Ann Thorac Surg 2004;78:1371-1381
© 2004 The Society of Thoracic Surgeons
Original article: cardiovascular
Reevaluation of Histomorphometric Analysis of Lung Tissue in Decision Making for Better Patient Selection for Fontan-Type Operations
Katsuhide Maeda, MDa,*,
Shigeo Yamaki, MDb,
Hideaki Kado, MDc,
Toshihide Asou, MDc,
Arata Murakami, MDa,
Shinichi Takamoto, MDa
a Department of Cardiothoracic Surgery, Graduate School of Medicine, University of Tokyo, Tokyo, Japan
b Japanese Research Institute of Pulmonary Vasculature, Shiroishi, Japan
c Department of Thoracic and Cardiovascular Surgery, Fukuoka-City Children's Hospital, Fukuoka, Japan
Accepted for publication February 3, 2004.
* Address reprint requests to Dr Maeda, Department of Cardiothoracic Surgery, Graduate School of Medicine, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8655, Japan
maedak-tky{at}umin.ac.jp
Presented at the Fortieth Annual Meeting of The Society of Thoracic Surgeons, San Antonio, TX, Jan 26–28, 2004.
 |
Abstract
|
|---|
BACKGROUND: The various types of cavopulmonary connection are occasionally unsuccessful even when the indications have been strictly fulfilled based on preoperative hemodynamic studies. We performed a detailed study of lung specimens from 60 patients who were judged to be candidates for the modified Fontan procedures based on a catheterizaton study in order to reevaluate the role of histologic studies of the pulmonary vasculature and to determine histologic criteria for the Fontan-type operation.
METHODS: We performed a histomorphometric analysis of specimens from 53 biopsy and 7 autopsy cases (0.5 to 23 years of age), with single ventricle physiology. Twenty-eight cases were treated with a bidirectional Glenn shunt (BDGS) and 32 cases were treated by means of total cavopulmonary connections (TCPC) with or without fenestration, on the basis of the clinical and hemodynamic findings. To evaluate the medial thickness of small pulmonary arteries (SPAs), we used a measurement, DR = 100 µm, that is unaffected by various degrees of vasoconstriction of the media or the age of the patients and that is representative of all SPAs in a section. Other variables, such as intimal lesions, SPA density, and the percentage of vessels containing a thrombus, were also measured.
RESULTS: There was a significant difference in DR = 100 µm values between the BDGS cases with good and bad outcomes at p = 0.0007 (8.9 ± 2.4 versus 13.4 ± 1.9 µm), and the cutoff point for the success of BDGS was 13.7 µm. The same was true of the TCPC cases at p less than 0.0001 (8.4 ± 1.7 versus 14.7 ± 1.5 µm), and the cutoff point was 11.6 µm. There were no significant differences in other histomorphometric variables. Investigation of the relationship to hemodynamic data revealed a correlation between DR = 100 µm and mean pulmonary artery pressure at p = 0.028. There were no statistically significant correlations between other variables.
CONCLUSIONS: The study revealed marked differences in DR = 100 µm values between patients with good and bad outcomes and provided DR = 100 µm cutoff points for BDGS and TCPC. In some cases, there were discrepancies between the results of the preoperative hemodynamic data and of the histomorphometric analysis; and because some patients were wrongly assessed based on clinical and hemodynamic criteria, histomorphometric study might be a useful method of supplementing the variety of clinical data used to determine the indications for this operation.
|
Introduction
|
|---|
Since Fontan and colleagues first described a novel surgical procedure for the repair of tricuspid atresia in 1971 [1], the Fontan procedure and modifications of it have been applied to many other types of congenital heart disease in which biventricular repair is not applicable [2, 3]. The selection of patients for these types of procedures is particularly important. Selection criteria based on the preoperative cardiac catheterizaton findings have been established by the work of Choussat and colleagues [4], and the indications have gradually been extended as major improvements in operative technique and postoperative intensive care have been made [3]. However, some patients who were ideal candidates from the standpoint of the hemodynamic data obtained in a catheter study have shown very low cardiac output and marked accumulation of extravascular fluid resistant to medical treatment postoperatively and did not survive [5, 6]. Some investigators have examined lung tissue from autopsy cases and reported medial hypertrophy of small pulmonary arteries (SPAs) in patients who had had normal pulmonary arterial pressure and resistance [7–9]. They suggested that these changes might be responsible for the high pulmonary arteriolar resistance in the postoperative period that leads to failure. Various studies have since been performed to ascertain the role of lung biopsy in determining the indications for the Fontan-type operation [10–13]; however, lung biopsy studies can be concluded to be of minimal benefit in decision making, because they are not strongly associated with variables that differentiate between cases with good and poor outcomes [9, 10].
