Dense Embryonal or Alveolar Rhabdomyosarcoma?
Dense Embryonal or Alveolar Rhabdomyosarcoma?
Of the 255 cases originally classified as ARMS, 84 (33%) were classified as ERMS on re-review Table 1. A diagnosis of ARMS was confirmed in 148 cases (58%), whereas 18 (7%) were reclassified as mixed RMS and 5 (2%) were called RMS-NOS.
Cases reclassified as ERMS showed heterogeneous histology. Botryoid morphology was predominant in 3 of 84 (5%) cases of ERMS, and typical ERMS morphology accounted for 5 of 84 (6%) cases. A dense pattern of ERMS was the most commonly reclassified group and the primary reason for review discrepancy. Fifty-five percent (46/84) of reclassified ERMS had uniformly dense cellularity superficially resembling solid ARMS; however, their angulated nuclei, variably prominent nucleoli, and moderate (2–3+) myogenin expression were more consistent with a revised diagnosis of ERMS Image 2. An additional 17 of 84 (20%) cases had a partial or complete sclerosing pattern with weak (1–2+) myogenin expression. Ten of 84 (12%) ERMS cases exhibited codominant patterns with a combination of sclerosing/dense or spindled/sclerosing patterns. Finally, 2 of 84 (2%) cases were reclassified as spindle cell RMS, and 1 (1%) was reclassified as epithelioid RMS18 Figure 1.
(Enlarge Image)
Figure 1.
Distribution of histologic patterns in reclassified embryonal rhabdomyosarcoma. Combined histology indicates tumors with codominant patterns, including sclerosing/dense or sclerosing/spindled patterns.
(Enlarge Image)
Image 2.
A, The solid pattern of alveolar rhabdomyosarcoma (ARMS) shows monomorphic round cell cytology with vesicular chromatin and inconspicuous nucleoli (H&E, ×400). B, Myogenin expression is strong and diffuse in solid-variant ARMS (×200). C, The dense pattern of embryonal rhabdomyosarcoma (ERMS) shows uniformly dense cellularity superficially resembling solid ARMS; however, the angulated nuclei and variably prominent nucleoli allow diagnosis of ERMS (H&E, ×400). D, Dense ERMS shows moderate myogenin expression (×400).
Of control ERMS cases, 25 of 38 (66%) showed typical histology and 4 of 38 (11%) showed botryoid morphology. Only 6 of 38 (16%) had a dense pattern, 1 of 38 (3%) had a sclerosing pattern, and 2 of 38 (5%) had a combination of spindled/sclerosing or dense/sclerosing patterns.
When categorized by primary tumor site, genitourinary (GU) tract tumors not arising from the bladder or prostate and perineum, retroperitoneum, and trunk tumors were overrepresented among reclassified ERMS Table 2. Dense ERMS showed no obvious association with sites of higher reclassification, however. In fact, the GU non–bladder/prostate site showed the highest rates of reclassification but the lowest incidence of dense ERMS. Most GU non–bladder/prostate primary tumors had a focal or diffuse sclerosing pattern.
Fusion data were available for 173 cases Table 3 but not for 23 re-review cases. Fifty-seven (34%) original ARMS cases classified as ERMS or ARMS on re-review were fusion negative. After our second review, 83 (66%) ARMS cases showed a P3F fusion, and 20 (16%) confirmed ARMS cases showed a P7F translocation. In the final analysis, only 23 (18%) confirmed ARMS cases were fusion negative. Importantly, no reclassified ERMS case, including all dense variants, contained a gene fusion.
Myogenin expression data were available for 231 ARMS and ERMS cases. The distribution of myogenin scores for 71 of 84 reclassified ERMS cases was as follows: 6%, 0+ expression; 24%, 1+ expression; 59%, 2+ expression; and 13%, 3+ expression. None had 4+ expression. Myogenin stains were not available for evaluation for 13 reclassified ERMS cases. This distribution was similar for the 34 original ERMS cases with myogenin stains: none showed 0+ expression, 41% had 1+ expression, 44% had 2+ expression, and 15% had 3+ expression. None had 4+ expression. In total, 87% of original and reclassified ERMS cases showed 0 to 2+ myogenin expression, whereas only 5% of ARMS cases had 0 to 2+ myogenin expression Table 4. Cases with 3+ myogenin expression were nearly equally divided between ARMS and ERMS. The majority (79%) of ARMS cases showed 4+ myogenin expression, but no original or reclassified ERMS cases showed 4+ myogenin expression. For the 54 dense ERMS cases, 50% (27/54) showed 2+ myogenin expression, 17% (9/54) showed 1+ myogenin expression, and 19% (10/54) showed 3+ myogenin expression. Myogenin stains were not available for the remaining 8 dense ERMS cases.
