Predicting histology in nonretroperitoneal residual masses

The ability to predict the histology of post-chemotherapy residual masses, both within and outside the retroperitoneum, is important because the prognosis and management differ for tumors composed of viable GCT, teratoma, and necrosis. For masses within the retroperitoneum, logistic regression analysis has attempted to identify parameters that are predictive of necrosis. This identification is critical because additional surgery, although therapeutic for teratoma and viable GCT, is not beneficial for those who have residual necrosis. Current investigations indicate that within the retroperitoneum necrosis composes approximately 40% of residual masses, teratoma another 40%, and viable GCT the remaining 20% [13]. Reported variables that predict necrosis are small (<1.5 cm) residual lesions, a greater than 90% reduction in the mass after chemotherapy, and the absence of teratomatous elements in the primary tumor [6,14-16]. These findings have been inconsistent in the literature. Because of the inability to accurately predict post-chemotherapy histology in the retroperitoneum, the most conservative approach is to resect all residual radiographic lesions. The histology of the postchemotherapy RPLND specimen subsequently dictates further treatment if necessary.

For residual masses outside of the retroperitoneum, research efforts have focused on the histologic concordance between residual masses in the retroperitoneum and other distant sites (eg, thoracic, hepatic, and cervical). If necrosis in the retroperitoneum could accurately predict necrosis at other sites, patients could be safely spared the morbidity of additional surgery.

Residual thoracic disease (pulmonary and mediastinal)

The most recent Memorial Sloan-Kettering Cancer Center (MSKCC) experience of 86 patients who underwent synchronous (within 2 months of each other) and 44 patients who underwent asynchronous (thoracotomy >2 months after RPLND) thoracotomy and RPLND procedures for residual disease revealed an overall histologic discordance rate of 28% and 57%, respectively [17]. A worse histology was found in the chest approximately one third of the time. Hartmann and colleagues [18] examined 27 patients who required at least two surgical interventions for multiple sites of residual disease after first-line cisplatin-based chemotherapy. Eight of 27 (30%) patients demonstrated dissimilar histologic findings at the multiple sites. In the subset of patients who underwent retroperitoneal and thoracic resections, less favorable histology was found in 2 patients in the retroperitoneum and in 1 patient in the thorax. In a study from Norway, investigators found discordant pathology in 7 of 15 (47%) patients who had metastatic testic-ular cancer who underwent at least two postche-motherapy resections for residual tumor [19]. Brenner and colleagues [20] reported on their experience of simultaneous resection for retroperito-neal, thoracic, and cervical residual masses after chemotherapy in 24 patients. Six patients (25%) demonstrated discordant pathology in the chest and retroperitoneum, 1 with viable tumor only in the chest and 2 with viable tumor only in the retroperitoneum. Another retrospective review of 31 patients undergoing thoracotomy for residual pulmonary lesions revealed dissimilarity between the retroperitoneal and pulmonary histology in

50% of the patients [21]. This study included 3 patients (15%) who had viable tumor found in the pulmonary specimen but not in the RPLND. Toner and colleagues [22] found that the size of postchemotherapy pulmonary metastases did not predict final histology, and that patients who have bilateral pulmonary lesions may have a different histology in each lung. Six of 14 patients who had pulmonary nodules smaller than 1 cm had either teratoma [5] or viable GCT [1]. Three of 8 patients who underwent bilateral thoracotomy had different histologies in each lung. Mandelbaum and colleagues [23] and Tiffany and colleagues [24] found dissimilar patterns of histology in thoracic and retroperitoneal specimens in 29% and 35% of patients, respectively. The above studies consistently demonstrate a significant discordance rate, on average of approximately one third of cases, between postchemotherapy RPLND and thoracic histology. The inability of the RPLND specimen to accurately predict histology outside of the retroperitoneum has led most investigators to conclude that all sites of disease should be resected.

In 1998, Tognini and colleagues [25] updated the Indiana University experience first presented by Mandelbaum and colleagues [23] in 1983. A total of 143 patients underwent postchemotherapy resection of retroperitoneal and thoracic masses under the same anesthetic. There was thoracic and retroperitoneal histologic concordance in 78% of those who had retroperitoneal necrosis, in 70% who had teratoma, and in 69% who had viable cancer. Analyzing only those deemed uncomplicated by Indiana University criteria (standard induction chemotherapy, normalization of serum tumor markers, and no previous attempts at surgical resection) demonstrated that 18 of 21 (86%) patients who had necrosis in the retroperi-toneum revealed the same pathology in the chest. This finding led the authors to conclude that in uncomplicated cases a finding of necrosis in the retroperitoneum allows for observation of residual thoracic tumors. In 1997, an international, multi-institutional study of 215 patients who had testicular cancer used logistic regression analysis to help predict residual pulmonary histology after chemotherapy [26]. The histology at RPLND was the strongest predictor of pulmonary histology. Necrosis was found at thoracotomy in 89% of patients who had necrosis in the RPLND specimen. When RPLND histology was necrosis and the primary tumor was teratoma negative, the probability of necrosis at thoracotomy increased to 93%.

