Günter Köhler, Lars Kaderali, Matthias Evert and Marek Zygmunt
6Prevention of inadequate sarcoma operations
6.1Inadequate surgical treatment of uterine sarcomas in leiomyoma surgery
Uterine LG-ESS, LMS and STUMP are very often incidental intraoperative and postoperative diagnoses reached during or following operations performed under the indication “LM”. This applies to 69–78% of LG-ESS (6, 85, 98, 116) and up to 65% of LMS (98). According to DKSM data (58), 51% of LMS and 68% of LG-ESS are subjected to surgery under that indication, as are 35% of HG-ESS, 26% of UUS, 14.3% of AS and 92% of STUMP (Tab. 6.1.1) (58).
Improper indications consequently result in a large number of sarcomas being subjected to modes of surgery that are inadequate for such tumors, like TE, SAH and SLH, as well as tumor and/or uterus injuries through morcellation or clamping with sharp instruments (cf. Chapters 1, 2, 4). In Germany, 21.4% of LG-ESS patients undergo TE, and 12.1% are subjected to SAH/SLH. Morcellation constitutes part of the surgical treatment performed on LG-ESS in 42% of cases. In total, around 70% of LG-ESS patients undergo inadequate surgery (58). According to one source, 12.4% of LG-ESS are subjected to TE without BSO (6). Five percent of LMS patients undergo TE, while 27% of LMS are morcellated. Overall, inadequate modes of surgery (any tumor or uterus injury including SAH and SLH) are opted for in 66% of LMS (58). Likewise, unacceptable proportions of HG-ESS (45.7%), UUS (36.4%) and AS (27.2%) are subjected to inadequate surgical procedures (58).
A large share of uterine sarcomas is subjected to inadequate forms of surgical treatment, mostly under the wrong indication of leiomyoma, because they are not preoperatively recognized as sarcoma. This applies to 70 resp. 66% of low-grade endometrial stromal sarcomas and leiomyosarcomas, but also to the other types of uterine sarcoma and atypical smooth muscle tumors as well.
6.2Inadequate surgical treatment of sarcomas and STUMP – impact on prognosis
THE without injuring the uterus constitutes the only form of surgical treatment that is adequate for all uterine sarcomas and STUMP. Surgery can be either open or endoscopic. The issue of whether or not to perform LNE or BSO (both of which apparently have no impact on prognosis) is discussed at length in the respective chapters (cf. Chapters 1, 2, 4 and 5 and Vol. 2, Chapter 6).
Tab. 6.1.1: Characteristics of uterine leiomyomas and sarcomas and the indication for primary surgery (DKSM data, 58).
An analysis of retroperitoneal STS that also covered LMS revealed that R1/2 resections and tumor injuries are significant independent predictors of poorer OS (13). These data can also be taken as surrogate evidence for the effects of tumor injury in the course of inadequate surgery performed on uterine sarcomas.
A negative impact on DFS and OS should always be expected whenever a uterine LMS is subjected to morcellation, though the data situation pertaining to OS remains very poor. One study on uterine sarcoma (including LMS) has shown that morcellation only results in a slight increase in the rate of pelvic recurrence without distant metastases, and without any measurable impact on OS (72). In an early study (86), outcomes were retrospectively compared between patients who underwent TAH without tumor morcellation (n = 31) and those who underwent surgery that included abdominal, vaginal or laparoscopic tumor morcellation (n = 25). The percentage of patients with abdominopelvic dissemination, as shown by peritoneal sarcomatosis or vaginal apex recurrence, was significantly greater in patients with than without tumor morcellation (86). In uni- and multivariate analysis, tumor morcellation was significantly associated with poorer DFS and OS. This has been the only study to date in which OS was also significantly shorter. Endoscopic follow-up examinations (82, 100) of LG-ESS, STUMP and LMS patients who underwent morcellation revealed intraperitoneal tumor dissemination in 64% of cases that occurred within 3mo in 43% of cases, and that had not been present at the time of primary surgery. There was no evidence for newly developed distant metastases in any of these cases (cf. Fig. 2.1.12 (A)–(C), 4.1.13 (C), (D), 4.1.14 (C)). Three of four LMS patients in whom there was peritoneal dissemination succumbed to their disease after a mean period of 24.3 mo. Given the poor overall OS with which LMS are associated (see above), it is indeed questionable that the aforementioned outcomes were directly associated with the present intraperitoneal dissemination. A recent analysis (37) compared intraperitoneal morcellation to TAH without morcellation in terms of their effects on patient’s prognosis. The study revealed a significantly shorter abdominopelvic RFI when morcellation was performed (10.9 vs. 39.6mo for TAH). The OS rate at 36mo was 64% in the morcellation cohort and 73% in the TAH cohort, but the difference was not statistically significant.