To evaluate the role of morphohistometric studies more precisely, we used DR = 100 µm as a measurement that represents the medial thickness of all muscular SPAs in a histologic section, from intraacinar SPAs to preacinar SPAs. This measurement was first reported in 1971 by Suwa and colleagues and has been applied to various types of congenital heart disease to clarify the characteristics of the pulmonary vasculature and determine indications for congenital heart repair. Although its validity and usefulness have been confirmed in previous studies [14–16, 21, 28], no studies had used it to assess Fontan-candidates. The measurement DR = 100 µm is an integrated one, calculated from all SPAs in a section and indicates the general status of the media. Moreover, since medial thickness is estimated by measuring the area of the media and the length of the internal elastic membrane, it is unaffected by variability of medial vasoconstriction [18], and the normal value is almost constant after 5 months of age [16, 21, 28]. In addition to the study of medial thickening, other histomorphometric factors that may influence outcome, such as intimal lesions, percentage of SPAs containing a thrombus and the corrected density of SPAs, were also measured.
|
Material and Methods
|
|---|
Material
Between 1993 and 2003, we obtained 60 lung specimens from patients with single-ventricle physiology during cavopulmonary connections and take down or at autopsy from patients who died early (within 1 month) after cavopulmonary connection at 8 institutions in Japan (age range, 6 months to 23 years; mean 5.9 ± 5.2 years; 53 biopsy specimens and 7 autopsy specimens). The informed consent was obtained from each subject or from the family. To evaluate the influence of pulmonary arterial structure on outcome, we excluded specimens of patients whose death was definitely attributable to technical problems. Eight cases with fenestrated Fontan were included in our study (in some cases, fenestration was added after the total cavopulmonary connections [TCPC] operation). All clinical and hemodynamic data and outcomes are summarized in Table 1. In our study, "good outcome" means survival for more than 30 days postoperatively and "bad outcome" means death or TCPC (or bidirectional Glenn shunt [BDGS]) take down within 30 postoperative days.
All patients underwent preoperative cardiac catheterizaton and were considered to have indications for a BDGS or TCPC. Each institution made these judgements based on the clinical and hemodynamic findings. Moderate or severe atrioventricular regurgitation was detected in 8 BDGS cases and 6 TCPC cases, and valve repair was performed in 5 BDGS cases and 3 TCPC cases. In the biopsy cases, open lung biopsies, mostly of the right middle lobe, were initially performed during the procedure. No complications related to the lung biopsies, such as pneumothorax or bleeding, were observed in any of the biopsy cases. In the autopsy cases, specimens were also taken from the right middle lobe without injecting fluid, such as barium sulfate, into the pulmonary artery.
Tissue Preparation
One block from each lobe or lung was fixed in 10% formalin, and paraffin histologic sections were prepared. In each case, 30 step-sections at 50-µm intervals, each 3 µm thick, were prepared and modified Elastica-Goldner staining was performed [17]. In all lungs, we assessed (1) medial thickness of SPAs, (2) the severity of the intimal lesions, (3) corrected density of SPAs, and (4) percentage of SPAs containing a thrombus.
Measurements
Medial thickness (DR = 100 µm) was determined as follows. Variable vasoconstriction interferes with direct measurement of the medial thickness of arteries at biopsy or autopsy (Fig 1, A) [18–20]. Medial thickness (D) and radius (R), the distance from the center of an artery to the midpoint of the media, were measured by a previously reported histometric method [21]. All muscular SPAs in a section were scanned with a digitizing camera (Olympus PDMCII/OL, Olympus, Tokyo, Japan), and the images were fed into a Macintosh computer (Apple PowerMacG4, Apple Japan, Inc, Tokyo, Japan). To exclude the influence of various conditions of SPAs, the D and R were calculated in a state in which the internal elastic lamina was distended to form a circle, as follows. With the cross sectional surface area of the media expressed as S and the length of the internal elastic lamina as L, assuming a uniform thickness, D and R, were calculated by using the following equations (Fig 1, B):
All the measurements were performed using software NIH Image version 1.53. The values of R and D were obtained from at least 30 samples from each lobe or lung in each case and plotted on a logarithmic coordinate system. Linear regressions of R and D were performed, and the D value at R = 100 µm (DR = 100 µm) was calculated such that medial thickness could be compared among cases (Fig 1, C).