Fusion status was available for a subset of 126 ARMS cases with myogenin expression data (Table 4). The fraction of ARMSn increased with decreased myogenin expression: 13% of ARMS cases with 4+ myogenin expression, 30% of ARMS cases with 3+ myogenin expression, and all ARMS cases with 2+ myogenin expression were fusion negative. Weak (0–1+) myogenin expression was found in a single ARMS case that harbored evidence of gene fusion; however, the weak staining appeared to be secondary to poor staining quality. Five of the 20 ARMS cases with 3+ myogenin expression and 15 of 99 ARMS cases with 4+ myogenin expression had unknown fusion status.
Analysis of outcome for D9803 cases was restricted to patients with stage 2/3, group III disease since these patients were eligible for the D9803 study irrespective of histology subtype. The estimated 5-year EFS was 77% (95% confidence interval [CI], 71%–82%) for original and confirmed ERMS (n = 238), 69% (95% CI, 50%–82%) for reclassified ERMS (n = 34), and 55% (95% CI, 44%–64%) for confirmed ARMS (n = 88) (P < .001) Figure 2A. The estimated 5-year OS was 81% (95% CI, 75%–85%) for original and confirmed ERMS, 81% (95% CI, 63%–91%) for reclassified ERMS, and 68% (95% CI, 57%–77%) for confirmed ARMS (P = .018) Figure 2B. The outcome for patients with ARMS was poorer than that for those with ERMS, and cases reclassified from ARMS to ERMS had an outcome more similar to ERMS than ARMS.
(Enlarge Image)
Figure 2.
Event-free survival (P < .001) (A) and overall survival (P = .018) (B) of stage 2/3, group III reclassified embryonal rhabdomyosarcoma (rcERMS) cases vs original and confirmed alveolar rhabdomyosarcoma (ARMS) and ERMS cases enrolled in the D9803 study.
Results
Re-review of D9803 Cases Previously Classified as ARMS
Of the 255 cases originally classified as ARMS, 84 (33%) were classified as ERMS on re-review Table 1. A diagnosis of ARMS was confirmed in 148 cases (58%), whereas 18 (7%) were reclassified as mixed RMS and 5 (2%) were called RMS-NOS.
Cases reclassified as ERMS showed heterogeneous histology. Botryoid morphology was predominant in 3 of 84 (5%) cases of ERMS, and typical ERMS morphology accounted for 5 of 84 (6%) cases. A dense pattern of ERMS was the most commonly reclassified group and the primary reason for review discrepancy. Fifty-five percent (46/84) of reclassified ERMS had uniformly dense cellularity superficially resembling solid ARMS; however, their angulated nuclei, variably prominent nucleoli, and moderate (2–3+) myogenin expression were more consistent with a revised diagnosis of ERMS Image 2. An additional 17 of 84 (20%) cases had a partial or complete sclerosing pattern with weak (1–2+) myogenin expression. Ten of 84 (12%) ERMS cases exhibited codominant patterns with a combination of sclerosing/dense or spindled/sclerosing patterns. Finally, 2 of 84 (2%) cases were reclassified as spindle cell RMS, and 1 (1%) was reclassified as epithelioid RMS18 Figure 1.
(Enlarge Image)
Figure 1.
Distribution of histologic patterns in reclassified embryonal rhabdomyosarcoma. Combined histology indicates tumors with codominant patterns, including sclerosing/dense or sclerosing/spindled patterns.
(Enlarge Image)
Image 2.
A, The solid pattern of alveolar rhabdomyosarcoma (ARMS) shows monomorphic round cell cytology with vesicular chromatin and inconspicuous nucleoli (H&E, ×400). B, Myogenin expression is strong and diffuse in solid-variant ARMS (×200). C, The dense pattern of embryonal rhabdomyosarcoma (ERMS) shows uniformly dense cellularity superficially resembling solid ARMS; however, the angulated nuclei and variably prominent nucleoli allow diagnosis of ERMS (H&E, ×400). D, Dense ERMS shows moderate myogenin expression (×400).