A meta-analysis (11) covering morcellated vs. non-morcellated LMS revealed an odds ratio of 3.16 for all recurrences, 4.11 for intraabdominal recurrences and 2.42 for the death rate. The cumulative relapse rate was only significantly elevated for intraabdominal disease. The authors themselves do conclude, however, that the numerous biases in the study do not allow any definitive conclusions to be drawn as to the impact on OS.
A further study (65) compared morcellation of LMS performed in the context of various different surgical (abdominal, vaginal, laparoscopic) methods to TAH without morcellation, and revealed no differences between the two groups in terms of OS. There was an (albeit statistically non-significant) increase in the rate of recurrence in the morcellation group, even though the two most important prognostic variables (tumor size and elevated MI) were less pronounced/lower in this group. Morcellation had no impact on the rate of distant metastasization.
In a recent retrospective analysis (79), morcellated LMS had better (!) 5-year RFI and 5-year OS than non-morcellated tumors (each group included abdominal, vaginal and laparoscopic operations), though the difference was not statistically significant. The statistically non-significant RFI-related findings in the morcellation arm (which in itself was relatively small) need to be regarded critically. Prognostically relevant factors (like stage of disease and patient’s age) were lower, while tumor size (esp. relevant for stage I) and MI were not devoted sufficient consideration in this group. The most important finding from this study was that the rate of distant metastasization was identical for both surgical methods. Since survival is primarily determined by the distant metastases, this finding at least lends support to the assumption that there is no impact on OS in LMS patients (which is already very short as is). No distant metastases were found in three cases in which there was (in part expansive) pelvic and intraperitoneal recurrence following morcellation of an LMS (40). By contrast, the causal connection between morcellation and a shorter RFI is further supported by the fact that, in these cases, locoregional and intraabdominal recurrences were considerably more frequent than in LMS patients who underwent adequate surgical procedures (69, 86). Nor can it be discerned from the DKSM data that morcellation or any form of tumor injury does result in increased distant metastasization. In contrast, intraperitoneal recurrences have been observed to develop repeatedly without distant metastases occurring (cf. Fig. 2.1.12 (A)–(C)).
In summary, it is possible that morcellation only increases the risk of intraabdominal recurrence, but not (at least in the short term) the risk of potentially lifethreatening pulmonary and hepatic metastases. Since patients with LMS already have a very poor prognosis, and death is usually a result of distant metastasization, it cannot be reliably claimed that morcellation actually shortens survival.
Another study (85) on LG-ESS compared outcomes of patients who underwent TAH or LAVH with morcellation to TAH or LAVH without morcellation. The results revealed that 5-year DFS and abdominal-pelvic DFS were significantly poorer in the morcellation arm (with a majority of endoscopic procedures). Interestingly, no intraperitoneal recurrences were observed in 7 cases in which incidentally detected LG-ESS were subjected to TLH with morcellation (which was necessary in order to be able to extract the uterus) (117). Within the data provided by a DKSM working group, among 3 cases of LG-ESS that were morcellated, 1 recurred once, another recurred 3 times, and the third developed 4 consecutive recurrences, all within up to 5 years. Despite the relatively long interval, there were no cases of distant metastasization or deaths (58).