View larger version (47K):
[in this window]
[in a new window]
|
Fig 1. (A) A variety of degrees of vasoconstriction on the same biopsy section. Two intraacinar small pulmonary arteries (SPAs) are seen branching from one preacinar SPA. The medial hypertrophy of the left artery is judged to be more severe by direct measurement of the medial wall. (Left, 29%; right, 14% by the percent wall thickness method.) However, the quantitative analysis (B) shows that this is due to vasoconstriction of the media. When the internal elastic membrane is stretched, there is no difference in the medial thickness between the two SPAs. (B) Histometrical determination of the radius (R) and medial thickness (D). The constricted artery (left) has been reduced to the hypothetical state of a perfectly stretched internal elastic membrane (right). This method minimizes the influence of differing states of vasoconstriction during biopsy or at autopsy and makes comparisons among cases more reliable. (L = length of internal elastic lamina; S = cross-sectional surface area of media.) (C) Significant correlation between medial thickness (D) and radius (R) of SPAs on a logarithmic coordinate system in SPAs. The D value at R = 100 µm (D[R = 100 µm]) was calculated from the linear regression and used for comparative analysis. The D(R = 100 µm) value was calculated from at least 30 SPAs with radii of 15 to 250 µm (from intraacinar to preacinar SPA) and can be used as representative of the medial thickness of SPAs.
|
|
The severity of the intimal lesions was evaluated by using the Heath-Edwards (HE) classification [22] and the index of pulmonary vascular disease (IPVD) score, as described by our previous report [13, 21]. The IPVD score was defined as follows. First, pulmonary vascular intimal lesion in each vessel was classified into four grades: (1) no intimal lesions, (2) cellular proliferation of the intima, (3) fibrous thickening of the intima, and (4) destruction of the media. A score of 1 to 4 was assigned to each SPA (at least 50 SPAs were evaluated), and the mean score was calculated for each case (IPVD score).
Corrected density of small pulmonary arteries (NA) was determined as follows. The number of SPAs per square centimeter having a diameter greater than 100 µm was counted by a method described by Takahashi and Wagenvoort and colleagues [23, 24]. This value indicates the density of SPAs in a histologic section and is corrected by the number of alveolar septa intersected by a test line of sufficient length drawn on the section to exclude the influence of the degree of lung collapse.
To obtain percentage of SPAs containing thrombus, all SPAs (at least 100) in sections were observed, and the arteries containing even a small amount of thrombus in the lumens were counted, then the percentage of these arteries in total SPAs were calculated.
Statistics
All values are presented as means ± SD. Statistical comparisons among groups were made by using Student's t test. Associations between variables were determined by linear regression analysis, and a coefficient was calculated for the correlations between the values. A p value of less than 0.05 was considered significant. All statistical analyses were performed using Statview version 5.0.
|
Results
|
|---|
Relationships Between Morphohistometric Measurements and Preoperative Hemodynamic Data
Value of DR = 100 µm: the DR = 100 µm values were calculated from a correlation curve plotted by using data from at least 30 vessels, and the p values of all correlation curves were less than 0.05. In a previously study of cases with normal cardiovascular and respiratory systems, DR = 100 µm declined after birth and reached a constant level of 6 to 8 µm by 5 months [21]. Ages in our case series ranged from 6 months to 23 years old, and their DR = 100 µm values had already reached a constant value. In 35 (BDGS, 17; TCPC, 18) of the 60 cases, DR = 100 µm exceeded 8 µm and indicated medial hypertrophy. The highest DR = 100 µm value, 16.9 µm, was in a case of failed single stage TCPC after palliative pulmonary artery banding (PAB). The lowest DR = 100 µm value, 5.9 µm, was in a successful two-stage TCPC case. As shown in Figure 2, there was a correlation between the DR = 100 µm values and mean pulmonary artery pressure (mPAP) mm Hg (p = 0.028). However, the DR = 100 µm values of some patients indicated greater thickening than expected from the correlation curve, and the results in those cases were unsatisfactory. As for other hemodynamic variables, there were no significant correlations between DR = 100 µm and pulmonary vascular resistance (PVR), Nakata's pulmonary artery index (PAI), or pulmonary/systemic flow ratio (Qp/Qs) (p = 0.71, p = 0.59, and p = 0.16, respectively).
View larger version (16K):
[in this window]
[in a new window]
|
Fig 2. The relationship between pulmonary artery pressure (PAP) and D(R = 100 µm) has been plotted. The correlation is statistically significant. Open circles = surviving TCPC; solid circles = decreased or takedown TCPC; open triangles = surviving BDG; solid triangles = decreased or takedown BDG; solid line = correlation curve between pulmonary artery pressure and D(R = 100 µm) in all cases. (BDG = bidirectional Glenn shunt; TCPC = total cavopulmonary connections.)