Of control ERMS cases, 25 of 38 (66%) showed typical histology and 4 of 38 (11%) showed botryoid morphology. Only 6 of 38 (16%) had a dense pattern, 1 of 38 (3%) had a sclerosing pattern, and 2 of 38 (5%) had a combination of spindled/sclerosing or dense/sclerosing patterns.
Histology and Primary Site
When categorized by primary tumor site, genitourinary (GU) tract tumors not arising from the bladder or prostate and perineum, retroperitoneum, and trunk tumors were overrepresented among reclassified ERMS Table 2. Dense ERMS showed no obvious association with sites of higher reclassification, however. In fact, the GU non–bladder/prostate site showed the highest rates of reclassification but the lowest incidence of dense ERMS. Most GU non–bladder/prostate primary tumors had a focal or diffuse sclerosing pattern.
Histology and Fusion Status
Fusion data were available for 173 cases Table 3 but not for 23 re-review cases. Fifty-seven (34%) original ARMS cases classified as ERMS or ARMS on re-review were fusion negative. After our second review, 83 (66%) ARMS cases showed a P3F fusion, and 20 (16%) confirmed ARMS cases showed a P7F translocation. In the final analysis, only 23 (18%) confirmed ARMS cases were fusion negative. Importantly, no reclassified ERMS case, including all dense variants, contained a gene fusion.
Myogenin Expression by Histology and Fusion Status
Myogenin expression data were available for 231 ARMS and ERMS cases. The distribution of myogenin scores for 71 of 84 reclassified ERMS cases was as follows: 6%, 0+ expression; 24%, 1+ expression; 59%, 2+ expression; and 13%, 3+ expression. None had 4+ expression. Myogenin stains were not available for evaluation for 13 reclassified ERMS cases. This distribution was similar for the 34 original ERMS cases with myogenin stains: none showed 0+ expression, 41% had 1+ expression, 44% had 2+ expression, and 15% had 3+ expression. None had 4+ expression. In total, 87% of original and reclassified ERMS cases showed 0 to 2+ myogenin expression, whereas only 5% of ARMS cases had 0 to 2+ myogenin expression Table 4. Cases with 3+ myogenin expression were nearly equally divided between ARMS and ERMS. The majority (79%) of ARMS cases showed 4+ myogenin expression, but no original or reclassified ERMS cases showed 4+ myogenin expression. For the 54 dense ERMS cases, 50% (27/54) showed 2+ myogenin expression, 17% (9/54) showed 1+ myogenin expression, and 19% (10/54) showed 3+ myogenin expression. Myogenin stains were not available for the remaining 8 dense ERMS cases.
Fusion status was available for a subset of 126 ARMS cases with myogenin expression data (Table 4). The fraction of ARMSn increased with decreased myogenin expression: 13% of ARMS cases with 4+ myogenin expression, 30% of ARMS cases with 3+ myogenin expression, and all ARMS cases with 2+ myogenin expression were fusion negative. Weak (0–1+) myogenin expression was found in a single ARMS case that harbored evidence of gene fusion; however, the weak staining appeared to be secondary to poor staining quality. Five of the 20 ARMS cases with 3+ myogenin expression and 15 of 99 ARMS cases with 4+ myogenin expression had unknown fusion status.
Outcome of Reclassified ERMS
Analysis of outcome for D9803 cases was restricted to patients with stage 2/3, group III disease since these patients were eligible for the D9803 study irrespective of histology subtype. The estimated 5-year EFS was 77% (95% confidence interval [CI], 71%–82%) for original and confirmed ERMS (n = 238), 69% (95% CI, 50%–82%) for reclassified ERMS (n = 34), and 55% (95% CI, 44%–64%) for confirmed ARMS (n = 88) (P < .001) Figure 2A. The estimated 5-year OS was 81% (95% CI, 75%–85%) for original and confirmed ERMS, 81% (95% CI, 63%–91%) for reclassified ERMS, and 68% (95% CI, 57%–77%) for confirmed ARMS (P = .018) Figure 2B. The outcome for patients with ARMS was poorer than that for those with ERMS, and cases reclassified from ARMS to ERMS had an outcome more similar to ERMS than ARMS.
(Enlarge Image)
Figure 2.
Event-free survival (P < .001) (A) and overall survival (P = .018) (B) of stage 2/3, group III reclassified embryonal rhabdomyosarcoma (rcERMS) cases vs original and confirmed alveolar rhabdomyosarcoma (ARMS) and ERMS cases enrolled in the D9803 study.
Source...