Resection margins that are not definitively disease-free are equally associated with an elevated risk of recurrence (113). A disease-free margin of < 1mm is deemed an independent risk factor for STS that is significantly associated with elevated recurrence risk and poorer OS (13). Since uterine sarcomas have no capsule, or have a pseudo-capsule that in fact consists largely of compressed tumor tissue, it is almost inevitable that TE will leave R1 and RX situations. In one study, for LMS, tumor injury and/or positive surgical margins had a hazard ratio of 5.6 for developing local recurrence, while an impact on OS and the rate of distant metastasization could not be confirmed (113). In another study, among LG-ESS patients who underwent TE, not one remained recurrence-free within 5 years, while there were no differences in terms of OS (6). Intraperitoneal masses (sometimes very large) have been observed to develop within 2wk to 3mo of performing endoscopic TE or morcellation with and without SLH on LMS and LG-ESS (19, 27, 58, 90). In a recent publication, a significant lowering of RFI compared to THE (8.2 vs. 46.6 mo) among all uterine sarcomas was observed (48). For these reasons, performing TE on uterine sarcomas needs to be deemed prognostically critical and regarded as inadequate surgical treatment (28, 31).
SLH and SAH are generally associated with the risk of leaving intravascular ormyometrial invasion/infiltration routes in the uterine stump. Even if the surgical report describes an R0 resection, from a clinical perspective, it must nonetheless be assumed that such operations will leave RX situations. This is further exacerbated by the fact that SLH is virtually always combined with morcellation. Early tumor recurrences are known to occur within 3mo following such procedures (3). The situation is apparently similar in cases in which the uterus is injured with a sharp instrument (90). Accordingly, clamping with sharp instruments must be categorically avoided whenever sarcoma is even suspected. According to a retrospective study, any form of tumor injury resulted in a 2.8-fold increase in mortality and significantly shorter RFS and OS (p = 0.03 resp. 0.04), though it needs to be mentioned that the study was based on a relatively small sample, there were numerous biases, and the important prognostic factor “tumor size” was not taken into consideration (90). In the DKSM materials database, LMS subjected to inadequate surgery (any form of tumor or uterus injury) had a recurrence rate of 28.3%, compared to 27.7% in cases in which adequate procedures were opted for. However, it is important to consider that the women who underwent adequate modes of surgery (n = 101) had an average age of 55 years and a mean tumor diameter of 9 cm, and thus a poorer primary prognosis than the women subjected to inadequate operations (age 47 years, tumor diameter 6.5 cm) (58). According to DKSM data, the median time period from surgery to first recurrence was 17mo among LMS patients subjected to adequate operations, and 7mo for women who underwent inadequate surgery (58). Among LG-ESS patients corresponding median time period for adequate surgery was 29.6mo (13 of 92 patients) and for inadequate surgery 7mo (14 of 67 patients) with a relapse rate of 14.1% and 21%, respectively (58).
The fact that intraperitoneal recurrences develop so quickly after LG-ESS have been subjected to inadequate surgery is rather surprising, considering that LG-ESS are otherwise associated with such promising courses and usually recur or metastasize rather late (for details on the prognosis of LG-ESS see Chapter 4). Their relatively promising prognosis is likely a result of the fact that, in contrast to HG-ESS, UUS and LMS, LG-ESS that infiltrate the blood and lymphatic vessels remain cohesively connected with the primary tumor for a very long time. The intravascular plugs contain both tumor cells and small vessels that are enclosed by endothelial cells. The tumor cells within the tumor thrombi are thus kept apart from blood circulation. In contrast, intravascular tumor foci in HG-ESS are composed of discohesive cell clusters, lacking the features described in LG-ESS (96) (Fig. 4.1.3 (A), (B), Fig. 5.1.3 (D) and Fig. 6.2.1 (A), (B)). Along with other factors, this feature could help to explain why LG-ESS metastasize much less frequently than HG-ESS do. From a mechanistic perspective, it can thus be assumed that tumor injuries (e.g. in the course ofmorcellation) cause the intravascular tumor plugs (96), that are originally cohesively attached to the primary tumor, to detach from their base, and to subsequently mimic or masquerade as purely intraperitoneally situated, discohesive lesions.
In summary, there is no evidence to suggest that subjecting LG-ESS to inadequate surgery, the negative impact on RFI notwithstanding, impacts negatively on 5-year OS. This can be largely explained by the fact that, according to the current data situation, the rate of distant metastasization apparently does not increase, that even extensive locoregional and intraabdominal recurrences remain in situ for a long time (Fig. 6.2.1 (D)) and are amenable to surgery, and that long-lasting PFI can be achieved under HT. Regarding LMS, the data situation is too weak to be able to definitively state that inadequate surgery has a negative impact on OS, which is already rather poor from the onset. Regardless, the shortening of RFS that inadequate surgery entails for LG-ESS and LMS has a considerable impact on quality of life.