|
|
Intimal lesions: in 54 of the 60 cases, the IPVD scores were 1.0 (maximum IPVD score in the successful BDGS and TCPC cases was 1.2 and 1.3, respectively). Of the 6 cases whose IPVD scores were greater than 1.0, 3 PAB had been performed as palliation in 3, an aortopulmonary shunt in 1 and no procedure in 2. According to the HE classification, 18 cases were grade 0, 35 were grade 1, 1 was grade 2, and 6 were grade 3. There were no significant correlations between these two variables and the preoperative hemodynamic data in our case series (HE versus PVR, p = 0.77; HE versus PAI, p = 0.94; HE versus Qp/Qs, p = 0.85; HE versus PAP, p = 0.12; IPVD versus PVR, p = 0.27; IPVD versus PAI, p = 0.32; IPVD versus Qp/Qs, 0.94; and IPVD versus PAP, p = 0.079)
Values for NA: NA values corrected for the number of alveoli have been reported to remain almost constant regardless of age, and the mean NA was 10.5/cm2 [23, 24]. The mean NA in our case series was 12.1 ± 6.0/cm2 [6], and in 24 of the 60 cases, NA was below than 10.5/cm2. There were no correlations between NA and preoperative variables (NA versus PVR, Qp/Qs, PAI, and mean PAP; p = 0.064, p = 0.089, p = 0.84, and p = 0.24, respectively).
Percentage of SPA containing a thrombus: in 25 cases, no SPAs containing a thrombus. In 25 cases, a thrombus was present in less than 50% of the SPAs, and in 10 cases, a thrombus was present in more than 50%. The BDGS was successful even in 2 cases in which the percentage of SPAs containing a thrombus was 100%. Again there was no statistically significant correlation (thrombi versus PVR, Qp/Qs, PAI, and mean PAP; p = 0.68, p = 0.42, p = 0.18, and p = 0.60, respectively).
Relationships Between Morphohistometric Measurements and Early Outcome
Bidirectional Glenn Shunt
Preoperative hemodynamic measurements: the mean PAP, Qp/Qs, PVR, and PAI values in the groups with good and bad outcomes were similar (mPAP: 14.8 ± 5.2 versus 15.8 ± 6.5 mm Hg, p = 0.72; QpQs: 1.4 ± 0.77 versus 1.8 ± 0.82, p = 0.99; PVR: 1.8 ± 0.79 versus 1.7 ± 0.82 units · m2, p = 0.85; PAI: 305 ± 109 versus 254 ± 91.0 mm2/m2, p = 0.40). However, there were differences in age and body surface area ([BSA] age: 4.4 ± 4.0 versus 12.3 ± 7.3 years, p = 0.023; BSA: 0.55 ± 0.29 versus 1.1 ± 0.46 m2, p = 0.001; Table 2).
Measurement of DR = 100 µm: there was a difference between the DR = 100 µm values of the patients with good and bad outcomes at p = 0.0007 (8.8 ± 2.4 µm versus 13.4 ± 1.9 µm; Fig 3). None of the deceased cases had a DR = 100 µm values below than 11.6 µm. The highest value in a surviving patient was 13.7 µm in a 4.9-year-old patient (case 21) with a single right ventricle who had been treated by palliative pulmonary artery banding (PAB).
View larger version (20K):
[in this window]
[in a new window]
|
Fig 3. Relationship between D(R = 100 µm) and outcome in bidirectional Glenn shunt groups. There is a marked difference between the two groups (p = 0.0004). Circles = bidirectional Glenn shunt cases; upper dotted line = cutoff point for the success of bidirectional Glenn shunt; lower dotted line = the lowest D(R = 100 µm) point in the deceased or takedown group.
|
|
Other histomorphometric analyses: there were no significant differences in IPVD, HE, NA, or percentages of SPA containing a thrombus (Figs 4–7).
View larger version (19K):
[in this window]
[in a new window]
|
Fig 4. Relationship between Heath-Edwards and outcome in bidirectional Glenn shunt (A) and total cavopulmonary connections (B). There are no differences between the two groups. The black dots are cases with fenestration. (NS = not significant.)
|
|
View larger version (15K):
[in this window]
[in a new window]
|
Fig 5. Relationship between index of pulmonary vascular disease (IPVD) and outcome in bidirectional Glenn shunt (A) and total cavopulmonary connections (B) groups. There are no differences between the two groups. The black dots are cases with fenestration. (NS = not significant.)
|
|
View larger version (21K):
[in this window]
[in a new window]
|
Fig 6. Relationship between corrected density of small pulmonary arteries (NA) and outcome in bidirectional Glenn shunt (A) and total cavopulmonary connections (B) groups. There are no differences between the two groups. The black dots are cases with fenestration. (NS = not significant.)
