
      ,
      • Kirchner J.
      • Obermann A.
      • Stuckradt S.
      • et al.
      Lung infarction following pulmonary embolism: a comparative study on clinical conditions and CT findings to identify predisposing factors.
      ]. We speculate that the lower lobes are compromised in alveolar oxygenation based upon lower ambient alveolar oxygen tension at the lung bases in the upright position, as well as an increased likelihood of basilar atelectasis due to the pleural pressure gradient (again in the upright lung) [
      • West J.B.
      Respiratory Physiology : the Essentials.
      ]. Other explanations underlying the observed phenomena may relate to the occlusion degree of emboli present in the lower lobes, as compared to the upper lobes.
      In our case series, we did identify a greater absolute number of emboli in the right lung relative to the left lung (116 right lobes affected versus 65 left lobes). Predilection of infarction for the right lung has been reported with multiple prior studies [
      • Kirchner J.
      • Obermann A.
      • Stuckradt S.
      • et al.
      Lung infarction following pulmonary embolism: a comparative study on clinical conditions and CT findings to identify predisposing factors.
      ,
      • He H.
      • Stein M.W.
      • Zalta B.
      • Haramati L.B.
      Pulmonary infarction: spectrum of findings on multidetector helical CT.
      ,
      • Ohtsubo M.
      Computerized tomography in pulmonary infarction.
      ], although no reason has yet been proposed regarding etiology. We speculate that this may be simply attributable to increased perfusion on the right side due to cardiac anatomy occupying the left thoracic space. However, once the greater prevalence of embolism in the right lung is taken into account, our results did not demonstrate emboli laterality to be associated with significantly increased likelihood of pulmonary infarction.
      One important limitation of this retrospective review is the use of historical controls to form conclusions. A survival comparison among all patients diagnosed with pulmonary embolism seen at our center would be ideal; however, the large number of emboli diagnosed in the same period made such an analysis unrealistic, and our main goal was focused on the characterization of pulmonary infarction. We also did not assess pulmonary function related to pulmonary infarction because of the difficulty in deciding the appropriate timing of testing compared to the development of an infarct.
      Patients dying from pulmonary embolism prior to diagnosis of pulmonary infarction were not included to our study; thus, our survival rates may be erroneously inflated. However, reports have suggested that symptoms and signs related to pulmonary infarction tend to manifest after 24 h or later [
      • Terry P.B.
      • Buescher P.C.
      Pulmonary infarction: in the beginning: the natural history of pulmonary infarction.
      ]. Therefore, early deaths after PE may not yet have had infarction occur given the requisite development time. We also excluded bacterial endocarditis or septic emboli from our study, as we distinctly wanted to focus upon venous thromboembolism as an entity separate from other emboli types. We are additionally unable to comment on the sensitivity and specificity of the radiographic criteria that we used to confirm the presence of pulmonary infarction by CT because there remains no gold standard for its identification beyond CT characteristics [
      • Bray T.J.
      • Mortensen K.H.
      • Gopalan D.
      Multimodality imaging of pulmonary infarction.
      ]. A recent retrospective review suggested that the reversed halo sign may strongly suggest pulmonary infarction, and a combination of radiographic and clinical factors was used to confirm the diagnosis, in similar fashion to this study [
      • Marchiori E.
      • Menna Barreto M.
      • Pereira Freitas H.M.
      • et al.
      Morphological characteristics of the reversed halo sign that may strongly suggest pulmonary infarction.
      ]. Of note, Parambil and colleagues published a series of infarction cases diagnosed via biopsy, which revealed necrotic tissue over an extended period of time [
      • Parambil J.G.
      • Savci C.D.
      • Tazelaar H.D.
      • Ryu J.H.
      Causes and presenting features of pulmonary infarctions in 43 cases identified by surgical lung biopsy.
      ]. Despite this impressive series, most clinicians do not rely upon biopsy for the diagnosis, as it is most typically established upon clinical and radiographic grounds. In our cohort, no patients were noted to undergo biopsy confirmation.
      Contrast-enhanced ultrasonography has recently been suggested as new method to estimate pulmonary embolism risk, with the detection of pulmonary infarction as one contributory factor [
      • Zhu R.
      • Ma X.C.
      Clinical value of ultrasonography in diagnosis of pulmonary embolism in critically ill patients.
      ]. Although it may have utility in certain cases, it is not yet widely performed, as it requires specific experience and expertise. By contrast, CT is universally available. Incidentally, in our cohort, one ultrasound of the chest was performed for line placement, although no patients underwent the use of diagnostic ultrasound for the evaluation of pulmonary embolism and/or pulmonary infarction.
      In our study, pulmonary infarctions were reported by description in the final interpretations of intravenous contrast-enhanced chest CT scan exams. Our investigation referred to one time point of the venous thromboembolic disease assessment, and did not include the continued surveillance of patients with recently diagnosed venous thromboembolism for the subsequent development of infarctions. We also did not include other imaging tests (such as chest X-ray, magnetic resonance imaging, ventilation-perfusion scan, or conventional pulmonary angiography), and it is therefore possible that we underestimated the true incidence of infarctions related to pulmonary embolism. We believe these unavoidable limitations have minimal negative impact upon the findings of the study, as the conduct of the study reflects directly the current practice patterns underlying diagnosis and management of patients with pulmonary venous thromboembolic disease.