Fig. 6.2.1: Different forms of vascular invasion in low-grade endometrial stromal sarcoma and undifferentiated uterine sarcoma. (A) wide, tongue-like invasion of low-grade endometrial stromal sarcoma into a vessel. The tumor plug is surrounded by red endothelium on all sides and is cohesive with the primary tumor. There are no visible individual, freely distributed tumor cells or cellular accumulations; (B) the intravascular tumor cell thrombus of undifferentiated uterine sarcoma is not surrounded by the endothelium, and there are numerous isolated tumor cell accumulations that provide the foundation for early metastasization. High-grade endometrial stromal sarcoma and leiomyosarcoma have a similar appearance and exhibit similar behavior in this regard; (C) pelvic-peritoneal recurrence of low-grade endometrial stromal sarcoma discovered in followup laparoscopy 6 months after endoscopic supracervical hysterectomy with morcellation; (D) large pelvic recurrence 30months after surgery identical to that performed in (C). There was no traceable distant metastasization, and the findings were amenable to R0 resection.
According to various sources, STUMP have a recurrence rate of 7.7 to 26.7%(41, 46, 91, 108). Recurrences and metastases apparently only occur after ME, morcellation and SHE (2, 10, 19, 108). While there have been accounts of recurrent disease following THE, it cannot be discerned from the reports whether or not morcellation had been performed (41, 67). STUMP are known to have recurred within only a few weeks after morcellation (58). In 3 of 51 cases (34 HE, 17 ME), subsequent THE or ME revealed two atypical LM in the residual uterus, while in the third case, there was an atypical LM in the retroperitoneum. Interestingly, the latter patient had probably in fact previously undergone THE with morcellation (67). It currently remains difficult to assess whether endoscopic ME more frequently results in intrauterine recurrences when performed on STUMP than when performed on ordinary LM. The cumulative intrauterine recurrence rate for ordinary LM following laparoscopic ME is 51% (80). This rate is possibly no lower than the equivalent rate for STUMP. What is decisive in this regard, though, is that recurrences and metastases of STUMP need to be assessed differently in terms of their malignant potential, and that STUMP can transform into LMS via malignant change (102) (for details see also Chapter 1). In the DKSM database (58), among 12 STUMP that metastasized, 42% did so as STUMP, while the remaining 58% metastasized as LMS. The same risk (recurring or metastasizing as LMS) generally also applies to CLM (cf. CLM Chapter 1 and Fig. 1.4.7 (A)–(C)). In the DKSM materials, three such cases became known within a period of 6 mo. In contrast to primary LMS, it must be assumed that OS will be shortened in these cases.
The prognostic assessments made for LMS, LG-ESS and STUMP also largely apply to HG-ESS, UUS and AS. The data situation for these latter entities is, however, rather weak, largely a result of the rarity of these neoplasms. Furthermore, these tumor types are far less frequently primarily misdiagnosed as LM. This is largely because patients are clearly older on average and often present with AUB in the form of intermenstrual bleeding or postmenopausal bleeding, and because these types of tumors are often detected rather early via curettage and on the basis of suspicious sonographic and HSC findings (cf. Chapter 5 and Vol. 2, Chapter 6).
Two more recent studies (97, 103) have shown that, in LMS patients, open TAH has a substantially higher mortality rate than TLH with morcellation (32 vs. 12 deaths/ 100,000 cases).However, open abdominal surgery has a lower LMS mortality rate than subjecting LMS patients to laparoscopic HE with morcellation (71 vs. 86/100,000). According to the authors, the risk of leiomyosarcoma morcellation is balanced by procedure-related complications that are associated with laparotomy, including death. Better quality of life and lower costs were further highlighted as advantages of endoscopic procedures. Such statements contradict (and in fact serve to thwart) efforts that are being made to avoid subjecting LMS (and thus also LG-ESS) patients to inadequate surgery, and to significantly improve diagnostics for these neoplasms. Furthermore, sarcomas can in fact also be treated via TLH without injuring the uterus, and the proportion of sarcomas that are subjected to inadequate surgery can be substantially reduced simply by adhering to the proposed flowchart (Fig. 6.6.1). Since morcellation might not necessarily impact negatively on OS (see above), harms and benefits will need to be carefully weighed against each other pending further developments in regard to this issue (81).