|
|
View larger version (23K):
[in this window]
[in a new window]
|
Fig 7. Relationship between percentage of small pulmonary artery containing a thrombus and outcome in bidirectional Glenn shunt (A) and total cavopulmonary connections (B) groups. There are no differences between the two groups. The black dots are cases with fenestration. (NS = not significant.)
|
|
Total Cavopulmonary Connection
Preoperative hemodynamic parameters: age, BSA, mean PAP, Qp/Qs, PVR, and PAI were similar in the groups with good and bad outcomes (age: 6.5 ± 5.4 versus 3.8 ± 2.0 years, p = 0.25; BSA: 0.74 ± 0.33 versus 0.58 ± 0.12 m2, p = 0.27; mPAP: 13.3 ± 3.5 versus 16.3 ± 2.3 mm Hg, p = 0.057; QpQs: 0.86 ± 0.53 versus 1.3 ± 0.69, p = 0.086; PVR: 2.1 ± 0.95 versus 1.7 ± 0.56 units · m2, p = 0.30; PAI: 277 ± 91.4 versus 335 ± 70 mm2/m2, p = 0.16; Table 2).
Measurement of DR = 100 µm: there was also a significant difference between the DR = 100 µm values of the patients with good and bad outcomes in at p less than 0.0001 (8.4 ± 2.0 µm versus 14.7 ± 1.5 µm). There were no deaths among the 32 TCPC patients whose DR = 100 µm was less than 11.6 µm; and no patients whose DR = 100 µm was greater than 12.9 µm had a good outcome (Fig 8). In 8 cases, fenestration was performed at the time of TCPC operation or within 1 month after TCPC. The number of fenestrated cases was too small to identify the difference between fenestrated and nonfenestrated cases in terms of a DR = 100 µm cutoff point.
View larger version (20K):
[in this window]
[in a new window]
|
Fig 8. Relationship between D(R = 100 µm) and outcome in total cavopulmonary connections. There is a significant difference between the two groups (p < 0.0001). The black dots are cases with fenestration. Upper dotted line = the lowest D(R = 100 µm) point in the deceased or takedown group; lower dotted line = cutoff point for the success of total cavopulmonary connections.
|
|
Other histomorphometric analyses: there were no significant differences in IPVD, HE, percentages of SPA containing a thrombus, or NA (Figs 4–7).
Cases in Which a Biopsy Was Obtained at Both TCPC and BDGS
In 3 cases, lung biopsies were performed both at the time of the BDGS and the TCPC in three cases. The biopsy specimens showed that DR = 100 µm significantly (p = 0.022) decreased in all 3 cases (from 9.0 to 6.9 µm in cases 2 and 36, from 12.2 to 10.8 in cases 7 and 39, and from 10.8 to 8.4 in cases 8 and 33).
|
Comment
|
|---|
Benefits and Validity of Methods
The morphohistometric method we used for the medial thickness of SPAs has three advantages. First, medial thickness was calculated from the area of the media and the length of the internal elastic membrane in a hypothetical state in which the internal elastic lamina is distended into a circle. In previous studies by other investigators, direct measurement of the media against the diameter of SPAs was used to measure medial thickness [5, 7, 10, 12]. Although that method is very useful and easy to perform [25], but it is easily influenced by the degree of medial vasoconstriction, and some error is possible in autopsy cases as well as biopsy cases [19, 20]. In addition, in autopsy cases the pulmonary arteries were usually injected with a barium sulfate suspension less than 25 cm hydrostatic pressure[9, 19], and that can cause unpredictable overextension or underextension of SPAs that results in their thickness being overestimated or underestimated, even when the tissue is treated in the same manner. The tissue was not injected with any fluid under pressure in either the autopsy or biopsy cases in our study, and the area and internal elastic membrane were easily measured by using free NIH Image software after the histologic section had been scanned with a digitizing camera. This technique have been used in many other studies, and the reproducibility of the data has been confirmed in several other papers [14–16, 18–21].
Second, to compare the cases we used DR = 100 µm, calculated from all of SPAs, both intraacinar and preacinar SPAs. One reason for some authors having concluded that lung biopsy studies are of minimal benefit in decision making is that, as Levy and colleagues [13] pointed out, they reported essentially status of the proximal arteries (preacinar SPAs) and did not mention the status of distal intraacinar SPAs. Because of this, lung biopsy specimens from patients with a single ventricle and low pulmonary blood flow are often considered normal based on the status of the preacinar SPAs alone. However, the medial thickening of intraacinar SPAs may be involved in postoperative elevations of pulmonary resistance, and it may have been responsible for the unfavorable outcome.
And third, DR = 100 µm values have been reported to start to decline just after birth and remain almost constant after 5 months of age [29]. Another reason for not concluding that lung biopsy is useful is that there no methods comparing in different age groups. By using DR = 100 µm values, comparisons among various age groups are possible.