      6. Conclusions

      We describe the characteristics of a cohort of 74 patients with pulmonary infarction over a 5 year period at our academic medical center in Philadelphia. The survival rates to discharge, at 3 months, and at 6 months were 97%, 93%, and 88%, respectively, and were not reduced compared to recently published pulmonary embolism mortality outcomes. Chest pain was present in almost half our cases, while hemoptysis was rare. Pulmonary infarction disproportionately affected the lower lobes in our cohort. While we speculate as to possible explanations for the increased odds of pulmonary infarcts occurring in the lower lobes, further study would be required to gain a more detailed understanding of the complex pathophysiological mechanisms in the lung where pulmonary alveolar tissue receives oxygen via three distinct sources.

      Conflicts of interest

      No authors report any potential conflicts of interest.

      Acknowledgements

      Author contributions: TC and GCK had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. TC, BS, WBL, and GCK were responsible for study conception and design. TC and BS were responsible for data acquisition. All authors were responsible for data analysis and interpretation, while SWK performed statistical analyses. TC was responsible for drafting the manuscript. All authors contributed revisions to the final manuscript.
      Other contributions: The authors are grateful to Suzanne Adams and Melissa McCarey at the Jefferson Clinical Research Institute for their support.

      References

        • Miniati M.
        Pulmonary infarction: an often unrecognized clinical entity.
        Semin. Thromb. Hemost. 2016; 42: 865-869https://doi.org/10.1055/s-0036-1592310
        • Wiener R.S.
        • Schwartz L.M.
        • Woloshin S.
        Time trends in pulmonary embolism in the United States: evidence of overdiagnosis.
        Arch. Intern. Med. 2011; 171: 831-837https://doi.org/10.1001/archinternmed.2011.178
        • Tsao M.S.
        • Schraufnagel D.
        • Wang N.S.
        Pathogenesis of pulmonary infarction.
        Am. J. Med. 1982; 72: 599-606
        • Parambil J.G.
        • Savci C.D.
        • Tazelaar H.D.
        • Ryu J.H.
        Causes and presenting features of pulmonary infarctions in 43 cases identified by surgical lung biopsy.
        Chest. 2005; 127 (S0012-3692(15)34464-0): 1178-1183
        • Schraufnagel D.E.
        • Tsao M.S.
        • Yao Y.T.
        • Wang N.S.
        Factors associated with pulmonary infarction. A discriminant analysis study.
        Am. J. Clin. Pathol. 1985; 84: 15-18
        • Dalen J.E.
        • Haffajee C.I.
        • Alpert 3rd, J.S.
        • Howe J.P.
        • Ockene I.S.
        • Paraskos J.A.
        Pulmonary embolism, pulmonary hemorrhage and pulmonary infarction.
        N. Engl. J. Med. 1977; 296: 1431-1435https://doi.org/10.1056/NEJM197706232962503
        • Cha S.I.
        • Shin K.M.
        • Lee J.
        • et al.
        Clinical relevance of pulmonary infarction in patients with pulmonary embolism.
        Thromb. Res. 2012; 130: e1-5https://doi.org/10.1016/j.thromres.2012.03.012