The authors concluded that the necessity of thora-cotomy is doubtful in some patients in whom the pulmonary histology is likely to be necrosis (ie, necrosis at RPLND and teratoma-negative primary tumor). Given the strong predictive value of the RPLND histology, the authors also recommended that RPLND be performed before thora-cotomy is considered. The above studies from Tognini and colleagues [25] and Steyerberg and colleagues [26] suggest that necrosis at RPLND in a subset of patients may be predictive enough to preclude thoracotomy. Moreover, using a logistic regression model with multiple variables may better predict which patients benefit from close follow-up rather than resection. It seems evident that despite the above compelling data, however, no regression model or set of variables can perfectly predict thoracic histology. We still advocate the resection of all residual disease in the thorax to ensure that no viable tumor or teratoma remains in situ.

Residual disease at other sites

There is a paucity of data comparing residual RPLND histology with histology at non-retro-peritoneal sites other than the thorax. In a report of 16 patients who had residual neck masses, histologic discordance was seen in 7 (44%) patients [27]. Rivoire and colleagues [28] reviewed 37 patients who underwent liver resection after chemotherapy for GCT. Twenty-seven of the patients had additional synchronous abdominal resections, 20 of which were RPLND. Eleven of the 27 (41%) patients demonstrated histologic discordance of the two sites. Six of these 11 individuals had the more adverse histology in the liver. These discordance rates further support the rationale for resecting all sites of residual disease if technically feasible. Table 1 summarizes the histo-logic discordance between postchemotherapy RPLND and non-retroperitoneal residual masses.

A small proportion of men presenting with advanced, life-threatening testicular cancer may be treated with platinum-based chemotherapy before radical orchiectomy. After the appropriate course of primary systemic chemotherapy, these patients subsequently undergo delayed orchiectomy. Because this is a rare clinical scenario, there is a paucity of data on the effect of chemotherapy on the primary tumor. There are, however, a few studies examining the histology of orchiectomy specimens after chemotherapy and the histologic concordance of the testicular primary with other

Table 1

Histologic discordance between retroperitoneal and non-retroperitoneal postchemotherapy residual masses

Table 1

Histologic discordance between retroperitoneal and non-retroperitoneal postchemotherapy residual masses

Source, year

patients % Discordance

Mandelbaum et al, 1983 [23]

24

29

Donohue et al, 1987 [6]

24

29

Tiffany et al, 1986 [24]

23

35

Qvist et al, 1991 [19]

15

47

Brenner et al, 1996 [20]

24

25

Hartmann et al, 1997 [18]

27

30

Gels et al, 1997 [21]

20

50

Mohseni et al, 2002 [27]

16

44

McGuire et al, 2003 [17]

105

38

postchemotherapy sites of disease. One of the largest studies on this topic from Indiana University examined 160 patients who underwent postchemotherapy orchiectomy and RPLND [29]. They found teratoma in 31% and viable GCT in 25% of the orchiectomy specimens. Of the patients who had viable GCT in the testis after chemotherapy, 42.5% demonstrated a complete response (necrosis/scar) in the RPLND specimen. Moreover, there was less than 50% histologic concordance between the testicular and RPLND specimens. These results confirm that, like residual disease at other sites, postchemotherapy orchi-ectomy is indicated regardless of the response to chemotherapy at metastatic sites.

The management of brain metastases has traditionally been recognized as a separate clinical entity. The blood-brain barrier and cerebral edema are unique features of intracranial GCT, and the major concern has been the ability to obtain cytotoxic levels of chemotherapeutic drugs in the brain. Prognosis is based on the features of brain involvement. Mahalati and colleagues [30] and Spears and colleagues [31] have divided patients into three groups: (1) brain metastases at presentation, (2) isolated relapses in the brain, and (3) brain metastases with progressive disease. Oftentimes, residual intracranial disease is managed with a multimodality approach, including surgery, radiotherapy, and additional chemotherapy. Ma-halati and colleagues [30] reviewed their experience of 11 patients who had nonseminomatous GCT brain metastases. Patients were divided into the three groups described above. Four of 5 patients in group 1 were still alive (mean follow-up 13 months, range 3-47 months). Three of 5 patients in group 1 received chemotherapy, craniotomy, and radiotherapy. Radiation therapy was given to those who had viable GCT in the specimen.

The authors concluded that surgery was indicated for single residual brain lesions amenable to resection in patients in groups 1 and 2. Surgery was not performed for patients who had multiple brain metastases and progressive disease.

10 Ways To Fight Off Cancer

10 Ways To Fight Off Cancer

Learning About 10 Ways Fight Off Cancer Can Have Amazing Benefits For Your Life The Best Tips On How To Keep This Killer At Bay Discovering that you or a loved one has cancer can be utterly terrifying. All the same, once you comprehend the causes of cancer and learn how to reverse those causes, you or your loved one may have more than a fighting chance of beating out cancer.

Get My Free Ebook


Post a comment