Other Measurements
In addition to medial thickness, we analyzed pulmonary vascular intimal lesions, percentage of SPAs containing a thrombus and the corrected density of SPAs. There were no significant differences in these variables between the cases with a good and poor outcome. One reason for this may be sample bias, namely, all cases in our series were considered to have indications for the Fontan-type operation on the basis of catheterization data, and thus, patients with severe intimal lesions, thrombi, and a low SPA density may have been excluded from our series. In addition, the HE classification was made based on the severest lesions in the sections, that is, if there was only one SPA with intimal fibrosis, the case was classified as grade 3, which may not have been representative of the general status of all SPAs [13, 21].
Benefits of Staged Fontan-Type Operations
The benefits of two-stage Fontan operations in cases with risk factors have already been reported [26, 27]. Biopsy was performed twice in three of the cases we have studied. The DR = 100 µm values decreased in all three cases, and this findings supported the benefits of staged Fontan procedures in cases with severe pulmonary conditions in which waiting for medial hypertrophy to regress sufficiently makes TCPC possible. In case 7, DR = 100 µm was 12.2 µm at the time of the BDGS (a value between the cutoff point of BDGS and that of TCPC), and it gradually decreased to 10.8 µm at the time of the TCPC.
Implications of Histologic Indications for Fontan-Type Operations
Since the group of cases we studied was small, we could not strictly determine the histologic indications for the Fontan-type operation. However, as the DR = 100 µm results indicate, we recognized a significant increase in risk of death after Fontan operations in patients with medial hypertrophy. Previously used methods, such as direct measurement of the media, were very useful for ascertaining which portion of the SPA was important, however, it might be difficult to use such measurements to represent the general (overall) medial thickness of the SPAs in each case and determine the criteria for the Fontan operation. In this study, BDGS and TCPC were successful when the DR = 100 µm values were less than13.7 µm and 11.6 µm, respectively.
Suggestions and Further Studies on Usefulness of Histologic Studies in Clinical Situations
We conclude that histomorphometric study is useful in surgical decision making, and that DR = 100 µm is very valuable as an integrated variable. Histologic studies of lung tissue obtained by biopsy may be useful in the following situations.
(1) Reevaluation of patients who are considered borderline candidates or inappropriate candidates for a Fontan-type operation based on the catheterization data. When we are uncertain as to indications of Fontan-type operations based on the hemodynamic data, lung biopsy may be useful in decision making. In addition, as long as the hemodynamic criteria are not perfectly reliable, some cases may be erroneously excluded from the Fontan-track. Based on the results of the present study, we are currently conducting study to determine whether hemodynamically poor candidates with low DR = 100 µm values in histologic specimens can survive the Fontan-type operation.
(2) Selection of a single stage or two-stage Fontan operation based on frozen section quick diagnosis made intraoperatively. The benefits of a two-stage Fontan procedure in cases with risk factors have already been reported, and in our study, its benefits were recognized in cases with medial hypertrophy. However, single-stage Fontan procedures still have the benefit of avoiding multiple operations. Intraoperative quick diagnosis by frozen section may be useful for selecting the procedure (single stage or two stage) in a cases in which pulmonary condition is the only risk factor for the one-stage procedure. Frozen section diagnosis was accomplished in a relatively short time, as reported in our previous paper on hypertensive pulmonary vascular changes (30 minutes were needed for specimen preparation in our previous study, another 60 minutes was adequate for microscopic examination and calculating the histomorphological variables if the system, including the computer and scan, was operated by skilled technician) [29].
(3) Decision on take down. To determine why failure occurred after the Fontan type operation, lung biopsy may be useful to ascertain the patient's condition and to determine whether take down of the Fontan operation is needed. For example, a low oxygen saturation and high central venous pressure after cavopulmonary connection may be temporary due to the influence of long-term cardiopulmonary bypass (or respiratory infection and so forth). In such cases, DR = 100 µm may be useful to determine between take down or just waiting for the influence of the cardiopulmonary bypass to be decreased.
Study Limitations
This study was retrospective, and since the cases (especially the cases in which the patient died) were collected from eight institutions in Japan, there may have been differences in treatment and methods of clinical diagnosis among the institutions. Moreover, our series of 60 patients does not represent a random sample of patients who underwent a modified Fontan procedure during the study period. As might be expected, autopsy cases are overrepresented in our series, We cannot rule out the possibility that our results may be attributable to some degree of sample bias [19]. A prospective study is needed to clarify the effectiveness of histomorphological study in improving the selection criteria.