The problems and challenges pertaining to the use of power morcellation are addressed at length in Chapter 2.
Tumor and uterus injury (tumor extirpation, supracervical hysterectomy with and without morcellation, tumor injury and/or morcellation during total hysterectomy) apparently has a negative iatrogenic impact on the prognosis of uterine sarcomas, particularly in the form of a shorter recurrence-free interval in the pelvis and the abdomen. There is currently no definitive evidence for an earlier occurrence of distant metastases. The available data are currently insufficient to allow any definitive statements to be made as to whether there is a strong likelihood of shorter overall survival in patients with uterine sarcoma.
6.3Prevalence of uterine sarcomas in hysterectomies and surgery performed under the indication of leiomyoma
The principal reason why sarcomas are subjected to inadequate forms of surgical treatment is that the majority of these neoplasms is not known to be sarcoma prior to surgery, and is incidentally discovered during or after surgery performed under an indication of LM (see above). According to older sources, the risk of incidentally detecting LMS in the course of HE performed under any indication is 0.07% (49), and 0.08% for LM operations (62, 88). For standalone ME, LMS risks of 0.09% (105) and 0.39% (103) have been reported. In another study, 1.2% of morcellations performed under the indication “LM” (1,091 operations) were in fact performed on LMS (100). Further analyses suggest that 0.22% of all types of morcellation performed under any indication in fact involved LMS (103). The LMS risk for all HE performed on symptomatic (!) LM is 0.49%, and is thus considerably greater. That risk increases from 0.2% at age 36, to 1.4% at age 50, all the way to 1.7% from age 62 onwards (63). However, these data are in part based on relatively small cohorts. According to another analysis, the risk of incidentally encountering uterine sarcoma during LM surgery in patients aged 50 and older is 0.62% (1 : 116), and 3.4% (1 : 29) for patients aged older than 60. In contrast, only 0.04% of LM operations performed on women younger than 40 incidentally reveal uterine sarcomas (114). 64.2% of all LMS arise in patients aged 50 and older (Tab. 6.1.1 and 6.4.1) (58).
In an older German publication, the rate at which LMS were incidentally detected during or following SLH was 1 : 774 (n = 1,585) (110),while a more recent German study revealed a significantly lower rate of only 1 : 5,555 (n = 10,731 SLH) (12). In 2015, a DKSM working group calculated a likelihood of 1 : 3,717 for LMS being incidentally diagnosed during or following HE performed under any indication (n = 3,717). A second DKSM working group calculated a ratio of 1 : 1,851 in a total sample of 31,267 LM operations (58). For postmenopausal women, the risk was 1 : 100 for all sarcomas. A recent comprehensive literature review assumes that 0.14% of all presumed LM are in fact uterine sarcomas (1 : 700) (15). This figure stands in stark contrast to the German results as well as to the findings from a recentmeta-analysis, according to which the rate of LMS among operations performed under the assumption of LM was 1 : 8,300 in prospective studies, and 1 : 2,000 in retrospective studies (94). In another recent study, there were no instances of occult uterine sarcoma in a total sample of 1,004 HE and TE involving power morcellation, but the risk of occult STUMP or atypical LM was 0.029%(1 : 3,448) (92). In a smaller morcellation cohort (352 cases), occult STUMP were detected at a ratio of 1 : 352, and variants of LM (CLM, mitotically active LM) had a likelihood of 1 : 88 (29). A combined look at the literature from 2015/2016, with large total case numbers, suggests that the prevalence of all uterine sarcomas among all LM operations or HE is most likely ≈ 1 : 1,200, and ≈ 1 : 2,000–2,300 for uterine LMS. These data notwithstanding, it needs to be pointed out that, among 3,031 cases in which TLH and TAH were performed under the indication “LM”, the risk of occult LG-ESS was 1 : 232, and the likelihood of LMS was 1 : 604 (117). Another study revealed a likelihood of 1 : 352 (n = 352 cases) for incidentally detecting LG-ESS in cases in which supposed LM are subjected to morcellation (29).