        • Miniati M.
        • Bottai M.
        • Ciccotosto C.
        • Roberto L.
        • Monti S.
        Predictors of pulmonary infarction.
        Medicine (Baltim.). 2015; 94: e1488https://doi.org/10.1097/MD.0000000000001488
        • Kirchner J.
        • Obermann A.
        • Stuckradt S.
        • et al.
        Lung infarction following pulmonary embolism: a comparative study on clinical conditions and CT findings to identify predisposing factors.
        Röfo. 2015; 187: 440-444https://doi.org/10.1055/s-0034-1399006
        • Revel M.P.
        • Triki R.
        • Chatellier G.
        • et al.
        Is it possible to recognize pulmonary infarction on multisection CT images?.
        Radiology. 2007; 244 (doi: 244/3/875): 875-882
        • Ng A.C.
        • Chung T.
        • Yong A.S.
        • et al.
        Long-term cardiovascular and noncardiovascular mortality of 1023 patients with confirmed acute pulmonary embolism.
        Circ Cardiovasc Qual Outcomes. 2011; 4: 122-128https://doi.org/10.1161/CIRCOUTCOMES.110.958397
        • Kniffin Jr., W.D.
        • Baron J.A.
        • Barrett J.
        • Birkmeyer J.D.
        • Anderson Jr., F.A.
        The epidemiology of diagnosed pulmonary embolism and deep venous thrombosis in the elderly.
        Arch. Intern. Med. 1994; 154: 861-866
        • van Beek E.J.
        • Kuijer P.M.
        • Buller H.R.
        • Brandjes D.P.
        • Bossuyt P.M.
        • ten Cate J.W.
        The clinical course of patients with suspected pulmonary embolism.
        Arch. Intern. Med. 1997; 157: 2593-2598
        • Anderson J.A.
        • Weitz J.I.
        Hypercoagulable states.
        Crit. Care Clin. 2011; 27 (vii): 933-952https://doi.org/10.1016/j.ccc.2011.09.007
        • West J.B.
        Respiratory Physiology : the Essentials.
        ninth ed. Lippincott Williams & Wilkins, 2011
        • He H.
        • Stein M.W.
        • Zalta B.
        • Haramati L.B.
        Pulmonary infarction: spectrum of findings on multidetector helical CT.
        J Thorac Imaging. 2006; 21: 1-7https://doi.org/10.1097/01.rti.0000187433.06762.fb
        • Ohtsubo M.
        Computerized tomography in pulmonary infarction.
        Nihon Igaku Hoshasen Gakkai Zasshi. 1992; 52: 600-610
        • Terry P.B.
        • Buescher P.C.
        Pulmonary infarction: in the beginning: the natural history of pulmonary infarction.
        Chest. 2017; (S0012-3692(17)31251-5)
        • Bray T.J.
        • Mortensen K.H.
        • Gopalan D.
        Multimodality imaging of pulmonary infarction.
        Eur. J. Radiol. 2014; 83: 2240-2254https://doi.org/10.1016/j.ejrad.2014.07.016
        • Marchiori E.
        • Menna Barreto M.
        • Pereira Freitas H.M.
        • et al.
        Morphological characteristics of the reversed halo sign that may strongly suggest pulmonary infarction.
        Clin. Radiol. 2018; 73 (e7-503.e13. doi: S0009-9260(17)30555-X): 503
        • Zhu R.
        • Ma X.C.
        Clinical value of ultrasonography in diagnosis of pulmonary embolism in critically ill patients.
        J Transl Int Med. 2017; 5: 200-204https://doi.org/10.1515/jtim-2017-0034