Conclusions
The outcome of right-heart bypass procedures such as BDGS and TCPC has improved greatly. Hemodynamic variables are influenced by the methods and the patient's condition, however, and they do not perfectly represent the status of the pulmonary vasculature. Because of this, we have occasionally encountered cases with disappointing outcomes in which the indications were based solely on the hemodynamic data. Morphohistologic study using DR = 100 µm may be a useful method of supplementing a variety of clinical data in cases in which a judgment is difficult to make.
|
Discussion
|
|---|
DR CARL BACKER (Chicago, IL): It appears that you have a very sophisticated analysis system here with the DR100 technique as compared to the old Heath-Edwards classification.
My question is: how much lung tissue do you need for a pre-Fontan biopsy to decide whether or not the patient is a good candidate for that operation?
DR MAEDA: I need just only this size (indicating).
DR BACKER: A portion the size of the tip of the thumb?
DR MAEDA: A thumb, yes, the size of a thumb.
DR BACKER: And where should we take the biopsy; left upper lobe, right upper lobe, right lower lobe?
DR MAEDA: Mainly, lung biopsies were taken from the right middle lobe.
DR BACKER: What is your evidence base for using the right middle lobe?
DR MAEDA: Well, all I can say is that's what we used in this study.
DR ERIC CEITHAML (Jacksonville, FL): In the old days when we used to use the Heath-Edwards classification, we noted that different areas of the same lung could have different Heath-Edwards classifications; and that's one of the reasons that lung biopsy was discarded as an unreliable modality to accurately assess the degree of pulmonary vascular disease. How did you account for this variability in your biopsy specimens?
DR MAEDA: Well, in our studies there is no relation between intimal lesion and outcome, including Heath-Edwards classification. One reason for this is that all of our cases were considered to be Fontan candidates based on hemodynamic data; and therefore, there is a possibility that patients with severe intimal lesions are excluded in your study.
And regarding Heath-Edward grade, this was determined by the severest lesion in the section. Therefore, if even one vessel is showing severe lesion, that case is just to be very severe Heath-Edwards grade. So as many researchers pointed out, Heath-Edwards grade does not represent the general status of the pulmonary vascular beds.
|
Acknowledgments
|
|---|
We thank the following physicians for supplying materials and clinical data: Yoshihisa Tanoue, MD, Yuichi Shiokawa, MD, and Hiroshi Imoto, Fukuoka-City Children's Hospital, Fukuoka; Michio Yokota, MD, Shizuoka Children's Hospital, Shizuoka; Yukihiro Takahashi, MD, Toshio Kikuchi, MD, and Katsuhiko Tatsuno, MD, Sakakibara Heart Institute, Tokyo; Hiroshi Kurosawa, MD, Jikei Medical College, Tokyo; Mitsuru Nakazato, MD, and Tetsu Sato, MD, Yamagata University, Yamagata; Masahiro Fujii, MD, Nihon Medical College, Tokyo; Takashi Miyamoto, MD, and Tetsuhiro Takaoka, MD, University of Tokyo, Tokyo, Japan. We also thank the following technical assistants for preparing the histologic sections: Takako Kato, Junko Kato, Miyuki Abe, and Mitsuru Sasaki, Sendai City Medical Center, Sendai City, Japan.
|
References
|
|---|
- Fontan F, Mounicot FB, Baudet E, Simonneau J, Gordo J, Gouffrant JM. Correction of tricupid atresia. 2 cases "corrected" using a new surgical technique. Ann Chir Thorac Cardiovasc. 1971;10:39–47[Medline]
- Stellin G, Mazzucco A, Bortolotti U, et al. Tricuspid atresia versus other complex lesions: comparison of results with a modified Fontan procedure. J Thorac Cardiovasc Surg. 1988;96:204–211[Abstract]
- Mayer JE Jr, Helgason H, Jonas RA, Lang P, Vargas FJ, Castaneda AR. Extending the limits for modified Fontan procedures. J Thorac Cardiovasc Surg. 1986;92:1021–1028[Abstract]
- Choussat A, Fontan F, Besse P, Vallot F, Chauve A, Bricaud H. Selection criteria for Fontan's procedure. Anderson RH, Shinebourne EA. Pediatric cardiology. Edinburgh, Scotland: Churchill-Livingstone; 1978. p. 559–566
- Knott-Kraig CJ, Danielson GK, Schaff HV, Puga FJ, Weaver AL, Driscoll DD. The modified Fontan operation: an analysis of risk factors for early postoperative death or take down in 702 consecutive patients from one institution. J Thorac Cardiovasc Surg. 1995;109:1237–1243