According to most recent research, the prevalence of occult uterine sarcomas among all leiomyoma operations is most likely around 1 : 1,200. For leiomyosarcoma, the likelihood is roughly 1 : 2,000, while for atypical smooth muscle tumors (excluding mitotically active and cellular leiomyoma) the risk is at around 1 : 1,000.
Tab. 6.3.1: Prevalence of incidental uterine sarcoma in leiomyoma operations and hysterectomies (12, 53, 58, 94).
* only supracervical laparoscopic hysterectomies
** all leiomyoma operations
*** all hysterectomies
Despite the relatively low risk of roughly one sarcoma per 1,200 LM operations, from a prognostic perspective it is nonetheless vital that the proportion of sarcomas that are subjected to inadequate surgery be clearly reduced, even in light of the possibility that OS may not be affected (see above).
6.4Anamnestic and clinical criteria suggestive of uterine sarcoma in cases of supposed leiomyoma
In order to prevent that uterine sarcomas are incidentally detected in the course of or following LM surgery, anamnestic and clinical criteria, symptoms and findings need to be given meticulous and consistent consideration, especially regarding the two most common types of uterine sarcoma (LG-ESS and LMS) (34). Differentiation from LM is the most important issue in this regard, as it constitutes the most common DD and also the most common misdiagnosis for these neoplasms. Studies by the DKSM (58) have identified several factors that can be drawn on in order to discern LMS (the most common form of uterine sarcoma) from LM (Tab. 6.4.1). It remains decisive, however, that a clinical suspected diagnosis of sarcoma can only be arrived at when multiple criteria are met.
In general, the possibility of sarcoma must always be considered whenever a patient aged > 50 years is suspected to have LM (see also Tab. 6.1.1). 66.1% ofLMSpatients are aged ≥ 50 years,while 80.4% of HG-ESS patients, 84% of UUS patients and 60.8% of AS patients meet this criterion (58). In contrast, only 25% of women with LG-ESS are aged ≥ 50 (median 46). Among 3,920 LM, only 16% of women were aged ≥ 50 years. The difference to the different types of sarcoma is statistically significant to a strong degree (58). In one study, 9 of 10,119 HE performed for benign gynecologic indications were incidentally revealed to have been performed on sarcomas (5 LMS, 2 LG-ESS, 2 AS) (53). Not one patient was postmenopausal, and only 2 patients were aged ≥ 50. The youngest patient was 25 years old. Thus, while a patient’s age of ≥ 50 years is by all means well suited for distinguishing uterine sarcomas from LM, this criterion alone is not enough to be able to reliably reduce the risk of occult sarcoma.
With the exception of low-grade endometrial stromal sarcoma, 65–84% of all women with uterine sarcoma are aged ≥ 50 years, dependent on the precise variant of uterine sarcoma. Patient age of ≥ 50 can thus be regarded as a criterion for differentiating uterine sarcoma from leiomyoma.
Virtually all uterine sarcomas exhibit suspicious or conspicuous characteristics and features in sonography (Tab. 6.4.2). However, for various reasons, the sonographic findings are only interpreted as being suggestive of sarcoma, even when they are suspicious, in 50.3% of LMS patients and 32% of LG-ESS patients. Sonography and Doppler sonography should be deemed suspicious if they exhibit criteria listed in Tab. 6.4.2. Details pertaining to the sonographic characteristics of the individual sarcoma types are presented at length in the corresponding subsections on imaging in the respective Chapters 1, 2, 4, 5 and Vol. 2, Chapter 6. However, after reevaluation sonography shows suspicious properties and features for LM in 15.9% and for LMS in 88.4% of cases (58). Besides patient’s age, sonographic findings and features are of pivotal importance when seeking to differentiate uterine sarcomas from LM.
