diff --git a/pdfplumber/table.py b/pdfplumber/table.py
index f189742e..102e48cd 100644
--- a/pdfplumber/table.py
+++ b/pdfplumber/table.py
@@ -357,12 +357,14 @@ def get_centroids(indices):
                 yield (vertices[index - 1] + vertices[index]) / 2
 
         def get_cell_by_centroid(centroid_x, centroid_y):
-            return next(
+            cells = list(
                 filter(
                     lambda c: c[0] < centroid_x < c[2] and c[1] < centroid_y < c[3],
                     self.cells,
                 )
             )
+            if cells:
+                return cells[0]
 
         x_centroids = list(get_centroids((0, 2)))
         y_centroids = list(get_centroids((1, 3)))
diff --git a/tests/pdfs/issue-420-example.pdf b/tests/pdfs/issue-420-example.pdf
index 7ecd215f..e719c289 100644
Binary files a/tests/pdfs/issue-420-example.pdf and b/tests/pdfs/issue-420-example.pdf differ
diff --git a/tests/test_table.py b/tests/test_table.py
index eb48c905..d31e0e57 100644
--- a/tests/test_table.py
+++ b/tests/test_table.py
@@ -102,13 +102,13 @@ def test_merge_cell(self):
         See issue #420
         """
         path = os.path.join(HERE, "pdfs/issue-420-example.pdf")
-        TABLE_0 = [
+        TABLE_0_0 = [
             ["Header 1", "Header 2", "Header 3", "Header 4"],
             ["Merged cell 1", "Merged cell 2", "Cell 3-1", "Cell 4-1"],
             ["Merged cell 1", "Merged cell 2", "Cell 3-2", "Cell 4-2"],
             ["Merged cell 1", "Merged cell 2", "Cell 3-2", "Cell 4-3"],
         ]
-        TABLE_1 = [
+        TABLE_0_1 = [
             ["1-1", "1-1", "1-3", "1-4", "1-5", "1-6"],
             ["2-1", "2-2", "2-3", "2-4", "1-5", "1-6"],
             ["3-1", "3-2", "3-3", "3-4", "1-5", "1-6"],
@@ -118,9 +118,14 @@ def test_merge_cell(self):
             ["7-1", "7-1", "7-1", "7-1", "1-5", "1-6"],
             ["7-1", "7-1", "7-1", "7-1", "7-5", "1-6"],
         ]
+        TABLE_1_0 = [["1", "2"], [None, "4"]]
 
         with pdfplumber.open(path) as pdf:
             tables = pdf.pages[0].extract_tables()
             assert len(tables) == 2
-            assert TABLE_0 == tables[0]
-            assert TABLE_1 == tables[1]
+            assert TABLE_0_0 == tables[0]
+            assert TABLE_0_1 == tables[1]
+
+            tables_missing_edges = pdf.pages[1].extract_tables()
+            assert len(tables_missing_edges) == 1
+            assert TABLE_1_0 == tables_missing_edges[0]