- DeLeon SY, Ilbawi MN, Idriss FS, et al. Persistent low cardiac output after the Fontan operation: should take down be considered? J Thorac Cardiovasc Surg. 1986;92:402–405[Abstract]
- Kirklin JW, Barratt-Boyes BG. Cardiac surgery. New York: Wiley; 1986. 45;83–91
- Ravinovitch M, Sanders SP, Castaneda AR, Reid L. Morphohistometric analysis of lung biopsy tissue in candidates for Fontan-type procedure. [Abstract]Am J Cardiol. 1981;47:447
- Juaneda E, Haworth S. Pulmonary vascular structure in patients dying after a Fontan procedure. The lung as a risk factor. Br Heart J. 1984;52:575–580[Abstract/Free Full Text]
- Geggel RL, Mayer JE Jr, Fried R, Helgason H, Cook EF, Reid LM. Role of lung biopsy in patients undergoing a modified Fontan procedure. J Thorac Cardiovasc Surg. 1990;99:451–459[Abstract]
- Caspi J, Coles JG, Rabinovitch M, et al. Morphological findings contributing to a failed Fontan procedure: twelve year experience. Circulation. 1990;82(Suppl):177–182
- Yamaki S, Ajiki H, Haneda K, Takanashi Y, Ban T, Takahashi T. Pulmonary arterial changes in patients dying after a modified Fontan procedure following pulmonary artery banding. Heart Vessel. 1994;9:263–268[Medline]
- Levy M, Daniel C, Tamisier D, et al. Histomorphometric analysis of pulmonary vessels in single ventricle for better selection of patients for the Fontan operation. J Thorac Cardiovasc Surg. 2002;123:263–270[Abstract/Free Full Text]
- Yamaki S, Horiuchi T, Ishizuka E, Mohri H, Fukuda M, Tezuka F. Indication for total correction of complete transposition of the great arteries with pulmonary hypertension. J Thorac Cardiovasc Surg. 1980;79:890–895[Abstract]
- Yamaki S, Yasui H, Kado H, et al. Pulmonary vascular disease and operative indications in complete atrioventricular canal defect in early infancy. J Thorac Cardiovasc Surg. 1993;106:398–405[Abstract]
- Yamaki S, Tsunemoto M, Shimada M, et al. Quantitative analysis of pulmonary vascular diseases in total anomalous pulmonary venous connection in sixty infants. J Thorac Cardiovasc Surg. 1992;104:728–735[Abstract]
- Goldner J. A modification of the Masson trichrome technique for the routine laboratory purpose. Am J Pathol. 1938;14:237–243
- Suwa N, Takahashi T. Morphological and morphometrical analysis of circulation in hypertension and ischemic kidney. Munchen: Urban & Schwarzenberg; 1971.
- Honda T. Histometrical study of the pulmonary arteries in patients with ventricular septal defect. J Jpn Assoc Thorac Surg. 1967;15:162
- Niwa T. Histometrical studies of the pulmonary artery in sudden death of young and apparently healthy subjects. Tohoku J Exp Med. 1971;105:365[Medline]
- Yamaki S, Tezuka F. Quantitative analysis of pulmonary vascular disease in complete transposition of the great arteries. Circulation. 1976;54:805–809[Abstract/Free Full Text]
- Heath D, Edwards JE. The pathology of hypertensive pulmonary vascular disease: a description of six grades of structural changes in the pulmonary arteries with special references to congenital cardiac septal defects. Circulation. 1958;18:533–547[Medline]
- Takahashi T, Wagenvoort N, Wagenvoort CA. The density of muscularized pulmonary arteries in normal lungs. Arch Pathol Lab Med. 1983;107:19–22[Medline]
- Takahashi T, Wagenvoort CA. Density of muscularized arteries in the lung. Arch Pathol Lab Med. 1983;107:23–28[Medline]
- Haworth SG, Reid L. A morphohistometric study of regional variation in lung structure in infants with pulmonary hypertension and congenital cardiac defects: a justification of lung biopsy. Br Heart J. 1978;40:825–831[Abstract/Free Full Text]
- Masuda M, Kado H, Shiokawa Y, et al. Clinical results of the staged Fontan procedure in high-risk patients. Ann Thorac Surg. 1998;65:1721–1725[Abstract/Free Full Text]
- Norwood WI, Jacobs ML. Fontan's procedure in two stage. Am J Surg. 1993;166:548–551[Medline]
- Yamaki S, Wagenvoort CA. Comparison of primary plexogenic arteriopathy in adults and children. A morphometric study in 40 patients. Br Heart J. 1985;54:428–434[Abstract/Free Full Text]
- Yamaki S, Tezuka F, Sasano N, Horiuchi T, Kagawa Y. Frozen section diagnosis of hypertensive pulmonary vascular changes at surgery for congenital heart anomalies. Acta Pathol Jpn. 1985;35:803–807[Medline]
Top
Abstract
Introduction