Tab. 6.4.1: Anamnestic and clinical criteria suitable for discerning between leiomyoma and leiomyosarcoma; p-value all criteria < 0.01 except prior medication-based therapy (compiled by Dennis, Lehr, Köller, Krichbaum, Riechmann, Trojnarska, Lehnhoff) (58).
|Criteria||LM (n = 3,920)||LMS (n = 224)|
|≥ 50 years||15.9||66.1|
|Failure of previous medication-based or interventional therapy||3.6**||1.9|
|≥ 50 years + ∅≥ 5 cm||6.6||61.6|
|Suspicious sonography + ∅≥ 5 cm||3.6||49.6|
|Conspicuous/rapid growth + symptoms||8.7||25.6|
|Conspicuous/rapid growth + ≥ 45 years||9.4||35.7|
|Conspicuous/rapid growth + ∅≥ 5 cm||13.4||41.5|
|∅≥ 8 cm||16.7||62.5|
|Tumor abnormally soft||0.5||36.8|
|Conspicuous growth + ≤ 25 years***||0.1||3.4|
* excludes menorrhagia and prolonged menstrual bleeding
** refers to 193 cases
*** esp. STUMP, cellular leiomyoma and mitotically active leiomyoma
Tab. 6.4.2: Sonographic and Doppler sonographic characteristics of uterine sarcomas, particularly leiomyosarcoma.
|Gray scale sonography|
|–||Boundaries to the myometrium are poorly defined, derounded, often bizarre or with pointed, tapered extensions|
|–||Tumor predominantly has heterogeneous echogenicity with large areas of strong hyperechogenicity|
|–||Patchy or predominating hypo- to anechoic regions distributed across the entire tumor|
|–||Margins between areas of differing echogenicity are derounded, often bizarre or with pointed, tapered extensions|
|–||Serosa reached or ruptured|
|–||Marked peripheral and irregular central hypervascularity|
|–||Central vascularity plus cystic sections plus tumor > 8 cm|
|–||RI < 0.4|
The sonographic characteristics of uterine sarcomas generally differ substantially from those of leiomyoma. Routine gray scale sonography plays a pivotal role in the detection of uterine sarcomas whenever leiomyomas are to be subjected to any form of therapy, and therapists must always be aware of the sonographic findings prior to administering any such treatment.
Assumed LM are suspicious of in fact being uterine sarcoma when patients present with AUB in the form of intermenstrual bleeding (73, 111) or postmenopausal bleeding (Tab. 6.4.1). Postmenopausal bleeding occurs in 42.2% (79) resp. 45% (58) of postmenopausal LMS patients, while this applies to only 21.9% of postmenopausal LM (58). In one study, among women found to have occult sarcoma during or following HE performed for assumed benign conditions, said surgery was performed as a primary indication for abnormal bleeding in 77.8% of cases (53).
Abnormal uterine bleeding in the form of intermenstrual bleeding or postmenopausal bleeding in patientswith suspected leiomyoma can be suggestive of uterine sarcoma. This particularly applies in cases in which the uterine cavum appears to be suspicious in sonography or is filled with tumor tissue.
According to DKSM data, 50% of LMS (n = 224) and 40.2% of LM (n = 3,920) arise as solitary tumors, i.e. without coincidence with (other) LM (58). LMS and LG-ESS per se only very rarely occur in multiples in the wall of the uterus. The trait “solitary tumor” without the presence of secondary LM does not suffice on its own to differentiate uterine sarcomas from LM. While a recent report suggests that 100% of LMS arise solitarily, the findings were based on only 8 cases (32).However, 38.0% of LMS are solitary tumors and have a size of ≥ 8 cm, while only 9.1% of ordinary LM meet both of these criteria (58). The combined occurrence of the features “solitary tumor” and “tumor size ≥ 8 cm” thus supports a diagnosis of LMS.
LMS (n = 224) have median and mean diameters of 9 resp. 9.7 cm. A more detailed breakdown reveals that 87.9% are ≥ 5 cm, 60.2% are ≥ 8 cm, and 41.1% are ≥ 10 cm in their largest dimension (58). According to DKSM data, the figures for LM are 54.6%, 16.6% and 7.0%, respectively (58). A tumor size of ≥ 5 cm, and ≥ 8 cm in particular, can therefore be drawn on as a criterion for differentiating LMS from LM. However, it is sensible to set the cut-off value for further, more extensive diagnostic procedures at 8 cm. While setting the cut-off value at 5 cm would increase the sensitivity of risk criteria for LG-ESS and AS (median tumor diameters of 5.5 resp. 4.5 cm), doing so would also unnecessarily encompass more than 50% of LM within the risk profile. However, the accuracy in discriminating between sarcomas (especially LMS) and LM can be significantly improved if “tumor size ≥ 5 cm” is combined with the criterion “suspicious sonography” or “age ≥ 50 years” or “conspicuous/rapid growth”. The differences between LM and LMS for these combinations are highly statistically significant in univariate analysis (Tab. 6.4.1). For AS, the fact that their median diameter (4.5 cm) is below the 5 cmcut-off value, is largely compensated for by the fact that patients are generally older, often present with AUB, and frequently return suspicious sonographic findings.
The fact that a lesion suspected to be leiomyoma is a solitary tumor alone is not enough to suspect uterine sarcoma. A tumor diameter of ≥ 8 cm, however, is an important criterion suggestive of sarcoma, especially when the tumor is solitary. Combinations of tumor size ≥ 5 cm plus suspicious sonography or plus patient age ≥ 50 years or plus conspicuous/rapid growth are further important features suggestive of uterine sarcoma.
The true significance of rapid or conspicuous uterine growth for properly diagnosing LMS has long been, and remains, a highly debated issue, not least because there are multiple definitions of it in the literature. According to one source, a tumor exhibits rapid growth if it doubles in size within 6mo (16), while in another study (7), a tumor is classified as growing rapidly if its volume increases by 30% or more within 1 year. In a third source, a volume increase of 20% within 6mo is defined as rapid growth (89), while a fourth publication defines rapid growth as growth equivalent to 6 weeks’ gestation within 1 year (87). For practical reasons, a tumor should be considered as exhibiting rapid or conspicuous growth if the attending medical physician clinically or sonographically describes either this condition or otherwise conspicuous progression in size.
Of the LMS covered in a recent publication, 21.9% exhibited rapid growth (79). In the DKSM materials, conspicuous/rapid uterine or tumor growth was reported in 42.6% of LMS cases (58). Two DKSM working groups covering a total of 3,920 otherwise regular LM retrospectively revealed rapid growth in 16.9% of cases (58). In an immediately subsequent prospective analysis on 193 LM conducted by the same working group, the share was 12.4%. The p-value between the LMS analysis and the two myoma analyses was strongly significant in the χ2 test (58). Another source found rapid growth in 31% of LM (26), while in a further prospective analysis (7), 26.7% of LM increased in volume by 30% within 1 year. These differences are likely a result of differences in the definitions used. Judging by the DKSM data (58) that were evaluated along identical criteria for both LMS and LM, conspicuous or rapid growth alone can apparently be sufficient for differentiating uterine sarcoma from LM.
LM with an initial diameter of ≥ 5 cm are said to only rarely exhibit growth spurts, and when they do, they are only minor (7, 26).One prospective study (89) revealed that 34% of LM of any diameter and 15% of LM with diameters of ≥ 5 cm exhibited rapid growth (> 20% in 6 mo). In the DKSM case database (58), 13.4% of LM and 41.5% of LMS had a diameter of ≥ 5 cm and exhibited conspicuously rapid growth.
Rapid growth was only recognizable in 2% of LM in patients aged 45 and older (89). According to DKSM calculations, the respective shares among LM and LMS were 9.4% and 35.7% (58). Taken together, the data (7, 58, 89) suggest that LMS should be considered a serious possibility when there is conspicuous growth, especially in combination with an initial tumor diameter of > 5 cm and a patient’s age of ≥ 45 years. Only 0.1% of LM in patients aged ≤ 25 years are observed to grow rapidly or at a suspicious pace (58). This combination of criteria is 40 times more common among patients with STUMP (including CLM and mitotically active LM), but nonetheless still rather rare (3.4% of cases). In a recent DKSM case (58), a 9 cm large, sonographically suspicious, rapidly growing “LM” in a 24-year-old symptomatic patient was found to actually be an extremely rare uterine melanoma (Fig. 6.4.1).